<%BANNER%>

Analysis, Testing and Verification of the Behavior of Composite Pavements Under Florida Conditions Using a Heavy Vehicle...

University of Florida Institutional Repository
Permanent Link: http://ufdc.ufl.edu/UFE0021631/00001

Material Information

Title: Analysis, Testing and Verification of the Behavior of Composite Pavements Under Florida Conditions Using a Heavy Vehicle Simulator
Physical Description: 1 online resource (247 p.)
Language: english
Creator: Tapia, Patricio E
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: asphalt, bond, calibration, composite, concrete, florida, hvs, interface, model, pavement, whitetopping
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre: Civil Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Whitetopping (WT) is a rehabilitation method to resurface deteriorated asphalt pavements. While some of these composite pavements have performed very well carrying heavy load, other have shown poor performance with early cracking. With the objective of analyzing the applicability of WT pavements under Florida conditions, a total of nine full-scale WT test sections were constructed and tested using a Heavy Vehicle Simulator (HVS) in the APT facility at the FDOT Material Research Park. The test sections were instrumented to monitor both strain and temperature. A 3-D finite element model was developed to analyze the WT test sections. The model was calibrated and verified using measured FWD deflections and HVS load-induced strains from the test sections. The model was then used to evaluate the potential performance of these test sections under critical temperature-load condition in Florida. Six of the WT pavement test sections had a bonded concrete-asphalt interface by milling, cleaning and spraying with water the asphalt surface. This method produced excellent bonding at the interface, with shear strength of 195 to 220 psi. Three of the test sections were intended to have an unbonded concrete-asphalt interface by applying a debonding agent in the asphalt surface. However, shear strengths between 119 and 135 psi and a careful analysis of the strain and the temperature data indicated a partial bond condition. The computer model was able to satisfactorily model the behavior of the composite pavement by mainly considering material properties from standard laboratory tests and calibrating the spring elements used to model the interface. Reasonable matches between the measured and the calculated strains were achieved when a temperature-dependent AC elastic modulus was included in the analytical model. The expected numbers of repetitions of the 24-kip single axle loads at critical thermal condition were computed for the nine test sections based on maximum tensile stresses and fatigue theory. The results showed that 4 inch slabs can be used for heavy loads only for low-volume traffic. To withstand the critical load without fear of fatigue failure, 6 inch slabs and 8 inch slabs would be needed for joint spacings of 4 feet and 6 feet, respectively.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Patricio E Tapia.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Tia, Mang.
Local: Co-adviser: Najafi, Fazil T.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021631:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021631/00001

Material Information

Title: Analysis, Testing and Verification of the Behavior of Composite Pavements Under Florida Conditions Using a Heavy Vehicle Simulator
Physical Description: 1 online resource (247 p.)
Language: english
Creator: Tapia, Patricio E
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: asphalt, bond, calibration, composite, concrete, florida, hvs, interface, model, pavement, whitetopping
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre: Civil Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Whitetopping (WT) is a rehabilitation method to resurface deteriorated asphalt pavements. While some of these composite pavements have performed very well carrying heavy load, other have shown poor performance with early cracking. With the objective of analyzing the applicability of WT pavements under Florida conditions, a total of nine full-scale WT test sections were constructed and tested using a Heavy Vehicle Simulator (HVS) in the APT facility at the FDOT Material Research Park. The test sections were instrumented to monitor both strain and temperature. A 3-D finite element model was developed to analyze the WT test sections. The model was calibrated and verified using measured FWD deflections and HVS load-induced strains from the test sections. The model was then used to evaluate the potential performance of these test sections under critical temperature-load condition in Florida. Six of the WT pavement test sections had a bonded concrete-asphalt interface by milling, cleaning and spraying with water the asphalt surface. This method produced excellent bonding at the interface, with shear strength of 195 to 220 psi. Three of the test sections were intended to have an unbonded concrete-asphalt interface by applying a debonding agent in the asphalt surface. However, shear strengths between 119 and 135 psi and a careful analysis of the strain and the temperature data indicated a partial bond condition. The computer model was able to satisfactorily model the behavior of the composite pavement by mainly considering material properties from standard laboratory tests and calibrating the spring elements used to model the interface. Reasonable matches between the measured and the calculated strains were achieved when a temperature-dependent AC elastic modulus was included in the analytical model. The expected numbers of repetitions of the 24-kip single axle loads at critical thermal condition were computed for the nine test sections based on maximum tensile stresses and fatigue theory. The results showed that 4 inch slabs can be used for heavy loads only for low-volume traffic. To withstand the critical load without fear of fatigue failure, 6 inch slabs and 8 inch slabs would be needed for joint spacings of 4 feet and 6 feet, respectively.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Patricio E Tapia.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Tia, Mang.
Local: Co-adviser: Najafi, Fazil T.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021631:00001


This item has the following downloads:


Full Text





ANALYSIS, TESTING AND VERIFICATION OF THE BEHAVIOR OF COMPOSITE
PAVEMENTS UNDER FLORIDA CONDITIONS USING A HEAVY VEHICLE
SIMULATOR




















By

PATRICIO ENRIQUE TAPIA GUTIERREZ


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007


































2007 Patricio Enrique Tapia Gutierrez



























To my wife Juana, my son Pablo, Gabriela, Ana, and my family in Chile









ACKNOWLEDGMENTS

I thank my adviser for his constant support during the course of this investigation. His

knowledge and dedication was crucial in the success of this research. I also thank to all members

of the committee for their valuable contribution to this dissertation. The Florida Department of

Transportation (FDOT) is also gratefully acknowledged for providing the financial support for

this research. The FDOT Material Office provided the necessary testing equipment, materials

and personnel for this investigation. Personnel of the Department of Civil and Coastal

Engineering who helped in conducting laboratory tests are also gratefully acknowledged. The

Universidad Catolica del Norte which provided financial support for my staying in the USA and

the Fulbright Commission in Chile which granted the Fulbright-Laspau Scholarship, are also

gratefully acknowledged. Finally I would like to thank my family here and in Chile and my

colleagues in the Department of Civil Engineering at UCN for their constant support.









TABLE OF CONTENTS
page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IST O F TA B LE S ......... .... ........................................................................... 9

LIST OF FIGURES .................................. .. .... ..... ................. 12

A B S T R A C T ............ ................... .................. ........................................ 2 2

CHAPTER

1 INTRODUCTION ............... .......................................................... 24

1.1 R research N eed ............................................................................24
1.2 Problem Statem ent .................. .................. .................. .......... .. ............ 26
1.3 R research H ypothesis......... .................................................................... .......... ....... 26
1.4 O bjectiv es of R research ...................................................................... ........................ 27
1.5 Approach and Scope of Research ...................................................................... 27
1.6 Significance of Research ........... .......................................................................... 28

2 LITERATURE REVIEW ON WHITETOPPING......................................30

2 .1 G en eral C on cepts ................................................................................... ....................30
2.1.1 U ltra-Thin W hitetopping ............................ ........... ..................... ............... 31
2.1.2 T hin W hitetopping ............. ............. .. ........ .......................... .. .......... .... 3 1
2.1.3 Conventional W hitetopping......................................................... ............... 32
2.2 Concrete Mixture Proportions and Properties ...................................... ............... 32
2.3 C construction P procedures ............................ ........... ................................. ............... 34
2.4 UTW and TW T Design Considerations ........................................ ........ ............... 35
2 .4 .1 S lab T h ic k n e ss ........................................... ...................................................... 3 5
2 .4 .2 Joint Sp acing ................................................................3 5
2.4.3 Interface B ending Strength................................ ........................................ 36
2.5 Design Procedure for UTW and TWT Pavements ................................. ...............37
2.6 Perform ance of U TW and TW T Projects ......... ............................... .....................40
2.7 Accelerated Pavement Testing and Field Testing of UTW ........................................42
2 .8 A n aly tical M o d els............................................................................................... .. 4 4

3 INSTRUMENTATION AND CONSTRUCTION OF TEST SECTIONS ............................48

3.1 D description of the T testing Phases ......................................................... .....................48
3.2 L ayout of the T est Sections ............................................................... ..... .................... 49
3 .2 .1 P h a se I ............................................................................. 4 9
3.2.2 Phase II ..................................................................... ......... 49
3.3 L ayout of the Instrum entation ............... .....................................................................50
3.3.1 W heatstone B ridge Circuits......................................................... ............... 50
3.3.2 Prelim inary Stress A nalysis......................................................... ............... 50









3.3.3 Instrum entation L ayout ................................................ .............................. 51
3 .3 .3 .1 P h ase I ........................ .. ........................ .. .. .. .. .. ...... .... .... 5 1
3 .3 .3 .2 P h a se II ..............................................................5 3
3 .4 H V S L o ad in g P lan ...................................................................................................... 54
3 .5 D ata C o lle ctio n ........................................................................................................... 5 4
3.6 Construction of the Test Tracks......... ................. ........................ ......... ............... 55
3.6.1 Construction of Concrete Test Tracks in Phase I .......... .................................55
3.6.1.1 Asphalt surface preparation and formwork................................................55
3.6.1.2 Concrete mix proportions........................ .................... ............... 55
3.6.1.3 Placem ent of concrete ............................................................................ 56
3.6.1.4 Placement of strain gages ......... ............... ............... ............... 56
3.6.1.5 Placement of thermocouples ............. ......... ... ...................... 57
3.6.2 Construction of Concrete Test Tracks in Phase II......................................... 57
3.6.2.1 Asphalt surface preparation and formwork...............................................57
3.6.2.2 Installation of strain gages and thermocouples and concrete placement......58

4 MATERIALS AND PAVEMENT CHARACTERIZATION.................. ...............74

4 .1 M materials C characterization ..................................................................... .....................74
4.1.1 Interface B ond Strength............................ ....... .. ................................. 74
4.1.1.1 Results from test sections in Phase I-a.............................. ..................... 74
4.1.1.2 Results from test sections in Phase I-b ............ ...............................75
4.1.1.3 R results from test sections in Phase II...................................... ..................75
4.1.1.4 Comparison before and after HVS loading ............................................76
4 .1.2 C concrete P properties ......................................... ...................................................76
4.1.2.1 Properties of concrete sampled from concrete trucks ................................76
4.1.2.2 Properties of concrete from core samples .......................................... 76
4.1.3 Asphalt Concrete Properties .......... ................................ 77
4.2 Measurement of Joint Movement...................... ......... ..... ............... 78
4.3 Measurement of Slab Profile Using a Dipstick ..................................... ...............78
4 .4 F W D T e sts .................................................................................................................. 7 9
4.4.1 FW D Tests in Phase I-a ...................................................................... 80
4.4.2 FW D Tests in Phase I-b ........................................................................... ...... 80
4.4.3 FW D Tests Phase II............... .... .... ................ ........ ...... .. 80

5 TESTING OF TEST SECTIONS AND DATA ANALYSIS ...........................................102

5.1 H V S L loading of T est Sections ......... ............... ......................................................... 102
5.1.1 HVS Loading of Test Sections in Phase I-a................................. ..................102
5.1.2 HVS Loading of Test Sections in Phase I-b ............... ................ ......... ........ 103
5.1.3 HVS Loading of Test Sections in Phase II................................. ............... 104
5.2 A analysis of Tem perature D ata........................... ........................... ............... 104
5.2.1 Tem perature D ifferential .............. ................ .............. ............................. 104
5.2.2 Temperature in the AC layer ...... ........... .............. ............... 107
5.2.3 Tem perature D distribution ............................................. ............................ 108
5.2.4 Summary of temperature analysis ............. ......................................................108
5.3 A analysis of Strain D ata...... ............... .............................................. ............... 109


6









5.3.1 Dynamic Strain versus Static Strain.................................................................. 109
5.3.2 M measured strain and calculated strain............... .... .. .... ..... ............... 110
5.3.3 Effect of Temperature on the Strain .......................... ......... .... ........... 112
5.3.4 Effect of the load magnitude on the strain...................................... ................ 113
5.3.5 Effect of the loading period on the strain............. ............................................ 114
5.3.6 Evaluation of the bond condition using strain ratios................. .....................116

6 DEVELOPMENT OF A 3-D FINITE ELEMENT MODEL................................158

6.1 F inite E lem ent Program ................... ... .. ..................................................................... 158
6.2 Six-Slab and Twelve-Slab 3-D Finite Element Models ...........................................158
6.3 Solid 20-N ode Finite Elem ent ................................................. ........................ 159
6.4 M odeling of Concrete Slab Joints ........................................................ ............... 160
6 .5 M odeling of M materials ......................................................................... ..................... 160
6.6 Modeling of Concrete-Asphalt Interface ........................ ............................160
6.7 Modeling of Loads and Temperature Effects............................................... 162

7 MODEL CALIBRATION AND VERIFICATION...........................................169

7.1 Overview of M odel Calibration............................................. ............................. 169
7.2 Deflection-Based Calibration of Model Parameters.................................................170
7.2 .1 P hases I-a and I-b ....................... .. ........................ .. ...... ............... 170
7.2.2 Phase II ......... ..... .. .......... ..... ..... ....... .......................... 172
7.3 Strain-Based Calibration of Model Parameters............. ............................. ........174
7.3.1 G general A approach ............. ........................................................ ...... .... ...... 174
7 .3 .2 P h a se I-a ...............................................................17 5
7 .3 .3 P h a se I-b ...............................................................17 7
7.3.4 Phase II .............. ..... ......... ........................179
7.4 Summary of Calibration Results ...................... ......... ..................180

8 EVALUATION OF POTENTIAL PERFORMANCE OF THE WT DESIGNS .................211

8.1 O verview ..................................................................... ...... ................. 2 11
8.2 Assumptions for the evaluation of the potential performance of the test sections .........211
8.2.1 Critical loading conditions ........... ......... ......... ..................211
8.2.2 M odel param eters ............................................................. 2 12
8.3 Results of Critical Stress Analysis........................................ 212
8.3.1 Maximum stresses in the concrete slabs.....................................212
8.3.1.1 Effects of elastic modulus of AC layer ................................................213
8.3.1.2 Effects of temperature differential ................... .. .......... ........ 213
8.3.1.3 Effects of panel size ................................ ......... ........................ ..............214
8.3.1.4 Effects of bonded versus partially bonded interface .............................215
8.3.2 Maximum shear stresses at the interface.....................................................216
8.3.3 M maximum stresses in the AC layer.................. ........... ............ ............... 217
8.4 Potential performance of the test sections ......... ................................. ..........218
8.5 Evaluation of the actual performance of the test sections .............................................219









9 CONCLUSION................ ..... .. .......... ........... ............... .. 237

9.1 Summary of Findings ........................................ ....................... ............... 237
9.1.1 Bond Strength at the Concrete-Asphalt Interface.....................................237
9.1.2 Development of the 3-D Finite Element Model........................... ............... 238
9.1.2.1 D eflection-based calibration ........................................... ............... 238
9.1.2.2 Strain-based calibration........................................................... ... .......... 239
9.1.2.3 L oad transfer ....................... ................ ................... ........ 239
9.1.2.4 Interface bond.................................................. .......... ............... 239
9.1.3 Stress analysis of the test sections ................. ..................... ..................... 239
9.1.3.1 Effects of elastic m odulus of AC .................................... ............... 239
9.1.3.2 Effects of concrete panel size.......................... .................................. 240
9.1.3.3 Effects of bonded versus partially bonded interface ...............................240
9.1.4 Performance of the Test Sections............... .................... .................240
9.1.4.1 Potential perform ance ........................................ .......................... 240
9.1.4.2 M mechanism of failure ............................................................. ...............241
9.2 Lim stations of the R research .................................. .......... ............... ............... 241
9.3 Recom m endations and future w ork......................................................... ............... 242

L IST O F R EFER EN CE S ....................................................... ....................... ............... 244

B IO G R A PH IC A L SK E T C H ............................................................................. ....................247































8









LIST OF TABLES


Table page

2-1 Concrete Mix Proportions Used in Louisville Experimental Project (Riser et al
1993) ......................................................... ....................................4 6

2-2 Concrete Mix Proportions Used in Leawood, Kansas (Wu et al. 1997)............................46

2-3 Concrete Mix Proportions Used in SNH, Tennessee (Saeed et al. 2002).....................46

2-4 Mix proportions for eight whitetopping projects in Illinois............................ ...........47

3-1 M ix Designs of Concrete Used in Phases I and II. ................................. .................59

4-1 Results of the Iowa Shear Tests on the cored samples from test sections in Phase I-a
after HV S loading. ....................... ........ .. ... ... .. ...... ............82

4-2 Results of Iowa Shear Tests on the cored samples from test sections in Phase I-b
after HVS loading. ....................... ........ .. ... ... .. .................. 83

4-3 Results of Iowa Shear Tests on cores from test sections in Phase II after HVS
loading .................. ........... .......................... ...........................84

4-4 Summary of the interface bond strength before and after HVS loading .........................84

4-5 Properties of fresh concrete used. .......................................... ...............85

4-6 Properties of hardened concrete sampled from truck in Phase I-a............................... 85

4-7 Properties of the hardened concrete sampled from truck in Phase II. ............................85

4-8 Results of Indirect Tensile Strength Test on the concrete samples taken from the test
sections in Phase I-a after HVS loading. ........................................ ....................... 86

4-9 Results of Resilient Modulus and Indirect Tensile Strength Tests on the asphalt
concrete samples obtained from test sections in all phases after HVS loading...............86

4-10 Results of Penetration and Absolute Viscosity Tests on the recovered asphalt binders
from cores from Phase I-a after HVS loading. ...................................... ............... 87

5-1 HVS loading period and number of 12-kip wheel passes on the test sections in Phase
I -b ....................................................................... ................ 1 2 0

5-2 Summary of HVS loading on test sections in Phase II ............. .... ...............120

5-3 Number of hours in a day when the temperature differential was in certain ranges for
the 4-inch slab in Phase I-a. ...................................................................... .................. 120









5-4 Number of hours in a day when the temperature differential was in certain ranges for
the 5-inch slab in Phase I-a. ...................................................................... .................. 120

5-5 Extreme values for temperature differential and temperature in the AC Layer. .............121

5-6 Measured static and dynamic strains for gages 1, 2 and 5 in the 4-inch slab in Phase
I-a caused by 9 and 12-kip loads............................................. ............................. 121

5-7 Measured static and dynamic strains for gages 2 and 4 in the 5-inch slab in Phase I-a
caused by 9 and 12-kip loads. ............................................... ................................ 122

5-8 Measured static and dynamic strains for gages 2 and 4 in the 5-inch slab in Phase I-a
caused by 15- and 18-kip loads............................................... ............................. 122

5-9 Verified depths of strain gages in Phase I-b ............................................ ..................123

5-10 Verified depths of strain gages in Phase II. ........................................ ............... 123

5-11 Strain Ratios to evaluate the degree of bond in the interface .......................................123

7-1 Extreme values of the AC resilient modulus based on extreme temperature during the
H V S test, using E q. 7.1 ............................................................. .............. .. .. ........ .. 182

7-2 Elastic modulus and Poisson's ratio of the pavement materials used in the 3-D finite
elem ent m odel................................... ................................. .......... 182

7-3 HV S loading periods for Phase I-a. ............................................................................ 182

7-4 HV S loading periods for Phase I-b. ........................................................................... 182

7-5 HVS loading periods for Phase II. ..... ......................................................................183

7-6 Summary of the best estimated parameters of the 3-D model for all test sections..........184

8-1 Model parameters of the 3-D model for each test section used in the analysis.............222

8-2 Maximum tensile stresses in the concrete slabs caused by a 24-kip single axle Load
at various critical loading conditions. ........................................ ......................... 223

8-3 Maximum shear stress at the concrete-asphalt interface caused by a 24-kip single
axle load at a temperature differential of +20 F for the bonded slabs .........................224

8-4 Maximum tensile stresses in the asphalt concrete layer caused by a 24-kip single
Axle load at various critical loading conditions. .................................. .................225

8-5 Computed stress ratio in the concrete and allowable number of 24-kip single axle
loads under critical loading conditions for the test sections evaluated in this study. ......226









8-6 Level of tensile stresses in the 4" slab that failed in Phase I-a. Load applied at the
c o rn er ...........................................................................................2 2 6









LIST OF FIGURES


Figure page

2-1 Transition between UTW and Adjoining Asphalt Pavement (ACPA, 1998)....................47

3-1 Layout of the Test Sections for Phase I. ........................... ......................... ............. 60

3-2 Layout of the Test Sections for Phase II............................ ....... .....................60

3-3 Strain Gage Arrangements in a Half Bridge Circuit....................................................61

3-4 Connection of the Active and Dummy Strain Gages in the Half Bridge Circuit...............61

3-5 Instrumentation Layout for the Test Slabs in Phase I-a..................................................62

3-6 Vertical Positions of the Strain Gages in Phase I-a. .................................. ... ..................63

3-7 Vertical Positions of the Thermocouples for the 4", 5" and 6" Slabs in Phase I...............64

3-8 Instrumentation Layout for the Test Slabs in Phase I-b...............................................65

3-9 Vertical Positions of the Strain Gages in Phase I-b. .......................................................66

3-10 Instrum entation layout for Phase II............................................... ............... 66

3-11 Vertical positions of the strain gages in Phase II........................................................... 67

3-12 Vertical positions of thermocouples in Phase II. ........................................67

3-13 Milled surface before concrete placement on Lane 6 in Phase I-a. ..................................68

3-14 Formwork prepared for Lane 7 in Phase I-b............. ....................................... 68

3-15 After placement of concrete on Lane 6 in Phase I-a.................. ................ ............... 69

3-16 Placement of top and bottom strain gages on Lane 7 in Phase I-b. ..................................69

3-17 Placement of surface strain gages at a joint in Phase I-a. .............................................70

3-18 Strain gages in a protective PVC pipe before placing concrete ...................................70

3-19 Removal of concrete slabs from Lane 6 in Phase I-a........ ..................................71

3-20 Form w ork for test slabs in Phase II. ........................................... ........................... 71

3-21 Grooves on asphalt surface for placement of strain gages and thermocouples cables
in P hase II. ................................................................................72









3-22 Asphalt surface with white curing compound before concrete placement in Phase II......72

3-23 Finishing of the concrete for the test track in Phase II. ................ ..............................73

3-24 Curing of concrete by sprinkling w ith w after ........................................ .....................73

4-1 Cores samples from Lane 6 in Phase I-a after HVS loading. .........................................88

4-2 Location of the cores taken after loading in Phase I-a.................................................89

4-3 Location of the cores taken after loading for each test section in Phase I-b....................90

4-4 Relationship between Temperature and Resilient Modulus for the AC layer in the
com posite pavem ent .................. ........................ .... ................ ..... 90

4-5 The W hitmore gage with invar bar. .................................. ..................................91

4-6 Measured gage spacing from a 4-inch slab in Lane 6..................... ................................. 92

4-7 Measured gage spacing from a 5-inch slab in Lane 6..................... ................................. 92

4-8 Measured gage spacing from a 6-inch slab in Lane 6..................... ................................. 93

4-9 Changes of joint spacing on a selected day. ........................................................ 93

4-10 Grid marked on slabs for the Dipstick measurement ......................................................94

4-11 T he D lipstick instrum ent .................................................................................. .......... 94

4-12 Dipstick measurements at two critical temperatures. .............................. ................95

4-13 FWD load and sensor locations for FWD Test at slab center in Phase I-a......................96

4-14 FWD load and sensor locations for FWD Test at slab corner in Phase I-a. ....................96

4-15 FWD load and sensor locations for FWD Test at the slab corer in Phase I-b ................97

4-16 FWD load and sensor locations for FWD Test at the slab edge in Phase I-b ..................97

4-17 FWD load and sensor locations for FWD Test at the slab center in Phase I-b..................98

4-18 FWD Test at the slab corner and measuring deflections on the opposite slab. ................98

4-19 FWD Test at the mid-edge and measuring deflections in the loaded slab.......................99

4-20 FWD Test at the center and measuring deflections along the transverse center line. .......99

4-21 FWD Testing Plan for the mid-edge and corer load in Phase II..................................100

4-22 FWD Testing Plan for the center load at the two ends of the test track in Phase II. .......100









4-23 Comparative analysis of load transfer factor for the test sections in Phase I-b and II.....101

5-1 Corner cracks on 4-inch slabs in Phase I-a after 21-kip wheel loads. ...........................124

5-2 Shrinkage cracks on a 4-inch test slab in Phase I-a. .............................. ......... ...... .124

5-3 Shrinkage cracks on a 5-inch test slab in Phase I-a. .............................. ......... ...... .125

5-4 Shrinkage cracks on a 6-inch concrete slab in Phase I-a. ..............................................125

5-5 Temperature differential variation in the 4-inch slab in Phase I-a ..............................126

5-6 Temperature differential variation in the 5-inch slab in Phase I-a ...........................126

5-7 Temperature differential variation in the 6-inch slab in Phase I-a ...........................127

5-8 Temperature differential variation in the 6-inch slabs in Phase I-b.............................127

5-9 Temperature differential variation in the 5-inch slabs in Phase I-b...........................128

5-10 Temperature differential variation in the 4-inch slabs in Phase I-b..............................128

5-11 Temperature differential variation in the 10-inch slab in Phase II ...............................129

5-12 Temperature differential variation in the 8-inch slab in Phase II ..................................129

5-13 Temperature differential variation in the 6-inch slab in Phase II ..................................130

5-14 Temperature variation on the surface of the asphalt layer for the 4" slab in Phase I-a. .130

5-15 Temperature variation on the surface of the asphalt layer for the 5" slab in Phase I-a. .131

5-16 Temperature variation on the surface of the asphalt layer for the 6" slab in Phase I-a. .131

5-17 Temperature on the surface of the AC layer for the 6-inch slab in Phase I-b................132

5-18 Temperature on the surface of the AC layer for the 5" slab in Phase I-b........................132

5-19 Temperature on the surface of the AC layer for the 4" slab in Phase I-b........................133

5-20 Temperature on the surface of the AC layer in the 10-inch slab in Phase II .................133

5-21 Temperature on the surface of the AC layer in the 8-inch slab in Phase II. ....................134

5-22 Temperature on the surface of the AC layer in the 6-inch slab in Phase II. ....................134

5-23 Temperature distribution along the depth of the 4 in concrete slab in Phase I-a at
maximum positive temperature differential ..........................................................135









5-24 Temperature distribution along the depth of the 5in concrete slab in Phase I-a at
maximum positive temperature differential ...........................................................135

5-25 Temperature distribution along the depth of the 6in concrete slab in Phase I-a at
m aximum positive temperature differential..................... ........ .................................. 136

5-26 Temperature distribution along the depth of the 4in concrete slab in Phase I-b at
maximum temperature differentials...................... ..... ............................. 136

5-27 Temperature distribution along the depth of the 5in concrete slab in Phase I-b at
maximum temperature differentials...................... ..... ............................. 137

5-28 Temperature distribution along the depth of the 6in concrete slab in Phase I-b at
maximum temperature differentials...................... ..... ............................. 137

5-29 Temperature distribution along the depth of the 6in concrete slab in Phase II at
maximum temperature differentials...................... ..... ............................. 138

5-30 Temperature distribution along the depth of the 8in concrete slab in Phase II at
maximum temperature differentials...................... ..... ............................. 138

5-31 Temperature distribution along the depth of the 10in concrete slab in Phase II at
maximum temperature differentials...................... ..... ............................. 139

5-32 Comparison of dynamic and static strain for gage 1 in the 4-inch slab in Phase I-a.......139

5-33 Comparison between static and dynamic strain for Gages 2 and 5 in the 4-inch slab
in P h a se I-a ...................................... ..................................................... 14 0

5-34 Measured dynamic and static strains at gage 2 in the 5-inch slab in Phase I-a .............140

5-35 Measured dynamic and static strains at gage 4 in the 5-inch slab in Phase I-a .............141

5-36 Measured strains at two different depths at the mid-edge of the 6-inch slab in Phase
I -a ....................................................................... ................ 1 4 1

5-37 Zeroed strains profile at two different depths at the mid-edge of the 6-inch slab in
P h ase I-a. ......... ......... ................ ...... ....................................................... 14 2

5-38 Strain in the composite pavement as a function of time. ...................... .................142

5-39 Effect of temperature differential on the peak strain for the 6" slab in Phase I-a............143

5-40 Effect of temperature differential on the peak strain for the 4" slab in Phase I-b .........143

5-41 Effect of temperature differential on the peak strain for the 5" slab in Phase I-b .........144

5-42 Effect of the temperature differential on the peak strain for the 6" slab in Phase I-b .....144









5-43 Effect of the temperature differential on the peak strain for the 6" slab in Phase II .......145

5-44 Effect of the temperature differential on the peak strain for the 8" slab in Phase II .......145

5-45 Effect of the temperature differential on the peak strain for the 8" slab in Phase II .......146

5-46. Effect of AC temperature on the peak strain for the 6" slab in Phase I-a.......................146

5-47 Effect of AC temperature on the peak strain for the 4" slab in Phase I-b.......................147

5-48 Effect of AC temperature on the peak strain for the 5" slab in Phase I-b.......................147

5-49 Effect of AC temperature on the peak strain for the 6" slab in Phase I-b.......................148

5-50 Effect of AC temperature on the peak strain for the 6" slab in Phase II.........................148

5-51 Effect of AC temperature on the peak strain for the 8" slab in Phase II.........................149

5-52 Effect of AC temperature on the peak strain for the 10" slab in Phase II.....................149

5-53 Relationship between strain and load in Phase II. Gages at the Mid-Edge ...................150

5-54 Relationship between strain and load in Phase II. Gages at the corer .........................150

5-55 Relationship between strain and load in Phase II. Gages in the AC layer.......................151

5-56 Variation of peak strains during the HVS test in the 6-inch slab in Phase I-a ................ 151

5-57 Variation of maximum strains during the HVS test in the 4-inch slab in Phase I-a........152

5-58 Variation of maximum strains in the 6-inch concrete slab and on the surface of the
asphalt layer at Location 1 (mid edge of the slab) during HVS test in Phase I-b............152

5-59 Variation of peak strains in the 6-inch concrete slab and on the surface of the asphalt
layer at Location 2 (corner of the slab) during HVS test in Phase I-b........................... 153

5-60 Variation of maximum strains in the 5-inch concrete slab and on the surface of the
asphalt layer at Location 1 (mid-edge of the slab) during HVS test in Phase I-b............153

5-61 Variation of peak strains in the 4-inch concrete slab and on the surface of the asphalt
layer at Location 2 (corner of the slab) during HVS test in Phase I-b............................. 154

5-62 Variation of peak strains in the 6-inch slab at Location 1 during HVS test in Phase II.. 154

5-63 Variation of peak strains in the 6" slab at Location 5 during HVS test in Phase II.........155

5-64 Variation of peak strains in the 8-inch slab at Location 1 during HVS test in Phase II. .155

5-65 Variation in peak strains in the 8-inch slab at Location 5 during HVS test in Phase II. .156









5-66 Variation of peak strains in the 10-inch slab at Location 1 during HVS test in Phase
I I ....................................................................... ................ 1 5 6

5-67 Variation of peak strains in the 10-inch Slab at Location 2 during HVS test in Phase
I I ....................................................................... ................ 1 5 7

6-1 Six-slab 3-D finite elem ent m odel. .............................................................................164

6-2 Twelve-slab 3-D finite element model. ........................................ ....................... 164

6-3 M esh pattern in the XY Plane for the 6-slab model........................................................ 165

6-4 M esh pattern in the XY plane for the 12-slab model................... ................... ................166

6-5 Twenty-node 3D solid element used in the analytical model.......................................166

6-6 Springs to model load transfer at concrete slab joints. ............. .................................... 167

6-7 Springs in the concrete-asphalt interface to model the partial bond condition................167

6-8 Non-linear springs to model the fully un-bonded condition in the interface.................68

7-1 Matching of deflection basin in the longitudinal direction caused by a 12-kip FWD
load applied to the center of a 4" slab in Phase I-a.......... ........................................185

7-2 Matching of deflection basin in the longitudinal direction caused by a 12-kip FWD
load applied to the center of a 6" slab in Phase I-b.......................... .... ............... 185

7-3 Matching of deflection basin in the transverse direction caused by a 12-kip FWD
load applied to the center of a 6" slab in Phase I-b.......................... .... ............... 186

7-4 Matching of deflection basin in the longitudinal direction caused by a 12-kip FWD
load applied to the center of a 4" slab in Phase I-b.......................... .... ............... 186

7-5 Matching of deflection basin in the transverse direction caused by a 12-kip FWD
load applied to the center of a 4" slab in Phase I-b.......................... .... ............... 187

7-6 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the comer of a 4" slab in Phase I-a. ...................................... .......... 187

7-7 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the corner of a 4" slab in Phase I-a ................................. 188

7-8 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the comer of a 4" slab in Phase I-b. ...................................... ......... 188

7-9 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the comer of a 5" slab in Phase I-b. ...................................... ......... 189









7-10 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the comer of a 6" slab in Phase II. ....................................... ........... ... 189

7-11 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the mid-edge of a 6" slab in Phase II. ......................... .................. 190

7-12 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the mid-edge of a 6" slab in Phase II. .........................................190

7-13 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the comer of an 8" slab in Phase II. ............................. ............... 191

7-14 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the corner of an 8" slab in Phase II. ............................................191

7-15 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the mid-edge of an 8" slab in Phase II. ................................................192

7-16 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the mid-edge of an 8" slab in Phase II. ........................................ 192

7-17 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the com er of a 10" slab in Phase II. ............. ...................................... 193

7-18 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the mid-edge of a 10" slab in Phase II. ..............................................193

7-19 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the mid-edge of a 10" slab in Phase II. ........... ...............194

7-20 Strain comparison at Gage 1 in the 6" slab in Phase I-a ............ ...............................194

7-21 Strain comparison at Gage 2 in the 6" slab in Phase I-a ............. ..............................195

7-22 Strain comparison at Gage 3 in the 6" slab in Phase I-a ............ ...............................195

7-23 Strain comparison at Gage 2 in the 5" slab in Phase I-a ............ ...............................196

7-24 Strain comparison at Gage 1 in the 4" slab in Phase I-a ............ ...............................196

7-25 Strain comparison at Gage 2 in the 4" slab in Phase I-a ............ ...............................197

7-26 Strain comparison at Gage 3 in the 4" slab in Phase I-a ............ ...............................197

7-27 Strain comparison at Gage 1 in the 6-inch slab in Phase I-b. ....................................198

7-28 Strain comparison at Gage 2 in the 6-inch slab in Phase I-b. ....................................198

7-29 Strain comparison at Gage 3 in the 6-inch slab in Phase I-b. ....................................199









7-30 Strain comparison at Gage 1 in the 5-inch slab in Phase I-b. ........................................ 199

7-31 Strain comparison at Gage 3 in the 5-inch slab in Phase I-b ....................................... 200

7-32 Strain comparison at Gage 2 in the 5-inch slab in Phase I-b. ........................................200

7-33 Strain comparison at Gage 2 in the 4-inch slab in Phase I-b. ........................................201

7-34 Strain comparison at Gage 3 in the 4-inch slab in Phase I-b. ........................................201

7-35 Strain comparison at top of Location 1 (mid edge) of the 10-inch slab in Phase II. .......202

7-36 Strain comparison at bottom of Location 1 (mid edge) of the 10-inch slab in Phase II..202

7-37 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 10-
inch slab in Phase II..................... ........................... ........ 203

7-38 Strain comparison at top of Location 5 (slab comer) of the 10-inch slab in Phase II. ....203

7-39 Strain comparison at bottom of Location 5 (slab corner) of the 10-inch slab in Phase
II ................. ....................................... 204

7-40 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the
10-inch slab in Phase II .............. ..................... ........ .... .. ............. 204

7-41 Strain comparison at top of Location 1 (mid edge) of the 8-inch slab in Phase II. .........205

7-42 Strain comparison at bottom of Location 1 (mid edge) of the 8-inch slab in Phase II. ..205

7-43 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 8-
inch slab in Phase II..................... ............................ ........ 206

7-44 Strain comparison at top of Location 5 (slab comer) of the 8-inch slab in Phase II. ......206

7-45 Strain comparison at bottom of Location 5 (slab corner) of the 8-inch slab in Phase
II ................. ....................................... 207

7-46 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the 8-
inch slab in P hase II .......................... ........................ .. .................. 207

7-47 Strain comparison at top of Location 1 (mid edge) of the 6-inch slab in Phase II. .........208

7-48 Strain comparison at bottom of Location 1 (mid edge) of the 6-inch slab in Phase II. ...208

7-49 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 6-
inch slab in Phase II ....................................... .... ..... ... ........ .... 209

7-50 Strain comparison at top of Location 5 (slab comer) of the 6-inch slab in Phase II. ......209









7-51 Strain comparison at bottom of Location 5 (slab corner) of the 6-inch slab in Phase
II................................. .......................... 210

7-52 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the 6-
inch slab in Phase II................... .. ......................... ........... 210

8-1 Axle load positioned on slabs with 4-ft joint spacing.....................................................227

8-2 Axle load positioned on slabs with 6-ft joint spacing.....................................................227

8-3 Effect of AC elastic modulus on maximum tensile stress in concrete caused by a 24-
kip axle load at mid-edge of 4-inch bonded concrete slabs with 6 ft joint spacing.........228

8-4 Effect of AC elastic modulus on maximum tensile stress in concrete caused by a 24-
kip axle load at mid-edge of 5-inch bonded concrete slabs with 4 ft joint spacing.........228

8-5 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at mid-edge of bonded slabs with 6 ft joint spacing..............229

8-6 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at corner of bonded slabs with 6 ft joint spacing.....................229

8-7 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at mid-edge of bonded slabs with 4 ft joint spacing..............230

8-8 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at corner of bonded slabs with 4 ft joint spacing.....................230

8-9 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at mid-edge of partially bonded slabs with 6 ft joint
sp a cin g .................. ................... ..................................................... .. 2 3 1

8-10 Effect of slab size on the maximum tensile stresses in concrete caused by a 24-kip
single axle load at m id-edge of bonded slabs. .............................................................. 231

8-11 Effects of slab size on the maximum tensile stresses in concrete caused by a 24-kip
single axle load at corner of bonded slabs. ........................................... ............... 232

8-12 Effects of interface condition on maximum tensile stresses in concrete caused by a
24-kip single axle load at mid-edge of 6-inch slabs with 6 ft joint spacing. ...................232

8-13 Effects of interface condition on maximum tensile stresses in concrete caused by a
24-kip single axle load at corner of 6-inch slabs with 6 ft joint spacing. ........................233

8-14 Effects of interface condition on maximum stresses in concrete caused by a 24-kip
single axle load at mid-edge of slab for the test sections in Phase II............................233









8-15 Maximum shear stresses at the interface caused by a 24-kip single load at a
temperature differential of +20 F for the bonded slabs with 6 ft joint spacing..............234

8-16 Maximum shear stresses at the interface caused by a 24-kip single load at a
temperature differential of +20 F for the bonded slabs with 4 ft joint spacing..............234

8-17 Maximum tensile stresses in the AC layer caused by a 24-kip single axle load at a
temperature differential of +20 F for the bonded slabs with 6 ft joint spacing..............235

8-18 Maximum tensile stresses in the AC layer caused by a 24-kip single axle load at a
temperature differential of +20 F for the bonded slabs with 4 ft joint spacing..............235

8-19 4" slab of the non-linear model curled up at covers due to temperature negative
tem perature differential.......................................................................... ....................236

8-20 4" slab of the non-linear model loaded with 21 kips in the corer and negative
tem peratu re differential.......................................................................... ....................236









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

ANALYSIS, TESTING AND VERIFICATION OF THE BEHAVIOR OF COMPOSITE
PAVEMENTS UNDER FLORIDA CONDITIONS USING A HEAVY VEHICLE
SIMULATOR

By

Patricio Enrique Tapia Gutierrez

December 2007

Chair: Mang Tia
Cochair: Fazil Najafi
Major: Civil Engineering

Whitetopping (WT) is a rehabilitation method to resurface deteriorated asphalt pavements.

While some of these composite pavements have performed very well carrying heavy load, other

have shown poor performance with early cracking. With the objective of analyzing the

applicability of WT pavements under Florida conditions, a total of nine full-scale WT test

sections were constructed and tested using a Heavy Vehicle Simulator (HVS) in the APT facility

at the FDOT Material Research Park. The test sections were instrumented to monitor both strain

and temperature. A 3-D finite element model was developed to analyze the WT test sections.

The model was calibrated and verified using measured FWD deflections and HVS load-induced

strains from the test sections. The model was then used to evaluate the potential performance of

these test sections under critical temperature-load condition in Florida.

Six of the WT pavement test sections had a bonded concrete-asphalt interface by milling,

cleaning and spraying with water the asphalt surface. This method produced excellent bonding at

the interface, with shear strength of 195 to 220 psi. Three of the test sections were intended to

have an unbonded concrete-asphalt interface by applying a debonding agent in the asphalt









surface. However, shear strengths between 119 and 135 psi and a careful analysis of the strain

and the temperature data indicated a partial bond condition.

The computer model was able to satisfactorily model the behavior of the composite

pavement by mainly considering material properties from standard laboratory tests and

calibrating the spring elements used to model the interface. Reasonable matches between the

measured and the calculated strains were achieved when a temperature-dependent AC elastic

modulus was included in the analytical model.

The expected numbers of repetitions of the 24-kip single axle loads at critical thermal

condition were computed for the nine test sections based on maximum tensile stresses and

fatigue theory. The results showed that 4" slabs can be used for heavy loads only for low-

volume traffic. To withstand the critical load without fear of fatigue failure, 6" slabs and 8" slabs

would be needed for joint spacings of 4' and 6', respectively.









CHAPTER 1
INTRODUCTION

1.1 Research Need

The increasing truck weights and tire pressures on our pavements in recent years have

pushed the demand on the performance of our pavements to a higher level. Many asphalt

pavements have experienced rutting while many others have experienced longitudinal cracking.

One of the possible solutions to this problem is the use of whitetopping (WT), which is placing a

concrete layer over an existing asphalt pavement. Whitetopping has an advantage over an

asphalt overlay in that the concrete surface is stronger and thus is more resistant to rutting and

surface-initiated cracking. The better durability and long-term performance characteristics of

concrete pavement surfaces can significantly reduce traffic delays associated with the frequent

maintenance of asphalt pavements. In addition, when concrete surfaces are used, skid resistance

and safety can be substantially improved, especially under wet conditions. In recent years, with

the sky-rocketing price of asphalt, concrete is becoming more competitive in cost with that of

asphalt. This makes the use of whitetopping a more economically viable alternative for

rehabilitation of asphalt pavements.

There are three types of WT pavements based on the thickness of the concrete

slab. Ultra-Thin Whitetopping (UTW) is a relatively new technique for resurfacing deteriorated

asphalt pavements. It involves placing very thin concrete slabs, 2 to 4 inches thick, on an old

asphalt pavement to create a bonded (or partially bonded) composite pavement. The reduction of

thickness is justified by the use of a high quality concrete, shorter joint spacing, and good bond

between the concrete and the existing asphalt pavement.

Thin Whitetopping (TWT) involves placing relatively thicker concrete slabs,

normally 5 to 8 inches thick, bonded (or partially bonded) over an existing asphalt pavement.









Similar to UTW pavements, TWT pavements use short joint spacing and good bond between the

concrete and the asphalt layer.

Conventional whitetopping (CWT) involves placing concrete slabs which are typically

greater than 8 inches in thickness. The concrete slabs are typically not bonded to the underlying

asphalt layer.

Experimental UTW pavements have been constructed in many states, including Colorado,

Georgia, Iowa, Kansas, Kentucky, Missouri, New Jersey, North Carolina, Pennsylvania and

Tennessee. Preliminary evaluations of these recently constructed UTW projects have shown

that UTW is a viable rehabilitation method for asphalt pavements. The Florida Department of

Transportation (FDOT) has also experimented with UTW in recent years. Three UTW test

tracks were constructed behind the FDOT State Materials Office in Gainesville in 1996. An

experimental UTW project was also constructed at the Ellaville Truck Weigh Station on 1-10 in

Northwest Florida in 1997. However, the performance of these test sections were less than ideal,

with the observation of some early cracking on the concrete surface. These problems were

attributed mainly to the fact that all of the UTW test sections were inadequately designed for the

traffic at the Ellaville Weigh Station (Tia et al, 2002). While the UTW technique may provide

durable wearing surface for normal traffic loads on residential and city streets, low-volume

roads, street intersections, general aviation airports, and parking areas, the UTW technique was

probably not an appropriate rehabilitation alternative for weigh stations subjected to frequent

applications of heavy truck traffic. The use of TWT or CWT might have been a more

appropriate choice in such an application.

With the potential economical and technical benefits of WT pavements, there is a need to

effectively evaluate the feasibility and proper application of UTW, TWT and CWT pavements in









Florida, so that the WT techniques can be properly and effectively utilized to achieve the

maximum benefits to the traveling public.

1.2 Problem Statement

To effectively evaluate WT pavements, it is necessary to have a reliable model or

methodology to analyze their behavior under the effects of loading, geometric and environmental

conditions. This model is to be developed based on existing knowledge on modeling and

behavior of WT pavements and taking advantage of the state-of-the-art tools in 3-D finite

elements method. The model will consider the effects of joints, interface bond, temperature, and

other pertinent material properties and pavement parameters. The model is to be calibrated and

verified using experimental results from full scale experiments considering most of the variables

affecting the behavior of WT pavements. The main model characteristic should be its capability

of modeling different interface conditions.

In addition to providing the necessary data for the model calibration, the full scale

experiments will allow for observation of relationships between pavement performance and

pavement response (measured strains and calculated stresses). In this way, a better

understanding of the behavior of the WT pavements can be accomplished.

1.3 Research Hypothesis

The following hypotheses were investigated in this research:

1. Conventional procedures for WT interface treatments provide enough bond strength for

the composite pavement to become bonded or partially bonded, and such interface condition can

be adequately modeled using linear springs connecting the layers.

2. Load transfer at the concrete slab joints due to the interlocking mechanism can be

modeled using vertical linear springs connecting the adjacent slabs.









3. Temperature is a critical factor affecting the behavior of WT pavements. It affects the

AC as a supporting layer and severe temperature conditions may cause a composite pavement to

become de-bonded, and to have lower load carrying capabilities.

1.4 Objectives of Research

The main objectives of this research are as follows:

(1) To develop analytical models for analysis of the behavior of UTW, TWT and CWT

pavements. These models were to be calibrated and verified using experimental results.

(2) To evaluate the effects of temperature and other variables in the behavior of WT

pavement and their role in the failure mechanism.

(3) To evaluate the applicability of UTW, TWT and CWT techniques for rehabilitation of

deteriorated asphalt pavements in Florida.

1.5 Approach and Scope of Research

The objectives of this research study were achieved through the following main

tasks:

1. A literature review on the state-of-the-art of WT pavements.

2. Development of an experimental design and instrumentation plan for evaluation of

several UTW, TWT and CTW pavement test sections by means of accelerated pavement testing.

3. Construction and testing of the UTW, TWT and CWT test sections located in the Florida

DOT Research Park by means of the HVS.

4. Characterization of the test sections by laboratory testing of cored samples and FWD

testing to obtain pavement parameters.

5. Development of analytical models for analysis of UTW, TWT and CWT pavements, and

calibration and verification of the models by comparing the analytical results with experimental

results from the HVS test sections.









6. Use of the developed analytical model to estimate the stresses in the WT pavements

sections under critical loading conditions to study the applicability of these composite pavements

under Florida conditions.

1.6 Significance of Research

In the past, there have been many studies where WT pavements were constructed and their

performance observed. There have also been some studies where WT pavements were modeled

and analyzed with respect to the various factors which may affect their performance. However,

there has been little work done where the WT pavements were instrumented and the measured

responses were compared to the analytical results to validate the models used.

This research evaluated WT pavements by taking into consideration all the pertinent

factors affecting their behavior. All the variables that have been previously shown as important

in the performance of WT pavements, such as thicknesses of the concrete and AC layers, joint

spacing, temperature and level of bond in the interface were assessed in this research. A 3-D

finite element method was used to model the behavior of the WT pavements, and a full-scale

experiment using full-scale instrumented pavement test sections and full-scale wheel loads was

used to evaluate the behavior and performance of various WT pavement designs. The measured

responses from the test sections were used to validate and fine-tune the analytical model. The

validated and fine-tuned model was used to predict the level of stresses and performance of the

WT pavement sections.

The significance of this research work is that the analytical model developed was validated

and fine-tuned using measured responses from full-scale and instrumented WT pavements,

which has not been done before. The 3-D finite element model has also further refinements from

previous work in this area. In the past research work, 2-D models using 4-node elements and 3-

D models using 8-node elements have been used to analyze this type of pavements. In this









research, a 3-D finite element model with 20-node 3D solid elements was used to model the

pavement structure. Previous models have used a 4-slab system to model the WT pavements. In

this research, a 12-slab system was used. Using a 12-slab system can possibly give better

modeling of the effects of the adjacent slabs. The bonded interface between the concrete slab

and the AC layer was also modeled with special elements to cover the range from a fully bonded

to fully un-bonded condition.









CHAPTER 2
LITERATURE REVIEW ON WHITETOPPING

2.1 General Concepts

The concept of resurfacing existing asphalt pavement using Portland cement concrete

(whitetopping) is not new. In fact, the first reported use of whitetopping dates back to 1918

(Hutchinson, 1982). However, this technology has improved over the years as the concrete

paving technology has improved. Plain concrete, reinforced concrete and fibrous (fiber

reinforced) concrete have been used to resurface flexible pavements (Hutchinson, 1982 and

McGhee, 1994). In the 1940's and 1950's, plain concrete was mainly used at civil and military

airports. Concrete thickness used in these projects ranged from 8 to 18 in. (200 to 460 mm).

Since 1960, plain concrete has been extensively used to resurface existing highway pavements in

states such as California, Utah, and Iowa. Concrete thickness of these resurfacing projects

ranged from 7 to 10 in. (175-250 mm). Continuously reinforced concrete and fiber-reinforced

concrete were also used on a limited number of projects. NCHRP synthesis 204 listed 189

whitetopping projects constructed in the United States between 1918 and 1992. This list

included streets, highways, and airfield projects.

There are several advantages of using whitetopping for rehabilitating asphalt pavements.

Whitetopping provides long-term benefits to the traveling public, and to roadway or airport

agencies. Concrete durability and long-term performance characteristics decrease the mainte-

nance required and life cycle costs of pavements. As a result, concrete surfaces significantly

reduce traffic delays associated with the frequent maintenance of asphalt pavements. In addition,

when concrete surfaces are used, skid resistance and safety are substantially improved, especially

under wet conditions. These advantages promote and contribute to the use of concrete pave-

ments over asphalt surfaces.









2.1.1 Ultra-Thin Whitetopping

Ultra-thin whitetopping (UTW) is a relatively new technique for resurfacing deteriorated

asphalt pavements. It involves placing very thin concrete slabs (2 to 4 in. thick) on top of an

asphalt pavement to form a bonded (or partially bonded) composite pavement. The reduction of

thickness is justified by the use of high quality concrete with relatively high strength, shorter

joint spacing, and bond between the concrete and the existing asphalt pavement.

The first UTW experimental project was constructed on the access road to a waste disposal

landfill in Louisville, Kentucky in September 1991(Cole and Mohsen, 1993, Brown, 1995, and

Risser et al. 1993). The concrete mixture was designed to provide relatively high early

compressive strength, 3,500 psi at 24 hours. A low water-cement ratio of 0.33 was selected to

achieve higher strength and reduce drying shrinkage. Two concrete slab thicknesses, 2 in. and

3.5 in., and two joint spacings, 2 ft and 6 ft were used. The Louisville UTW pavement has per-

formed well, carrying many more traffic loads than predicted by design procedures available at

that time.

Following the success of the Louisville UTW project, many other states, including

Tennessee, Georgia, North Carolina, Kansas, Iowa, Pennsylvania, New Jersey, Colorado,

Missouri, Mississippi, Virginia and Florida, have constructed and are currently evaluating UTW

projects. Over 200 UTW pavements have been built in the last decade. The development of a

mechanistic design procedure for UTW pavements in 1997 represented another major step in

advancing this promising technique (Wu et al. 1997, Mack et al. 1997, and ACPA, 1997).

2.1.2 Thin Whitetopping

Thin Whitetopping (TWT) is a variation of the UTW where thicker concrete slabs are

used. Slab thicknesses in the range of 5 to 7 inches are normal for this type of pavements. TWT

may have a bonded or an unbonded interface between the concrete slab and the AC layer. While









more attention has been paid to investigate the behavior of UTW pavements, few studies have

focused in the alternative of using TWT when the conditions for the thinner slabs cannot be met.

Among the states that have undertaking projects involving TWT are Colorado (Tarr et al. 1997),

Minnesota (Vandenbossche et al. 2002), and Mississippi.

2.1.3 Conventional Whitetopping

Whitetopping pavements with a slab thickness greater than 8 inches are commonly known

as Conventional Whitetopping (CWT). CWT pavements are generally used for pavements

subjected to heavier traffic loads, and have been designed based on the assumption that the

existing asphalt concrete (AC) layer does not contribute directly to the load-carrying capacity of

the pavement structure. Rather, the AC layer is considered to serve as a base layer for the new

concrete overlay, and no bond is considered to exist between the overlay and the existing asphalt.

Longer joint spacing (comparable to those of conventional jointed concrete pavements (JCP)) is

generally incorporated in CWT.

2.2 Concrete Mixture Proportions and Properties

The concrete mix for a particular UTW and TWT project is often selected based on the

requirements for early opening to traffic. A normal mix design includes cement, coarse and fine

aggregates, air-entraining agent, admixtures, and a lower water-cement ratio. Fibers have been

used in many UTW projects; however, the effects of fiber have not been well documented.

Compared to aggregate used for thicker concrete pavements, the top-size of coarse aggregate for

UTW and TWT is reduced. Materials and mix proportions selected for the first experimental

project in Louisville, Kentucky are shown in Table 2-1.

The concrete mixture was designed to provide relatively high strength at early ages (3500

psi or 24.2 MPa at 36 hours). A low water-to-cement ratio (0.33) was selected to achieve higher

strength and reduce drying shrinkage. Polypropylene fibers were used to enhance the flexural









strength, increase impact and freeze-thaw resistance, and further reduce drying and plastic

shrinkage cracking.

Similar mix proportions were used in the Tennessee (Speakman et al. 1996) and Georgia

(Cown, 1993) projects. In the Leawood, Kansas site, the mixture proportions were slightly

different; cement content was less (611 lb/yd4), but a setting-accelerating admixture was used

(Dumitru et al. 2002). Compressive strength of 3000 psi (20.7 MPa) at 24 hours was achieved.

Three pounds of polypropylene fibers were also used in this mixture. The mix design used in the

Kansas UTW project is presented in Table 2-2.

In April 2000, an old asphalt runway was rehabilitated using the UTW technique at the

Savannah-Hardin County Airport (SNH) in Tennessee (Saeed et al. 2001). The runway was

original constructed in 1962 and was subsequently overlaid and extended in 1975. The original

pavement consisted of an AC surface of 3.5 in. and a crushed aggregate base of 6.5 in. A 3-in.

AC overlay was added during the 1975 rehabilitation.

At the time of UTW construction, the AC surface had exhibited significant fatigue and

thermo cracking. Design UTW thickness was 4 in. with ajoint spacing of 48 in. Design

concrete flexural strength was 700 psi. The concrete mix design is shown in Table 2-3.

In 1997, the Illinois Department of Transportation (IDOT) began investigating the use of

whitetopping as an intersection repair method (Winkelman, 2003). Eight projects, including main

lines and intersections, were selected to be analyzed in this research. The projects are identified

in Table 2-4 along with the mixture design used in each case. All of the mixture designs

contained an air entraining admixture. The mixture designs also included a water reducer except

for the project on Clay County Highway 3. The project listed in Decatur contained a super-









plasticizer. The water to cement ratios for all of these projects ranged from 0.34 to 0.36 except

for the Clay County project, which was 0.46.

The performances of the thin whitetopping on the four mainline pavements and the

ultrathin whitetopping sections on the intersections have been reported as excellent.

Middleton et al. 2005, investigated the impact of different cement materials, synthetic fiber

types, and curing procedures on compressive strength, flexural strength, shrinkage and scaling

durability of concrete. According to the results, high early strength concrete containing ordinary

or rapid-hardening cement along with a low water-cement ratio gave excellent compressive and

flexural strength at the age of one day. It was also noted that it provided high resistance to

scaling from de-icing agent. The use of Class C fly ash gave acceptable final strength, but the

early strength was lower than the one obtained in mixes without fly ash. On the other hand the

fly ash concrete showed poor scaling resistance.

2.3 Construction Procedures

UTW pavements are constructed with slipform or fixed form pavers in essentially the same

way as conventional concrete pavements, with some special provisions. The construction

procedures consist of the following steps: preparing asphalt surface, placing the concrete,

finishing, surface texturing, curing, and sawing the joints.

Asphalt pavement surface preparation prior to concrete placement is a very important

procedure to achieve better bond and good performance of UTW. Milling, followed by cleaning

with compressed air to remove all laitance, dust, grit and all foreign materials, is the best way to

prepare the asphalt surface. It is recommended that an adequate asphalt thickness of a minimum

of 3 in. (75 mm) after milling, be used, if possible (Mack et al. 1997).

Concrete used in UTW can be produced at a ready-mix plant and delivered to the site by

ready-mix trucks. Normal slipform pavers can be used to spread, screed, and consolidate the









concrete in an efficient manner. After the surface is finished and textured, a curing compound is

immediately sprayed on the entire surface to achieve adequate curing. The curing compound is

generally applied at a rate twice the normal application rate for thicker concrete pavements

because thin concrete slabs can lose water rapidly (ACPA, 1998). Joint sawing must be

performed as soon as surface conditions permit, or when the concrete is able to support the

equipment and the operator. Usually joints are not sealed because joint openings are generally

narrow due to the short joint spacing.

2.4 UTW and TWT Design Considerations

A parametric investigation of the variables affecting the performance of whitetopping

pavements was performed in Gainesville, Florida. Confirming the results from previous studies,

Tia el al. 2003, concluded that the main factors affecting the behavior of UTW and TWT are the

thickness of the concrete slab, the joint spacing, the bond in the interface and the thickness of the

asphalt layer.

2.4.1 Slab Thickness

A major benefit of UTW is the reduction in concrete thickness. As defined above, the

recommended thickness for ultra-thin whitetopping is not more than 4 in. (100 mm). For most of

the experimental projects, thickness ranges from 2 in. (50 mm) to 4 in. (100 mm). The reduction

in thickness is justified by the use of high quality concrete with relatively high compressive and

flexural strength, closely-spaced joints, and good bonding between UTW and the existing asphalt

base.

2.4.2 Joint Spacing

Spacing between joints is an important factor controlling the performance of UTW. In the

Louisville experimental project, 6-ft transverse and longitudinal joint spacings (6-ft panels) were

compared with 2-ft panels. Although the 6-ft panels exhibited cracking, no signs of distress were









observed on the 2-ft panels. This performance was attributed to the smaller panels (2-ft)

transferring the load completely to the flexible base and the concrete being mostly under

compression. Comparatively, when larger panels are used, some of the load is absorbed by slab

bending. The contribution of the existing asphalt is based on the assumption that the overlay is

bonded to the flexible base. The American Concrete Pavement Association (ACPA) has

recommended that joint spacing be about 12 to 15 times of the slab thickness. For example,

spacing for UTW of 2 in. (50 mm) of thickness should be between 2 and 3 ft (0.6 and 0.9 m).

Spacing at a Georgia site was 2-ft (0.6 m) and in Kansas 3-ft (0.9 m) in one section, and 4-ft (1.2

m) in the other.

It is not practical to install dowel bars, tie bars, or keyway in UTW pavements because of

the very thin slabs. Field evaluation has indicated that load transfer provided by aggregate

interlock is generally high because of the short joint spacing and the support provided by the

asphalt layer. For UTW pavements, field performance demonstrated the need for thicker slabs at

the transition areas between the UTW and the asphalt roadways. Figure 2-1 shows transition

details for UTW pavements, as recommended by ACPA.

2.4.3 Interface Bonding Strength

Bond between UTW and existing asphalt pavement is a key factor controlling the perfor-

mance of the UTW composite pavement. The existence of bond strength not only significantly

reduces the stresses in the concrete section, but also allows the section to perform and be

analyzed as a composite section. A wide range of bond strength (shear strength between the two

layers) was measured at some of the experimental sites; from the average strength of 50 psi (0.34

MPa) measured at the Swedish site to the over 200 psi (1.38 MPa) measured at some Florida

UTW test sections. Bond strength can be improved by milling the old asphalt pavement; the









roughness in the surface and the exposed aggregates lock the layers together thus increasing the

bond.

An experiment intended to evaluate the bonded condition in the interface of concrete

overlay was performed in Iowa by Nishiyama et al. 2005. This experiment included field and

laboratory testing to evaluate the bond at different times after concrete placement. The project

concluded that bond strength between an overlay and the existing pavement gradually increases

over the time regardless of the initial bond strength.

To evaluate the interface condition, other authors have proposed the use of the vertical

tensile strength in the interface as an indicator of the bond, in addition to the shear strength. By

performing pull-off tests on the composite samples, it is also possible to characterize the vertical

interaction between layers.

Rasmussen et al., 2000, presented a method to characterize the axial slab support restraint

by running full-scale push-off test to obtain the stiffness and strength of the interface bond.

These parameters can be used in analytical models to better represent the interface condition in

whitetopping pavements.

2.5 Design Procedure for UTW and TWT Pavements

Because of the thin slabs, the Portland Cement Association's (PCA) thickness design pro-

cedure (PCA, 1984) does not completely apply to UTW. Furthermore, the AASHTO thickness

design method (AASHTO, 1993) does not account for the bond between the two layers. As the

development of the UTW technique continued, it became apparent that a design procedure for

selecting the optimum UTW thickness and joint spacing subjected to anticipated traffic and

environmental loading was needed.

In 1994 the PCA sponsored a comprehensive research effort aimed at developing a

mechanistic based design procedure for UTW (Wu et al. 1997 and Mack et al. 1997). The









PCA's UTW pavement design procedure was developed in 1997. Work conducted in developing

the design procedure included a thorough literature review of past and current work; condition

surveys of UTW sites in Georgia and Tennessee; instrumentation and load testing of several test

sections; development of a 3-dimensional finite element model for UTW analysis; and devel-

opment of a design procedure and construction guidelines.

Development of this design procedure included the following elements:

Verification of the 3-dimensional finite element model. Strain (stress) data collected

from the Spirit of St. Louis Airport UTW test sections were used to calibrate and verify the

three-dimensional model that was developed in this study and was used as an analytical tool to

analyze the UTW pavement behavior and to develop the design procedure. To calibrate and

verify the model, stresses were computed under each loading and temperature condition, and

were compared with the load testing results obtained from the test pavements.

Identification of degree of bonding existed in the field. It has been established from the

results of field testing that UTW pavements behaved as partially bonded composite pavements.

Using field data and the 3-D model, an effort was made to quantify the additional structural

capacity (or load carrying capacity) that could be offered by the asphalt layer to the UTW

pavements.

Correlation between stresses calculated from ILSL2 and the 3-dimensional model.

Since many computer runs would be required to develop the design guidelines, the use of the 3-

dimensional model was not feasible due to the time needed for each run. Instead, correlations

were developed between stresses computed by these two models, and were used in the

development of the design guidelines.









Development of design guidelines. Following are the processes involved in the

development of the UTW design guidelines:

Stresses induced by loads and temperature were separately computed for fully bonded
UTW pavements using the 2-dimensional model, ILSL2. A wide range of pavement
parameters and material properties were covered.

The 2-D model stresses were converted to 3-D model stresses using the conversion
equations derived in the previous step.

The converted 3-D model stresses were increased by 36% to account for the partially
bonded condition, as observed in the field testing.

Equations were developed to correlate the converted and adjusted stresses to different
pavement parameters. Stresses and strains were then calculated for typical param-
eters for UTW pavements under different loading and temperature conditions, and
were tabulated.

UTW pavement thickness design was accomplished by limiting both the concrete and
asphalt strains within safe limits under anticipated traffic and environmental loadings
in the pavement's design life.

A mechanistic design procedure for TWT was developed by the state of Colorado (Tarr et

al. 1998). The procedure followed to develop this design method was very similar to the one

used in the PCA method. Experiments using slab thicknesses between 5 and 7 inches were per-

formed and the measured strain values were used to calibrate and verify the computer model.

Theoretical design equations to predict critical stresses and strain were identified. Correction

factors for stresses due to load position and differences between the calculated and the measured

stresses were included in the model. Also correction factors for the strains at the interface

between the concrete slab and the asphalt layer were evaluated to account for the bond condition.

An additional factor to consider the effect of temperature was added to the model, which

modifies the stress in the concrete. The prediction equations were developed for load varying

from 20 kips to 40 kips single-axle load. Failure criteria for both materials were assumed from

fatigue relationships. As considered in the PCA design procedure, the number of load repetitions









for the concrete slab is a function of the flexural stress-to-strength ratio. The failure criterion for

the asphalt concrete is based on the allowable number of repetitions, which was considered as a

function of the asphalt elastic modulus and the volume of binder and voids. In this method the

number of load repetitions that the asphalt concrete has already carried is also considered.

2.6 Performance of UTW and TWT Projects

Most of the experimental UTW and TWT project sites have performed very satisfactorily.

At the Louisville site, two overlay thicknesses were evaluated (2 and 3.5 in.), and joint spacings

of 6 ft and 2 ft. were used. UTW test sections with 2-ft joint spacing showed much less cracking

than those with 6-ft joint spacing and 2-in. thick slabs.

Evaluations in Georgia, conducted two years after construction, indicated good perfor-

mance (Wu et al. 2001). The UTW test sections were located in a truck weigh station on 1-85.

The test sections had a design UTW thickness of 2.5 in. (64 mm) and the existing asphalt

thickness was about 11 in. (279 mm). It was reported that the test sections had been subjected to

351,000 18-kip (80-kN) Equivalent Single Axle Loads (ESAL) in two years. Only 2% of the

slabs had cracks for the test sections with fiber reinforced concrete and 5% of slabs had cracks

for the non-fiber sections. A design evaluation using the PCA's design procedure also indicated

that the UTW test sections were designed adequately.

A wealth of information on the behavior of UTW was generated from the Tennessee

experience (Wu et al. 2001). After the first experimental project was constructed in Nashville in

May of 1992, six more sites were built in Maryville, Chattanooga, McMinnville and Athens.

Cracks observed on the first site (Nashville) were mainly attributed to the asphalt base. It was

observed that during milling, the asphalt was completely removed and the concrete overlay was

supported by a cobblestone base. A design evaluation also indicated that the test section was

severely under-designed.









A UTW pavement project was constructed at the Spirit of St. Louis Airport to carry light-

load aircraft (gross weight of 12,500 lb or 5,670 kg) traffic in 1995. The design thickness of the

UTW pavements was 3 V2 in. (89 mm) with ajoint spacing of 50 in. (1.3 m). The existing

asphalt pavement was milled before concrete placement to create a rough surface. The asphalt

thickness after milling was about 3.1 in. (79 mm). A visual condition survey of the UTW pave-

ments was performed on July 2001 (Wu et al. 2002). It was observed that, after over six years of

service, the UTW pavements performed extremely well, with very little distresses observed on

the entire site. Out of the more than 7,200 panels, only 18 panels (0.25%) have exhibited

distresses, with majority of them in the form of corer cracking.

The Iowa and Minnesota Departments of Transportation (DOT) have also been actively

involved in the UTW technique development and evaluation. In 1994, Iowa constructed 7.2

miles (11.6 km) of UTW pavements on a segment of Highway 21 (Cable et al. 1997 and Cable et

al. 2001). The research was designed to evaluate the long-term performance of UTW pavements

and their applicability in Iowa. Four major variables were included for evaluation, resulting in a

total of 41 test sections. The four variables were overlay thickness, joint spacing, the use of

fiber, and the asphalt surface preparation. The pavements were subjected to an estimated

Average Daily Traffic (ADT) of 1,350 (or 40 ESAL's per day). Through continuous monitoring

of the test sections, after seven years in service, the test sections have performed well and have

exhibited minimal distresses.

In October 1997, the Minnesota DOT constructed several thin and ultra-thin whitetopping

pavements on 1-94 at the Minnesota Road Research (Mn/Road) (Vandenbossche et al. 2002).

The existing asphalt pavement was in fairly good condition, with minor cracking and rutting.

The asphalt pavements were milled to the depth of the overlay thickness to maintain the original









pavement surface elevation. The UTW test sections had two different thicknesses, 3 in. and 4 in.

(75 mm and 100 mm), with two different joint patterns, 4 ft by 4 ft and 5 ft by 6 ft (1.2 m by 1.2

m and 1.5 m by 1.8 m). After 312 years and over 4.7 million ESALs, cracking (transverse and

corner cracking) was observed in the UTW test sections. The majority of the cracking was in the

truck lane. It was indicated in the research study that most of the corner cracking occurred along

the inside longitudinal joint due to its location directly in the inside wheel path. Transverse

cracking often occurred in the outside wheel path near a transverse joint. This indicated that

using a joint layout that keeps the longitudinal joints outside the wheel paths could improve

UTW pavement performance.

A forensic investigation of the UTW constructed in the Ellaville Weigh Station in Florida

(Tia et al. 2002) showed that the poor performance of most of the six test sections was mainly

due to an inadequate design of the overlay. The thickness of the test slabs were 3 to 4 inches,

with joint spacing of 4 and 6 feet. The premature cracking, the extensiveness and the severity of

the cracking and the rapid progress of the cracking are also attributed to the lack of control on the

layer thickness especially in the AC, which presented no thickness at all in some sections. The

problem was aggravated by the loss of bond between the concrete slab and the asphalt layer,

which had a great effect on the rapid progress of cracking and the large percentage of shattered

slabs.

2.7 Accelerated Pavement Testing and Field Testing of UTW

In the spring of 1998, the Federal Highway Administration (FHWA), in partnership with

concrete industry groups, undertook a study to evaluate the performance of ultra-thin white-

topping under accelerated traffic load. Eight (8) sections of existing asphalt pavements were

whitetopped and subjected to FHWA's accelerated loading facility (ALF) in McLean, Virginia.

The experimental results indicate the bond between AC and concrete decreases the critical









tensile stresses in the concrete overlay as the UTW section acts in a composite manner.

Dynamic strain measurement of longitudinal strain indicates the concrete overlay experiences

significant stress reversal as the wheel rolls over the pavements (Cole et al. 1999).

The Indiana Department of Transportation (INDOT) conducted an experimental study at

the APT facility to investigate the performance of UTW pavement in Fall 1999 (Raj an et al.

2001). Four concrete mixtures were tested under slow moving loads. The results indicated the

joint spacing of 1.2 m (4') in all the lanes was sufficient. The measured strains in the overlay

were proportional to the applied load. The results showed that the pavement response was linear

within the overlay. However the increase of temperature in one of the lanes affected the linearity

of the pavement response. Strains at the asphalt surface increased considerably because of the

temperature gradient. The researchers have found that the concrete overlay experienced

significant stress reversal as the wheel rolled over the pavement. It was also observed that the

overlay thickness and the asphalt stiffness significantly affect the strains because of their

influence on the location of the neutral axis (Rajan et al. 2001).

An UTW pavement with 3-inch thick concrete on 6 inch asphalt was constructed in the

APT facility in Lancaster, Ohio for the purpose of measuring response under controlled loading

and environmental conditions (Edwards et al. 1999). The measured strain was relatively propor-

tional to the magnitude of the applied load over the range of 6 tolO kips at 50 OF and 5 mph. The

measured tensile strains were higher on the AC surface than would be expected by projecting

strains measured in the PCC layer down to the AC/PCC interface. The placement of the 3-inch

thick PCC layer on 6-inch thick layer of AC lowered the neutral axis of this UTW pavement

structure to the lower portion of the PCC where tensile strains were minimal.









The researchers at Purdue University and INDOT conducted an UTW pavement experi-

ment to investigate the applicability of PCA design guidelines for UTW designs (Galal et al.

2004). Preliminary results indicate that the PCA design equations may be able to be used if the

concept of an equivalent thickness is employed. Equivalent section was hypothesized to take

into account the additional layer in the existing composite pavement structure. There was a good

agreement between the measured strain in the composite UTW section and the computed strains

from the ESLYM5 program.

Nishizama et al. 2003 performed full-scale experiments to evaluate the mechanical

behavior of UTW pavements in Japan. The experiment included the appropriate instrumentation

with strain gages embedded in the concrete slabs and thermocouples to monitor strain and tem-

perature respectively. Two joint spacings were investigated (4' and 6'). The loading periods

included summer and winter time. The load was applied in two ways: stationary and moving. In

the stationary load test, a load was applied on the edge of the slab through a 30 cm diameter plate

by a jack and a crane truck. The load was applied at increments of 9.8 kN up to 49 kN and strains

were measured at each load increment. In the moving load test, the crane truck traveled on the

test pavement at a low speed and dynamic strains were measured every 0.1 sec. A single

thickness of 4 inches for both the concrete slab and AC layer was used in the test. The concrete

slab was placed bonded to the AC layer. The measured strains were compared with the calcu-

lated ones obtained from a 3D FE model and the comparison showed a good match.

2.8 Analytical Models

Many efforts have been made to model whitetopping pavements. Most of them have

focused on modeling the interface interaction between the concrete slab and the AC layer. In

addition to the bond interface evaluation, the analytical models have to consider also the load









transfer at joints. The general approach to these two types of interaction is the use of springs to

represent the stiffness in different directions.

A 2D finite element model, NSLIP (Nelson et al. 2002) was developed to model the

interface condition in composite pavements to study the delamination. In this model a 4-node

slip finite element with displacement in the normal and shear direction was used. The stiffness in

the normal direction was considered as infinitely large when the interface was in compression. It

was hypothesized that when the interfacial strength is exceeded, delamination occurs in the

interface.

The concept of friction factor was used in combination with the ISLAB2000 FE program

(Khazanovich at al. 2002) to characterize the bond in the interface of composite pavements. The

ISLAB2000 program contains two models to analyze the interface: the modified Coulomb

friction model and the simplified friction model. For the analysis, Khazanovich utilized the

simplified friction approach in conjunction with the transformed section concept. The simplified

friction model allowed for modeling of intermediate degrees of interaction between the fully

bonded and un-bonded and the friction factor was shown to be a good indicator of level of bond.

Nishizawa et al. 2003, developed PAVED3D, a 3D FE model using 3D solid elements with

8 nodes per element. This model has the capability to model several slabs with the respective

joints. The interaction between concrete slabs is modeled through spring elements that represent

the stiffness in the plane of the joint (Ks and Kt), and in the normal direction (Kn). Springs were

also used to model the interface interaction between the concrete slab and the AC layer. Similar

to the joint case, three springs were utilized. The findings of this research were that by varying

the stiffness of the springs used in the interface it is possible to model different levels of bond.









Table 2-1 Concrete Mix Proportions Used in Louisville Experimental Project (Riser et al 1993).
Constituents Quantity
Cement (ASTM C 150 Type I) 800 lb/cy
Coarse Aggregate 1800 lb/cy
Fine Aggregate 1150 lb/cy
Water 260 lb/cy
Polypropylene Fibers 3 lb/cy
High Range Water Reducer 14 oz/100 lb cement

Table 2-2 Concrete Mix Proportions Used in Leawood, Kansas (Wu et al. 1997).
Constituents Quantity
Cement (ASTM C 150 Type I) 611 lb/cy
Coarse Aggregate SSD (Crushed Limestone) 1730 lb/cy
Fine Aggregate SSD (Natural Sand) 1345 lb/cy
Total Water 225 lb/cy
Pave Air 5 oz/cy
Pozzutec 65 oz/cy
Rheobuild 43 oz/cy


Table 2-3 Concrete Mix Proportions Used in SNH, Tennessee (Saeed et al. 2002).
Constituents Attributes
Cement Type I, 573 lb/cy
Fly Ash Maximum 15% by weight of cementitious materials
Strength 700 psi flexural
Slump Between 1/2 and 2 in.
Water/Cement Ratio 0.35
Air Content 6% by volume
Synthetic Fiber 3 lb/cv









Table 2-4 Mix proportions for eight whitetopping projects in Illinois.


Project-Location


Coarse
Aggregate
(lb)


Decatur Intersection of US-36 and
Oakland Avenue
Carbondale Intersection of US-51
and Pleasant Hill Road
Harrisburg Intersection of US-45 and
Illinois Rt. 13
Anna Illinois Rt. 146 (Intersection of
Vienna and Main Streets)
Tuscola US-36
Clay County Highway 3
Piatt County Highway 4
Cumberland County Highway 2


1713

1805

1811

1811

1704
1814
1957
1836


Fine
Aggregate
(lb)

1210

1008

975

975

1035
1286
1220
1256


Cement
(lb)


Polypropylene
Water
We Fibers
(lb) (b)
(lb)


705 239

755 273

755 302

755 302


755
534
534
575


255
244
179
197


N/A


3.0

N/A
N/A
N/A
N/A


/ sBw C lrce


a aO
6h 0 Cocente Ovaiy%
t~ le ^ t


W ppms 6 (2 m) emierd overneral slubs if amm ry
I h + i in. (75 mm), minimum of6 in. (s10 mrm)

Figure 2-1 Transition between UTW and Adjoining Asphalt Pavement (ACPA, 1998)









CHAPTER 3
INSTRUMENTATION AND CONSTRUCTION OF TEST SECTIONS

3.1 Description of the Testing Phases

The first step in investigating the applicability of WT pavement in Florida was the design

and construction of a full scale experiment for Accelerated Pavement Testing (APT). The FDOT

Materials Office has a Heavy Vehicle Simulator (HVS) and an Accelerated Pavement Testing

(APT) facility for the operation of this HVS. The HVS can apply realistic full-size wheel loads

to full-size pavements to assess their behavior and performance directly. The HVS has the

capability to simulate 20 years of interstate traffic on a pavement test section within a period of 1

to 4 months. This accelerated pavement testing facility provided an excellent opportunity for

evaluating the long-term performance of WT pavements in Florida in a direct and effective

manner.

The HVS testing of the test sections in this study were divided into two main phases,

namely Phase I using bonded composite pavements, and Phase II using un-bonded composite

pavements. Phase I was divided in two sub phases by using two different joint spacings. The

description of the phases is as follows:

(1) Phase I-a. It involved three test sections on Lane 6 of the APT test area at the FDOT

State Materials Research Park. The concrete slabs were placed bonded to the top of an asphalt

concrete layer, and had a panel size of 6 feet by 6 feet. The three test sections had concrete slab

thicknesses of 4, 5 and 6 inches. The thickness of the underlying AC varied from 4 to 5 inches.

(2) Phase I-b. It involved three test sections on Lane 7 of the APT test area. Similar to

Lane 6 in Phase I-a, Lane 7 also had three test sections with concrete slab thickness of 4, 5 and 6

inches placed bonded to the top of a similar AC layer. The panel size in this case was 4 feet by 4

feet.









(3) Phase II. It involved three test sections on Lane 6 constructed after the test sections

from Phase I-a were tested and removed. In this case, the concrete slabs were 6, 8 and 10 inches

thick, placed un-bonded to the top of the asphalt layer. The concrete panel size was 6 feet by 6

feet.

Although there was some variation in the asphalt thickness in the test track, the thickness

of the asphalt layer was not considered as a variable in this experiment and an average thickness

of 4.5 inches was used in the analysis.

3.2 Layout of the Test Sections

3.2.1 Phase I

The test track in Phase I-a (on Lane 6) consisted of three test sections of 4, 5 and 6 inches

of concrete placed bonded to the existing AC layer, with 6 ft by 6 ft joint spacing. The test track

in Phase I-b (on Lane 7) consisted of three test sections with the same thicknesses as those used

on Lane 6, but with 4 ft by 4 ft joint spacing. The concrete overlay in Lane 7 was also bonded to

the existing AC layer. While the 4-inch concrete slabs are considered UTW, the 5- and 6-inch

slabs fall in the category of TWT. Figure 3-1 shows the layout of the test sections in Lanes 6 and

7. The test sections are confined by two ends and transition concrete slabs constructed to support

the HVS. To ensure a bonded condition in the interface in these test sections, the AC surface was

milled and cleaned prior to the concrete placement.

3.2.2 Phase II

After the removal of the test sections for Phase I-a, Lane 6 was overlaid with 6, 8 and 10

inches of concrete placed un-bonded to the existing AC layer, with 6 ft by 6 ft joint spacing.

Because it was not possible to remove the concrete slab without damaging the AC layer (a very

strong bond was observed), the existing asphalt layer was also removed and replaced with a new

one with the same thickness and properties. To ensure an un-bonded condition in the interface in









these test sections, a white pigmented curing compound was sprayed on the asphalt surface prior

to the concrete placement. Figure 3-2 shows the test section layout for Phase II.

3.3 Layout of the Instrumentation

3.3.1 Wheatstone Bridge Circuits

To monitor the strains in the test track, Wheatstone half-bridge circuits were used. In this

configuration one strain gage was used as an active gage to monitor the load-induced strain,

while another one was used as a dummy gage for temperature compensation. The Wheatstone

half-bridge circuit used is shown in Figures 3-3 and 3-4. The active gage with a resistance of RA

is subjected to a temperature-induced strain (y) and a load-induced strain (x) simultaneously.

The dummy gage with a resistance of RD, is subjected only to a temperature-induced strain (y).

The effect of the temperature-induced strain "(1+y)" is canceled out in this half bridge circuit,

and only the load-induced strain is measured.

3.3.2 Preliminary Stress Analysis

To determine the instrumentation layout, a stress analysis was performed to estimate the

maximum stresses. The capability of the ADINA program to consider a bonded condition

between layers was used to model the composite pavement for the preliminary stress analysis in

Phase I. A 3D model considering four slabs was built to evaluate the stresses under critical

combinations of load and temperature. Two critical load conditions were considered in the stress

analysis: at the mid edge and at the corner of the slab. Also three cases of temperature differen-

tial were applied to the model: -10, 0, and 10 C. The temperature differential is defined as the

difference between the temperature at the top of the concrete slab and the temperature at the

bottom.

The FEACONS IV (Finite Element Analysis of CONcrete Slabs version IV) program was

used to calculate the anticipated stresses on the test slabs for the un-bonded condition in Phase II.









The FEACONS program was developed at the University of Florida for the FDOT for analysis

of concrete pavements subject to load and thermal effects. This program was chosen for use

since both the University of Florida and FDOT have extensive experience with this program and

its reliability has been demonstrated in previous studies.

The FEACONS program was used to analyze the stresses in the test slabs when subjected

to a 12-kip (53-kN) single wheel load with a tire pressure of 120 psi (827 kPa) and a contact area

of 100 in2 (645 cm2), and applied along the edge of the slab, which represents the most critical

loading location. Similar to the case of bonded interface (Phase I), the analysis was performed

for two different load positions, at the corner of the slab and at the middle of the edge, for the

same temperature differentials in the concrete slabs. No load transfer at the joints was assumed

in the analysis, which represents the worst condition.

3.3.3 Instrumentation Layout

With the results from the stress analysis, it was possible to identify the locations where the

maximum stresses and strains in the test slab would occur so that strain gages could be placed to

monitor these maximum induced strains. The following sections describe the instrumentation

layout for Phases I and II.

3.3.3.1 Phase I

Figure 3-5 shows the instrumentation layout for the 6 ft by 6 ft test section (Phase I-a),

which were placed on Lane 6 of the APT test area. Three locations (Location 1, 2 and 3) were

identified to have the maximum anticipated strains due to the HVS load. Thus, strain gages were

placed at these three locations. Figure 3-6 shows the vertical positions of the gages at these three

locations along with the gage identification. While Location 1 had two gages, the other two loca-

tions had only one. At Location 1, one embedded strain gage was placed at a depth of 1 inch

from the concrete surface, while the other embedded strain gage was placed 0.5 inch from the









bottom of the concrete layer. Location 2 had a strain gage embedded 1 inch from the surface of

the concrete slab. Location 3 had a strain gage embedded 0.5 inch from the bottom of the con-

crete slab. Two surface gages were also used to monitor any micro cracks that may occur in the

concrete surface. These two gages were located next to the transversal joint at the middle of the

slab, in the adjacent panels. These surface gages, located in the adjacent slabs, were also used to

evaluate the load transfer at the joints.

Figure 3-5 also shows the locations of thermocouples to monitor the temperature in the

slab. Two positions were considered for the thermocouples, one at the center of the slab and the

other in the corer. Figure 3-7 depicts the vertical position of the thermocouples. They were

placed 1" apart along the depth of the slab with the first starting at 1" from the surface. An

additional thermocouple was placed on the surface of the AC layer to monitor daily variation of

temperature in the asphalt layer.

Figure 3-8 shows the instrumentation layout for the 4 ft x 4 ft slabs in Phase I-b. In this

case only two locations (namely Loci and Loc2) were identified as locations with maximum

stresses and for placement of embedded gages. Four strain gages were placed on the surface of

the concrete slab to monitor any micro cracks occurring in the slabs and load transfer at the

joints. The vertical positions for the gages are indicated in Figure 3-9 along with the gage identi-

fication. At Location 1, three gages were used -- one at 1 inch from the surface of the concrete

slab, the second at 0.5 inch from the bottom of the concrete slab, and the third one at 0.5 inch

below the surface of the AC layer. The placement of these three gages allowed for not only the

monitoring of the maximum strain at the top and bottom of the concrete slab, but also for the

comparison of the strain values near the interface of the layers. By comparing the strain values









near the interface, it would be possible to know how well the bonding condition would be in the

composite pavement.

Location 2 had two gages -- one at 0.5 inch above the bottom of the concrete slab, and the

other at 0.5 inch below the surface of the AC layer.

Figure 3-8 also shows the two locations for the thermocouples in Phase I-b. The vertical

positions of the thermocouples in Phase I-b were the same as those used in Phase I-a, which are

shown in Figure 3-7.

3.3.3.2 Phase II

Similar to Phase I, the strain gages were placed at the locations of maximum anticipated

stresses due to the HVS loads. Figure 3-10 shows the instrumentation plan adopted. Taking

advantage of the upgrade to the data collection equipment that allowed for more channels for

data acquisition, five locations were selected for placement of strain gages. For each of these

locations, a set of three strain gages were installed to monitor maximum strains in the concrete

slab and the strain at the surface of the AC layer. One strain gage was placed at one inch under

the surface of the concrete slab. A second strain gage was placed 1 inch from the bottom of the

concrete slab. The third strain gage was located 1/2 inch below the top of the asphalt layer. The

vertical positions of the strain gages are shown in Figure 3-11. Unlike the previous phase,

surface gages were not used in Phase II.

The locations for the thermocouples are shown in Figure 3-10. In Phase II, three locations

were used to monitor the temperature in the slabs. For each of the three thermocouple locations,

a set of thermocouples were placed at depths of 1, 3, 5, 7 & 9 inches for the 10-inch slabs, at

depths of 1, 3, 5 & 7 inches for the 8-inch slabs, and at depths of 1, 3 & 5 inches for the 6-inch

slabs. At each of the locations, a thermocouple was also placed in the asphalt layer at a depth of









1/2 inch from the top of the asphalt layer. The vertical positions of the thermocouples are shown

in Figure 3-12.

3.4 HVS Loading Plan

Testing of the composite pavements was performed using a Heavy Vehicle Simulator

(HVS), Mark IV model. HVS loading was scheduled to start 28 days after concrete placement

with an initial load of 9,000 lb, super single tire, with a contact pressure of 120 psi. The wheel

load traveled at a speed of 8 mph, in a uni-directional mode with no wander, and along the

longitudinal edge of the test slab. Loading along the edge was chosen because it represents the

most critical loading condition for a concrete slab.

If the composite pavement test sections could withstand the 9-kip load with no visible or

detectable cracks for a certain period of time, the load would be increased to 12 kips, 15 kips, 18

kips, and 21 kips, to observe the behavior of the test sections under different loads, and the load

at which cracking would occur.

3.5 Data Collection

For each test slab, the strain gages were connected to a strain indicator unit, Vishay System

6000 (Model 6100) for strain reading and data acquisition. This system has the ability to take

individual strain readings at a very high frequency. This enabled the recording of dynamic

strains as the wheel passed over the pavement. Data collection for load-induced strain was

started immediately after the start of HVS loading. Strain data were collected for 30 seconds at

one-hour intervals. The rate of data collection was 100 strain values per second. This rate

allowed for the capture of the progression of the strain and to especially observe the strain

reversal phenomenon. Strain gage readings due to a static wheel load were also taken for two

wheel loading positions, namely corer (pt 1), and mid-edge (pt 2). Static readings were

recorded while the wheel was traveling at slow speed towards the static loading position and









while it stayed at the two load positions (pt 1 and pt 2) for 20 seconds each. Static strains were

measured only for Phase I-a for comparison purposes with the dynamic load application. Phase

I-b and Phase II included only dynamic strain data collection. While in Phase I, strain data were

collected only for the loading period, strain data collection in Phase II was started a few days

after concrete placement to monitor strain due to temperature changes.

All the thermocouples were connected to the same data acquisition system. Temperature

data were collected during the entire day at 5-minute intervals. In both phases, data collection

for temperature was started before the loading period, especially in Phase II where strain due to

temperature changes was monitored.

3.6 Construction of the Test Tracks

3.6.1 Construction of Concrete Test Tracks in Phase I

3.6.1.1 Asphalt surface preparation and formwork

The concrete test track for Phase I-a was constructed on Lane 6 on June 10, 2004, and that

for Phase I-b was constructed on Lane 7 on August 10, 2004. These two concrete test tracks

were constructed over an existing four-inch thick asphalt surface (two 2 in lifts of asphalt) at the

APT test area at the FDOT State Materials Research Park. The asphalt surface was milled and

cleaned prior to the placement of formworks for the test track. Figure 3-13 shows the milled

asphalt surface in Lane 6. In the construction of Lane 7, a tapered formwork was used at the

transition from one thickness to the other to make the placement of the concrete easier. Figure 3-

14 shows the formwork used for Lane 7.

3.6.1.2 Concrete mix proportions

A minimum 24-hour compressive strength of 2500 psi and minimum 28-day strength of

5800 psi were specified for the concrete for the test tracks. The mix designs used for Lanes 6 and

7 in Phases I-a and I-b are shown in Table 3-1.









3.6.1.3 Placement of concrete

Before placing the concrete on top of the asphalt layer, water was sprayed on the asphalt

surface to promote a good bond between the concrete and the asphalt and to prevent the reduc-

tion of water from the concrete. Samples of concrete were taken from a selected truck during

concrete placement. The slump, air content and temperature of the fresh concrete were measured.

Samples were fabricated for compressive strength and elastic modulus, maturity test, and flexural

strength tests. Figure 3-15 shows the concrete test track after placement of the concrete. After

placement and finishing of the concrete on the test track, saw cuts were made to a one third (1/3)

of the thickness to form the joints for the slabs. A diamond-bladed saw was used for these cuts to

ensure a smooth, straight vertical surface.

3.6.1.4 Placement of strain gages

Embedded strain gages were installed in the test slabs at the location described in Section

3.3.3. Each strain gage was fixed between two steel rods fixed to the base layer. At the locations

where both the top and bottom embedded gages needed to be placed, one gage was fixed at the

top of two rods and the other gage was fixed at the bottom of the two rods using nuts and bolts.

Figure 3-16 shows the placement of the top and bottom strain gages in a 4-inch test slab in Lane

7. The strain gages were placed in the concrete with a distance of 1 inch from the top and 0.5

inch from the bottom of the concrete. Surface gages were placed before loading was started on

the test slab. The surface of the concrete where surface gages were placed was cleaned using a

sand paper and applied with the recommended glue to bond the strain gages to the concrete.

Figure 3-17 shows a picture of the surface gages at a joint on the 4-inch test slab in Lane 6.









A PVC cylinder was placed around the embedded strain gages during placement of

concrete, as shown in Figure 3-18. The concrete was placed in the cylinder manually to prevent

disturbance from the concrete handling instruments. After the concrete was placed to the same

thickness both inside and outside the PVC pipe, the PVC pipe was then removed by pulling it out

vertically.

3.6.1.5 Placement of thermocouples

Thermocouples were placed at various depths in the test slabs to monitor temperature

variation. This was achieved by fixing the thermocouples to a PVC rod at different heights. The

thermocouple-attached rods were fixed to the asphalt layer. Similar to the case of the strain

gages, a PVC cylinder was placed around each of the rods to protect it from the concrete

handling instruments during the concrete placement. The concrete was placed manually in the

PVC pipe to prevent any disturbance from the concrete handling instrument. The PVC pipe was

removed after the concrete was placed to the same thickness inside and outside of the PVC pipe.

This procedure ensured the proper position of the thermocouples by preventing any disturbance

from the concrete placement.

3.6.2 Construction of Concrete Test Tracks in Phase II

3.6.2.1 Asphalt surface preparation and formwork

After testing on the concrete slabs in Phase I-a was finished, the concrete slabs were

removed from Lane 6, and the test sections in Phase II were placed on Lane 6. When attempts

were made to remove the concrete slabs without removing the underlying asphalt layer, it was

found that the concrete slabs were bonded so well to the asphalt layer that a lot of the underlying

asphalt concrete was also removed at the same time. Figure 3-19 shows the condition of Lane 6

during the slab removal process. Due to this situation, a new asphalt layer was placed on Lane 6

before the concrete slabs in Phase II were placed.









Figure 3-20 shows the formwork for the concrete slabs in Phase II. Each test section was

18 feet long, with transition zones separating one test section from another. A tapered formwork

was used at the transition from one thickness to the other to make the placement of the concrete

easier.

3.6.2.2 Installation of strain gages and thermocouples and concrete placement

The method of installation of strain gages and thermocouples was similar to that used in

Phase I. Grooves were cut with a diamond saw on the surface of the asphalt surface for place-

ment of the strain gage and thermocouple wires (Figure 3-21). Thermocouples and strain gages

were installed at specified locations on the test slabs and covered by PVC pipes before placement

of concrete, as described in Section 3.6.1.

Before the concrete was placed, a white pigmented curing compound was applied to the

asphalt surface to act as a debonding agent between the asphalt and the concrete slab. The

prepared asphalt surface is shown in Figure 3-22.

The concrete in Phase II was placed on October 11, 2005. The mix design of the concrete

used is shown in Table 3.1. The same procedure used in Phase I to protect the instrumentation

during concrete placement was followed in this case.

Figure 3-23 shows the finishing of the concrete test track in Phase II. The joints for the

concrete slabs were sawed the following day. Each test section was sawed into six 6 ft x 6 ft

panels. The concrete slabs were kept moist by sprinkling with water for at least 3 days to ensure

adequate curing (Figure 3-24).









Table 3-1 Mix Designs of Concrete Used in Phases I and II.
Lane No. Material Target
Cement 508 lb
D57 Stone 1801 lb
Lane 6 DOT Sand 1328 lb
PHASE I-a Air entrainment, MBAE 90 1 oz
(06/08/2004) Admixture, MBL 80 45 oz
Water 15.6 Gal
W/C


Actual
506 lb
1798 lb
1316 lb
1.1 oz
40.2 oz
17.2 Gal


Moist, % Remarks


Pit # 08-012
Pit # 76-349


Lane 7
PHASE I-b
(10/08/2004)


Lane 6
PHASE II


Cement
D57 Stone
DOT Sand
Air entrainment, MBAE 90
Admixture, MBL 80
Water
W/C
Cement
D57 Stone
Silica Sand
Air entrainment, MBAE 90
Admixture, MBL 80
Water
W/C


508 lb
1801 lb
1328 lb
1 oz
45 oz
15.6 Gal


508 lb
1750 lb
1265 lb
2 oz
65.0 oz
255 lb
0.502


504 lb
1810 lb
1346 lb
1.1 oz
45 oz
16.5 Gal


6.99%
5.84%


Pit # 08-012
Pit # 76-349


Pit # 08-004
Pit # 76-349










6ftX6ft slabs

--- ---ft


12ft


4ftX4ft slabs


Plan View


Concrete


Asphalt
Cross section
Figure 3-1 Layout of the Test Sections for Phase I.


6ftX6ft slabs


l1ft


Plan View

Concrete


- Asphalt

Cross section


Figure 3-2 Layout of the Test Sections for Phase II.


tt~


tt~














Voc


R4 Ro(1+y)



RA= R(1-4-x)(1 +y)


Figure 3-3 Strain Gage Arrangements in a Half Bridge Circuit.




SMember under test


Dummy
Gauge
Gaue Active Gauge




Same material as
member under test


Figure 3-4 Connection of the Active and Dummy Strain Gages in the Half Bridge Circuit.











3 27"

_6"


- Embedded Gauge
- Surface Gauge
* Thermocouple


72"

Figure 3-5 Instrumentation Layout for the Test Slabs in Phase I-a.




















Asphalt


Location 1





Gauge 3
Gauge3 3


Concrete


Asphalt


Location 2


Gauge 4 -


-0.5"


Concrete


Asphalt


Location 3


Figure 3-6 Vertical Positions of the Strain Gages in Phase I-a.




































Asphalt


4. 1E
El 1Concrete
1"

Asphalt


Figure 3-7 Vertical Positions of the Thermocouples for the 4", 5" and 6" Slabs in Phase I.














































Embedded Gauge
-Surface Gauge
Thermocouple

48"


Figure 3-8 Instrumentation Layout for the Test Slabs in Phase I-b.









































Figure 3-9 Vertical Positions of the Strain Gages in Phase I-b.


* Thermocouple
- Gauge


Figure 3-10 Instrumentation layout for Phase II.


























Locations 1,2,3,4 and 5


Figure 3-11 Vertical positions of the strain gages in Phase II.


2
6"

n3"
Concrete


Asphalt


1"




1 I


SConcrete


Asphalt


Figure 3-12 Vertical positions of thermocouples in Phase II.


Concrete



Asphalt

































Figure 3-13 Milled surface before concrete placement on Lane 6 in Phase I-a.


Figure 3-14 Formwork prepared for Lane 7 in Phase I-b.

































concrete on Lane 6


Figure 3-16 Placement of top and bottom strain gages on Lane 7 in Phase I-b.


Figure 3-15
































Figure 3-17 Placement of surface strain gages at a joint in Phase I-a.


Figure 3-18 Strain gages in a protective PVC pipe before placing concrete.




















Figure 3-19 Removal of concrete slabs from Lane 6 in Phase I-a.


w .
U-
i*.. "


Figure 3-20 Formwork for test slabs in Phase II.































Figure 3-21 Grooves on asphalt surface for placement of strain gages and thermocouples
cables in Phase II.


Figure 3-22 Asphalt surface with white curing compound before concrete placement in Phase
II.
































Figure 3-23 Finishing of the concrete for the test track in Phase II.


Figure 3-24 Curing of concrete by sprinkling with water









CHAPTER 4
MATERIALS AND PAVEMENT CHARACTERIZATION

4.1 Materials Characterization

Tests were performed to characterize the pavement materials used in the test sections in

this study. The properties measured include the concrete-asphalt interface bond strength, com-

pressive strength, splitting tensile strength and elastic modulus of the concrete, resilient modulus

and indirect tensile strength of the asphalt concrete, and the penetration and absolute viscosity of

the recovered asphalt from the asphalt cores.

4.1.1 Interface Bond Strength

4.1.1.1 Results from test sections in Phase I-a

Iowa shear tests were performed on both the 4-inch and the 6-inch diameter core samples

extracted from the 4-inch slabs in Phase I-a before loading started (at 28 days or later). The

average shear strength for two 6-inch diameter core samples was 207.5 psi, while the shear

strength for one 4-inch diameter sample was 165 psi.

Six 6-inch diameter cores of the concrete/asphalt composite layer were extracted from each

of the test sections at the end of the HVS testing in Phase I-a. The locations of the cores were

selected so that the bond strength for different conditions could be evaluated. For each test

section, two cores were obtained from the wheel path (loaded area) one from the corner and one

from the mid edge of the slab. Four cores were obtained outside the wheel path one from the

center of an unloaded slab, one from the center of the loaded slab, one from the mid edge of the

longitudinal joint, and one from the mid edge of the transverse joint. Figure 4.1 shows the core

samples from the 6-inch slabs. Figure 4.2 shows the locations of the cores with the measured

bond strengths displayed next to them. Table 4-1 shows the results of the Iowa tests on the cores

extracted from the test lane in Phase I-a (Lane 6) after HVS loading.









An examination of the data shows that the measured bond strength was not affected by

their locations on the pavement. No loss of bonding due to repeated loading was observed. The

loaded area had equally high bond strength as the unloaded area. The average bond strength was

about 220 psi.

4.1.1.2 Results from test sections in Phase I-b

For the test sections in Phase I-b, a total of 7 cores of the concrete/asphalt composite layer

were extracted from each test section. Six of these samples were tested for interface bond

condition after loading. Table 4-2 summarizes the results of the Iowa Shear Test run in these

samples. It can be observed that the average shear strength was 195 psi. Similar to the case for

Phase I-a, no significant difference in the shear strength was observed among the cores taken

from the corner, mid edge and center of the slabs both in and out the wheel path. This means that

the area of the slab interface loaded for a short period of time did not experience more deteriora-

tion than that out of the wheel path. Figure 4-3 shows the locations of the cores taken after HVS

loading for each test section in Phase I-b.

4.1.1.3 Results from test sections in Phase II

Four cores were taken from the 10-inch slabs in Phase II before the HVS loading was

started, and Iowa shear tests were run on these cores to determine the interface shear strength.

Though the concrete-asphalt interface was intended to be unbonded, the cores indicated that the

concrete was partially bonded to the asphalt layer. The average interface shear strength from the

four cores was 118.6 psi.

After HVS loading, six cores were taken from each test slab, and Iowa shear tests were run

on these cores for comparison purpose. The locations of the cores in the 6' x 6' slabs were the

same as the one used for the Phase I-a, which is shown in Figure 4-2. Table 4-3 displays the









results of the Iowa shear tests on these cores. The average shear strength for these cores was 135

psi.

4.1.1.4 Comparison before and after HVS loading

Table 4-4 summarizes the measured interface bond strength before and after HVS loading

for both Phase I and Phase II. It can be noted that for the test sections in Phases I-a and I-b,

where a bonded condition was intended, there was a small increase in bond strength after HVS

loading. For the test sections in Phase II, where an unbonded condition was originally intended,

there was a larger gain in bond strength after HVS loading.

4.1.2 Concrete Properties

4.1.2.1 Properties of concrete sampled from concrete trucks

Samples of concrete were taken from a selected truck during the placement of the test

slabs. The slump, air content and temperature of the fresh concrete were measured. Samples

were fabricated for compressive strength and elastic modulus, and flexural strength tests. The

properties of fresh concrete used in the construction of the test sections in Phases I-a, I-b and II

are shown in Table 4-5.

Additional concrete samples from Phase II were prepared to evaluate its coefficient of

thermal expansion. The average value of this parameter was determined to be 6.5 X 10-6 1/F.

Tables 4-6 and 4-7 show the compressive strength, elastic modulus and flexural strength at

various curing times of the concrete sampled from the truck in Phase I-a and Phase II,

respectively. All these tests were performed by FDOT personnel at the FDOT facility.

4.1.2.2 Properties of concrete from core samples

Splitting tensile strength test was run on the concrete portion of the core samples after the

Iowa bond strength test. From the core samples obtained from the test sections in Phase I-a

before the start of the HVS loading, the average indirect tensile strength of concrete from three









samples was 610 psi. As described earlier, 18 core samples (6 from each test section) were taken

from the test sections in Phase I-a after the HVS loading. Table 4-8 displays the results of the

indirect tensile strength test on the concrete portion of these 18 core samples. The average

indirect tensile strength ranged from 473 psi for the 4-inch concrete slabs to 509 psi for the 5-

inch slabs.

4.1.3 Asphalt Concrete Properties

Resilient modulus test was performed on the asphalt portion of a core that was obtained

from a 4-inch slab in Phase I-a before HVS loading. This test was run by FDOT personnel at the

FDOT facility. The resilient modulus at 10 OC was determined to be 1,263 ksi.

Resilient modulus and indirect tensile strength tests were also run on the asphalt portion of

the cores taken from the test sections in Phase I-a, Phase I-b and Phase II at the end of each HVS

loading period. These tests were performed in the Asphalt Lab of the Department of Civil and

Coastal Engineering at UF. The results of these tests are summarized in Table 4-9.

Figure 4-4 shows the effect of the temperature in the resilient modulus of the AC layer.

From this graph it can be observed that the MR of the AC layer can be drastically reduce from

1800 ksi at 5 C to a value lower than 400 psi at 40 C. The adjusted curve and the equation

shown in this Figure 4-4 will be used in the next chapters for both calibrating the analytical

model and estimating the level of stresses.

The asphalt binders were extracted and recovered from these asphalt concrete samples.

Penetration tests at 25 C and the absolute viscosity test at 60 oC were run on the recovered

asphalt binders. These tests were performed by FDOT personnel at the FDOT facility. The test

results are shown in Table 4-10. The recovered asphalt binders were shown to be fairly

consistent in properties with the penetration ranging from 20 to 26, and the viscosity at 60 C

varying from 61,000 to 73,000 Poises. The viscosity at 60 oC of a recovered asphalt binder from









a new pavement in Florida is generally in the range of 6,000 to 10,000 Poises. Thus, the

recovered asphalt represented an asphalt binder which had been substantially aged.

4.2 Measurement of Joint Movement

Two pairs of Whitmore plugs were placed at the joints of each test slab to monitor joint

movement. Each pair of Whitmore plugs were placed at a distance of 6 inches apart from one

another, and one on each side of the joint. These plugs were fixed to concrete before the fresh

concrete stiffened during placement. Figure 4-5 shows the Whitmore plugs fixed at a joint. The

Whitmore gage with the standard Invar bar is shown in Figure 4-6. The invar bar is a reference

bar which was used to calibrate the Whitmore gage. The distance between the gage points was

measured in early morning before 7 AM and in mid afternoon around 3 PM, which represent the

two extreme temperature conditions in a day. In addition, joint movements were measured every

two hours from 6 AM to 5 PM on some selected days to monitor the slab movement throughout

the day. Figures 4-7 through 4.9 show the measured Whitmore plug spacing from a 4-inch, 5-

inch and 6-inch slab in Lane 6 (with 6 ft by 6 ft joint spacing) at 7 AM and 3 PM in some

selected days. Significant joint movement was observed at the joints with extended cracks. The

joint movement was minimal until the extended cracks were formed at the joints. Figure 4-10

shows the change of gage spacing for all three slabs on a selected day. Here, the gage spacing is

shifted to an original gage length of 10 inches for comparison purpose. Thus, only the changes

in gage spacing, rather than the absolute gage spacing, are to be read from this plot.

4.3 Measurement of Slab Profile Using a Dipstick

A grid was drawn on four test slabs with grid points spaced 12 inches apart. A Dipstick

profiler was used to measure the elevation of the grid points with respect to a reference point

established close to the test slab. Figure 4-10 shows the grid that covers four slabs. The

coordinates (1, 0) represent the reference point. Figure 4-11 shows the dipstick profiler with one









leg of it placed on the reference point and the other on a point in concrete slab. The elevation

collected at 7 AM and 3 PM were plotted as shown in Figure 4-12. It can be observed that this

4-slab unit curled up at the middle of the slab in the afternoon and curled up at the edges in the

morning. The curling of these 4 slabs together agrees with the typical rigid pavement behavior at

positive and negative temperature differentials. This observation indicates that the 4 slabs act as

one continuous slab before cracking occurred at the joints.

4.4 FWD Tests

Falling Weight Deflectometer (FWD) tests were performed on the composite pavement

test sections in all phases of the study. The measured FWD deflection basins were used to

estimate the elastic modulus of the pavement materials and the stiffness of the springs used to

model the load transfer at the joints and concrete-asphalt interface through a back-calculation

process. This back-calculation process also allowed for the verification of the elastic modulus of

the concrete and the asphalt layer, previously evaluated from laboratory testing.

FWD tests were run at midday between 1:30 PM and 3:30 PM and at early morning

between 7 AM and 8 AM. At mid day, the temperature differential tends to be positive and slab

tends to curl down at the edges and joints. This is the best time to run the FWD test for

evaluation of joints because the slab is more likely to be in full contact with the layer underneath

at both the edges and joints. From midnight to early morning, the temperature differential tends

to be negative and the slab tends to curl down at the center of the slab. This is an ideal time to

run the FWD test at the center of the slab for evaluation of the condition of the concrete slab and

the layer underneath. In order to reduce the effects of the joints, the FWD test run in the morning

was performed on a transition slab with the same slab thickness but with a larger panel size.









4.4.1 FWD Tests in Phase I-a

FWD tests were run on the 4" slab of the test section in Phase I-a. Figures 4-13 and 4.14

show the schemes of the morning and afternoon tests, respectively. The morning test was run at

7AM with an average pavement temperature of 65 F. The afternoon test was run between 2 PM

and 4 PM with an average pavement temperature of 99 F. In the case of the afternoon test, a

load of 12 kips was applied to both the comer and at the mid-edge of the slab. In all cases the

distance between sensors was 12 inches.

4.4.2 FWD Tests in Phase I-b

FWD tests were run on the test sections in Phase I-b. Three sets of reading were taken for

three different loads (at around 9, 12, and 15 kips). A replicate test was run right after each test

was completed to check for consistency. Figures 4-15 and 4-16 show the FWD load and sensor

positions used for the FWD test at the slab corner and slab edge, respectively. These schemes

correspond to those applied in the 4' x 4' slabs (in Phase I-b). The average pavement

temperatures were 56 and 75 F for the morning and afternoon tests, respectively. Figure 4-17

shows the FWD load and sensor positions used for the FWD test at the slab center. Figures 4-18

through 4-20 show pictures of the FWD tests run at the slab corner, slab edge and slab center,

respectively.

4.4.3 FWD Tests Phase II

Figures 4-21 and 4-22 show the FWD testing plan for the test sections in Phase II. This

plan is very similar to the one used for Phase I-a since both used 6' x 6' slabs. Morning and

afternoon tests were performed to evaluate the elastic modulus of the layers and load transfer

characteristics at the joints, respectively. Three sets of reading were taken for three different

loads (at around 9, 12, and 15 kips). The average pavement temperatures were 78 and 108 F

during the morning and afternoon tests, respectively.









Using the deflections caused by the FWD load, an analysis of the load transfer was

undertaken. The following load transfer factors where defined to represent the load transfer at

both longitudinal and transversal joints:

FLon = (DOU)ME/(DOL)ME x 100 (4-1
FTra= (D-12U)C/(DoL)C x 100 (4-2

Where:
FLon = Longitudinal load transfer factor
FTra = Transversal load transfer factor
(DOL)ME = Deflection under the FWD load on the loaded slab, when the test was run a
the mid-edge
(DOU)ME = Deflection across the longitudinal joint, in front of DoL, on the unloaded sla
when the test was at the mid-edge
(DL)c = Deflection under the FWD load on the loaded slab, when the test was run a
the corer
(D-12U)c = Deflection across the transversal joint, 12 inches apart of Do, in the direction


)
)


t

b

t

n


of the sensors, on the unloaded slab when the test was run at the corner.


Figure 4-23 shows a comparative analysis of the load transfer using the factors defined

above. According with the definition, as the factor approach to 1.0, more load is transferred from

one slab to the adjacent slab.

While these factors cannot be used to determine the degree of load transfer for composite

pavements, they can be used to compare the load transfer based on different pavement

characteristics. It is clear from the graph that the load transfer at both the longitudinal and the

transversal joint increases as the thickness of the slab increases. Similarly, there is more load

transfer at the transversal joints than at the longitudinal joints. By comparing the 6" slab in both

phases, it seems that the shorter slabs can transfer more load. However this result might be

influenced by the fact that Phase I-b considered a bonded condition in the interface while Phase

II was intended to be unbonded. It may be possible that the bond in the interface collaborated

with the load transfer since the AC layer is continuous.









Table 4-1 Results of the Iowa Shear Tests on the cored samples from test sections in Phase I-a


after HVS loading.
Sample
Testing Date S

6/28/2005 1-4"
6/28/2005 2-4"
6/28/2005 3-4"
6/28/2005 4-4"
6/28/2005 5-4"
6/28/2005 6-4"
6/28/2005 1-5"
6/28/2005 2-5"
6/30/2005 3-5"
6/28/2005 4-5"
6/30/2005 5-5"
6/30/2005 6-5"
6/30/2005 1-6"
6/30/2005 2-6"
6/30/2005 3-6"
6/30/2005 4-6"
6/30/2005 5-6"
6/30/2005 6-6"


Diameter
(mm)
151.02
151.02
151.33
151.00
151.24
150.98
151.24
151.21
151.26
151.14
151.27
150.95
151.19
150.99
151.12
151.23
151.05
151.22


Diameter
(in)
5.946
5.946
5.958
5.945
5.954
5.944
5.954
5.953
5.955
5.950
5.955
5.943
5.952
5.944
5.950
5.954
5.947
5.954


Area
(in2)
27.76
27.76
27.88
27.76
27.85
27.75
27.85
27.83
27.85
27.81
27.86
27.74
27.83
27.75
27.80
27.84
27.78
27.84
Average
St. Dev.


Average without specimen 5-6"
St. Dev. without specimen 5-6"


Load
(lbs.)
5860
5800
6000
5720
5660
6440
5680
5980
5940
6860
7360
5992
6000
6140
6040
5020
3900
7520
5995
799
6118
622


* Outlier, excluded in the computation of average and standard deviation.


Shear Strength
(psi)
211.1
208.9
215.2
206.1
203.3
232.1
204.0
214.8
213.3
246.7
264.2
216.0
215.6
221.2
217.3
180.3
140.4*
270.1
215.6
28.7
220.0
22.3










Table 4-2 Results of Iowa Shear Tests on the cored samples from test sections in Phase I-b after
HVS loading.


Cross Sectional
Area (in2)
28.18
28.09
28.09
28.09
27.99
27.81


Load
(lb.)
5340
5230
5450
5030
5500
7040


Core ID

L7-1A
L7-2A
L7-3A
L7-4A
L7-5A
L7-6A

L7-1B
L7-2B
L7-3B
L7-4B
L7-5B
L7-6B

L7-1C
L7-2C
L7-3C
L7-4C
L7-5C
L7-6C


7040
5590
5650
6270
4520
5050

7080
5770
5680
4500
5130
4780


Average:
Standard Deviation:
Minimum:
Maximum:


Diameter
(in.)
5.99
5.98
5.98
5.98
5.97
5.95


Shear Stress
(psi)
189.5
186.2
194.0
179.1
196.5
251.5

251.5
199.0
201.8
224.0
161.5
180.4

253.8
206.1
201.6
160.2
183.3
170.8
194.9
26.6
160.2
253.8


27.99
28.09
27.99
27.99
27.99
27.99

27.90
27.99
28.18
28.09
27.99
27.99


5.97
5.98
5.97
5.97
5.97
5.97

5.96
5.97
5.99
5.98
5.97
5.97









Table 4-3 Results of Iowa Shear Tests on cores from test sections in Phase II after HVS loading.
Diameter Cross Sectional Load Shear Stress
Core ID
(in.) Area (in2) (lb.) (psi)
L6-1A 5.94 27.71 4720 170.3
L6-2A 6.00 28.27 3980 140.8
L6-3A 5.97 27.99 4140 147.9
L6-4A 5.98 28.09 n/a n/a
L6-5A 5.97 27.99 3800 135.8
L6-6A 5.97 27.99 4250 151.8

L6-1B 5.99 28.18 4300 152.6
L6-2B 5.95 27.81 n/a n/a
L6-3B 5.91 27.43 2140 78.0
L6-4B 5.94 27.71 3550 128.1
L6-5B 5.94 27.71 3190 115.1
L6-6B 5.96 27.90 4200 150.5

L6-1C 5.95 27.81 4990 179.5
L6-2C 5.94 27.71 4280 154.4
L6-3C 5.94 27.71 1730 62.4
L6-4C 5.94 27.71 4270 154.1
L6-5C 5.95 27.81 2640 94.9
L6-6C 5.95 27.81 n/a n/a
Average: 134.4
Standard Deviation: 33.3
Minimum: 62.4
Maximum: 179.5


Table 4-4 Summary of the interface bond strength before and after HVS loading.
Shear Strength Before Shear Strength After
Intended Bond Condition Slab Size Loading Loading
(psi) (psi)
Bonded (Phase I-a) 6' x 6' 207.5 220
Bonded (Phase I-b) 4' x 4' 194.5
Un-bonded (Phase II) 6' x 6' 118.6 134.4









Table 4-5 Properties of fresh concrete used.
Properties Phase I-a Phase I-b Phase II
Slump, inch 3.5 5.75 3.0
Air, % 2.50 2.00 2.4
Unit Weight, pcf 145.8 142.9 143.2
Temperature, OF 93 92 78

Table 4-6 Properties of hardened concrete sampled from truck in Phase I-a.
Curing Time, Compressive Elastic Modulus, Flexural Strength,
Days Strength, psi ksi psi


1,690
2,940
3,930
4,750
5,980
6,750


3,440
3,737
3,940
4,380


Table 4-7 Properties of the hardened concrete sampled from truck in Phase II.
Curing Time, Compressive Elastic Modulus, Flexural Strength,
days Strength, psi ksi psi
1 1,933
3 3,608


4,651


6,083
6,612


3,307
3,875
4,004
4,272


808
855










Table 4-8 Results of Indirect Tensile Strength Test on the concrete samples taken from the test
sections in Phase I-a after HVS loading.


L2 L3 L4


Sample

4"-1
4"-2
4"-3
4"-4
4"-5
4"-6

5"-1
5"-2
5"-3
5"-4
5"-5
5"-6

6"-1
6"-2
6"-3
6"-4
6"-5
6"-6

* Outlier


in
3.81
3.94
4.11
4.02
4.43
4.44


in
3.86
4.04
4.18
4.09
4.26
4.49

5.25
5.18
5.31
5.46
5.47
5.44

6.23
6.20
6.18
6.20
6.00
5.91


L Load Strength
in lb psi
3.78 14610 410.4
4.03 14900 392.8
4.12 17199 440.6
4.09 24220 628.7
4.33 2510*
4.50 20880 492.3


in
3.66
3.92
4.14
4.10
4.30
4.55

5.38
5.36
5.23
5.45
5.47
5.89

6.25
6.15
6.00
6.10
6.01
5.94


in
3.78
4.20
4.04
4.14
4.32
4.52

5.34
5.18
5.23
5.49
5.52
5.47

6.25
6.16
6.00
6.11
6.03
6.00


473.0
379.7
449.9
699.1
585.0
498.2
440.5
508.7
431.2
461.4
477.5
538.7
540.9
491.8
490.3


Table 4-9 Results of Resilient Modulus and Indirect Tensile Strength Tests on the asphalt
concrete samples obtained from test sections in all phases after HVS loading.
T (C) MR (ksi) Strength (psi)
5 1,780
10 1,390
15 1,170


Diam


5.34
5.25
5.27
5.45
5.48
5.62

6.26
6.17
6.08
6.14
6.01
5.95


Average
19121
22270
34690
30050
25720
23320
Average
25440
26840
27340
31150
30640
27570
Average


5.40
5.29
5.29
5.40
5.45
5.67

6.31
6.18
6.12
6.13
6.00
5.94









Table 4-10 Results of Penetration and Absolute Viscosity Tests on the recovered asphalt binders
from cores from Phase I-a after HVS loading.


Needle Pen
ampl ID (@ 25 C)


A-498
Core 4" B-171
B-464
A-498
Core 5" B-171
B-464
A-498
Core 6" B-171
B-464


Avg.
Pen


Viscosity
Tube Bulb Constant Seconds (Poises)
(@ 60 C)


21.00 400/R164 D 810.00 90.30 73,143



23.00 400/E171 D 684.00 96.30 65,869


25 25.00 400R/E357 D


872.00 69.60 60,691

























Figure 4-1 Cores samples from Lane 6 in Phase I-a after HVS loading.


4a Kite^ K^.7













Shear Strength (psi)
4" 6'x6'


5
203.3


211.1 215.2
2

208 9


206.1

4


Shear Strength (psi)
5" 6'x6'


5
264.2


Shear Strength (psi)
6" 6'x6'


5
140.4


6

210.1


215.6 217.3
2 180.3

2212 4




Figure 4-2 Location of the cores taken after loading in Phase I-a.


232.1










Location of cores (4'x4' Slabs)


Figure 4-3 Location of the cores taken after loading for each test section in Phase I-b.




Effect of the Temperature in the AC Elastic Modulus


2000
1800
1600
1400
1200
1000
800
600
400
200
0


20 30 40
Temperature (C)


Figure 4-4 Relationship between Temperature and Resilient Modulus for the AC layer in the
composite pavement.


MR= 2329e-O 0473T
R- 2 = 0.9929





















































Figure 4-5 The Whitmore gage with invar bar.










91












4 Inch slab


North-7.00AM m South-7.00AM A North-3.00PM

Figure 4-6 Measured gage spacing from a 4-inch slab in Lane 6.


0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
_n n3


x South-3.OOPM


xi
I I - 7


x

S--- ---- ---- I --- --- ----

6 A 6 A 66 i A




36 6/8 6/10 6/12 6/14 6/16 6/18 6/20 6/22 6/24 6/26 6/28 6/30 7/2 7
1 1 I I


North-7.00AM m South-7.00AM A North-3.00PM x South-3.00PM

Figure 4-7 Measured gage spacing from a 5-inch slab in Lane 6.


0.26


0.21


0.16

0.11


0.06

0.01


-n nA


0 ** 0- *.








6 6/8 6/10 6/12 6/14 6/16 6/18 6/20 6/22 6/24 6/26 6/28 6/0 -7/2 7
I- x I E* 5w- It ii ---- x IPIt I a
'6 618 6/10 6/12 6/14 6/16 6/18 6/20 6/22 6/24 6/26' 6/28 6/ 712 7














0.09


0.07


0.05


0.03


0.01


-0.01


_n n3


North-7.00AM m South-7.00AM A North-3.00PM x South-3.00PM

Figure 4-8 Measured gage spacing from a 6-inch slab in Lane 6.


10.06

10.05

10.04

10.03

10.02

10.01

10

9.99
6:00 AM


8:00 AM 10:00 AM 12:00 PM 2:00 PM 4:00 PM


Figure 4-9 Changes of joint spacing on a selected day.


1 It .. R


x
x x
x x

x
S- A- -
66-


3 6 618 6/10 6/12 6/14 6/16 6/18 6/20 6/22 6/24 6/26 6/28 6/30O 7/2 7


I I I I I I I I I I_


6:00 PM


















12


11


10


9





c 7

6
o



o 5


4.


3


2


1

S5
3


0 1 2 3 4 5 Xbire&ioWft 9 10 11 12 13 14



Slab Joints

Figure 4-10 Grid marked on slabs for the Dipstick measurement.


Figure 4-11 The Dipstick instrument.








- ---- --- ---- ---- ---- -
S---- ---- ---- - -- -





S---- _-- --- -1 -- -_- _-


- -- - -- -- -- -- -- -

------------
-- --------------------- ---- -



:- - -:- :. -1: ',.-


- ---- ---- -- --- ---- ----


-- ... .. ..-- -------- ---- -






-- 1- --1 -_ ---- _-- -- -_

















-.s 2 L -



0.8
II
0.6



1 0.2


i A 0'- --"



-0.4
.-0.6 --I-


-0.8 i

-1.0 --- -




iiiil 7.00AM
3, t PM

Figure 4-12 Dipstick measurements at two critical temperatures.


































6ft


- 8' 3" 8'- 3" "*






SI S --S


Figure 4-13 FWD load and sensor locations for FWD Test at slab center in Phase I-a.


6Ft






I l- a '-
eFt -0


.--



c--.--
-U U



A. *A

c-rn c--
;-i t^


Figure 4-14 FWD load and sensor locations for FWD Test at slab corer in Phase I-a.













4 ft-


CORNER LOADING


4ft


7 2 3 4 5 6

7 1 2 3 4 5 6





7 2 3 4 5 6

7 1 7 3 4 f


Figure 4-15 FWD load and sensor locations for FWD Test at the slab corner in Phase I-b.


-4 ft-


CENTER EDGE LOADING


4ft

7 2 3 4 5 6
0 () 0 0 0 0


7 1 2 3


* 0 !
4 5 6


7 2 3 4 5 6

7 1 2 3 4 5 6







Figure 4-16 FWD load and sensor locations for FWD Test at the slab edge in Phase I-b.













10' 2"


7 2 3 4 5 6





Sft "A


-o


Figure 4-17


FWD load and sensor locations for FWD Test at the slab center in Phase I-b.


Figure 4-18 FWD Test at the slab corner and measuring deflections on the opposite slab.


-10'- 2"
































Figure 4-19 FWD Test at the mid-edge and measuring deflections in the loaded slab.


Figure 4-20 FWD Test at the center and measuring deflections along the transverse center line.
















CENTER EDGE LOADING (PM TESTING PLAN)
6 ft


D1 positioned beneath applied load

6 ft

30:"-


CORNER LOADING (PM TESTING PLAN)
6 ft


D1 positioned beneath applied load
6ft



2 S 4 S 6 S.55


Figure 4-21 FWD Testing Plan for the mid-edge and corer load in Phase II.


6' 10" 10'- 9." 10' -3" 6' -3"


() Load Load

D1 positioned beneath applied load
5 ft 6 ft


-* -*-*- -*--


t"
7 I 2 3 4 5 6
-e---o<- -0O--- 4- -O'

-


6" Slab End 10" Slab End

CENTER LOADING (AM TESTING PLAN)



Figure 4-22 FWD Testing Plan for the center load at the two ends of the test track in Phase II.












Load Transfer at Joints


90

75 ---------- ---- -- ----------- --------
85




0---- ---- ------- -
80






65
70 --- --- .... ---..

4 (Phase I-b) 5 (Phase I-b) 6 (Phase I-b) 6 (Phase II) 8 (Phase II) 10 (Phase II)

Thickness (in)

F Transversal U Longitudinal


Figure 4-23 Comparative analysis of load transfer factor for the test sections in Phase I-b and II.









CHAPTER 5
TESTING OF TEST SECTIONS AND DATA ANALYSIS

5.1 HVS Loading of Test Sections

The HVS testing has the primarily objective of providing the necessary data for the

calibration of the computer model. HVS testing of the test sections was performed to measure

strain, evaluate temperature effect and eventually observe performance of the composite

pavement under load repetition. The following paragraphs are a description of the HVS test for

each phase.

5.1.1 HVS Loading of Test Sections in Phase I-a

The three test sections on Lane 6 (in Phase I-a), with a panel size of 6 ft x 6 ft and concrete

slab thickness of 4, 5 and 6 inches, were loaded by the HVS according to the plan as described in

Section 3.4.

HVS loading of the test section with 4-inch slabs was started on July 11, 2004. This test

section performed well with no observed cracking after 3 days of loading at 9 kips with a total of

36,407 passes, followed by 12 days of loading at 12 kips with a total of 146,748 passes. Then the

load was increased to 15 kips for 3 days with 35,918 passes and then to 18 kips for two days with

21,727 passes. Finally, the load was increased to 21 kips, and corer cracks developed after

12,187 passes, as shown in Figure 5-1. This was the only load-induced crack observed during

this phase and in the entire experiment. It is to be noted that the HVS loading had to be

suspended for two times during this entire test duration due to mechanical problems. As

indicated, the application of a very large wheel load and a high number of load repetitions were

necessary to break the thinnest slab in the full scale experiment. At that time, it was realized that

for the thicker slabs it would take much more time to break them and the experiment was









confined to study the level of strains and stresses due to normal wheel load, which is 12 kips per

wheel (24 kips per axle).

HVS loading of the test section with 5-inch slabs was performed from November 1, 2004

through January 14, 2005. This test section was loaded with 79,014 passes of 9-kip, 130,186

passes of 12-kip, 25,638 passes of 15-kip and finally 128,817 passes of 18-kip HVS wheel load.

No load-induced cracks were observed during and at the end of this testing period. Only a few

small shrinkage cracks were observed on the surface of the concrete.

HVS testing of the test section with 6-inch slabs was performed from May 23, 2005

through May 26, 2005. This test section was loaded for three days with a total 41,239 passes of

12-kip wheel load. No load-induced cracks were observed, and only a few small shrinkage

cracks were observed on the surface of the concrete.

Shrinkage cracks were observed in most of the slabs in Phase I-a before the loading period.

Shrinkage cracks on the test slabs of the 4, 5 and 6-inch test sections are shown in Figures 5-2

through 5-4, respectively. In addition, a lot of shrinkage cracks in the transverse direction were

observed in the transition zones where the thickness changed from 4 inches to 5 inches and from

5 inches to 6 inches.

5.1.2 HVS Loading of Test Sections in Phase I-b

The three test sections on Lane 7 (in Phase I-b), with a panel size of 4 ft x 4 ft and concrete

slab thickness of 4, 5 and 6 inches, were loaded with 12-kip HVS wheel loads at 120 psi tire

pressure. Table 5.1 shows the dates of the test and the number of wheel passes applied to each of

the three test sections. No load-induced or shrinkage crack was observed in any of the three test

sections at the end of each loading period.









5.1.3 HVS Loading of Test Sections in Phase II

The test sections in Phase II with a panel size of 6 ft x 6 ft, concrete slab thickness of 6, 8

and 10 inches, and un-bonded concrete-asphalt interface were loaded by three levels of HVS

wheel loads to study the linearity of the load-strain relationship and eventually observe any load-

induced crack. Table 5.2 summarizes the HVS loads, numbers of wheel passes and loading

periods for the test sections in Phase II. Unfortunately no load-induced or shrinkage cracks were

observed in any of the slabs at the end of the tests.

5.2 Analysis of Temperature Data

Temperature data were analyzed in three ways. First the temperature differential was

estimated based on the values from the thermocouple located near or on the surface of the

concrete slab and the thermocouple located near the bottom of the slab or the one located in the

surface of the asphalt layer. This value was to be used to estimate the level of stresses under the

combination of wheel load and temperature effect using the analytical model in later chapters.

Secondly, the absolute temperature in the AC layer was estimated directly from the

thermocouple located in the surface of the layer. This value was used to estimate the effect of the

temperature on the AC's elastic modulus and thus its effect on the composite pavement as a

supporting layer. Finally the distribution of the temperature along the depth of the concrete slab

was analyzed to better address its effects on the stresses. An investigation of the effect of both

the temperature differential and the temperature in the AC layer on the peak strain was also

undertaken, which is shown in the next section in this chapter.

5.2.1 Temperature Differential

In Phase I-a, two sets of thermocouple wires were used to monitor the temperature in the

slabs. Thermocouples were installed at various depths, at one-inch increments, in the test

sections to monitor the temperature distribution. One set of thermocouples was installed at the









slab corner at the side that would be loaded by the HVS wheel and the other set of thermocouples

was installed at the slab center. Temperature differentials between the top and the bottom of the

slab were computed and plotted against time for the 4-inch, 5-inch, and 6-inch slabs in Figures 5-

5, 5-6 and 5-7, respectively.

It was observed that for the 4-inch slab, the temperature differential at the slab corner was

relatively low compared with that at the slab center. In all cases, the slab corner was under the

shade of the HVS during the day time, and thus did not get as much heating from the sun as the

slab center did. This explains why the slab corner had lower positive temperature differentials

than the slab center, as the slab corner was less heated by direct sunlight in the day time.

However, the temperature differentials computed for the 5-inch slab did not show a significant

difference between slab center and slab corer. It is to be noted that the temperature readings

from the 5-inch slab were taken during the winter, while those from the 4-inch slab were taken in

the summer. For the 6-inch concrete slabs, temperature was recorded between May 20 and May

25, 2005. In this case, temperature data were recorded at one minute intervals for the first three

days and at 5-minute intervals for the other days.

For the 6-inch slabs, the maximum positive temperature differential at the slab corer was

+18.5 OF (+10.28 oC), while the maximum negative at the slab corner was -8.6 OF (-4.8 oC) during

the HVS loading of these slabs. For the slab center, the maximum positive and negative values

were +25 OF (13.9 oC) and -5.2 OF (-2.9 oC), respectively

The temperature differentials for the 4-inch and 5-inch slabs, which were monitored and

loaded a longer time compared with the 6" slab, were analyzed. Statistical analyses were

performed to determine the number of hours in a day when the temperature differential was in a

certain range. Table 5.3 shows the results of the statistical analysis for the 4-inch slab in









different months during the testing period. The 4-inch slab was tested during the summer of 2004

and the thermocouple used to do this analysis was located under the shade of the HVS. That is

the reason why most of the time the temperature differential was between -5 C and +5C. Table

5.4 shows the results of the statistical analysis for the 5-inch slab. The 5-inch slab was tested

during the winter of 2004 and it can be observed that most of the time the temperature

differential was between -5 C and +5 C

Similar to the case in Phase I-a, temperature in the 4'x 4' concrete slabs in Phase I-b was

measured by means of thermocouples installed in the slabs. They were placed at one inch apart

vertically in the slab at two different locations, namely at the center and at the corner of the slab.

The temperature data were collected (1) between June 17 and June 22, 2005 for the 6-inch slabs,

(2) between June 13 to June 17, 2005 for the 5-inch slabs, and (3) between June 9 and June 12,

2005 for the 4-inch slabs.

Figures 5-8 through 5-10 show the plots of temperature differentials versus time in Phase

I-b, as measured at the slab edge and at the slab center of the 6-, 5- and 4-inch slabs,

respectively. The maximum positive and negative temperature differentials were (1) +28.8 F

(+16 C) and -6.3 F (-3.5 C) for the 6-inch slabs, (2) +27 F (+15 C) and -5.4 F (-3 C) for

the 5-inch slabs, and (3) +18 F (+10 C) and -2.7 F (-1.5 C) for the 4-inch slabs during their

respective test periods.

Temperature data in Phase II were collected every 5 minutes during the loading period.

Thermocouples were installed in three different locations in the slab. At each location,

thermocouples were installed at 1 inch from the top of the slabs and then at 2-inch increments

along the thickness of the slab. One thermocouple was installed in the interface between the

concrete slab and the asphalt surface. Figures 5-11 through 5-13 show the temperature









differential for the 10-, 8- and 6-inch slabs, respectively. The slabs in Phase II were loaded

during winter time, and for this reason the maximum positive temperature differential was less

than 15 F. The maximum negative temperature differential reached -10 F in the 10-inch slab.

The actual temperature differential should be slightly higher since the thermocouple used to

estimate the temperature at the top was located at 1 inch below the surface of the slab.

5.2.2 Temperature in the AC layer

Figures 5-14 through 5-16 show the variation of the temperature on the surface of the

asphalt concrete layer during the loading period in Phase I-a. This is an important variable to be

considered in the analysis since the elastic modulus of the asphalt layer varies substantially with

the changes in temperature as shown in previous chapter, affecting the behavior of the composite

pavement.

In the cases of 4- and 5-in test sections, which were loaded and monitored for a longer time

compared with the 6" slab, a seasonal variation of the temperature in the AC layer could be

observed. While the 4" slab kept a temperature between 25 and 35 C during the whole summer

of 2004, the 5" slab varied from 30 C in the Fall of 2004 to 10 C in the Winter of 2004-2005.

The 6" slab in Phase I-a was monitored for only 4 days and a variation of no more than 12 C

was observed.

Figures 5-17 through 5-19 show the temperature on the surface of the asphalt layer in

Phase I-b. All these temperatures were measured during the very hot summer of 2005, and

temperatures in the range of 78-106 F (25 41 C) were observed. The thinnest test section in

this phase (4") was the one that showed less variation in the AC temperature. On the other hand,

the thicker slab (6") showed the greatest AC temperature variation.









Figures 5-20 to 5-22 show the temperature in the AC layer for Phase II. Since the loading

period was during the winter time, the temperature in the asphalt layer was in the range of 50 to

73 F (10 to 23 C). Unlike the test sections in Phase I-b, all test sections in Phase II showed a

similar variation in the AC temperature.

5.2.3 Temperature Distribution

Figures 5-23 through 5-25 show the temperature distribution along the depth of the

concrete slab for the cases of maximum positive and maximum negative temperature differential.

A significant difference can be observed in the temperature distribution when comparing the

cases of positive and negative temperature differentials. The distribution tends to be non-linear

under a negative temperature differential condition. Also the distribution tends to be non-linear

for thicker slabs and the non-linearity is more severe when considering the temperature measured

in the center of the slab compared with the one measured in the corner. It seems that the shade

provided by the HVS has also an effect on the temperature distribution.

Figures 5-26 through 5-28 show the temperature distribution along the depth of the

concrete slab in Phase I-b. Similar to the previous phase, some differences can be observed from

the temperature values measured by thermocouples at different locations. Again the nonlinearity

tends to increase as the thickness of the slab increases and that is also affected by the shade of

the HVS.

Figures 5-29 through 5-31 show the temperature distribution along the slabs for the tests

sections in Phase II, which show the same trends as in the previous phases.

5.2.4 Summary of temperature analysis

Table 6-5 shows a summary of the extreme values for temperature differential and the

absolute temperature on the AC layer for all the slabs in the different phases. It can be noticed









that a maximum temperature of 106 F (41 C) and a minimum of 48 F (8.9 C) were reached on

the surface of the asphalt concrete layer depending on the season.

After analyzing all the cases of temperature distribution along the depth of the concrete

slab, the following can be stated:

(1) Temperature distribution tends to be linear when the slab is directly heated by the
sunlight. During the day, light sun heats the slab uniformly so that the heat flows
rapidly along the depth of the slab allowing the temperature distribution to be linear for
positive temperature differential

(2) Non-linearity of the temperature distribution is commonly observed when the slab is
affected by negative temperature differentials. It seems that the surface of the concrete
slab cools faster than the rest of the slab

(3) Non-linearity of the temperature distribution increases as the thickness of the slab
increases

5.3 Analysis of Strain Data

5.3.1 Dynamic Strain versus Static Strain

The first type of analysis using the measured strains collected from the test sections was a

comparison between maximum strains caused by dynamic and static loads. The maximum strains

(compression or tension) measured when the slab was loaded by a moving HVS wheel load were

compared to the corresponding maximum strains when the same HVS load was applied

statically. Table 5.6 shows the maximum static and dynamic strains as measured by Gages 1, 2

and 5 of the 4-inch slab in Phase I-a, when the wheel load of various magnitudes was applied at

the mid-edge or corner of the slab. Figure 5-32 shows the comparison of static and dynamic

strain for Gage 1. Comparison of measured strains caused by dynamic loads and static loads

indicated that they were very close to one another for this gage, which was in compression.

However, the difference between static strain and dynamic strain increased with time due

possibly to the micro cracks induced in the concrete. There were significant differences between

the strains measured for dynamic and static loads at Gages 2 and 5, which were in tension. It









appears that the effects of micro cracks were more significant when the concrete was in tension

than when it was in compression. Figure 5-33 shows the comparison of static and dynamic

strains for Gages 2 and 5.

Comparisons were also made for the static and dynamic strains for the strain gages in the

5-inch slab in Phase I-a. Dynamic strains caused by slow moving HVS loads at 1 mph were also

measured to detect any positioning errors and to determine the strains measured at slow speed.

Table 5.7 shows the static and dynamic strains for Gages 2 and 4 caused by 9 and 12 kip loads.

Dynamic strain measurement at slow speed was made at the end of the testing. Table 5.8 shows

the strains data obtained with slow-speed loads at 1 mph, along with the strains obtained with

static loads and loads at 8 mph, caused by 15 and 18 kips wheel loads. The plots of dynamic and

static strains for Gage 2 and 4 are shown in Figures 5-34 and 5-35, respectively.

Static and low-speed load produce higher strains compared with dynamic 8mph load since

in the first two cases the loading time period allows the pavement to completely develop the

strain. During the dynamic load, any point of the pavement is loaded for a very short period of

time.

5.3.2 Measured strain and calculated strain

Dynamic strain data were collected at each hour for 30 seconds during the loading period.

Figure 5-36 shows a plot of typical strain data for two gages at the mid-edge of the slab. One

gage was located at a depth of 1 inch from the top of the slab, and the other gage was located 1"

or 0.5" above the bottom of the concrete slab. The positions of the strain gages for the different

phases have been previously shown in Figures 3-5 and 3-6, Chapter 3.

Figure 5-36 is an example taken from Phase I-a and it clearly depicts the strain reversal

that was observed during the passing of the wheel load in all test sections. For the gage located

near the top of the slab in the mid edge, as the wheel approached the gage location, tensile strain









was measured by the gage. When the wheel was directly above the gage location, the strain

reversed to a high value of compressive strain, and when the wheel moved away from the gage

location, the strain again reversed to a tensile strain. Strains of opposite signs can be observed for

the strain gage located near the bottom of the slab. The peak strain reversal happens at the time

the wheel load is positioned at the corner of the slab.

To isolate the effect of the wheel load and to overcome the fact that the strain gages

cumulated a lot of strain at each reading, the strain profile shown in Figure 5-36 was zeroed so

that it looked like the one depicted in Figure 5-37. Then the peak load-induced strain in

compression (at the top gage) and tension (at the bottom gage) were extracted from each 30-sec

data file. Where a strain gage was installed, the peak strain on the surface of the AC layer was

also extracted. Peak load-induced strain here refers to the highest value of strain observed in the

30-sec strain data as shown in Figure 5-37, which happened when the wheel load passed directly

above the strain gage location.

Maximum strains correspond to the strains calculated at the very top and at the very

bottom of the concrete slab. These strains were calculated based on the peak strain as measured

by gages near the top and near the bottom of the slab and assuming a linear variation of the strain

along the depth of the slab.

Verification of the vertical position of the embedded strain gages was undertaken to

improve the accuracy in the calculation of maximum strains. Small variation in the position of

the gages in thin slabs in Phase I can drastically affect the result of the strain analysis. Some

vertical positions of the gages were verified through cores taken at the locations of the strain

gages after HVS testing. A total of 7 gage positions were verified in Phase I and 9 were verified

in Phase II.









Knowing the exact position of the strain gages allows for a better estimation of the

maximum strain at the top and at the bottom of the slab. Most of the analysis involving strain

was done with the data collected for the smaller load in the experiment (12 kips) so that the

concrete slab behaved as a linear elastic material. In the cases where the exact position of the

strain gages was not known, the target position was used. In Phase I the top gage was planned to

be 1" below the top surface of the slab and the bottom gage was intended to be 0.5" above the

bottom surface. In Phase II each gage was 1" away from the respective concrete surface.

5.3.3 Effect of Temperature on the Strain

The peak strain extracted from the data file as explained in the previous section did not

completely consider the effect of the temperature differential. As indicated before the strain

profile for each 30-sec data set was zeroed not only to isolate the load-induced strain but also to

delete the strain cumulated in the gages after each reading. Each strain data set was 30 sec long

with 4 or 5 peaks that lasted only 0.6 sec each (Figure 5-36). The fastest temperature differential

change measured during the HVS test was 4 F in 10 minutes. Because both the temperature

differential and the correspondent temperature-induced strain remained almost constant within

the 30-sec loading period, the effect of the temperature was lost at the end of the zeroing process.

The wheel load-induced strain can be a function of the temperature, since its magnitude

can be affected by the slab curling due to a temperature differential. When the slab curls the

wheel load might induce more strain in the slab.

In the composite pavement, if the bond between the layers is perfect, the slab is completely

restricted to curl, then the load-induced strain is no longer a function of temperature.

The effect of the temperature on the load-induced strain was investigated using the data

collected during the HVS test. Plots of strain versus both temperature differential in the concrete

slab and temperature in the AC layer were developed. The temperature measured for each









thermocouple was obtained at the time when the strain data was collected. As mentioned in

Chapter 3, temperatures values were measured at 5 minutes intervals and strain values were

recorded at 60 minutes intervals. The strain extracted from each data file (Figure 5-37) is the

load-induced strain. Figures 5-39 through 5-45 show plots of measured strain versus temperature

differential in the concrete slab. It can be noticed that the measured load-induced strain did not

vary too much with temperature differential. This can be interpreted as the slabs were bonded

well enough to the asphalt layer so that very little or no curling occurred. The slabs in Phase II

were intended to be fully un-bonded, however the result from this analysis showed that they have

a considerable degree of bond.

Temperature in the AC layer was found to have an effect on the load-induced strain, as it

can be seen from Figures 5-46 through 5-52. This result demonstrates the fact that temperature in

the asphalt layer affects the AC's elastic modulus, and therefore the AC as a supporting layer. As

the temperature increases, the elastic modulus of the AC layer decreases and therefore both the

compressive strain and the tensile strain in the concrete slab increases. These figures also show

the variation of the strain in the surface of the AC layer with respect to the temperature in that

layer. The AC's elastic modulus tends to decrease when temperature increases. At the same time

stresses in the AC layer tend to decrease since the higher stiffness of the concrete makes the slab

to take more stresses compared to the AC layer. Therefore the strain in the AC layer can increase

or decrease depending on how small the new stress in the AC layer is. That is probably the

reason why a trend cannot be observed in these plots for the tensile strain in the asphalt layer.

5.3.4 Effect of the load magnitude on the strain

Test sections in Phase II were also loaded with higher wheel loads to investigate both the

performance of the composite pavement and the relationship between strain and load magnitude.

In addition to the 12 kips wheel load, the test sections in Phase II were loaded with 15 kips for 5









days followed by another five days with a 18-kips wheel load. As mentioned earlier in this

chapter, no load-induced cracks were observed in any of the loading period in any of the test

slabs. In Figures 5.53, 5.54 and 5.55, the average strains were calculated for each loading period

and plotted versus load to show the strain-load relationship.

Figures 5-53 and 5-54 show a linear relationship between strain and load in both locations

at the top and at the bottom of the concrete slab. The strain gage at the bottom of the 6" slab in

the corner did not work properly and thus was not included in the plots. Figure 5-55 shows the

case of the tensile strain in the AC layer as a function of the load. Though a fairly linear

relationship between load and strain can be observed for the gages located at the mid-edge, a

certain degree of non-linearity can be observed in the gages located at the corer. In Figure 5-55,

it can also be observed that as the slab becomes thicker and the load increases, the strain tends to

remain constant or decrease. By inspecting the plots involving the strain in the AC layer, it seems

that this layer become more rigid as the load increases.

5.3.5 Effect of the loading period on the strain

Plots of strain versus number of passes of the wheel load were developed to show the

effect of the loading period. The strains in these plots (peaks and maximum) were extracted from

the strain data files in the same way as it was done in the previous section. These strains can be

considered as pure load-induce strains since they do not include the temperature-induced strain

component (eT) and very little curling occurred in the slabs as shown in previous sections.

Figure 5-56 shows a plot of the peak load-induced strains measured at different gage

locations as a function of the number of passes of the wheel load for the 6-inch slabs in Phase I-

a. Gages 1 and 2 were located at the mid edge (Location 1) at the top and bottom of the slab,

respectively. Gage 3 was located at the top of the concrete slab in Location 2. Gage 4 was









located at the bottom of Location 3, in the corer of the slab. Similarly, Figure 5-57 shows a plot

of the maximum strains at Location 1 for the 4-inch slab in Phase I-a.

For the 5-inch slabs in Phase I-a, the strain gage at the top of the slab at Location 1 (Gage

1) was out of order from the beginning of the test, and thus maximum strains could not be

estimated from the peak strains. It can be observed from Figures 5-56 and 5-57 that no

appreciable change in the load-induced strains occurred over the testing period. This may

indicate that no crack had occurred near the locations of the strain gauges during the testing

period for the specific load shown in the plot.

Similar to the procedure used in Phase I-a, strain data in Phase I-b were collected every

hour during HVS loading of each test section, at the rate of 100 measurements per second for 30

seconds. In addition to the peak strain at the top and at the bottom of the concrete slab as

indicated in Phase I-a, the strains in the AC surface were also extracted from each strain record.

Figure 5-58 depicts the maximum load-induced strains at the top and at the bottom of the

concrete at Location 1 in the 6-inch slabs as a function of the number of HVS wheel passes. This

figure also shows the strains at the surface of the asphalt layer. Figure 5-59 shows the strain

measured by the gage located at the bottom of the concrete slab and the gage located on the

surface of the asphalt, both of them at the corner of the 6" slab. Figure 5-60 shows similar plots

of load-induced strains versus number of HVS wheel passes for the 5-inch concrete slabs in

Phase I-b.

Figure 5-61 shows the peak strains at the bottom of the concrete slab and on the surface of

the AC layer for the 4" slab in Phase I-b. In this case, the analysis of maximum strain was not

possible to perform due to the failure of the gage located at the top of the slab. The values shown









in this figure cannot be used to check the bonded condition in the interface, since the strains

plotted here were measured by gages close to the interface but at different depths.

From all the figures depicting load-induced strains in Phase I-b, no significant variation of

the strain can be observed as the number of load repetitions increases. The only exception to this

is perhaps the gage located at the corner of the 6" slab, where there is a noticeable increase in the

strain at the end of the loading period. This might be the effect of micro cracks developed near

the location of the strain gages.

In Phase II, a total of 15 strain gages were installed in the slabs to record the strain at

different locations. Figures 5-62 through 5-67 show the maximum measured HVS load-induced

strains in Location 1 (at mid-edge) and Location 5 (at slab corer) as a function of the number of

passes of the 12-kip wheel load. The measured strains shown include (1) the strain near the top

of the concrete layer, (2) the strain near the bottom of the concrete layer, and (3) the strain on top

of the asphalt layer.

Using the measured strain near the top of the concrete layer (Gage 1 and 13) and that near

the bottom of the concrete layer (Gage 2 and 14), the maximum strain at the top and the bottom

of the concrete layer were computed by linear extrapolation. These maximum calculated strains

are also shown in Figures 5-62 through 5-67 along with the gage measurements.

Similar to the results from Phase I-a and Phase I-b, no significant variation in the strain

was observed as a function of the number of passes for the 8- and 6-inch slabs. However for the

10-inch slab, a slight increase in strain was observed at the end of the 12-kips loading period at

the corner of the concrete slab.

5.3.6 Evaluation of the bond condition using strain ratios

To better evaluate the bond condition using the strain data collected during the HVS

loading, strain ratios were developed using averaged values from plots in section 5.3.5. Two









types of strain ratios were used to quantitatively evaluate the degree of bond in the interface of

the composite pavement:

R1 = et/ec = ratio between the maximum tensile strain and the maximum compressive

strain in the concrete slab

R2 = eAc/et = ratio between the tensile strain in the surface of the AC layer and the

maximum tensile strain at the bottom of the concrete slab

These ratios were calculated using the measured peak strains in both the concrete slab and

the AC layer and the maximum calculated strains in the slab. Theoretically, R1 should be close to

1.0 when the bond in the interface is very weak, and the composite pavement can be considered

as fully un-bonded. The value is also close to 1.0 when the composite pavement is bonded but

the asphalt layer is very thin compared to the concrete slab. In both cases there is no

collaboration of the asphalt layer in reducing the tensile strain in the concrete. Theoretically, if

the concrete remain linearly elastic, the compressive strain should be the same as the tensile

strain. In general, it can be said that the lower the value of R1, the better the bond in the interface.

Strain ratio R2 should be close to 1.0 if the composite pavement is fully bonded. In this

case there is full bonding between the concrete and the asphalt layer, and therefore the tensile

strain in the interface should be the same for both materials. Values of R2 far from 1.0 mean that

the composite pavement has a partial bond condition in the interface.

Table 5-11 summarizes the strain ratio values for the different phases in the HVS

experiment. In some cases, Ri could not be calculated because the gage located either at the top

or at the bottom of the concrete slab failed, making impossible to estimate the maximum strains.

In other cases, R2 could not be calculated because there was no gage in the asphalt layer as it was

the case in Phase I-a.









Few things could can be inferred with respect to the interface bond condition in Phase I-a

since few strain gages were placed to monitor the strain and no gage was installed on the asphalt

surface. However from the strain ratio at the mid-edge of the slab, it can be seen that the

composite pavement behaved monolithically since the strain at the bottom was lower than the

strain at the top (Ri<1.0). The low value of R1 for the thinnest slab is a consequence of the

comparable thicknesses between the concrete slab and the AC layer.

For the 5" slab in Phase I-b, the R1 value was lower than 1.0, what agrees with the

predicted behavior. This value indicates that a good bond was achieved between the concrete and

the asphalt layer, which resulted in the reduction of the tensile strains at the bottom of the

concrete layer. However, for the 6" slab in Phase I-b, the maximum compressive strains were

only slightly higher than the maximum tensile strains (Ri very close to 1.0). This might be

caused by either (1) the loss of bonding between the concrete and the asphalt layer, or (2)

inappropriate placement or functioning of the strain gages. By inspecting the R2 values that also

indicates the degree of bond in the interface, both the 5" and the 6" slabs seem not to have a fully

bond condition in the interface (R2<1.0). One possible explanation for this is that the gage

located at the bottom of the 5" slab could have been displaced up during concrete placement, and

thus the measured strain was not only lower than what was expected, but also different compared

with the one measured in the AC layer. Unfortunately, the exact position of gages in the 5" slab

was not possible to check.

The 10" slab in Phase II shows characteristics of an un-bonded composite pavement. The

RI value at both the mid-edge and the corner of the slab are very close to 1.0 and the R2 value is

far from 1.0 (-0.6). For the 8" and 6" slabs the R1 and R2 values at the mid-edge of the slab are

lower than 1.0. Because the Ri values are still high for these two slabs, the interface condition in









these slabs can be considered as a partially bonded. However, the high value of R1 in Phase II

can also be due to the difference in thickness between the two layers and therefore there is little

collaboration of the AC layer in taking the bending. A difference of more than 15 micro-strains

was observed between the two layers at the interface in the 10-inch slab, while the difference

was much lower for the 8- and 6-inch slabs.

The fact that the 10-inch slab had a relatively poorer bond at the interface as compared

with the 8- and 6-inch slabs can be partially explained in terms of the loading period. The 10-

inch slab was the first slab to be loaded in Phase II, at 28 days after the concrete was placed. It

has been reported in other studies that the bond in the interface tends to increase with time due to

mainly the effect of the slab weight. The 8-inch and the 6-inch slabs were loaded at 2 and 3

months after concrete placement. In these two cases, the bond at the interface might have gained

somewhat during this time.

A summary review of Table 5.11 indicates that, in general, the bond condition in the

interface ranged from fully bonded to partially bonded. By looking at the R1 value of the test

sections in Phase I, a high level of bond in the interface was achieved, with the exception of the

6" slab with a joint spacing of 4 ft (Phase I-b). The bond in Phase II seems to be lower than

Phase I as it was meant to be. By looking at the value of Ri and R2 in the corner of the slab, it

looks like the composite pavement in Phase II developed a low level of bond in that particular

location.









Table 5-1 HVS loading period and number of 12-kip wheel passes on the test sections in Phase
I-b.
Slab thickness Starting date Ending date # of 12-kip wheel passes
4" June 9, 2005 June 12, 2005 40,650
5" June 14, 2005 June 17, 2005 38,800
6" June 20 2005 June 22 2005 26040


Table 5-2 Summary of HVS loading on test sections in Phase II.


Starting date
November 14, 2005
December 4, 2005
December 11, 2005
January 9, 2006
January 15, 2006
January 21, 2006
January 30, 2006
February 5, 2006
February 10, 2006


Ending date
December 4, 2005
December 11, 2005
December 16, 2005
January 15, 2006
January 21, 2006
January 28, 2006
February 5, 2006
February 10, 2006
February 15, 2006


# of passes
87,508
86,954
72,554
73,662
60,923
67,015
73,108
67,015
65,908


Table 5-3 Number of hours in a day when the temperature differential was in certain ranges for
the 4-inch slab in Phase I-a.
T<-50F -5FT<+50F 50F>T<10F 100F>T<150F
Month
Ave Max Min Ave Max Min Ave Max Min Ave Max Min Ave Max Min
July 0.0 0.0 0.0 15.2 21.2 11.8 6.8 10.0 2.7 1.3 2.5 0.0 0.6 2 0.0
Aug. 0.4 1.2 0.0 14.7 16.7 12.8 7.5 9.8 3.8 1.1 2.3 0.0 0.3 1.2 0.0
Sep. 0.0 0.0 0.0 15.1 16.8 13.0 6.8 7.8 4.8 1.3 2.3 0.7 0.8 1.7 0.0
Oct. 0.1 0.7 0.0 15.3 17.7 13.2 6.0 8.2 3.7 1.4 2.3 0.8 1.1 1.8 0.0
Season 0.1 1.2 0.0 15.1 21.2 11.8 6.7 10.0 2.7 1.3 2.5 0.0 0.7 2.0 0.0
Time in hours

Table 5-4 Number of hours in a day when the temperature differential was in certain ranges for
the 5-inch slab in Phase I-a.
ST<-50F -5FT<+50F 50F>T<10F 100F>T<150F
Month
Ave Max Min Ave Max Min Ave Max Min Ave Max Min Ave Max Min
Nov. 0.8 7.7 0.0 15.3 24.0 8.3 7.0 11.8 0.0 0.9 1.7 0.0 0.0 0.0 0.0
Jan. 0.0 0.0 0.0 13.8 15.8 12.0 9.7 11.0 8.2 0.5 1.3 0.0 0.0 0.0 0.0
Season 0.6 7.7 0.0 14.9 24.0 8.3 7.7 11.8 0.0 0.9 2.0 0.0 0.0 0.0 0.0
Time in hours


Slab

10"


Load
12 kips
15 kips
18 kips
12 kips
15 kips
18 kips
12 kips
15 kips
18 kips


I


I


I


)










Table 5-5 Extreme values for temperature differential and temperature in the AC Layer.


Temperature Differential
OF (OC)


Temperature in AC Layer
OF (oC)


Min
78.1 (25.6)
50.7 (10.37)
78.5 (23.6)

81.1 (27.3)
78.8 (26.0)
78.3 (25.7)

48 (8.9)
53.6(12.0)
56.7(13.7)


Max
97.7 (36.5)
86.2(30.1)
89.2(31.8)

93.0(33.9)
99.9 (37.7)
105.6 (40.9)

71.6(22.0)
72.9 (22.7)
69.3 (20.7)


Table 5-6 Measured static and dynamic strains for gages 1, 2 and 5 in the 4-inch slab in Phase I-
a caused by 9 and 12-kip loads.
Measured Strain, 10-6 in/in


Gage 1
Static Dynamic
22 24
22 22
23 28
30 36
39 32
35 32
36 37
43 38


Gage 2
Static Dynamic
30 25
20 22
35 27
30 29
50 35
47 31
47 34
53 37
57 38
50 34
53 42
54 41
63 47


Gage 5
Static Dynamic
21 14
18 15
25 15
24 16
22 16
24 15
29 18
31 18
34 17
28 19
32 24
34 23
35 28


Phase


Slab


Min
-4.6 (-2.5)
-5.7 (-3.2)
-8.4 (-4.7)

-3.1 (-1.7)
-5.6 (-3.1)
-5.2 (-2.9)

-6.9 (-3.8)
-7.8 (-4.3)
-9.3 (-5.2)


Max
15.1 (8.4)
9.7(5.4)
24.7(13.7)

15.5(8.6)
26.8(14.9)
29.9 (16.6)

15.0(8.3)
14.0(7.8)
12.2(6.8)


Date

7/8/2004
7/9/2004
7/10/2004
7/12/2004
8/4/2004
9/23/2004
9/24/2004
10/1/2004
10/2/2004
10/4/2004
10/5/2004
10/6/2004
10/7/2004


Load

9
9
9
12
12
12
12
12
12
12
15
15
15











Table 5-7 Measured static and dynamic strains for gages 2 and 4 in the 5-inch slab in Phase I-a
caused by 9 and 12-kip loads.


Dynamic Strain (x10-)
Gage No 2 Gage No 4
21 -6 -4
21 -5 -3
21 -5 -4
20 -6 -4
17 -5 -4
18 -7 -4.5
21 -6 -4
20 -8 -5
20 -7 -7
20 -6 -7
18 -5 -6
20 -7 -7
20 -7 -7
21 -7 -6
21 -8 -6
23 -8 -7


Static Strain (x 10-)
Gage No 2 Gage No 4
21
20
17
25


30
27
22
35
33
23.5
31
30
32
34
31


Table 5-8 Measured static and dynamic strains for gages 2 and 4 in the 5-inch slab in Phase I-a


caused by 15- and


Date


Load


18-kip loads
Dynamic 8 mph


Gage No 2


12/28/2004
12/29/2004
12/30/2004
1/3/2005
1/4/2005
1/5/2005
1/6/2005
1/7/2005
1/9/2005
1/10/2005
1/11/2005
1/13/2005
1/14/2005


Gage
No 4
-5
-6
-6
-7
-9
-9
-9
-9
-11
-9
-10
-9
-9


Static


Gage
No 2
26
37
36
29
43
39
40
46
45
38
41
50
51


Gage
No 4


Dynamic-1 mph
Gage
Gage No 2 Gage
No 4


23
25
28
-2 32
33
-4 35
-4 36
-4 37
-7 40
-3 38
-4 39
-6 40
-6 39


Date


Load


11/1/2004
11/2/2004
11/3/2004
11/4/2004
11/6/2004
11/7/2004
11/8/2004
11/9/2004
11/10/2004
11/15/2004
11/16/2004
11/17/2004
11/18/2004
11/19/2004
11/22/2004
11/23/2004










Table 5-9 Verified depths of strain gages in Phase I-b.

Intended Position Location 1 (mid-edge) Location 2 (corner)
Intended Position 6 4
6" 5" 4" 6" 5" 4"


Top Gage
Bottom Gage


1" from top
0.5" from bottom


Table 5-10 Verified depths of strain gages in Phase II.
Intended Position Location 1 (mid-edge)
Intended Position 10
6" 8" 10"


Top Gage
Bottom Gage


1" (from top)
1" (from bottom)


0.7"
5.0"


0.8"
6.65"


1.0'


Location 2 (corner)
6" 8" 10"
" 0.87" 1.2"
7.0" 9.0"


Table 5-11 Strain Ratios to evaluate the degree of bond in the interface
Thickness Mid-edge Corner
Phase In R1 = et/ec R2 = eAc/et R = et/ec R2 = eAc/et
4 0.75
I-a 5 -
6 0.89

4
I-b 5 0.66 0.66
6 0.96 0.83
6 0.81 0.83 0.83 0.65
II 8 0.88 0.69 1.00 0.91
10 0.99 0.64 1.13 0.60


1.25
5.5"


- 0.75"
- 3.2"


N/A
5.3"


N/A
4.3"


N/A
3.3"

































Figure 5-1 Corner cracks on 4-inch slabs in Phase I-a after 21-kip wheel loads.


Figure 5-2 Shrinkage cracks on a 4-inch test slab in Phase I-a.






























Figure 5-3 Shrinkage cracks on a 5-inch test slab in Phase I-a.


Figure 5-4 Shrinkage cracks on a 6-inch concrete slab in Phase I-a.
















Terrperatrwe Diffential- Phase I-a 4"-6'x6' slab
Jul; B -Jul, 31 2004
16







-2: -- --- -- :--- ------ -----; r---- --

-- - - -






Ti me
F igure 5-5Tm raturedt ial varia----tion ----- in --the--4-inch slab in Phase-I-a.





----Ck t r. orme r



Figure 5-5 Temperature differential variation in the 4-inch slab in Phase I-a.


Temperature Differentia Phase I-a 5" 6'x6 stab
S rmber 12004 -Janusar; 13 2005
12


-6 ---- ----------------------- ------ ---
-i-




-------------------------------



Ti rre

C enter Corner



Figure 5-6 Temperature differential variation in the 5-inch slab in Phase I-a.






















126













Temperature Differential Phase I-a -- 6'x6' slab
May 20"1 May 251'
30
25
S20
= 15-4

10



-10
-15
Time (min)

Center Corer


Figure 5-7 Temperature differential variation in the 6-inch slab in Phase I-a.


Temperature Differentia I P hase I-b -6" -4'x4' slab
June 17t1 June 22nd

40

40 -----;-- ----- ----- I----- I------ -----

S20 ------- ------ ---- -- -- -

Figure 0 5-8 Te ------mperature differential variation in the 6 -inch slabs in Phase I-b.











127
|- o "oi o" o y o o, L o s o, o, o,
-10P 0: (P a J

Time

-- Center -Comer


Figure 5-8 Temperature differential variation in the 6-inch slabs in Phase I-b.









127











Temperature Differential Phase I-b -5" 4'x4' slab
June 13th- Jun 17th, 2005


30

20 -1----:-------------.- u



i CD CVi



Time

-- Center- Comer


Figure 5-9 Temperature differential variation in the 5-inch slabs in Phase I-b.


Temperature D rferenual Pnase I-b 4" -4'x 4' slab
June 901 June 1211, 2005


13 -------- --------- -- --- ---j -
.- --------- ------ -- ------- --------,-------- --------


77 1 -L-------- ---- ---- I - -- -- --- -
5- ------------------ --- __J -- ------- ------------- ----'---- --------- L
S 1 iI



S1 -L ------------------------------------ ------ -------`-----------I
7 -- ----- r- --- r------.--- -- ----

S. --. .--- ---- ------------------


Time

--Center .Correr


Figure 5-10 Temperature differential variation in the 4-inch slabs in Phase I-b.









128


r4














Temperature Differential Phase II -10" -6'x6' slab
Dec I- 161, 2005

15








: I ::, .. : '
-10 ---- ---- --------- -------------------
L 7.









-1 5

Time

-- -- TC 1 TC 2 -----TC 3


Figure 5-11 Temperature differential variation in the 10-inch slab in Phase II.


Temperature D differential Phase II '8 6'x6' slab
20 Jan 10- Jan 28. 20o16

15 ---------------------------------------------------------------------------------

-10 -- -------- ------------- ---- --------- -----








-10

















129
-1
~lnn-

I- IC q C .. s














Temperature D differential Ph ase II- 6" 6'G' slab
arn 31- Feb 15, 2006



I I I I I I I-

Io . f .. r .- - -- -





0-o

Time

-TC 1 TC 2 -- -TC3


Figure 5-13 Temperature differential variation in the 6-inch slab in Phase II.


Temperature on the surface of the AC layer Phase a 4" 6'x6' slab
Jul; B' -Ju l 31, 2S04




*2S. ----------- --------------------- ---- -- -- --- ----- -----------------------














Figure 5-14 Temperature variation on the surface of the asphalt layer for the 4" slab in Phase
I-a.



















130
130













Temperature onthe surface of the AC layer Phase 1-a 5"- 6'x6' slab
Nov m ber 1 20 04 -Janua ry 13h 20 05
35 -

26D -- ---









Time

-- Center -Corner


Figure 5-15 Temperature variation on the surface of the asphalt layer for the 5" slab in Phase
I-a.


Temperature in the surface of the AC layer Phase I-a 6"- 6'x6'
slab
May 20th May 25th, 2005
35
30 -


20


1-.



0 1000 2000 300C 4000 6000 6COO 7000 8000
Time (m in)


Center - Corner


Figure 5-16 Temperature variation on the surface of the asphalt layer for the 6" slab in Phase
I-a.














131












Temperature in the surface of AC Phase I-b 6" 4'x4' slab
June 17th June 22d

50

5 40 --------------------------f- -----------




S0 -------------.--------------. -------. -- ---- ............ .... .......
I-
0
oT c5T & ox ox ox o c oI
CN V4 fl 0 0 0 04 l < 0C 04 0c 0f 0a 04

Time

-- Center - Corner


Figure 5-17 Temperature on the surface of the AC layer for the 6-inch slab in Phase I-b.


Temperature in the surface of AC Phase I-l 5" 4'x4' slab
June 13h -Jun 17h, 2005




20 ---0 ..
40
Lo 20 -*-------- ----------- ----- --------- -----* I- V--- --------------- *- -- '------ ------------------ .-= -
E 10 --- ------ ----- ---

vs 20 : : : :



H0




Time


-Certer - Comer
,-.'n I':--: b-.; ,-ir 55 E 4= 6 ,<. <


Time






Figure 5-18 Temperature on the surface of the AC layer for the 5" slab in Phase I-b.








132

















Temperature on the surface of the AC Phase I-b 4" 4'x 4' Slab

S- e 9th June 1h, 2005


< CD
M< <0- fi


Time


< 'C g< <
O] iM :


-- Center ---- Corner




Figure 5-19 Temperature on the surface of the AC layer for the 4" slab in Phase I-b.


Temperatu re in the AC Layer Phase II -10" -6'x6' slab
Dec IIt -16", 2005


25

20


S15

10
0C
E
i5

0


0- fl
a i


0 0aa
a 8 05
a ~ ~ ~


Time

- TC1 TC2 -----TC 3


Figure 5-20 Temperature on the surface of the AC layer in the 10-inch slab in Phase II.


40
6a 35
30
3 25
20
15
a.
E 10
5
0


N


----------- -------- --------- ----- -
--------------- ---- ---'---' -4 . .- ---------- ----....




_ -_ .- -_ -_ --- - _- -_ _ _ _- --_ -. --__ --- --__ _- -__ -
_i _ _ _,"__ _,__ ,,__ _ _,__ i _.__ _-__ _ _. _


--------------------_ ----------_----------_ _-
- - ~ - - i - -


-


------__ _- ---- -- --- -.- -- ;-- -- _- ; ----
------ ., .... ------------ -- -- -

i i i i-- -- -- -- +- -- -


- - - -.... .. .... -. . . --- --. .. -L-- - - - - - - - - -
- - - - I - - - - L - - - -


lwr













Temperature in the AC Layer -Plhase II 8" O'x' slab
25 Jan 10' Jan 28At 2006



5 1- ----------------N~----






10 *-.---------------------------------------.------T------T--C--3-------------------------
e 5

0
8 = = =

F1 a a s U ) s U ) a a B B B s s
N rJ N N 4 M M n g N N f f M M
Time

-I--- TC 1 TC 2 ---TC 3


Figure 5-21 Temperature on the surface of the AC layer in the 8-inch slab in Phase II.


Temperature in the surface of AC layer- Phase II -6" 6'x6 slab
Jan 31 Feb 15 ,2006
25
20 -]------ ----,---- --,-------------,----------------,-------
20



1 6' K --_ _ ^ - -_ -_ - _-- _ _ _ _
S----------i-----i-----i-----i- ----------------




S- r-- r-- r-- r-- r-- r-- r-- r-- r-- r-- r-- ---
E






Time


TC 1 TC2 TC 3



Figure 5-22 Temperature on the surface of the AC layer in the 6-inch slab in Phase II.

















134














Temperature Distribution along the Depth of the Concrete Slab
4" Slab Phase I-a


Temperature (C)
25 27 29 31 33 35


37 39 41 43


-*--Center (+DT) --Corner (+DT) Center (-DT) Corner (-DT)


Figure 5-23 Temperature distribution along the depth of the 4 in concrete slab in Phase I-a at
maximum positive temperature differential






Temperature Distribution along the Depth of the Concrete Slab
5" Slab Phasa I-a


Temperature (C)
15 17 19 21 23 25 27 29 31 33


-- Center (+DT) Corner (+DT) Center (-DT) 0 Corner (-DT)


Figure 5-24 Temperature distribution along the depth of the 5in concrete slab in Phase I-a at
maximum positive temperature differential






135


*-4 .


-
I
I l ^ B
I s 1
II 1




Si_
I I "1 ^












Temperature Distribution along the Depth of the Concrete Slab
6" Slab Phase I-a


Temperature (C)


- Center (+DT) Corner (+DT) + Center (-DT) G Corner (-DT)


Figure 5-25 Temperature distribution along the depth of the 6in concrete slab in Phase I-a at
maximum positive temperature differential


Temperature Distribution along the Depth of the Concrete Slab
4" Slab Phase 1-b


25 27


29 31


Temperature (C)
33


35 37 39
1 1


--- Center (+DT) Corner (+DT) --* Center (-DT) -. Corner (-DT)


Figure 5-26 Temperature distribution along the depth of the 4in concrete slab in Phase I-b at
maximum temperature differentials


-- ------------------- -------- -------- -
t*,1


It











Temperature Distribution along the depth of the Concrete Slab
5" slab Phase I-b


Temperature (C)
35 40


-*- Center (+DT) Corner (+DT) Center (-DT) 0 Corner (-DT)


Figure 5-27 Temperature distribution along the depth of the 5in concrete slab in Phase I-b at
maximum temperature differentials


Temperature Distribution along the Depth of the Concrete Slab
6" slab Phase I-b
Temperature (C)
25 30 35 40 45 50


--Center (+DT) --Corner (+DT) ** Center (-DT) Corner (-DT)


Figure 5-28 Temperature distribution along the depth of the 6in concrete slab in Phase I-b at
maximum temperature differentials










137


-t----------- T-- --)-- -^-- -------- --------

-<- --- -- --- -^ ------------- --------
I+ I




------------------
- ^ -- -- ----- -1----------^-------












Temperature Distribution along the Depth
6" Slab- Phase II


Temperature (C)
7 9 11 13 15




i- i i i-i
J 'i i i
\' ', i i
- 1- --I- I / I -
I I I
'- '*- -/ <


of the Concrete Slab





17 19 21


S-*-Center (+DT) Corner (+DT) -* Center (-DT) Corner (-DT)



Figure 5-29 Temperature distribution along the depth of the 6in concrete slab in Phase II at
maximum temperature differentials


Temperature Distribution along the Depth of the Concrete Slab
8" slab Phase II


Temperature (C)
10 15 20 25


0
1
2
S3
S4
. 5



8
Q


-- Center (+DT) Corner (+DT) Center (-DT) - Corner (-DT)



Figure 5-30 Temperature distribution along the depth of the 8in concrete slab in Phase II at
maximum temperature differentials





138


5
0

1

2
S2-

3
C4


5

6

7 --


* --
S


i B
















Temperature Distribution along the Depth of the Concrete Slab
10" Slab- Phase II


Temperature (C)

15


---- *Center (+DT) Corner (+DT) -Center (-DT) Corner (-DT)



Figure 5-31 Temperature distribution along the depth of the 10in concrete slab in Phase II at
maximum temperature differentials


e Gauge No 1- static load
-.o... Gauge No 1 dynamic load


Q.


0a


C -


Time, days

Figure 5-32 Comparison of dynamic and static strain for gage 1 in the 4-inch slab in Phase I-a.






139


**


- -- -
4 -

-1


q

:4













-4 G auge No 2 -static load
-0 Gauge No 2 dynamic Load
-w Gauge No 5 -static load
-7 Gauge No 5 -dynamic load


70

00-


50

0 -
oO -

30


o.

S'V .


o -
_V


W


'V. Co


G a. -ge No 2 Static Load


Gauge- o 2 Dynamic Load


G a g.e -o 5 -Static Load


V
. Gauge ..o
.7'


5 Dynamic Load


-7 "7... .


6 8 10 12 14
Tim e, days


Figure 5-33 Comparison between static and dynamic strain for Gages 2 and
slab in Phase I-a.


5 in the 4-inch


60


50 --


40 A Dynaric


1rrph G2 corrpression
x
x


x x

X X A iAA


x. "A .
x Ax *A
1 D x A *

20
1140


0 .

10111111 0 15 50. 25m 30-
-10 7 -


-20

No of days tested




Figure 5-34 Measured dynamic and static strains at gage 2 in the 5-inch slab in Phase I-a














140


* .0.


+ Dynanic G2

SDynanic G2-comrpression

x Static G2

A Dynaric 1rrph G2


"x


, A


x ,















2



-2 1 510 15 20 25 30 3
A
-4
S A

A- A A A A A


10 AI DyDamic G4
-1 -8

SStatic G4 L
-12 Dynamic 1mph G4

-14

No of days tested



Figure 5-35 Measured dynamic and static strains at gage 4 in the 5-inch slab in Phase I-a





Strain Data collected during HVS loading
Original data set Gage at the top


5 10 15 20 25
Time (sec)

Strain Data collected during HVS loading
Original data set Gage at the bottom


5 10 15 20 25
Time (sec)


Figure 5-36 Measured strains at two different depths at the mid-edge of the 6-inch slab in
Phase I-a.





141












Strain Data collected during HVS loading
Zeroed data set Gage at the top


Time (sec)


Strain Data collected during HVS loading
Zeroed data set Gage at the bottom


Time (sec)


Figure 5-37 Zeroed strains profile at two different depths at the mid-edge of the 6-inch slab in
Phase I-a.


< 60 sec


Time


Strain in the composite pavement as a function of time.


Figure 5-38





























E -
C

Uo


Effect of Temperature Differential on the Peak Strain
Phase 1-a 6" slab 12 kips


Temperature Differential (F)


------ ------ 20*--- -20-,------u----



-10-
*- --I -2 -' -


0 0 5 10 1


0
--------------------0-----------------------------

------_------ --------- -20 --------- ------------ ------------
A i i i A
i AiAi


STop (Mid-edge) Bottom (Mid-edge) -- Bottom (Corner)



Figure 5-39 Effect of temperature differential on the peak strain for the 6" slab in Phase I-a.


Effect of the Temperature Differential on the Peak Strain
Phase I-b 4in slab 12 kips


-5 0 5 10 15 20

Temperature Differential (F)

Mid-Edge (Bottom) A Mid-Edge (AC Layer) Corner (Bottom)



Figure 5-40 Effect of temperature differential on the peak strain for the 4" slab in Phase I-b.


A A






* I
-L 50 -i-- - A A









- 20 ------------- -------------


to ---------------------- -- --------------------------
1-0













Effect of the Temperature Differential on the Peak Strain
Phase I-b 5in slab -12 kips

Temperature Differential (F)


i -4 --- --- -


- - 2- - - -

-0 -5 5 10 15 2025



A 20 -
-i a I I :
-,-----------4- ----- -30 ------------



Mid-Edge (Top) Mid-Edge (Bottom) A Mid-Edge (AC Layer) Corner (Bottom) X Corner (AC Layer)



Figure 5-41 Effect of temperature differential on the peak strain for the 5" slab in Phase I-b.


Effect of Temperature Differential on the peak Strain
Phase I-b 6in Slab 12kips


Temperatrure Differential (F)

-40

- ----- 30--- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -
3* **
,
+ *
-- -- --* -----20 -- -- -- -- -- -- -- -- -- -*- -- -- -- -- -- -- -- -- -- -- -

-- -- -- -- -- -- -- 10--- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- ---




-20
Sto




I---------------4--------------------------------------
40

Mid-Edge (Top) m Mid-Edge (Bottom) A Mid-Edge (AC Layer)



Figure 5-42 Effect of the temperature differential on the peak strain for the 6" slab in Phase I-b













Effect of the Temperature Differential on the Peak Strain
Phase II 6in slab- 12 kips


Temperature Differential (F)
|-----------------------------49--3-----------
S 9, *
00 -3I0 I IoI

3 (-20

i i 0 0 0 i
-------------- ----------------2G-------------------------------------

-------g-------i- -- g----t--- G-------------g--------------- ----!- -


-- -----------^-------- A AG----A- & -- -----------A ---- -----------A ----A----^
IA A-AA 30 4 A A A
I I I I I A


* Mid-edge-Bottom
E Corner-Bottom


A Mid-edge- Asphalt Surface
A Corner-Asphalt Surface


Figure 5-43 Effect of the temperature differential on the peak strain for the 6" slab in Phase II






Effect of the Temperature Differential in the Peak Strain
Phase II 10in slab -12 kips



20
|-----------------------------9-------------------------






0
S0 -6 -4 -2 2 4 6


-- T - -- - ^ -- -- -- -- -- -- - -- -- -- --- -- -
I-I I- -- -I-- ------- 2---- ------------------


Temperature Differential (F)


Mid-edge-Top -- Mid-edge-Bottom A Mid-edge-Asphalt Surface




Figure 5-44 Effect of the temperature differential on the peak strain for the 8" slab in Phase II












145
E I
I C
-6 -6 -4 2 4 I6
I I I I -tI-

-I I I- 1
ii I I I -20
Temperatur Diferetia (F


* Mid-edge-Top
o Corner-Top













Effect of Temperature Differential on the Peak Strain
Phase II 8in Slab 12 kips


Temperature Differential (F)



-39-1---------------

------- -------F------------ 1)--------4-------- ------4-----------------
40
-4 -2 4 -

------ ------ ------ --------- ---- ------ ------ -- ------


__ -
a A | i -
A I I I II AA I


* Mid-edge-Top
o Corner-Top


SMid-edge-Bottom
E Corner-Bottom


A Mid-edge-Asphalt Surface
A Corner-Asphalt Surface


Figure 5-45 Effect of the temperature differential on the peak strain for the 8" slab in Phase II







Effect of AC Temperature on the Peak Strain
Phase 1-a 6" slab 12 kips


Temperature (C)
-,n


-20 A


20 A


--- Top (Mid-edge) -+ Bottom (Mid-edge) -A- Bottom (Corner)

Figure 5-46. Effect of AC temperature on the peak strain for the 6" slab in Phase I-a.


S ----- *-- --------- ------* -
- - - - - - - - -
*
7 28 29 30 31 32







A
-


-10
E 2



O 10















Effect of the AC Temperature on the Peak Strain
Phase I-b 4in slab 12 kips


28 28.5 29 29.5 30 30.5 31

AC Temperature (C)

SMid-Edge (Bottom) A Mid-Edge (AC Layer) Corner (Bottom)



Figure 5-47 Effect of AC temperature on the peak strain for the 4" slab in Phase I-b.


Effect of the AC Temperature on the Peak Strain
Phase I-b 5in slab -12 kips


AC Temperature (C)
-50

-40 -*- - -

-30

-20

E -10
25 27 29 31 33 35 37
0

10

20 A A

30 ------- -- -

40

Mid-Edge (Top) Mid-Edge (Bottom) A Mid-Edge (AC Layer)



Figure 5-48 Effect of AC temperature on the peak strain for the 5" slab in Phase I-b.


A











I I


~

m B
i U

i


31.5














Effect of AC Temperature on the peak Strain
Phase I-b 6in Slab 12kips


AC Temperatrure (C)

-40

-30

-20 ------------ --------------------- --------

1 --------- ------------------------------------------------ ---------
-20



S25 27 29 31 33 35 37 39
.E 0

S10 -------
r -------- T -------- -------I ----I --------I-------




20 -------


30 ------

40

Mid-edge (Top) m Mid-edge (Bottom) A Mid-edge (AC Layer) e Corner (Bottom)



Figure 5-49 Effect of AC temperature on the peak strain for the 6" slab in Phase I-b.


-40

-30

-20

-10
E




20

30
An


Effect of the AC Temperature on the Peak Strain
Phase II 6in slab- 12 kips


AC Temperature (C)

4* -
----------- -_ -- -_ _



114 10 10 20
D---------- ---------- ---------- -_________-- ___ ___ _---___ ___- _
I I I 1





----------- ---- ----- -| A------ ------- ----- ----------

-- -- ------- A ---- A- --- -----


Mid-edge-Top Mid-edge-Bottom A Mid-edge- Asphalt Surface
o Corner-Top 0 Corner-Bottom A Corner-Asphal Surface



Figure 5-50 Effect of AC temperature on the peak strain for the 6" slab in Phase II.













Effect of AC Temperature on the Peak Strain
Phase II 8in Slab 12 kips


AC Temperature


* Mid-edge-Top
o Corner-Top


SMid-edge-Bottom
E Corner-Bottom


A Mid-edge-Asphalt Surface
A Corner-Asphalt Surface


Figure 5-51 Effect of AC temperature on the peak strain for the 8" slab in Phase II.






Effect of the AC Temperature on the Peak Strain
Phase II 10in slab -12 kips


AC Temperature (C)


Mid-edge-Top -- Mid-edge-Bottom -- Mid-edge-Asphalt Surface



Figure 5-52 Effect of AC temperature on the peak strain for the 10" slab in Phase II.







149


-20 -


-,---- --
*
-------- ---------- --

----------- ---------- ----------- ----------- ---------- ---------_-
0 1? 14 16 18 20





-------- ------- -- --- ----


20 -


0 12 14 16 18 20 22 24 26 2





A A A A
T- ---_--
I I II I A
I I I












Strain v/s Wheel Load Phase II
Gages at mid-edge


-50
-40
-30
-20
-10
0
10
20
30


Load (kips)


-- 6in-top 8in-top -A 1Oin-top
6in-bottom 8in-bottom A 1Oin-bottom


Figure 5-53 Relationship between strain and load in Phase II. Gages at the Mid-Edge.





Strain v/s Wheel Load Phase II
Gages at the corner
Load (kips)


-40 -


S-20

c 0

20

40


-- 6in-Top 8in-Top -A- 1in-Top
6in-Bottom 8in-Bottom A lin-Bottom


Figure 5-54 Relationship between strain and load in Phase II. Gages at the corer.


10 12 1 16 1 -----
- -----
1 0 - 1 2 - 14 - 1 f - - 2

- - - - r - - - -
6: : : I -----
-- - -m
I I I7


8 14 16 18 2



- - - - -
A, ---
- -I













Strain v/s Wheel Load Phase II
Gages in AC Layer


45
40
35
S30
25
20
S 15
10
5
0


10 11 12 13 14 15 16 17 18
Load (kip
-*-6in-Edge -- 8in-Edge --10in-Edge
-* in-Corner 8in-Corner -A 1in-Corner


Figure 5-55 Relationship between strain and load in Phase II. Gages in the AC layer.


19
>s)


Variation of Strain (12 kips)


30

20


r 0 ....
S-105000 10000 15000 20000250

-20 *I
A & A A A A
-30

# of Passes


Gaugel Gauge2 ---Gauge3 s Gauge4



Figure 5-56 Variation of peak strains during the HVS test in the 6-inch slab in Phase I-a.


---- ---- -- -
I I I












Variation of Maximum Strain (9 kips)


-100
-80
-60
-40
-20
0
20
40
60
80


*+Top Eottom


Figure 5-57 Variation of maximum strains during the HVS test in the 4-inch slab in Phase I-a.


Strain v/s # of passes -6"-4'x4' slab -Phase I-b
Location 1


-4n1 ---1 __ *. -- - - - # -- -- W --- 2k .. .. ---- +.. -
40 ---- ------ ----- ----
.20 --- -- ---------------------^ ------------- ----------------.--------.----



10l # ofpasses
I n -1 0 - - - -- - - I- - - - 0 0 0- - - - --- - - 2 3- 0

20 ------------------------------------------------------------------------

30 ------------- ---------------- A--------- ^ .. ^ .

50

0Top conc-ete Bot:om concrete e AAsphat Surface


Figure 5-58 Variation of maximum strains in the 6-inch concrete slab and on the surface of the
asphalt layer at Location 1 (mid edge of the slab) during HVS test in Phase I-b.









152


-


- -------------- ----------------- ---------- --- -4-



-- I 3 I
L --- -U----- LLL --- ------ ------ ----- -- - --I -

# of Passes


1- - - - - - -












Strain v/s # of passes 6"- 4'x4' slab Phase I-b
Location 2


10000


15000


20000


25000


* 5.5" Depth A Asphalt Surface


Figure 5-59 Variation of peak strains in the 6-inch concrete slab and on the surface of the
asphalt layer at Location 2 (corner of the slab) during HVS test in Phase I-b.


Strain vis # of passes -5"-4'x4' slab -Phase I-b
Location 1
-60 0 *

-40 ----------------------------------------------------------------------

-30 ----------------------------------------------------------------------
-3 -0 ---------------------------------------------------------------------
2 -20 -------------------------0-------------5---------------------------------
-40
-30
-20
-10 ------- 5000- -- ----4 0 ----1 - 200 --2500g I30 ---3~0 ----4g 00

10 # ofp ases
V 1 0 --------------------------------------------------------- #- of4 nass s.


5n 0 ------- -- ------------------- -- c------- ------------------------------
20 --------------


50 --- -------------------------------------------------------------------
60

Top concrete U Bottom concrete AAsphalt surface


Figure 5-60 Variation of maximum strains in the 5-inch concrete slab and on the surface of the
asphalt layer at Location 1 (mid-edge of the slab) during HVS test in Phase I-b.


5000


30 +


A

AA A passes


AA














Strain vis # of passes-4" -4'x4' Slab Phase I-b
Location 2

-60
-50 ----------- ---- ------------- ------------- ---- ------- --
-40 --------- ----------- ---- ----------- ------ --- -----------
-30 --------. --------- ------------------ ------------------ -------- ----
S-20 ------------------------
10 ------ __Q ._ IQo, .-QOD 1 5 Q. ---.2Q 00- .--25 .---- 3_QQO0, 0- 35.QQ _ DaQo
E 0




OeI -p o I A h II
10 ------------------- ------------------ -------------------I------#--- i_-asse


CA50-------'-








Figure 5-61 Variation of peak strains in the 4-inch concrete slab and on the surface of the
asphalt layer at Location 2 (corner of the slab) during HVS test in Phase I-b.




Strain v/s # of passes 12 kips 6" slab Phase II
(^ 2 0 -, . . . . . .....L - - L - L - - -


















Location 1 (mid-edge, longitudinal)

-50
S0 0 0 00 0 0 0
-40 --- --- --- ---- -- ---- ---- --- ----
5o 3- -- -- ---- --- -- ----- ---- -----




















-30
-20
E -------, ---- 20,00 ---30, ----40,poo ---- ---- 6000
35" Deep Concrepte h Asphalt Surface


Figure 5-61 Variation of peak strains in the 4-inch concrete slab and on the surface of the

















St n v of passes 1
10 .....
20 1 o" ----- 0 g- o,
3 0 -- -- -L------- --------A -- A- --- A- - -- A-- A -- -- -----


x x x x x xx x x X
40 .
50

[ Gage 1 (0.7" Depth) Gage 2 (5" Depth) A AC surface x Bottom 0 Top
,- 0 I-------------------------





50 --------------------------






Figure 5-62 Variation of peak strains in the 6-inch slab at Location 1 during HVS test in Phase
II.












Strain v/s #


Figure 5-63


of passes 12 kips 6" slab Phase II
Location 5 (corner, transversal)


S- -- -------0 -- 00- 0 -- --


S-------Q-----,pOO 30,p00 -----40,900 50,-00 -----60,)00
Sof passes
----------- --------- --------- --------- ---------^----------
S1 *a 1-
L I I





Gage 13 (1" Depth) 0 Gage 14 (5" Depth) A AC surface x Bottom 0 Top

Variation of peak strains in the 6" slab at Location 5 during HVS test in Phase II.


Strain v/s # of passes 12 kips 8" slab Phase II
Location 1 (mid-edge, longitudinal)




o -- -
*

S10,Q00 20000 30000 40,000 50,00 6000 7-0-00 0 80,
# of passes
_ _- i - -- - i - - - -
I I I I u-I I
i< x < X X x X X


Gage 1 (0.8" Depth) m Gage 2 (6.65" Depth) A AC Surface x Bottom o Top


Figure 5-64 Variation of peak strains in the 8-inch slab at Location 1 during HVS test in Phase
II.

















-3

-2

I -1
E
-c-
1

2

3





Figure 5-65
II.


-25
-20
-15
-10
E -5
- 0
2 5
23 10
15
20


Strain v/s # of passes 12 kips 8" slab Phase II
Location 5 (corner, transversal)



0
00 0 0 ; 0- 0 0 0 o
0* *O
0 10,00 0 20,00 30,000 40,000 50,000 60,000 70,000 80,)

0 # of passes
0 -------- ------- ------- ------ -------------- ------ -------
0

S ------ -- --- I

0


* Gage 13 (0.87" Depth) m Gage 14 (7.0" Depth) A AC surface x Bottom o T


Variation in peak strains in the 8-inch slab at Location 5 during HVS test






Strain v/s # of passes 12 Kips 10" slab Phase II
Location 1 (mid-edge, longitudinal)







#o f passes

-- 1 --------
X x X X x


Gage 1 (1" Depth) Gage 2 (9" Depth) A AC Surface x Bottom o Top


Figure 5-66 Variation of peak strains in the 10-inch slab at Location 1 during HVS test in
Phase II.


100










in Phase
in Phase












Strain v/s # of passes 12 Kips 10" slab Phase II
Location 5 (corner, transversal)


-30 0
-30 --------------------------------------------------------
-25 --- 0 1-00
-2 ------- 0 ------- -----------------------0----I ------ -- II-*--o

-2015 ----- -------- --------- --------
-10 ^ ------------ 20 --O----- -^---- G --------- ^ ^ ------- ------ 80F -O -
-10 2 ------------ -o 80o-o
5
c 0
0.-! --------------- - ^ -- - ----------^ -
5 --------------- -- ------------- -------- 4ofpase- -- -
'i 5 # of passes
10 ---- ------- A
) 15 ------- ------ --------
20 X x -- ---
25 x --------------I
30
35

Gage 13 (1.2" Depth) Gage 14 (9" Depth) A AC Surface x Bottom 0 Top


Figure 5-67 Variation of peak strains in the 10-inch Slab at Location 2 during HVS test in
Phase II.










































157









CHAPTER 6
DEVELOPMENT OF A 3-D FINITE ELEMENT MODEL

6.1 Finite Element Program

The multi-purpose finite element program ADINA version 8.2 was used to build the model

for the analysis of whitetopping pavements in this study. The capability of the ADINA program

for 3-D finite element analysis, its versatility in modeling materials behaviors under load and

temperature effects, and its capability in modeling the interface condition between two layers

make this program very appropriate to model composite pavements.

The ADINA program has a very friendly user interface to build the needed models for

specific applications. It has routines to automatically create finite element meshes based on the

boundary definitions and density specifications. The program also has a complete post-process

routine to generate the results both numerically and graphically.

6.2 Six-Slab and Twelve-Slab 3-D Finite Element Models

Figures 6-1 shows a 6-slab 3-D FE model developed for the analysis of the composite

pavement test sections with a joint spacing of 6 feet. The 6-slab model was used to analyze the

test sections in Phase I-a and Phase II. The number of slabs used in this model corresponds to

the actual number of slabs in each test section in these two phases of the study.

Figure 6-2 shows a 12-slab 3-D FE model developed for analysis of the composite

pavement test sections with a joint spacing of 4 feet, which were used in Phase I-b. The number

of slabs used in this model also corresponds to the actual number of slabs in each test section in

Phase I-b of the study.

As shown in Figures 6-1 and 6-2, the sub-grade is modeled by a 100-inch thick layer at the

bottom of the model. This layer is modeled as an assemblage of 3-D solid elements whose









vertical dimension decreases towards the top. The bottom of this layer is modeled as fixed with

no rotation or translation allowed.

On top of the sub-grade layer is a 12-inch thick layer modeling the lime rock base, which

is modeled as bonded to the sub-grade layer. The vertical dimension of the 3-D elements in this

layer also decreases towards the top.

On top of the base layer is a 4.5-inch layer modeling the AC layer, which is modeled as

bonded to the base layer. The AC layer was modeled by three layers of 3-D elements, 1.5 inches

thick each.

The layer at the top models the concrete slabs. The thickness of this layer is variable to

represent the 4, 5, 6, 8 and 10 inches thickness of the concrete slabs used in the test sections in

this study. This layer is modeled by four equal layers of 3-D element. The thicknesses of the

finite elements in this layer are 1, 1.25, 1.5, 2 and 2.5 inches for the 4-, 5-, 6-, 8- and 10- inch

concrete slabs, respectively.

The mesh pattern in the XY plane is the same for all four layers. Figures 6-3 and 6-4 show

the mesh patterns in the XY plane used for the 6-slab model and the 12-slab model, respectively.

Finer meshes are used in the areas of maximum anticipated stresses in the test sections where the

strain gages were placed, in order to obtain more accurate computed strains in these areas for

verification with the measured strains.

6.3 Solid 20-Node Finite Element

All 3-D solid elements were modeled as 20-node elements. Figure 6-5 shows the 3-D solid

element, along with the node configuration. The 20-node element has a hexahedral shape, with

one node at each of its 8 vertices and one node at the middle of each of its 12 edges. Each node

has three degrees of freedom (translations along three perpendicular directions). This type of

node configuration has been shown to give a high level of accuracy in combination with an









acceptable computing time demand. Considering the mesh patterns as shown in Figures 6-3 and

6-4 and the node configuration in the 3-D solid elements as shown in Figure 6-5, the model

response in terms of strains and stresses can be obtained at locations as close as 2" horizontally

and 1.25" vertically from one another.

6.4 Modeling of Concrete Slab Joints

Load transfer across the joints between two adjoining concrete slabs is modeled by trans-

lational springs connecting the slabs at the nodes of the finite elements along the joint. Three

values of spring constants are used to represent the stiffnesses along three different directions.

Figure 6-6 shows the load transfer elements used in the model. Kt, Ks and Kn are stiffness

components in the Z, Y and X direction respectively.

6.5 Modeling of Materials

The concrete, AC, base and subgrade materials are modeled as isotropic and linearly

elastic, and are characterized by their elastic modulus and Poisson's ratio. The AC material is

also modeled as a temperature dependent material with an elastic modulus which varies with

temperature. The contraction and expansion of the concrete due to temperature effects is also

considered in the analysis and characterized by the coefficient of thermal expansion of the

concrete.

6.6 Modeling of Concrete-Asphalt Interface

The concrete-asphalt interface in the composite pavement can be modeled as fully bonded,

partially bonded or un-bonded.

By default, the ADINA program treats adjacent nodes as rigidly connected to one another,

so that the fully bonded condition would be implicitly considered in the model with no special

treatment of the interface between the concrete slab and the AC layer. The composite pavement









test sections in Phase I-a and Phase I-b, which were constructed to have bonded interface

between the concrete and the AC, were modeled to have a fully bonded interface in this fashion.

To model the partial bond condition in the interface, translational spring elements were

used to connect the bottom of the concrete layer with the top of the AC layer at the nodes, with

zero distance between the two layers. Each spring was modeled with three spring constants to

represent stiffness along the three different directions. Figure 6-7 shows the configuration of the

springs used at the interface. Kt and Ks represent the stiffness in the interface plane (X and Y

directions), while Kn represents stiffness perpendicular to that plane (Z direction).

When the concrete-asphalt interface is fully un-bonded, there can not be any tensile load

transfer across the interface, but there can still be transfer of compressive load across the

interface. This is modeled by a special non-linear spring connecting the concrete and the asphalt

layer at the interface. This special spring has an infinite stiffness value when the spring is in

compression; however when the spring is in tension, the stiffness value is zero.

Cores taken from the test sections in Phase II indicated that, though the concrete-asphalt

interface was intended to be un-bonded, there was partial bonding between the concrete and

asphalt layers in these test sections. Analysis of the results of FWD tests on these test sections

also indicated that the computed FWD deflections matched better with the measured deflections

when the concrete-asphalt interface was modeled as partially un-bonded rather than fully un-

bonded. Similarly, the strains calculated using the fully un-bonded model resulted in strain

values 25% higher than the measured strains. Thus, the model for the fully un-bonded condition

was not used in the analysis of the strain data in this study. However the non-linear model was

used to evaluate the performance of the 4" slab in Phase I-a, which presented a corer crack after

the application of 9, 12, 15, 18 and 21 kips wheel load. Figure 6-8 shows the stiffness parameter









for the non-linear springs. As mentioned before this spring has the characteristic of being

infinitely rigid when it is subjected to compression, but it stiffness drop to zero when it is

tension. An intermediate case was also considered in the model with the springs presenting

certain stiffness before they reach certain level of tension force. This threshold can be estimated

based on laboratory test (pull-up and shear test). Pull-up test can help to determine normal

tension at which the interface bond breaks. Shear tests allow to determine the shear strength that

break the bond in the interface. Because these tests usually do not measure deformation, a force-

deflection relationship cannot be accurately determined.

The use of non-linear springs converts the static analysis into a time-dependent analysis.

Time steps have to be specified for the model to evaluate the stress state on the springs at each

time interval. The iteration method and the tolerance criteria have to be specified for the model

to converge. The model strain-energy was used as the tolerance criteria in all analysis involving

time dependant runs. The wheel load pressure was divided in five time increments while the

temperature differential was considered a constant load. Chapter 7 shows more details about the

use of the non-linear model.

6.7 Modeling of Loads and Temperature Effects

To model a moving HVS wheel load on a test section in this study, computations were

done for the responses of the pavement subjected to static loads placed at different consecutive

locations along the wheel path. The calculated response, such as strain at a particular location,

could then be plotted versus time, as the wheel passed over the various locations at the various

times. This computed strain versus time plot could then be compared to the measured strain

versus time plot as obtained from the strain gage measurements during HVS loading.

The applied HVS wheel load was modeled as a uniformly distributed load over a square

area. The contact area was taken to be the wheel load divided by the contact tire pressure. For









example, for a 12-kip wheel load with a tire pressure of 120 psi, the contact area would be 100

in2. Because the model is discrete and the pressure load in the ADINA program has to be

applied on element faces, the final contact area was adjusted considering the geometry of the

finite elements.

The concrete is modeled as a material which contracts or expands according to its tempera-

ture change and coefficient of thermal expansion. An initial temperature is first specified for the

entire concrete layer. The effects of the change in temperature from the initial temperature were

then duly considered in the analysis when the information on the temperature distribution in the

concrete layer was provided in the input. The temperature at the bottom of the concrete slab is

set to be unchanged from the initial temperature. The temperature in the concrete layer was

assumed to vary linearly from the top to the bottom of the slab. From the temperature at the top

and the bottom of the concrete slabs, the temperature of the concrete at all the nodes of the finite

element mesh for the concrete slabs were computed and entered as inputs to the model.











_ Wheel Load
Transversal Joint
Longitudinal Joint

SConcrete


AC Layer
r-- (4.5")
- Base




Subgrade


Figure 6-1 Six-slab 3-D finite element model.


Wheel Load
Transversal Joint
Longitudinal Joint
Concrete


AC Layer
(4.5")

Base



-Subgrade


Figure 6-2 Twelve-slab 3-D finite element model.



























Er


C:
nn

























32 -0" 72" 40

4 elements 10 elements 18 elements 10 elements 4 elements
length 8" length 4" length 4" length 4" length 8"

S Location of main gages Wheel Path

Figure 6-3 Mesh pattern in the XY Plane for the 6-slab model.
-------- -------- -------- -------- __________________________ ________________________________________________ ___________________________ -------- -------- ------- ------- j






-- -- -- - - - - - - - -- -- ^
__ __ __CD It-___c
















-- 32--- --40" --- ----- 72" -' ------------ 40" --- ---32"--







Figure 6-3 Mesh pattern in the XY Plane for the 6-slab model.














































12 elements
length 4"


--- 48 -----

12 elements
length 4"


* Location of main gages


--- 4-8 -----

12 elements
length 4"


-------------S----Q.- -

12 elements
length 4"


Wheel Path


Figure 6-4 Mesh pattern in the XY plane for the 12-slab model.




10,


Figure 6-5 Twenty-node 3D solid element used in the analytical model.






166













Y










/4.










Figure 6-6 Springs to model load transfer at concrete


Ks, Kn


x


Springs in the concrete-asphalt interface to model the partial bond condition.


Figure 6-7

























K sO


Deformation (compression) Deformation + (tension)

Figure 6-8 Non-linear springs to model the fully un-bonded condition in the interface.









CHAPTER 7
MODEL CALIBRATION AND VERIFICATION

7.1 Overview of Model Calibration

In order for the 3-D analytical model to accurately analyze the behavior of WT pavements,

it needs to have the correct properties of the pavement materials and the correct values of spring

stiffness for modeling the behavior of joints and concrete-asphalt interface. The elastic moduli

of the concrete and asphalt materials were initially estimated from the results of laboratory tests

on these materials, as described in Chapter 4. The results of the FWD tests on the composite

pavement test sections were used to estimate the elastic moduli of the other pavement materials

and the joint and interface spring stiffnesses by back-calculation method (of matching the

analytically computed deflections with the measured FWD deflections). FWD test results were

also used to verify the values of elastic moduli of the concrete and asphalt materials. This

process is referred to as "deflection-based calibration" of the model in this research.

The analytical model was further refined by matching the analytically computed strains

with the measured strains in the test sections caused by HVS wheel loads. This process is

referred to as "strain-based calibration" in this research.

Chapter 4 showed that AC Elastic Modulus of the test sections can be estimated based on

temperature using the following expression:

MR = 2321.9 e-00473T (7.1)

Where:
MR= Resilient Modulus of the AC layer (ksi)
T = Temperature ( C)

To effectively calibrate the model and because there are too many variables involved in the

calibration process, the AC elastic modulus first estimated from the Eq.7-1, for the range of

temperatures at the time of the HVS testing. During the HVS testing, the temperature in the AC









layer was measured in several points using thermocouples as indicated in Chapter 3.

Temperature in the asphalt layer was well known during the HVS testing period, since it was

collected every 5 minutes during the entire experiment. Unfortunately the temperature in the AC

layer at the time of the FWD test is unknown. However it can be estimated based on the surface

temperature measured during the test and the relationship between surface temperature and AC

layer temperature estimated from the full scale experiment.

Table 7-1 shows the maximum and minimum values of the AC Resilient Modulus using

Eq.7.1, based on the minimum and maximum temperatures during the HVS test, respectively.

7.2 Deflection-Based Calibration of Model Parameters

7.2.1 Phases I-a and I-b

As mentioned in Chapter 5, FWD tests were run on the test sections to estimate the values

of the elastic moduli for the pavement layers and the stiffness of the springs used to model the

load transfer at the joints and concrete-asphalt interface (for the partially-bonded condition).

To better estimate the elastic moduli of the pavement layers independently of the effects of

joints, FWD tests were run on large slabs which were located at both ends of the test track, and

had a size of 12 ft x 18 ft. Pavement surface deflection basins caused by a 12-kip FWD load

were used to estimate the elastic moduli of the pavement layers by the back-calculation method.

A special analytical model with a 12 ft x 18 ft slab was developed to estimate the

deflections generated by the FWD. Figures 7-1 through 7-5 show examples of the matched

deflection basins from the back-calculation process in which the elastic moduli of the concrete

and asphalt layer from material testing were verified and the elastic moduli of the limerock base

and subgrade materials were estimated. The estimated elastic modulus had a range of 130,000 -

160,000 psi and 28,000 30,000 psi for the lime rock base and the subgrade, respectively. As

mentioned before, the value of the elastic modulus of the AC layer was seen to significantly vary









with temperature. From the results of laboratory tests on the AC samples (as presented in

Chapter 4), the resilient modulus of the AC varied from 325 ksi at 40 OC (representing a very hot

summer condition) to 1,780 ksi at 5 OC (representing winter condition). FWD tests for Phase I-a

and Phase II were run in the summer time, while FWD test for Phase I-b were run in the Fall (at

the end of October of 2005). The elastic modulus of the AC layer, estimated from the back-

calculation method using the FWD data from Phase I-a, was very close to that obtained in the

laboratory (750 ksi). By using the data from the FWD tests run in a colder condition, the back-

calculation process gave a value of 1,100 ksi for the elastic modulus of the AC. The elastic

modulus of the concrete was estimated as 4,350 ksi, which compares very well with the one

obtained by laboratory testing (see Table 4-6, Chapter 4).

After the elastic moduli of the pavement layers had been defined, the joint spring

stiffnesses, which were used to model the load transfer at the joints, were estimated by matching

the analytical with the measured deflection basins caused by a FWD load. In this case the FWD

test (12 kips load) was run in two locations, at the corer and the mid-edge of the test slab.

Deflection basins were recorded along the edge of the slab on both the loaded and the un-loaded

slab, as described in Chapter 4.

Figures 7-6 through 7-9 show examples of the matched deflection basins from the back-

calculation process for the estimation of the joint spring stiffnesses. From the results of FWD

tests run on a 4-inch slab on Lane 6 (Phase I-a) as seen from Figures 7-6 and 7-7, an appropriate

match between the measured and the calculated deflection basin was achieved with a single

vertical stiffness in the order of 100,000 lb/in, and with stiffnesses of zero in the other directions.

Back-calculations performed using FWD data from the tests run in Phase I-b (using 4' x 4'

slabs) showed that the effect of the load transfer can be properly modeled with a spring constant









in the range of 100,000 to 1,000,000 lb/in for the vertical spring, and that it is not necessary to

use the springs in the other two directions, similar to what happened in Phase I-a.

7.2.2 Phase II

The elastic moduli of the pavement layers for the test sections in Phase II were assumed to

be similar to those in Phase I-a and Phase I-b, due to the fact that the same pavement materials

were used.

The deflection basins along the edge of the slabs caused by a 12-kip FWD load applied to

the corner and the mid-edge of the slabs were used to estimate the joint spring stiffnesses, and

the interface spring stiffnesses by back-calculation method. In this phase, many spring constants

had to be calibrated since three springs were used to model the load transfer at the joints and

three springs were used to model the interaction between the concrete slab and the AC layer in

the interface. Also, the springs used in both the transverse and longitudinal joints were con-

sidered separately (as they might have different stiffness values), which gave more flexibility

when matching the FWD deflection basins first and the measured HVS load-induced strains

later.

Figures 7-10 through 7-19 show examples of the matched deflection basins from the back-

calculation process for the estimation of the joint spring stiffnesses and interface spring

stiffnesses for the test sections in Phase II. The analytical deflections matched well with the

measured deflections in all the cases shown in these figures.

During the deflection-based calibration process, it was found that the parameter that had

the greatest effect on the FWD deflection of the composite pavement was the stiffness of the

vertical springs modeling the vertical load transfers at the concrete-asphalt interface. Using a

vertical spring stiffness ranging between 1018 and 1019 lb/in appeared to produce calculated

deflections that reasonably match the measured FWD deflections.









The deflection basin from the case when the sensors were located on the un-loaded slab

was used to calibrate the load transfer at the longitudinal joint. In this test, the FWD load was

applied at the corner and at the mid-edge of the slab and all the sensors measured deflections on

the adjacent slab. Transversal load transfer were calibrated using the FWD test with the drops at

the corer, with one of the sensor (d-12) located in an adjacent slab.

The horizontal springs (in the X and Y directions) modeling the horizontal load transfers at

the interface were found to have no effect on the calculated deflections. It is to be noted that a

value of 500,000 lb/in was used for the interface horizontal springs in the calculation of the FWD

deflection basins, which are presented in Figures 7-10 through 7-19. If a different value of hori-

zontal spring stiffness were used, the calculated FWD deflections would essentially be the same.

However, it is to be pointed out that in the strain-based calibration process, which is presented in

the next section, the horizontal springs at the interface were found to have significant effects on

the load-induced strains, and thus need to be properly calibrated.

Similar to the cases in Phases I-a and I-b, a vertical spring stiffness of 100,000 lb/in

appeared to work well in modeling the joint behavior of the composite pavement test sections in

Phase II. In fact the value of this parameter showed to have little effect in matching the

deflection basin. For some test sections, the value of this parameter ranged form 0 to 1,000,000

kips/in with no significant effect in the load transfer. It seems that the load transfer in bonded

pavements is also affected by bond between the layers in addition to the lateral interaction

between slabs. Because the asphalt layer is continuous at slabs joints, each time a slab is loaded

near the edge, the asphalt under the portion of the slab being loaded transfer the load to the

adjacent asphalt, which is under the unloaded slab. Due to the bond interface the asphalt under

the unloaded slab pushes (pulls) the slab generating the same effect as if there was a lateral load









transfer between adjacent slabs. As a result, the bond condition in the interface may be

responsible of a significant part of the load transfer at joints, collaborating with the interlock

mechanism. The large range of the spring values with little effect in matching the deflection

basin at joints can be explained by this issue, since the full bond in the interface could have

already taken care of the load transfer.

7.3 Strain-Based Calibration of Model Parameters

7.3.1 General Approach

The pavement parameters and spring stiffness values of the 3-D models for analysis of the

composite pavement test sections were further calibrated by matching the computed strains

caused by the HVS wheel loads to the strains measured by the strain gages. In the HVS experi-

ment, the wheel load traveled at a speed of 8 mph (140 in/sec). In the analysis for the load-

induced strains in the test sections using the analytical model, a static load was positioned in

different locations along the wheel path to represent a moving load. The distances between the

load positions were converted to time using this speed. The computed strains at a particular

strain gage location under different load positions were then plotted on a time scale, and

compared with the measured strain versus time plot.

Table 7-2 presents the elastic moduli and Poisson's ratios of the pavement materials of the

test sections as determined from the material characterization in Chapter 4 and from the

deflection-based calibration procedure as presented in the previous section. A range of values

was given for the elastic modulus of the AC since it varied with temperature. The lower and the

upper values correspond to the elastic modulus of the AC at 40 and 5 C, respectively.

In Phase I, after estimating and verifying the material pavement properties and calibrating

the springs located at the joints using FWD tests, there were no more unknown variables in the

model. The elastic moduli of the subgrade, base and concrete seem not to vary during the









experiment since they are basically not temperature-dependent material. For this reason the

strain-based calibration of the model in Phase I was in fact a verification process. When the

model could not appropriately replicate the strain measured in the test sections using the average

Elastic Modulus of the AC shown in Table 7-1, a fine-tune calibration process was undertaken

by slightly varying the AC elastic modulus within the range shown in Table 7-1.

The elastic moduli of the concrete, base and subgrade materials as given in Table 7-2 were

used in the 3-D models in computing the load-induced strains. As mentioned before, the elastic

modulus of the AC layer was chosen to be in the ranges shown in Table 7-1, which represent the

Elastic Modulus values during the HVS test. The stiffness of the joint vertical spring modeling

the load transfer at the joint was fixed at a value of 100,000 lb/in in the analysis.

7.3.2 Phase I-a

Table 7-3 shows the HVS loading periods for the 4-, 5- and 6-inch slabs when they were

loaded with 12 kips in Phase I-a. While the 4- and 6-inch slabs were loaded mainly during hot

weather time, the 5-inch slabs were loaded in the winter time.

Figures 7-20 through 7-26 show the comparison of the computed strains with the measured

strains for various test sections and gage locations. The values of the pavement parameters used

in the analyses are also given in these figures. It can be seen that the elastic modulus of the AC

were in the range and very close to the average shown in Table 7-1. Values of 300,000 and

500,000 psi gave a reasonable match for the 4-inch slab (which was tested in the summer time),

and a value of 950,000 psi worked well for the 5-inch slab (which was tested in the winter time),

representing the two extreme conditions.

It is to be pointed out that in Phase I-a the intended vertical positions of the strain gages

were used in the analysis. As indicated in Chapter 5, some gage locations in Phase I-b and Phase

II were verified by mean of cores taken from the test sections after the HVS loading. In that









process it was found that most of the strain gages had shifted a little bit during concrete

placement. The differences in the strains can be more substantial for thinner slabs, as a small

variation in vertical position in the slab could result in large percentage change in strain.

Unfortunately, cores were not able to be taken to check the positions of the strain gages in the

test slabs in Phase I-a. The possible shifting of the positions of the strain gages could explain

why some of the computed strains did not match very well with the measured strains. For

example, in the case of the 6" slab, two very different values of the AC elastic modulus were

necessary to match the measured strains and one of these values was out of the range shown in

Table 7-1.

Even though the model could replicate the measured peak strain, the strain reversal

phenomenon was not completely replicated by the model. The peak strain reversal at the mid-

edge occurs when the wheel load is located in the slab corer and the peak strain reversal at the

corner happens when the load is at the mid-edge. It seems that the magnitude of the strain

reversal is affected also by the dynamic behavior of the slab under the passing of the wheel load.

The analytical model computes the strains due to static loads applied at different locations, and

does not analyze the possible dynamic effects. For the strain gages located at the mid-edge, the

analytical strain versus time plots were all symmetrical around the point of maximum strain,

while the measured strains plots were not entirely symmetrical. The measured strains were

observed to have higher reversal strains when the wheel load moved away from the gage location

as compared with the strains when the wheel approached the location. In many researches

involving UTW and TWT pavement, it has been hypothesized that shorter slabs have to be used

to minimize the effect of bending in thin slabs. The fact that in all cases a strain reversal was

observed during the passing of the wheel load at the mid-edge indicates that the slab bends and









thus the flexural strain is still important. Considering the calculated strain from the figures, it can

be noticed that the value of the AC elastic modulus has more effect on the tensile strain at the

bottom rather than on the compressive strain at the top.

7.3.3 Phase I-b

The strain-based calibration of the model parameters in Phase I-b was undertaken in the

same way as in Phase I-a, except that in this case, some vertical positions of the gages were

verified through cores taken at the locations of the strain gages after HVS testing. The seven

known gages locations indicated in Table 5-9 in Chapter 5 were first used in the calibration

process. When the exact vertical position of a strain gage is known, then more attempts to

exactly match the strains can be done, since the uncertainties in this variable are now excluded.

In all those cases where the exact vertical position of a gage was known, the gage position was

corrected in the analytical model and the strain in the right place was calculated. In the cases

where the positions of the gages were not verified, the original intended positions of the gages

were used in the analysis.

HVS loading for all the test sections in Phase I-b were run at the beginning of the summer

of 2005 as shown in Table 7-4. Thus, the values of the elastic modulus of the AC layer were

lower than those for the winter time.

Figures 7-27 to 7-32 show the comparison of the computed strains with the measured

strains for various test sections and gage locations, for the cases where the exact gage locations

were known. In each graph, the exact vertical position of the gage is shown at the top of the

figure and the values of the pavement parameters used in the analyses are shown in the boxes.

From the strain-based calibration process considering only the cases where the exact position of

the gage was known, the estimated AC Elastic Moduli were between 450,000 psi and 550,000

psi for the 4" slab, 400,000 psi for the 5" slab, and between 300,000 psi and 400,000 for the 6"









slab. All these values fall within the ranges shown in Table 7-1. It can be seen that the analytical

strains match well with the measured strain at their maximum values by only considering the

right value of the AC elastic moduli, which were first estimated based on temperature.

For the cases where the exact location of the strain gages where unknown, a different

approach was undertaken to match the calculated strain and the measured strain, which also

served as a mean to demonstrate the important effect of the position of the gages. Figures 7-33

and 7-34 show the comparison of the computed strains with the measured strains for two gages

in the 5" slab where the exact gage locations were unknown. These two gages were located at

the same horizontal location in the same slab, with one gage at the top and the other at the

bottom. Because the exact positions of the gages were not known, the figures also include the

cases if the gages had been displaced up by 0.2". These figures show examples of the effects of

the vertical position of the gage on the load-induced strains at the gage. For the case of the gage

at the top (as shown in Figure 7-33), the best match was achieved with an elastic modulus of the

AC layer of 400,000 psi, whereas for the case of the gage at the bottom (as shown in Figure 7-

34), the best match was achieved with a value of 700,000 psi. When a different vertical position

of the gage at the top was used, an exact match in maximum strain was obtained, as shown in

Figure 7-33. However, the consideration of a different vertical position of the gage at the bottom

made the match worse, as shown in Figure 7-34. That means that the gage at the bottom was

probably displaced down rather than up. This demonstrates the importance of exactly knowing

the position of the gages when considering very thin slabs (UTW pavements). In any case, the

range of the AC elastic modulus found in this analysis reasonably agrees with the range shown in

Table 7-1 for this particular test section.









As a summary for this phase, the model could replicate very well the strain measured in the

test sections by only using the appropriate AC elastic modulus depending on the temperature of

the AC layer. The strain reversal was not completely captured by the model since the calculated

strain profile was symmetric with respect to the peak strain when matching the measured strain

at the mid-edge of the slab. The non-symmetrical shape of the strain profile for the gages located

at the slab corners was reasonably replicated by the model but the calculated strain reversal was

still lower than the measured one. Similar to Phase I-a, a large variation of the AC elastic

modulus showed to have a significant effect on the analytical tensile strain compared to its effect

on the analytical compressive strain.

7.3.4 Phase II

The strain-based calibration of the model parameters in Phase II was performed in a similar

manner as in Phases I-a and I-b, except that in this case the springs modeling the partial bonding

condition of the concrete-asphalt interface had to be calibrated. Similar to the case in Phase I-b,

the vertical positions of a few strain gages were verified through cores taken at the strain gages

location after HVS testing. Those locations were previously shown in Table 5-10 in Chapter 5.

In the case where the exact vertical position of a gage was known, the verified gage position was

used in the analytical model to calculate the strain. In the cases where the positions of the gages

were not verified, the original intended positions of the gages were used in the analysis.

HVS loading for all the slabs were run during the winter of 2005-2006, as shown in Table

7-5. Thus, the elastic moduli of the AC layer considered in the analysis were higher than those

estimated for the summer time according with Eq.7.1 and Table 7-1

Figures 7-35 through 7-52 show the comparison of the computed strains with the measured

strains for various test sections and gage locations. The value of the AC elastic modulus was

estimated based on Eq-7.1 and using the range of temperatures during the HVS testing.









According with Table 7-1 this range was between 800 and 1,500 ksi, representing a cold

condition. The verified vertical positions of the gages are shown at the top of the figures and the

model parameters are shown in the boxes. It can be seen that in general, the analytical strains

matched well with the measured strain at their maximum values.

As it happened in the previous phases, the model underestimated the strain reversal in all

analyzed cases and presented a symmetrical shape around the peak strain. It can also be observed

that the model could reasonably replicate the strain in the asphalt layer. In Phase II there is no

strain reversal at the corner since the strain gages at that location were measuring strain in the

direction perpendicular to the traveling wheel load. In that direction the slab deflection has a

single curvature at any moment during the passing of the load.

From the results of the calibration process, it can be observed that the values of the

interface horizontal springs modeling the concrete-asphalt interface were consistently higher at

the mid-edge than at the slab corners. This indicates that there was less interface bond at the slab

corner than at the slab edge. The values of these springs also varied from one test slab to

another, indicating the non-uniformity of the partial bond condition.

The calibrated model parameters for the three test sections in Phase II are presented in the

next section, along with those for the test sections in Phases I-a and I-b.

7.4 Summary of Calibration Results

The best estimated model parameters of the 3-D model for all the test sections in this

study, based on the results of the deflection-based and strain-based calibration, are summarized

in Table 7-6. It is to be pointed out that these model parameters are only applicable to the

conditions at the time of the HVS loading when the strain data were taken. The variation in the

elastic modulus of the AC layer from one test section to another, or within the same test section

was due to the different temperatures at the time of the tests. From the analytical results in all









test sections, the value of the AC elastic modulus was found to have more effect on the tensile

strain at the bottom of the slab compared to its effect on the compressive strain at the top. For the

test sections in Phase II, the stiffness values of the horizontal springs modeling the interface can

be observed to decrease as the slab thickness increases from 6 inches to 10 inches. For each test

section, the range of stiffness values for the joint and interface horizontal springs are given. For

the joint horizontal springs, the higher values generally represent the condition at the slab

corners, while the lower values represent the condition at the edge of the slabs. Conversely, for

the interface horizontal springs, the higher values generally represent the condition at the edge of

the joints, while the lower values represent the condition at the slab covers. In Phase II the

variables that control the strain in the analytical model are, in order of importance, the horizontal

springs at the interface, the horizontal springs at the joints, and the AC Elastic Modulus.

Since the model could satisfactorily replicate the strain at many different locations in the

test sections and considering many different slab conditions (thickness, joint spacing, interface

bond), it is possible to use the model to estimate stresses in both the concrete slab and the AC

layer. The model and the parameters indicated in Table 7-6 will be used in Chapter 8 to estimate

the level of stresses in the test section and the potential performance of the composite pavement.









Table 7-1 Extreme values of the AC resilient modulus based on extreme temperature during the
HVS test, using Eq. 7.1
Temperature (C) MR (ksi)
Thickness
Phase (in) Min Max MR(max) MR(min) Average
4 25.60 36.50 692 413 552
I-a 5 10.27 30.10 1428 559 993
6 23.60 31.80 760 516 638
4 27.30 33.90 638 467 553
I-b 5 26.00 37.70 679 390 534
6 25.70 40.90 688 335 512
6 8.90 22.00 1524 820 1172
II 8 12.00 22.70 1316 793 1054
10 13.70 20.70 1214 872 1043


Table 7-2 Elastic modulus and Poisson's ratio of the pavement materials used in the 3-D finite
element model.
Material Modulus of Elasticity Poisson's Ratio
Material (psi

Sub-grade 30,000 0.35
Base 160,000 0.35
Asphalt 325,000- 1,780,000 0.35
Concrete 4,350,000 0.20


Table 7-3 HVS loading periods for Phase I-a.


Thickness Joint Spacing
4" 6' x 6'
5" 6' x 6'
6" 6' x 6'


Table 7-4 HVS loading periods for Phase I-b.
Thickness Joint Spacing
4" 4' x 4'
5" 4' x 4'
6" 4' x 4'


From
07/11/04
11/01/04
05/23/05




From
06/09/05
06/14/05
06/20/05


To
10/03/04
11/24/04
05/26/05




To
06/12/05
06/17/05
06/22/05









Table 7-5 HVS loading periods for Phase II.
Thickness Joint Spacing From To
6" 6' x 6' 01/30/06 02/15/06
8" 6' x 6' 01/09/06 01/28/06
10" 6' x 6' 11/14/05 12/16/05











Table 7-6 Summary of the best estimated parameters
Phase I-a
4" 5"
Material Concrete 4,350 4,350
Elastic AC Layer 300-500 950-1,000
Moduli Base 160 160
(ksi) Subgrade 30 30


Spring
Constants
(lb/in 106)


Interface X
Interface Y
Interface Z
Trans. Joint X
Trans. Joint Y
Trans. Joint Z
Long. Joint X
Long. Joint Y
Long. Joint Z


of the 3-D model for all test sections.
Phase I-b
6" 4" 5" 6"
4,350 4,350 4,350 4,350
750-1,400 450-550 400-700 300-400
160 130 130 130
30 28 28 28






0.1 0.1 0.1 0.1



0.1 0.1 0.1 0.1


Phase II
6" 8"'
4,200-4,350 4,350
800-1,000 1,000
160 160
30 30
3-3.5 2-3
3-3.5 2-3
5x1012 1013

0-0.1 0
0.1 0.1-1
0.01-1 0.01-10
0 0
0.1 0.1-5


10"
4,350
800-1,000
160
30
1-3
1-3
1013

1-0
0.1
0.01-1
0-0.1
3
















Determination of Elastic Modulus Phase I-a
FWD test run on the center of the 4" 12' x 18' slab, sensors in the longitudinal direction

Distance (in)
0 6 12 18 24 30 36 42 48 54 60
0

1 ------
2

c 3

4 Concrete 4,350,000 psi
S_ .5 Base 160,000 psi
S- r Subgrade 30,000 psi
6 ------ ------- ------- ----------------------
7

--- FWD 3D Model


Figure 7-1 Matching of deflection basin in the longitudinal direction caused by a 12-kip FWD
load applied to the center of a 4" slab in Phase I-a.






Determination of Elastic Modulus Phase I-b
FWD test run on the center of the 6"- 12' x 18' slab, sensors in the longitudinal direction
0 10 20 30 40 50 60 70 80
0
0 ------------------------------------------------
0.5 Cae 1 Distance (in)
0.5 Case 1 -------T---------------------------------
1 Concrete 4,350,000 psir ----
AC 950,000 psi
S1.5 Base 160,000 psi -------- I-
Subgrade 28,00 pst t s -

3 ---------- 1-------- ---------------- AC 1,100,000 psi ----
Q 3.5 --------- Base 180,000 psi
SSubgrade 27,000 psi
4-----------------
4.5
-----FWD- .Case 1 Case 2


Figure 7-2 Matching of deflection basin in the longitudinal direction caused by a 12-kip FWD
load applied to the center of a 6" slab in Phase I-b.















Determination of Elastic modulus Phase I-b
FWD test run on the center of the 6"- 12' x 18' slab, sensors in the transversal direction
10 20 30 40 50 60 70
1 1 1 1 1


Case 1
Concrete 4,350,000 psi
-AC 1,100,000 psi
Base 160,000 psi
Subgrade 27,000 psi


Distance (in)


-Case 2
Concrete 4,350,000 psi
AC 1,100,000 psi
Base 180,000 psi
Subgrade 27,000 psi


-- FWD Case 1 -Case 2 Case3


Figure 7-3 Matching of deflection basin in the transverse direction caused by a 12-kip FWD load
applied to the center of a 6" slab in Phase I-b.


Determination of Elastic Modulus Phase I-b
FWD Test run on the center of the 4"- 12' x 18' slab, sensors in the longitudinal direction
Distance (in)
10 20 30 40 50 60 70


2

2 3

4

5

6


-- FWD A Case 1 *- Case 2


Figure 7-4 Matching of deflection basin in the longitudinal direction caused by a 12-kip FWD
load applied to the center of a 4" slab in Phase I-b.















Determination of Elastic Modulus Phase I-b
FWD test run on the center of the 4" 12' x 18' slab, sensors in the transversal direction

)10 20 30 40 50 60 70Distance(in0


--- Cs1:- --------------------------------------- --__-------
Case 1:
Concrete 4,350,000 psi
AC 600,000 psi
Base 115,000 psi
Subgrade 28,000 psi
Case 2:
Concrete 3,900,000 psi
AC 600,000 psi
Base 125,000 psi
--- ------- -^ - ------- --Subgrade 28,000 psi ---------


0

1

' 2
E




S5

6


- -FWD --Case 1 -A-Case 2


Figure 7-5 Matching of deflection basin in the transverse direction caused by a 12-kip FWD load
applied to the center of a 4" slab in Phase I-b.







Calibration of springs at joints Phase I-a
FWD test run on the 4" 6' x 6' slab, drops at the corner, sensors along the edge of the
loaded slab Distance (in)


-6 -10 -4 2 8 14 20 26 32 38 44 50 56

"-2 --

g --3,

-4 Concrete 4,350,000 psi
i I *AC 700,000 psi
-5 L- -- -- Base 160,000 psi L
S-* Subgrade 30,000 psi
S LI K-spring: 100,000 Iblin

-7

--- FWD 3D Model



Figure 7-6 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the corner of a 4" slab in Phase I-a.













Calibration of springs at joints Phase I-a
FWD test run on the 4" 6' x 6' slab, drops at the corner, sensors along the edge of the
unloaded slab Distance (in)


- --FWD 3D Model


Figure 7-7 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the corner of a 4" slab in Phase I-a.






Calibration of springs at joints Phase I-b
FWD test run on 4" 4' x 4' slab, drops at the corner, sensors along the edge of the loaded slab
-6 0 6 12 18 24 30 36 42 48 54 60 66

case 1: Distance (in)
SConcrete 4,300,000 psi I I
AC Layer 850,000 psi I I
SBase 160,000 psi I t I d
2- _Subgrade 30,000 psi I
E K-joint= 10,000 In 7


- -m ..Case 1 Case 2 *- -Case 3 FWD


Figure 7-8 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the corer of a 4" slab in Phase I-b.















Calibration of Springs at joints Phase I-b
FWD test run on 5" 4' x 4' slab, drops at the corner, sensors along the edge of the loaded slab
Distance (in)
-6 0 6 12 18 24 30 36 42 48 54 60 66


1 -- Case 1:-
Concrete 4,350,000 psi
AC 850,000 psi .
Base 130,000 psi
E Subgrade 28,000 psi II I
3-- K-spring 0,00,0 IbAn --- --------- -,- -
2 ~1I 1Case 2:
4 ------------- --- ---- Concrete 4,350,000 psi
S1 AC 1,100,000 psi
S 5- -------- --- Base 130,000 psi
Subgrade 28,000 psi
K-spring 1,000 Ibln




-"-FWD-1 -*- Case 1 A- Case 2

Figure 7-9 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the corer of a 5" slab in Phase I-b.







Calibration of vertical springs Phase II
FWD test run on 6" slab, drops at the corner, sensors along the edge of the loaded slab
Distance (in)
-6 0 6 12 18 24 30 36 42 48 54 60 66

Case 1:
nterfaceZ 25x 101 IbAn
InterfaceX 500,0001b/n *b
Interface Y 500,000 Ib/n I- -
2 Longitudinal JointZ 100000 IbAn 4
/ Transversal JointZ 100,000 Ibn bAn
s 3----- ---------- ---- .- "' Case 2:
z :ICase 2:
SIji IInterfaceZ 50 x 1018 Ib/in
a ------ - - - InterfaceX 500,000 IbAn -
O - Interface Y 500,000 Ib/in
i -Longitudinal JointZ 100,000 IbAn
Transversal JointZ 100,000 IbAn




[I IIFWD *- Case 1 -A -Case 2



Figure 7-10 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the comer of a 6" slab in Phase II.














Calibration of vertical springs Phase II
FWD test run on the 6" slab, drops at the mid-edge, sensors along the edge of the loaded slab

-12 -6 0 6 12 18 24 30 36 42 48 54
0 I
Case 1: Distance (ir
InterfaceZ 25x 10"b Ibn
- 1 InterfaceX 500,000 IbAn --- -
E- InterfaceY 500,000MIbn "
S _Longitudinal JointZ 1 ,00000 IbAn '" -' "
S-. ITransversal JointZ 1O,0O001bAn "

Interface X 500,O001IbAn
-* Interface Y 500,000 IbAn
."- -- - - Longitudinal JointZ 10,0001bO n -
S -- Transversal JointZ 100,000 IbAn



FWD -*- *Case 1 -A- Case 2

Figure 7-11 Matching of deflection basin along the edge of loaded slab caused by a 12-kip
FWD load applied to the mid-edge of a 6" slab in Phase II.






Calibration of vertical springs Phase II
FWD test run on the 6" slab, drops at the mid-edge, sensors along the edge of the unloaded slab
-12 -6 0 6 12 18 24 30 36 42 48 54


--- FWD Case 1 A Case 2


Figure 7-12 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the mid-edge of a 6" slab in Phase II.














Calibration of vertical springs Phase II
FWD test run on 8" slab, drops at the corner, sensors along the edge of the loaded slab


-6 0 6 12 18 24


30 36 42 48 54 60


--FWD *- Case 1 Case 2

Figure 7-13 Matching of deflection basin along the edge of loaded slab caused by a 12-kip
FWD load applied to the corner of an 8" slab in Phase II.






Calibration of vertical springs Phase II
FWD test run on the 8" slab, drops at the corner, sensors along the edge of the unloaded slab

-6 0 6 12 18 24 30 36 42 48 54 60 66

Case 1: Distance (in)
InterfaceZ 50x in Ox -b
-- InterfaceX 500,000 Ibn i_ .
InterfaceY 500,000IMbn L -' -
E 5 Longitudinal JointZ 1 ,00,0lbAn -- -
Transversal JointZ 100,000o b/n b .
-2 .2- Case2:
2 -- I -. LIe_ -._ InterfaceZ 1 Ox 10"9 IbAn I
S. InterfaceX 500,OlbAn
.. '" - --- -- InterfaceY 500,000 bIM n _I
.. S r Longitudinal JointZ 1DOOMOIbAn
35 Transversal JointZ 100,MO IbAn



--FWD *-Case1 Case2


Figure 7-14 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the corner of an 8" slab in Phase II.


,---,----- i i i Distance (in)
-0-5- Case 1: ---------- 4
InterfaceZ 50x 10" IbAn I b"
1 InterfaceX 500000 IbAn -. ---
Interface Y 5000001bn *
Longitudinal JointZ 100,000 b/in -
2 Transversal Joint Z 100,000 IbAn -
25- ---- 4-- --- Case 2:
Interface Z 10x 10" IbAn
- T InterfaceX 500,000 IbAn
S. '' ------- Interface Y 500,000 bAn
S*Longitudinal JointZ 100,000 IbAn
4 --------------- r Transversal JointZ 100,000 IbAin
--.1C-------------------------------------------------













Calibration of vertical springs Phase II
FWD test run on the 8" slab, drops at the mid-edge, sensors along the edge of the loaded slab

-12 -6 0 6 12 18 24 30 36 42 48 54 60


- FWD -* Case 1 -A Case2


Figure 7-15 Matching of deflection basin along the edge of loaded slab caused by a 12-kip
FWD load applied to the mid-edge of an 8" slab in Phase II.


-12


Calibration of vertical springs Phase II
FWD test run on the 8" slab, drops at the mid-edge, sensors along the edge of the unloaded slab
-6 0 6 12 18 24 30 36 42 48 54


-* FWD -A -3D Model


Figure 7-16 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the mid-edge of an 8" slab in Phase II.


3D Model: Distance in)
-- -5- InterfaceZ 10x 109 IbAn -- T
InterfaceX 500,000 IbAn
InterfaceY 500,000 Ibn n - '
Longitudinal JointZ 10,00,000 Ib .n
Transversal JointZ 00,000 IbAn -

2- 1 --
2 -.-Jl I--~I I
I I I I2- ^ - - -. S ^ ^ ^ -" I I I-- - - -
; 2 I I + I I

I------- a-----------------------------------------------


4














Calibration of vertical springs Phase II
FWD Test run on the 10" slab, drops at the corner, sensors along the edge of the loaded slab

-6 0 6 12 18 24 30 36 42 48 54 60 66

Distance (in)
-5 3D Model:- -
InterfaceZ 10x1c n bA *n
1 InterfaceX 500,00 IbAn -LL -. -- "
E InterfaceY 500M,00l1bn -
Longitudinal JointZ 1D,00 bn _ _
2 Transversal JointZ 100,0D0 Ibn
2

-. 2- --- -- --- -
I -I --- --- ----- I ------------ L---- -I- -I-
3.5


-U- FWD -- 3D Model

Figure 7-17 Matching of deflection basin along the edge of loaded slab caused by a 12-kip
FWD load applied to the corner of a 10" slab in Phase II.








Calibration of vertical springs- Phase II
FWD test run on the 10" slab, drops at the mid-edge, sensors along the edge of the loaded slab


-12 -6


0 6 12 18 24 30 36 42 48 54 60


-- FWD A 3D Model


Figure 7-18 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD
load applied to the mid-edge of a 10" slab in Phase II.


05
Distance (in)
- -r 0 .-5 -- -- ~, ~ r -- --~" ~ ~ ~ ~- ~ ~ ~ ~-~~ ~ ~ --~ ~ ~ -~ ~ ~ -~ ~ ~ ~
3D Model:
InterfaceZ 5 Ox 10A8 Ibn
InterfaceX 500,001bn bA
InterfaceY 500,001bAn "
5 Longitudinal JointZ 100,000 IbAn
Transversal JointZ 100 IbAn b I. *
L --- 2------- -L

2.-5------- ------------------
-- -- *-^' ^ - -- ----- t----- t----- t----- t----- t----- I --
I------- C I-----------------------------------------------------------















Calibration of vertical springs Phase II
FWD test run on the 10" slab, drops at the mid-edge, sensors along the edge of the unloaded slab


0 6 12 18 24 30 36 42 48 54 60


-* FWD test A 3D Model


Figure 7-19 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip
FWD load applied to the mid-edge of a 10" slab in Phase II.








Strain comparison 6" 6' x 6 slab Phase I-a
Location 1 Gage 1 (top), 12 kips load


17

Case 1:
Concrete 4,350,000 psi
AC 750,000 psi
Base 160,000 psi
Subgrade 30,000 psi
K-joint 100,0001bAn


Case 2 (Best Match):
Concrete 4,350,000 psi
AC 1,400,000 psi
Base 160,000 psi
Subgrade 30,000 psi
K-joint 100,000 bAn


Time (sec)


Gagel --3D Model (Case 1) -A 3D Model (Case 2)



Figure 7-20 Strain comparison at Gage 1 in the 6" slab in Phase I-a.


-12 -6


3i .... Distance (in)
3D Model:
~4-O Interface Z 50x 108 IbAn
Interface X 500,000 IbAn
S InterfaceY 500,000 IbAn
Longitudinal JointZ 1000,000 IbAn -
Transversal JointZ 1000,000 bAn i "
15 .. .
1L5-

2 --- -- -
-
i i i i-


-20


-30













Strain comparison 6" 6' x 6 slab Phase I-a
Location 1 Gage 2 (bottom), 12 kips load


Case 1 (best Match): I- Case 2:
Concrete 4,350,000 psi Concrete 4,350,000 psi
AC 750,000 psi AC 1,400,000 psi
Base 160,000 psi Base 160,000 psi
Subgrade 30,000 psi Subgrade 30,000 psi
K-joint 100,000 lbn K-joint 100,000 Ibn


.-/^A^^-^A. ~ ~ ~ ~ ~ ~ ~ X V'Q_ il >J-/ \i


_17 75_ 1' l_ 5 19


- 9 V i5


Time (sec)


Gage2 3D Model (Case 1) -A -3D Model (Case 2)


Figure 7-21 Strain comparison at Gage 2 in the 6" slab in Phase I-a.







Strain comparison 6" 6' x 6 slab Phase I-a
Location 2 Gage 3 (top), 12 kips load


15

10 -
15 T------------------/--------------- ------------------------


0
E _51.5 17 17.5 18 18.5 19 19.5 2

-10
Case 1: Case 2 (Best Match):
-15 Concrete 4,350,000 psi ------------ Concrete 4,350,000 psi
AC 750,000 psi AC 1,400,000 psi
Base 160,000 psi Base 160,000 psi
-20 Subgrade 30,000 psi Subgrade 30,000 psi
K-joint 100,000 Ibn K-joint 100,000 Ibn
-25

-30

Time (sec)

-- Gage3 3D Model (Case 1) 3D Model (Case 2)



Figure 7-22 Strain comparison at Gage 3 in the 6" slab in Phase I-a.













Strain comparison 5" 6' x 6' slab Phase I-a
Location 1 Gage 2 (bottom), 12 kips load


Case 1: Best Match
Concrete: 4,350,000 psi
AC: 950,000 psi
Base: 160,000 psi
Subgrade: 30,000 psi
K-joint: 100,000 psi


Case 2:
Concrete: 4,350,000 psi
AC: 1,100,000 psi
Base: 160,000 psi
Subgrade: 30,000 psi
K-joint: 100,000 psi


12.5 13 1


Time (sec)


-- Gage 2 3D Model (Case 1) A 3D Model (Case 2)

Figure 7-23 Strain comparison at Gage 2 in the 5" slab in Phase I-a.






Strain comparison 4"- 6'x6' Slab Phase I-a
Location 1, Gage 1 (top), 12 kips load


V
1.5


Case 1 (best Match):
Concrete 4,350,000 psi
-AC 300,000 psi
Base 160,000 psi
Subgrade 30,000 psi
-K-joint 10,00000bAn


2


Case 2:
Concrete 4,350,000 psi
AC 500,000 psi
Base 160,000 psi
Subgrade 30,000 psi
K-joint 10,00000bAn


Time (sec)


-3D Model (Case 1)


- Gage 1 A 3D Model (Case 2)


Figure 7-24 Strain comparison at Gage 1 in the 4" slab in Phase I-a.


0 --

C -10-

' -20


-30


~7 ~U


4.=,


01











Strain comparison 4" 6'x6' slab Phase I-a
Location 1, Gage 2 (bottom), 12 kips load


Time (sec)


-3D Model (Case 1)


-Gage 2 A 3D Model (Case 2)


Figure 7-25 Strain comparison at Gage 2 in the 4" slab in Phase I-a.





Strain comparison 4" 6'x6' slab Phase I-a
Location 2, Gage 3 (top), 12 kips load


Time (sec)

-- Gage 3 --3D Model (Case 1) -A 3D Model (Case 2)

Figure 7-26 Strain comparison at Gage 3 in the 4" slab in Phase I-a.

















Strain comparison 6" 4'x4' slab Phase I-b
Location Gage 1 (top, depth 1.25"), 12 kips load


~-. -~


17-2 1-7.4
Case 1:
Concrete 4,350,000psi
AC 700,000psi
Base 130,000 psi
Subgrade 28,000 psi
K-joint 100,0001bAn


17.6 -- 8


t&2 1 4 .6
___ Case 2 (Best Match):
Concrete 4,350,000 psi
-____ AC 400,000 psi
Base 130,000psi
___ Subgrade 28,000psi
K-iant DO,000IbAn


Time (sec)


-- Gage 1 3D Model (Casel) 3D Model (Case2)

Figure 7-27 Strain comparison at Gage 1 in the 6-inch slab in Phase I-b.









Strain comparison 6" 4'x4' slab Phase I-b
Location Gage 2 (bottom, depth 5.5"), 12 kips load


Time (sec)


-- Gage 2 3D Model (Casel) 3D Model (Case2)


Figure 7-28 Strain comparison at Gage 2 in the 6-inch slab in Phase I-b.

















Strain comparison 6" 4'x4' slab Phase I-b
Location Gage 3 (bottom, depth 5.3"), 12 kips load


Case 1:
Concrete 4,350,000psi
AC 700,000 psi
Base 30,000psi
Subgrade 28,000 psi
K-jant 100,00I0bAn


Case 2 (Best Match):
Concrete 4,350,000psi
AC 300,000psi
Base 130,000 psi
Subgrade 28,000psi
K-joint 100,000IbAn


- .


18.5


Time (sec)

-- Gage 3 3D Model (Casel) -* 3D Model (Case2)

Figure 7-29 Strain comparison at Gage 3 in the 6-inch slab in Phase I-b.


Strain comparison 5" 4'x4' slab Phase I-b
Location Gage 1 (top), 12 kips load


Time (sec)


l-- Gage 1 3D Model (Casel, 1") ----- 3D Model (Case2, 1") --*-- 3D Model (Case 2, 0.8")

Figure 7-30 Strain comparison at Gage 1 in the 5-inch slab in Phase I-b.















199


12 12.2 12.4 12.6 2.8 13 13.2 13.4 13.6 13.8 1


-- Case 1 -- --- Case 2 (Best Match) -----
Concrete 4,350,000 psi Concrete 4,350,000 psi
AC 700,000 psi __ I _____ AC 400,000 psi
Base 130,000 psi Base 130,000 psi
Subgrade 28,000 psi Subgrade 28,000 psi
SK-joint 100,000 IbAn ----- ------- K-joint 1000,000 Ibn


1













Strain comparison 5" 4'x4' slab Phase I-b
Location Gage 3 (bottom, depth 4.3"), 12 kips load


Time (sec)

Gauge 3 3D Model (Casel) A 3D Model (Case2)


Figure 7-31 Strain comparison at Gage 3 in the 5-inch slab in Phase I-b







Strain comparison 5" 4'x4' slab Phase I-b
Location Gage 2 (bottom), 12 kips load


Time (sec)


-- Gage 2 3D Model (Case1, 4.5") -... *.... 3D Model (Case2, 4.5") -- --- 3D Model (Case2, 4.3")

Figure 7-32 Strain comparison at Gage 2 in the 5-inch slab in Phase I-b.


Case 1 Case 2: Best Match
Concrete 4,350,000 psi / \ Concrete 4,350,000 psi
AC 700,000 psi - - - - - - - AC 400,000 ps - -
Base 130,000 psi I Base 130,000 psi
Subgrade 28,000 psi -I - -- -------------- Subgrade 28,000 psi -
K-joint 100,000 Ib/n f K-joint 100,000 IbAn



12- 122 2 -.- 128 13_ 13-2 13_4
















Strain comparison 4" 4'x4' slab Phase I-b
Location 1 Gage 2 (bottom, depth 3.2), 12 kips load


Time (sec)


-- Gage 2 3D Model (Casel) A 3D Model (Case2)


Figure 7-33 Strain comparison at Gage 2 in the 4-inch slab in Phase I-b.







Strain comparison 4" 4'x4' slab Phase I-b
Location 2 Gage 3 (bottom, 3.3" depth), 12 kips load


Case 1:
Concrete 4,350,000 psi
AC 700,000 psi
Base 130,000 psi
Subgrade 28,000 psi
K-joint 100,000 IbAn


Case 2 (Best Match):
Concrete 4,350,000 psi
AC 450,000 psi
Base 130,000 psi
Subgrade 28,000 psi
K-joint 100,000 IbAn


, 7>


1 144 -- 14.9 1


Time (sec)


-- Gage 3 3D Model (Casel) -- 3D Model (Case2)


Figure 7-34 Strain comparison at Gage 3 in the 4-inch slab in Phase I-b.


r
r
I


- 14.1--- 14.2 14.3 --- 4.4"


















Strain comparison 10" 6'x6' slab Phase II
CASE 1 Location 1 Top (depth 1"), 12 kips load CASE 2 (Best Match)
Material properties: Material properties:
Concrete 4,350,000 psi Concrete 4,350,000 psi
AC 1,000,000 psi AC 1,000,000 psi
Base 160,000 psi Base 160,000 psi
Subgrade 30,000 psi _________ _______ ___ ______ Subgrade 30,000psi


CASE
Sprin
Interfa
Interfa
Interfa
Trans


17.2 17.4 17.6 1 18


g constants:
iceX 3,000,000 bn b
ceY 3,000,001 bn --
iceZ 101l bAn
v JointX OlbAn


Transv JointY OlbAn
Transv JointZ 100,0001bAn
Long JointX 10,000bAnn
Long JointY OlbAn
Long JointZ 100,000 IbAn


Time (sec)


18.2 18.4 18.6 18.8 1

CASE 2 (Best Match)
Spring constants:
InterfaceX 3,000,000 IbAn
InterfaceY 3,000,00 IbAn
InterfaceZ 10 IbAn
Transv JointX OlbAn
Transv JointY 1,000,000
IbAn
Transv JointZ 100,000IbAn
Long JointX 10,000b/in
Long JointY 100,000 bAn
Long JointZ 3,000,00 IbAn


-- Gage 1 3D Model (Case 1) A 3D Model (Case 2)



Figure 7-35 Strain comparison at top of Location 1 (mid edge) of the 10-inch slab in Phase II.


CASE 1 (Best Match)
Material properties:
Concrete 4,350,000 psi
20 AC 1,000,000psi
Base 160,000 psi
Subgrade 30,000 psi
15


Strain comparison 10" 6'x6' slab Phase II

1 Location 1 Bottom, 12 kips load


CASE 1(Best Match)
Spring constants:
InterfaceX 3,000,000 IbAn
InterfaceY 3,000,000 IbAn
InterfaceZ 10Ilbn - -
Transv JointX OlbAn
Transv JointY Olb0 n
Transv JointZ 100,000 - -
IbAn t ,1
Long JointX 10,0001b/in
Long JointY Olb/in /
Long JointZ 100,000 IbO n


CASE 2
Material properties:
Concrete 4,350,000 psi
AC 1,000,000 psi
Base 160,000 psi
Subgrade 30,000 psi


Time (sec)


Gage 2 3D Model (Case 1) A 3D Model (Case 2)


Figure 7-36 Strain comparison at bottom of Location 1 (mid edge) of the 10-inch slab in Phase

II.













202


5

0

-5

-10

-15

-20

-25


7 1 17.6 17.8 18 18.2 18. 18.6 18.8


CASE 2
Spring constants:
Interface X 3,000,000 Ibin
- InterfaceY 3,000,00 Ib/in
InterfaceZ 101 Ib/in
Transv JointX OlbAn
- Transv JointY 1,000,000 IbAn
Transv JointZ 100,0001bAn
Long JointX 10,0001bAn
- Long JointY 100,0001b/in
Long JointZ 3,000,000,000 IbAn















Strain comparison 10" 6'x6' slab Phase II


CASE 1 Location 1 AC surface, 12 kips load
Material properties:
Concrete 4,350,000 psi CASE 2 (
-AC 1,000,000 psi Spring co
Base 160,000 psi Interface X
Subgrade 30,000 psi Interface Y
Interface Z
CASE 1 Transv Joi
Spring constants: - Transv Joi
Interface X 3,000,000 IbAn IbAn
Interface Y 3,000,000 bAn Transv Joi
InterfaceZ 10l1bAn -Long Joint
Transv JointX OlbAn Long Joint
Transv JointY OlbAn Long Joint
Transv JointZ 100,000 ibAn A.
Long JointX 10,000lbAn
7 Long JointY Obn 7.6 17.8 18 18.2
-Long JointZ 100,000 IbAn


Best Match)
nstants:
3,000,000 IbAn
3,000,000 IbAn
1091bAIn
ntX OlbAn
ntY 1,000,000

ntZ 100,000 IbAn
X 10,000 IbAn
Y 10,00,000 IbAn
Z 3,000,000 IbAn


CASE 2 (Best Match)
Material properties:
Concrete 4,350,000 psi
AC 1,00,000 psi
Base 160,000 psi
Subgrade 30,000 psi


18. 18.6 18.8


Time (sec)



Gage 3 3D Model (Case 1) -A 3D Model (Case 2)


Figure 7-37 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 10-

inch slab in Phase II.


CASE 1
Material properties:
Concrete 4,350,000 psi
AC 1,000,000 psi
Base 160,000 psi
Subgrade 30,000 psi

0 1/7\

17 17.2
(-


0)


S -10

M -10
Cn


Strain comparison 10" 6'x6' slab Phase II

Location 5 Bottom (depth 1.25"), 12 kips load


17.4 17.6 7

CASE 1
Spring constants:
InterfaceX 1,000,000 Ibin
InterfaceY 1,000,000 Ibin
InterfaceZ 10 IbAn
Transv JointX Olb/in
Transv JointY Olb/in
Transv JointZ 100,00o0b/in
Long JointX 10,000,0001bAn _
Long JointY OIbAn
Long JointZ 5,000,000 IbAn


CASE 2 (Best Match)
Material properties:
Concrete 4,350,000 psi
AC 800,000 psi
Base 160,000 psi
Subgrade 30,000 psi


18 1 82 18.6 18.8 1

SA CASE 2 (Best Match)
S Spring constants:
S_ -- - InterfaceX 1,000,000IbAn
\ /. InterfaceY 1,000,000IbAn
S'/ InterfaceZ 10 IbAn
--- ---- Transv JointX Olbin
Transv JointY Olbin
Transv JointZ 100,000b/in
Long JointX 10000,000M IbAn
Long JointY Olb/in
S(sec) Long JointZ 3,000,000 Ibn


-- Gage 13 3D Model (Case 1) A 3D Model (Case 2)


Figure 7-38 Strain comparison at top of Location 5 (slab corer) of the 10-inch slab in Phase II.














CASE 1
Material properties:
Concrete 4,350,000 psi
AC 1,000,000 psi
Base 160,000 psi
Subgrade 30,000 psi


7 17.2


Strain comparison -10" 6'x6' slab Phase II
Location 5 Bottom (depth 9"), 12 kips load
CASE 1
Spring constants:
Interface X 1000,000 Ib/in
Interface Y 1000,000 Ib/in -
InterfaceZ 1091lb/in
Transv Joint X 0 b/in
Transv Joint Y 0 Ib/in
Transv JointZ 10O,0001b/in
Long JointX 10,000,000 b/in
Long Joint Y 0 lb/in
Long Joint Z 5,000,000 Ib/in - -



17.4 17.6 17.8 18 18.2 18.4


CASE 2
Material properties:
Concrete 4,350,000 psi
AC 800,000 psi
Base 160,000 psi
Subgrade 30,000 psi

CASE 2
Spring constants:
InterfaceX 1000,000 Ib/in
Interface Y 1000,000 Ib/in
InterfaceZ 1091b/in
Transv JointX Olb/in
Transv JointY Olb/in
Transv JointZ 1000,000b/in
Long Joint X 10 000,000 Ib/in
Long JointY Olb/in
Long Joint Z 3,000,000 b/in


Time (sec)


-- Gage 14 --3D Model -A- Series3



Figure 7-39 Strain comparison at bottom of Location 5 (slab corner) of the 10-inch slab in Phase

II.


:ASE 2
material properties:
concrete 4,350,000 psi
\C 800,000 psi
lase 160,000 psi
ubgrade 30,000 psi


SE 2
ing constants:
faceX 1,000,000 IbAn
faceY 1,000,000 IbAn
faceZ 10 Ib/in
Isv JointX OlbAn
Isv JointY OlbAn
Isv JointZ 100,0001b/in
g JointX 10000,000IbAn
g JointY OlbAn
g JointZ 3,000,000 IbAn


C
CASE 1(Best Match) Strain comparison 10" 6'x6' slab Phase II M
Material properties: C
Concrete 4,350,000 psi Location 5 AC surface, 12 kips load A
AC 1,000,000 psi B
Base 160,000 psi S
Subgrade 30,000psi


CASE 1 (Best Match) __ CA
Spring constants: Spr
InterfaceY 1,000,000 Ibn Inter

InterfaceZ 101IlbAn Inter
Transv JointX OlbAn Trar
Transv JointY OlbAn _ _- _ __ Trar
Transv JointZ 100,001bAn Trar
Long JointX 10,000,0001bAn Lon!
Long JointY OlbAn Lon
Long JointZ 5,000,000 IbAn Lon!
7 17. 17.8 18 18.2 18.4


Time (sec)


Gage 15 -M 3D Model (Case 1) -A 3D Model (Case 2)

Figure 7-40 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the

10-inch slab in Phase II.


25

20

" 15
E
10

33 5

0

-5


20


15


E 10
C
5' 5


0 -
1
-5















Strain comparison 8" 6'x6' slab Phase II

Location 1 Top (depth 0.8"), 12 kips load

CASE 1 (Best Match)
Material properties:
Concrete 4,350,000 psi
AC 1,000,000 psi
Base 160,000 psi
Subgrade 30,000 psi --i


CASE 1 (Best Match)
Spring constants:
InterfaceX 3,000,000 bAn
Interface Y 3,000,000 bAn
InterfaceZ 1091bAn
Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,000MbAn
Long JointX 10,0001bAn
Long JointY OlbAn
Long JointZ 100,000 IbAn


Time (sec)


CASE 2
Material properties:
Concrete 4,200,000 psi
-AC 1,000,000 psi
Base 160,000 psi
Subgrade 30,000 psi


13 CASE 2 1
Spring constants:
Interface X 1,000,000 bAn
Interface Y 1,000,000 IbAn
InterfaceZ 10 DbAn
Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,0001bAn
Long JointX 10,0001bAn
Long JointY OlbAn
Long JointZ 1,000,000 IbAn


Gage 1 --3D Model (Case 1) A 3D Model (Case 2)



Figure 7-41 Strain comparison at top of Location 1 (mid edge) of the 8-inch slab in Phase II.


Strain comparison 8" 6'x6' slab Phase II

Location 1 Bottom (depth 6.65"), 12 kips load


CASE 1 (Best Match)
Spring constants:
Interface X 3,000,000 IbAn
InterfaceY 3,000,0001bAn
InterfaceZ 10llbAn
-Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,0001bAn
Long JointX 10,0001bAn
Long JointY OlbAn
Long JointZ 100,000IbAn


CASE 1 (Best Match)
Material properties:
SConcrete 4,350,000 psi
AC 1,000,000 psi
Base 160,000 psi
Subgrade 30,000 psi


1 12.5


13


Time (sec)


CASE 2
Spring constants:
InterfaceX 1,000,000 IbAn
InterfaceY 1,000,000 IbAn
InterfaceZ 10 DbAn
Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,0001bAn
Long JointX 10,0001bAn
Long JointY OlbAn
Long JointZ 1,000,000 IbAn




13.5 1
SE 2
erial properties:
create 4,200,000 psi -
1,000,000 psi
e 160,000 psi
grade 30,000 psi


Gage 2 3D Model (Case 1) A 3D Model (Case 2)


Figure 7-42 Strain comparison at bottom of Location 1 (mid edge) of the 8-inch slab in Phase II.


10 -


E

2 -10
5)


-20


-30 '


20 -


~


i (-,,














Strain comparison 8" 6'x6' slab Phase II
Location 1 AC surface, 12 kips load


CASE 1 (Best Match)
Spring constants:
Interface X 3,000,000 IbAn
Interface Y 3,000,000 IbAn
InterfaceZ 10D9bAn
Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,0001bAn
Long JointX 10,000IbAn
Long JointY OlbAn
Long JointZ 100,000 IbAn


CASE 2
Spring constants:
Interface X 1,000,000 IbAn
Interface Y 1,000,000 IbAn
InterfaceZ 10l bAn
Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,0001bAn
Long JointX 10,0001bAn
Long JointY OlbAn
Long JointZ 1,00 0 IbAn


-Gage 3 ---3D Model (Case 1) A 3D Model (Case 2)


Figure 7-43 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 8-
inch slab in Phase II.


Strain comparison 8" 6'x6' slab Phase II CASE 2
Location 5 Top (depth 0.87"), 12 kips load Material properties:
Concrete 4,200,000 psi
SAC 1,000,000 psi




51 .5 Subgrade 30,000 psi _. ,'1_ -- 135 1
E
c -10 -CASE 1(Best Match) CASE2
SSpring constants: \ Spring constants:
nterfaceX 3,000,000 Ib/ \T n \ Interface X 2,00000,0001 bA n
InterfaceX 3,00,000IbAn
U) -15 InterfaceY 3,000,0001b/n ----, .- InterfaceY 2,000,0001bAn
InterfaceZ 10lbAn 'InterfaceZ 10l1bAn
-20 Transv JointX OlbAn Transv JointX 0 bAn
Transv JointY 0Ib/n Transv JontY O0bAn
Transv JointZ 100,000 IbAn Transv JomtZ 100,000 lbAn
-25 Long JointX 10,000,000 bAn Long JointX 10,000 bAn
Long JointY OIbAn Long JointY OlbAn
Long JointZ 5,000,000 Ibn Time (sec) Long JointZ 5,000,000 IbAn


-- Gage 13 3D Model (Case 1) ----A---- 3D Model (Case 2)



Figure 7-44 Strain comparison at top of Location 5 (slab corner) of the 8-inch slab in Phase II.
















Strain comparison 8" 6'x6' slab Phase II

CASE 1(Bt Location 5 Bottom (depth 7"), 12 kips load
CASE 1(Best Match)
Material properties: CASE 2
20 Concrete 4,350,000 psi Material properties:
AC 1,000,000 psi Concrete 4,200,000 psi
Base 160,000 psi AC 1,000,000 psi
15 Subgrade 30,000 psi --- Base 160,000 psi
Subgrade 30,000 psi

E 10 CASE 1 (Best Match) - -- CASE 2
Spring constants: Spring constants:
InterfaceX 3,000,000 IbAn InterfaceX 2,000,000 IbAn
5 Interface Y 3,000,000 IbAn Interface Y 2,000,000 IbAn
.i InterfaceZ 1091bAn InterfaceZ 10l1bAn
S Transv JontX O bAn Transv JointX OlbAn
STransv JointY OlbAn Transv JointY OlbAn
1 .5 Transv JointZ 100,0001bAn 12.5 13 Transv JointZ 100,000bAn 1
-5 __ Long JointX 10,000,0001bAn Long JointX 10,0001bAn
Long JointY OlbAn Long JointY OlbAn
Long JointZ 5,000,000 bAn Time (sec) Long JointZ 5,000,000 IbAn



-- Gage 14 --3D Model (Case 1) -A 3D Model (Case 2)

Figure 7-45 Strain comparison at bottom of Location 5 (slab corner) of the 8-inch slab in Phase
II.


Strain comparison 8" 6'x6' slab Phase II


CASE 1 (Best Match)
Material properties:
Concrete 4,350,000 psi
AC 1,000,000 psi
Base 160,000 psi
Subgrade 30,000 psi

CASE 1 (Best Match)
Spring constants:
Interface X 3,000,000 IbAn
Interface Y 3,000,000 IbAn
InterfaceZ 10 9bAn
Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,000IbAn
Long JointX 10,000,0001bAn
Long JointY OlbAn
Long JointZ 5,000,000 IbAn


Location 5 AC surface, 12 kips load












--- 12.5 13

12.5 13


CASE 2
Material properties:
Concrete 4,200,000 psi
AC 1,000,000 psi
Base 160,000 psi
Subgrade 30,000 psi


CASE 2
Spring constants:
Interface X 2,000,000 IbAn
Interface Y 2,000,000 IbAn
InterfaceZ 10I1bAn
Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,000lbAn
Long JointX 10,000 IbAn
Long JointY OlbAn
Long Joint Z 5,000,000 IbAn


Time (sec)


-- Gage 15 3D Model (Case 1) A 3D Model (Case 2)


Figure 7-46 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the
8-inch slab in Phase II.















Strain comparison 6" 6'x6' slab Phase II

Location 1 Top (depth 0.75"), 12 kips load CASE 2 (Best Match)
Material properties:
Concrete 4,200,000 psi
AC 800,000 psi
--- ------- Base 160,000 psi
Subgrade 30,000 psi


t1.2 1.4


Time (sec)


- -1-1- --H-- -
CASE 2 (Best Match)
Spring constants:
Interface X 3,500,000 Ibn
Interface Y 3,500,000 Ibn
InterfaceZ 5x1018lbAn
Transv JointX 0 Ibin
Transv JointY 0 Ibin
Transv JointZ 100,0001bAn
Long JointX 10,0001bAn
Long JointY Olbin
Long JointZ 100,000 IbAn


-- Gage 1 3D Model (Case 1) -A -.3D Model (Case 2)


Figure 7-47 Strain comparison at top of Location 1 (mid edge) of the 6-inch slab in Phase II.









Strain comparison 6" 6'x6' slab Phase II

Location 1 Bottom (depth 5.5"), 12 kips load


CASE 1
25 Material properties:
Concrete 4,350,000 psi
20 AC 1,000,000 psi
Base 160,000 psi
15 Subgrade 30,000 psi

10 --CASE 1
Spring constants:
5 Interface X 3,000,000 IbAn
Interface Y 3,000,000 IbAin
o InterfaceZ 5x1018IbAn
S Transv JointX OlbAn
1Transv JointY 0 bAn
-5 Transv JointZ 100,0001bAn
Long JointX 10,0001bAin
-10 Long JointY OlbAn
Long Joint Z 100,000 Ibin


~~ -- -/ T- -\ -
--------------^ ---------------'-------1





1 11.2 11.4 1
L-

L
L
Time (sec)


CASE 2 (Best Match)
Material properties:
Concrete 4,200,000 psi
AC 800,000 psi
Base 160,000 psi
Subgrade 30,000 psi

:ASE 2 (Best Match)
Spring constants:
nterface X 3,500,000 IbAn
nterface Y 3,500,000 IbAn
nterfaceZ 5x10181b/in
ransv JointX OlbAn
ransv JointY OlbAn 1
ransv JointZ 100,0001b/in
ong JointX 10,0001bin
ong JointY 01b/in
ong JointZ 100,000 Ibin


-- Gage 2 --3D Model (Case 1) -A -3D Model (Case 2)



Figure 7-48 Strain comparison at bottom of Location 1 (mid edge) of the 6-inch slab in Phase II.


















208


I


1














Strain comparison 6" 6'x6' slab Phase II

CASE 1 Location 1 AC surface, 12 kips load CASE 2
Material properties: Material properties:
Concrete 4,350,000 psi Concrete 4,200,000 psi
35 AC 1,000,000 psi AC 800,000 psi
30 Base 160,000 psi __ Base 160,000 psi
Subgrade 30,000 psi Subgrade 30,000 psi
25
20 C ASE 1 CASE 2
E 15 -Springconstants: --- Spring constants:
Interface X 3,000,000 IbAn Interface X 3,500,000 IbAn
c 10 Interface Y 3,000,000 Ibin Interface Y 3,500,000 IbAn
5 InterfaceZ 5x1081bAn InterfaceZ 5x1011bAn/
Transv JointX OlbAn Transv JointX OlbAn
0 Transv JointY OlbAn Transv JointY OlbAn
Transv JolntZ 100,0001bAin & Transv JointZ 1000,0001bAn
Long JointX 10,0001bAin 10 1.2 .4 Long JointX 10,0001bAn -
-10 -Long JointY OlbAn Long JointY OlbAn
LongJointZ 100,000 Ibn Long Joint Z 100,000 IbAn
-15 -
Time (sec)

Gage 3 3D Model A Series3

Figure 7-49 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 6-
inch slab in Phase II.


Strain comparison 6" 6'x6' slab Phase II
__- Location 5 Top, 12 kips load


Gage 13 -- 3D Model (Case 1) -A 3D Model (Case 2)


Figure 7-50 Strain comparison at top of Location 5 (slab corner) of the 6-inch slab in Phase II.















Strain comparison 6" 6'x6' slab Phase II
Location 5 Bottom, 12 kips load


25 CASE 1 (Best Match)
Material properties:
Concrete 4,350,000 psi
20 -AC 1,000,000 psi
Base 160,000 psi
1 Subgrade 30,000 psi

CASE 1(Best Match)
10 Spring constants:
InterfaceX 3,000,000 bAn
S InterfaceY 3,000,000 IbAn
InterfaceZ 5x1081bAn
Transv JointX Olb/in
0 Transv JointY Olb/in
Transv JointZ 100,0001biAn
10 Long JointX 10,0001bAn
-5 Long JointY OlbiAn
Long JointZ 5,000,000 IbAn


------ -- -------- ------ ------







---- ----- --------



11 11.5


Time (sec)


CASE 2
Material properties:
Concrete 4,200,000 psi
AC 800,000 psi
-Base 160,000 psi
Subgrade 30,000 psi


CASE 2
Spring constants:
InterfaceX 3,000,000 IbAn
InterfaceY 3,000,000 bAn
InterfaceZ 5x10181bAn
Transv JointX Olbin
Transv JointY Olbin
Transv JointZ 100,000 bAn
Long JointX 10,0001bAn
Long JointY Olbin
Long JointZ 5,000,000 IbAn


-- Gage 14 3D Model (Case 1) A 3D Model (Case 2)J


Figure 7-51 Strain comparison at bottom of Location 5 (slab corner) of the 6-inch slab in Phase

II.


Strain comparison 6" 6'x6' slab Phase II
Location 5 -AC surface, 12 kips load


CASE 1 (Best Match)
Spring constants:
Interface X 3,000,000 IbAn
Interface Y 3,000,000 IbAn
InterfaceZ 5x1018 bAn
Transv JointX OlbAn
Transv JointY OlbAn
Transv JointZ 100,0001bAn
Long JointX 10,0001bAn
Long JointY OlbAn
Long JointZ 5,000,000 IbAn


Time (sec)


CASE 2
Material properties:
-Concrete 4,200,000 psi
AC 800,000 psi
Base 160,000 psi
Subgrade 30,000 psi


CASE 2
Spring constants:
InterfaceX 3,000,000 IbAn
InterfaceY 3,000,000 IbAn
InterfaceZ 5x1018 bAn
Transv JointX Olbin
Transv JointY Olbin
Transv JointZ 100,0001bin
Long JointX 10,0001biAn
Long JointY Olbin
Long JointZ 5,000,000 bAn


Gage 15 3D Model (Case 1) A 3D Model (Case 2)


Figure 7-52 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the

6-inch slab in Phase II.













210









CHAPTER 8
EVALUATION OF POTENTIAL PERFORMANCE OF THE WT DESIGNS

8.1 Overview

This chapter presents the evaluation of the potential performance of WT pavements with

the same designs as those used in the test sections in this study.

The 3-D finite element model with the model parameters for each test section, as deter-

mined from the deflection-based and strain-based calibration (as presented in Chapter 7), was

used to perform a stress analysis to determine the maximum stresses in each WT pavement under

typical critical temperature-load conditions in Florida. The potential performance of each WT

pavement was assessed based on (1) the maximum tensile stress in the concrete, (2) the maxi-

mum shear stress at the concrete-asphalt interface, and (3) the maximum tensile stress in the AC.

The non-linear version of the 3D analytical model, which was shown at the end of Chapter

6, was used in combination with the linear model to estimate the level of stresses in the 4" slab in

Phase I-a. This was the only test sections that had a corner crack during the HVS loading period.

The purpose of this analysis was to determine if the analytical model can also predict failure in

the composite pavement.

8.2 Assumptions for the evaluation of the potential performance of the test sections

The potential performance of the test sections was evaluated using the 3D analytical

model calibrated in the previous chapter. After the model could satisfactorily predict deflections

and strains for different conditions and slab characteristics, it is possible to use the model to

estimate the level of stresses in the composite pavement.

8.2.1 Critical loading conditions

A 24-kip single axle load, which is slightly higher than the maximum legal single axle load

of 22 kips in Florida, was used as the applied load in the analysis. The two critical loading









positions used in the analysis were (1) the mid-edge and (2) the corner of the slab. Figures 8.1

and 8.2 show the positions of the axle load used for the slabs with joint spacings of 4 ft and 6 ft,

respectively. These figures also show the comparison between a typical roadway load position

and the critical load position considered in the analysis.

The minimum and maximum temperature differential in the concrete slab as observed

during the HVS loading were around -10 F and +20 F, respectively. These two extreme

temperature differentials were used in the critical stress analysis.

8.2.2 Model parameters

The model parameters of the 3-D model for each test section used in the critical stress

analysis are displayed in Table 8.1. It is to be noted that all the model parameters, except the

joint spring stiffnesses, are from the results of deflection-based and strain-based calibration as

presented in Chapter 7. All the joint spring stiffnesses were set to zero in this analysis, based on

the expectation that all the joints will eventually crack all the way through, and there will

eventually be less load transfer across the joints as compared with their initial conditions.

To represent different temperature conditions, which affect the elastic modulus of the AC

layer, all the test sections were analyzed using three different values of the AC elastic modulus,

namely 300,000 psi, 700,000 psi and 1,100,000 psi.

For the interface horizontal springs, two spring stiffness values were used. The higher

values were used for the condition at the edge of the joints, while the lower values were used for

the condition at the slab covers.

8.3 Results of Critical Stress Analysis

8.3.1 Maximum stresses in the concrete slabs

The maximum computed tensile stresses in the various bonded concrete slabs (in Phases I-

a and I-b) and partially bonded concrete slabs (in Phase II) caused by a 24-kip single axle load









placed at two different critical positions (mid-edge or corner), and for three different temperature

differentials in the concrete slab (-10, 0 or +20 F) are shown in Table 8.2. Three different AC

moduli, namely 300,000 psi, 700,000 psi or 1,100,000 psi, which represent the condition of the

AC at different temperatures, were used in the analysis. The values shown in Table 8.2 are

principal stresses and they can be located at the top or bottom of the concrete slab.

From table 8.2, the relationship between many pavement parameters and pavement

behavior can be observed. Following paragraphs describe the effect of the AC Elastic Modulus,

Temperature differential, Slab size and bond in the interface.

8.3.1.1 Effects of elastic modulus of AC layer

Figure 8.3 shows the effects of the elastic modulus of the AC layer on the maximum

stresses in the concrete caused by a 24-kip axle load applied at mid edge for 4-inch bonded

concrete slabs with 6 ft joint spacing. It can be seen that at the condition of temperature differen-

tial of +20 F, an increase of 55% in tensile stress (from 367 to 568 psi) in the concrete was

obtained when the elastic modulus of the AC layer dropped from 1,100,000 psi to 300,000 psi.

However, at the condition of temperature differential of -10 F, a decrease of 35% in tensile

stress (from 380 to 246 psi) in the concrete was obtained when the elastic modulus of the AC

layer changed from 1,100,000 psi to 300,000 psi.

Figure 8.4 shows similar plots for 5-inch bonded concrete slabs with 4 ft joint spacing. In

this case, a decrease in the elastic modulus of the AC caused an increase in the tensile stress in

the concrete for all temperature conditions.

8.3.1.2 Effects of temperature differential

Figures 8.5 and 8.6 show the effects of temperature differentials on the maximum stresses

in the bonded slabs with 6 ft joint spacing (test slabs in Phase I-a), caused by a 24-kip single axle

load placed at mid-edge, and corer of the slab, respectively. An AC elastic modulus of 300,000









psi was used in these analyses. Figures 8.7 and 8.8 show similar plots for the bonded slabs with

4 ft joint spacing (test slabs in Phase I-b). It can be seen from these figures that higher stresses in

the concrete were obtained at a temperature differential of +20 F than at a temperature

differential of -10 F. For the condition of temperature differential of +20 F, loads at slab mid-

edge produced higher stresses than those produced by loads at slab corner.

Similar observation about the effects of temperature differential and loading positions can

be made for the partially bonded slabs in Phase II. Figure 8.9 shows the effects of temperature

differentials on the maximum stresses in the partially bonded slabs with 6 ft joint spacing (test

slabs in Phase II), caused by a 24-kip single axle load placed at mid-edge of the slab. It can be

seen that the condition of temperature differential of +20 F produced much higher stresses than

a temperature differential of -10 F.

8.3.1.3 Effects of panel size

Figures 8.10 and 8.11 show the effects of panel size on the maximum stresses in the

bonded concrete slabs caused by a 24-kip single axle load at mid-edge and corer of the slabs,

respectively. An AC elastic modulus of 300,000 psi was used in these analyses. It can be seen

that at the most critical temperature and load condition (when the temperature differential was

+20 F and the load was applied at mid-edge of slab), the 4 ft X 4 ft panels had slightly lower

stresses than the 6 ft X 6 ft panels. The reduction in stress ranges from 2% for the 4-inch slabs to

15% for the 6-inch slabs.

When the load was applied to the slab corner, the 4 ft X 4 ft panels had stresses

significantly higher than the 6 ft X 6 ft panels. For example, when the temperature differential

was +20 F and the load was applied at the slab corner, the 4 ft X 4 ft panels had higher stresses

than the 6 ft X 6 ft panels by 26% for the 5-inch slabs.









8.3.1.4 Effects of bonded versus partially bonded interface

A direct comparison of the effects of a bonded concrete-asphalt interface versus a partially

bonded interface can be made by comparing the computed maximum tensile stresses in the 6-

inch bonded slabs in Phase I-a with those in the 6-inch partially bonded slabs in Phase II.

Figures 8.12 and 8.13 show the comparison of maximum computed tensile stresses in concrete

for these two test slabs under a 24-kip single axle load applied at the mid-edge and corer of the

slabs, respectively. An AC elastic modulus of 300,000 psi was used in these analyses.

From these two figures, it can be seen that, for the most critical condition of a tempera-

ture differential of +20 F, the bonded slabs have about the same maximum tensile stresses as

those in the partially bonded slabs. However, for the condition of a temperature differential of

-10 F, the bonded slabs have slightly lower maximum tensile stresses than the partially

bonded slabs.

Analyses were also performed to determine the maximum stresses in concrete under

critical loading conditions for the hypothetical cases if the test slabs in Phase II were constructed

as fully bonded to the asphalt layer. These computed stresses are also shown in Table 8.2.

Figure 8.14 shows the comparison of the maximum stresses in concrete caused by a 24-kip single

axle load at mid-edge of slab for the test sections in Phase II with those for the hypothetical cases

if the same slabs were constructed bonded to the asphalt layer. Similar trends can be observed

here. For the condition of a temperature differential of +20 F, the bonded slabs have about the

same maximum tensile stresses as those in the partially bonded slabs. However, for the

condition of a temperature differential of -10 F, the bonded slabs have slightly lower maximum

tensile stresses than the partially bonded slabs.









8.3.2 Maximum shear stresses at the interface

The 3-D finite element model was also used to calculate the shear stresses at the concrete-

asphalt interface under critical loading conditions. Table 8.3 displays the maximum shear

stresses at the interface caused by a 24-kip single axle load at a temperature differential of +20

F for the bonded slabs in Phases I-a and I-b. Figures 8.15 and 8.16 show the plots of these

maximum shear stresses at the interface for the bonded slabs with 6 ft joint spacing and 4 ft joint

spacing, respectively. Three AC elastic moduli, namely 300,000, 700,000 and 1,100,000 psi,

were used in the analyses.

From these two figures, it can be observed that for both load locations (mid-edge and

corner), the shear stress is higher when the AC layer is stiffer, which is a consequence of the

degree of restriction that the AC layer represent for the concrete slab. It can also be observed

that the smaller 4 ft X 4 ft slabs (Phase I-b) had lower maximum shear stresses.

In all cases, the maximum computed shear stress at the interface are very low compared

with the shear strength measured from the core samples from the test sections using the Iowa

Shear test. From Table 4.4 in Chapter 4, it can be seen that the average shear strength for the 4 ft

x 4 ft slabs was 194.5 psi, and the average shear strength for the 6 ft X 6 ft slabs was 220 psi.

The maximum shear stress among all cases was only 84.8 psi. The fact that the shear

stress developed at the interface is much lower than the shear strength indicates that the fully

bonded condition at the interface may remain in place for a long time during the life time of the

composite pavement. It has to be pointed out that the Iowa Shear test is performed in such a way

that no tensile stress is developed in the interface. In real composite pavements, some critical

combination of shear stress and vertical tensile stress can negatively affect the shear strength in

the interface. However, the application of the analytical model showed that that particular









combination does not happen in the composite pavements investigated in this research.

Effectively, vertical stresses calculated using the analytical model resulted in maximum values of

50 psi in places where shear stress was less than 30 psi. In places where the shear stress was

maximum, the vertical stress in the interface was generally compression stress or a very low

tensile stress.

8.3.3 Maximum stresses in the AC layer

The 3-D analytical model was also used to determine the tensile stresses in the AC layer

under critical loading conditions. Table 8.4 displays the maximum tensile stresses in the asphalt

concrete layer caused by a 24-kip single axle load at various critical loading conditions. Figures

8.17 and 8.18 show the maximum tensile stresses in the AC layer caused by a 24-kip single axle

load at a temperature differential of +20 F for the bonded slabs with 6 ft joint spacing and 4 ft

joint spacing, respectively. Two AC elastic moduli, namely 300,000 psi and 1,100,000 psi, were

used in the analyses.

From these two figures, it can be observed that the tensile stress in the AC layer increases

as the elastic modulus of the AC increases and that the size of the slab concrete has little effect

on the tensile stresses in the AC layer. The tensile stress in the AC layer increases since the

relative stiffness between concrete and asphalt change as the asphalt become more rigid. In other

words the asphalt layer is helping the concrete slab to carry the effect of the load.

The maximum calculated tensile stresses in the AC layer are lower than the tensile strength

of the AC measured in the IDT test (Table 4.9, Chapter 4). The stress calculated using an AC

elastic modulus of 1,100 ksi is lower than strength measured at 15 oC, which can be considered a

low temperature for the AC layer. Similarly, the stress calculated using an AC elastic modulus of

300 ksi is lower than the strength measured at 40 oC, which can be considered as a very high

temperature for the AC layer. The highest tensile stresses in the AC layer were obtained when a









high value of the AC elastic modulus was used. Since high values of the AC elastic modulus

occur at low temperatures when the tensile strength of the AC layer is high, this creates a

favorable condition where the AC is less likely to crack. Thus, stresses in the AC layer should

not be a controlling factor in the performance of these WT pavements.

8.4 Potential performance of the test sections

The maximum computed stresses in the concrete slab caused by the critical loading

condition (of a 24-kip single axle load, placed at the mid-edge of the slab, and at a temperature

differential of +20 F in the concrete slab) were used to assess the potential performance of the

WT pavement test sections evaluated in this study. The fatigue curve given by the PCA, which

relates the stress/strength ratios with the number of repetitions to produce fatigue failure in

concrete, was used to estimate the number of load repetitions to failure. The following equations

were used to calculate the maximum number of load repetitions as a function of the

stress/strength ratio:

NLR = 10(96.5 100 r)/8.1 If r > 0.5 (8-1)
NLR = infinite If r < 0.5 (8-2)

Where
NLR = number of load repetitions to failure, and
r = stress/strength

The average flexural strength at 56 days of the concrete used in the test sections was 842

psi, as presented in Tables 4.6 and 4.7 in Chapter 4. This flexural strength value was used in the

computation of the stress to flexural strength ratios for the WT pavement test sections.

Table 8.5 displays the computed maximum stresses and stress to flexural strength ratios for

all the WT pavement test sections evaluated in this study. The allowable number of 24-kip

single axle loads under critical loading conditions were also computed and shown in this table. It









is to be stressed that the results of these analyses are only applicable to the condition of the test

sections, which had 4.5 inches of AC layer over 12 inches of limerock base.

It can be seen that the maximum computed stresses were all below the flexural strength of

the concrete for all the test sections. This means that all the WT pavement test sections with a

concrete slab thickness of 4 inches or higher can withstand certain number of repetitions of the

24-kip single axle load under the critical loading condition without cracking. The allowable

number of repetitions of this critical load increases with slab thickness. The allowable number of

load repetitions also increases with smaller joint spacing.

In order to be able to withstand the critical load without fear of fatigue failure (for an

infinite number of critical load repetitions), a minimum slab thickness of 6 inches would be

needed for a joint spacing of 4 ft, and a minimum slab thickness of 8 inches would be needed for

a joint spacing of 6 ft.

8.5 Evaluation of the actual performance of the test sections

During the HVS testing, only one slab failed. The 4" slab in Phase I-a was loaded with

36,407 passes of a 9-kips wheel load, followed by 146,748 passes of a 12-kips wheel load. Then

the load was increased to 15 kips with a total of 35,918 passes and then to 18 kips with 21,727

passes. Finally, the load was increased to 21 kips, and corer cracks developed after 12,187

passes. The crack was generated at the slab corer, which represents a typical type of failure for

the WT pavements. Using both the linear and the non-linear model, an investigation of the

failure mechanism was undertaken.

According with the results of WT experiments performed nationwide and in the current

research, if this type of composite pavement is constructed accordingly, with the concrete placed

bonded to the AC layer, which is at least 3" thick, then the pavement is able to carry a significant

number of load repetitions without failure. It has been hypothesized that prior to failure the bond









interface has to be broken, so that the composite pavement becomes unbonded. Temperature

differential plays an important role in making the pavement unbonded. Negative temperature

differential tend to curl the slab up at corners and edges. Because the pavement is bonded, a high

vertical tensile strain is generated at that location that eventually will break the bond in the

interface. Since the shear strength measured at the interface was in the order of 200 psi, it is

expectable that the tensile strength at the interface had a similarly high value.

When the temperature differential is positive, another critical stress condition appears in

the concrete. At positive temperature differential, the slab tends to curl down at the corners and

edges, and because the slab is restricted to curl due to the bond in the interface, tensile stress is

generated at the bottom of the slab. The wheel load positioned at the corner or edge will also

generate tensile stress at the bottom, making the combination of load and positive temperature

differential the worst scenario for tensile stress. According to this, a high value of the tensile

stress can be reached at the bottom of the slab, even though the composite pavement is bonded

and the slab does not curl.

The linear model for a slab thickness of 4" was analyzed for a load at the corner with

wheel loads of 12, 15, 18 and 21 kips for three temperature differentials (-10, 0, and +20 F).

Table 8.6 shows the maximum stresses due to these loads. The model was also analyzed for the

condition of only temperature differential to estimate the level of tensile vertical stress in the

interface. The vertical tensile stress in the corner interface was in the order of 200 psi for a -10 F

temperature differential. This means that critical points at the corner of the slab might become

unbonded as soon as the slab reaches a critical temperature differential. Using Eq-8.2 it can be

seen that the slab can withstand the 12-kip, 15-kip, and 18-kip wheel load without experiencing

fatigue cracking (the stress/strength ratio is less than 0.5), while it would fail after a limited









number of repetition of the 21-kip load. Table 8.6 also shows the maximum tensile stress in the

concrete due to the application of a 21-kip wheel load along with a temperature differential using

the non-linear model. As indicated in Chapter 6, the non-linear model has the characteristic that

as soon as the interface is in a tensile state of stresses, the connection between the layers is

broken and the layers are no longer bonded. Figure 8-19 shows the case when the slab is affected

by only negative temperature differential using the non-linear model. The entire temperature

differential was applied at the beginning of the computational analysis while the wheel load was

applied at 20% load intervals. The total energy was used as a convergence criterion in the

computational analysis. At the beginning of the analysis, the corners of the slab curl up and the

slab is no longer in contact with the asphalt. The incremental application of the 21 kips at the

corner put the slab immediately back in contact with the asphalt (Figure 8-20) not allowing

complete development of tensile stress at the top of the slab as normally happens in a cantilever

situation. Maximum tensile stress appears at the bottom of the slab (corners) similar to the case

of the bonded linear model.

Using Eq-8.1, the maximum number of load repetitions due to the application of the 21-

kips wheel load in both the linear and the non-linear model can be estimated.

NLR (21 kips, linear) = 112,201

NLR (21 kips, non-linear) = 14,791

It is to be pointed out that the 4" slab failed after 12,187 applications of a load of 21 kips.

This result shows that the composite pavement must have become unbonded before it fails since

the non-linear model representing the fully unbonded condition could satisfactorily predict the

number of load repetitions to failure in the HVS testing.










Table 8-1 Model parameters of the 3-D model for each test section used in the analysis.
Phase I-a Phase I-b Phase II
4" 5" 6" 4" 5" 6" 6" 8" 10"


Concrete 4,350
3-11
AC Layer 2
x 10
Base 160
Subgrade 30
Interface X
Interface Y
Interface Z
Trans. Joint X 0
Trans. Joint Y 0
Trans. Joint Z 0
Long. Joint X 0
Long. Joint Y 0
Long. Joint Z 0


4,350
3-11
x 102
160
30


4,350
3-11
x 102
160
30


4,350
3-11
x 102
130
28


4,350
3-11
x 102
130
28


4,350
3-11
x 102
130
28


4,200 4,350
3-11 x 3-11 x
102 102
160 160
30 30
3-3.5 2-3
3-3.5 2-3
5x1012 1013


0 0 0 0 0 0 0 0


Material
Elastic
Moduli
Ksi





Spring
Constants
lb/in 106


4,350
3-11 x
102
160
30
1-3
1-3
1013










Table 8-2 Maximum tensile stresses in the concrete slabs caused by a 24-kip single axle Load at
various critical loading conditions.
AC Elastic Modulus (psi)


Tensile Stress (psi)
Phase Slab Temp
-10
4" 0
20
I-a -10
(bonded) 5" 0
(6' x 6') 20
-10
6" 0
20
-10
4" 0
20
I-b -10
(bonded) 5" 0
(4' x 4') 20
-10
6" 0
20
-10
6" 0
20
II
-10
(partially 0
bonded) 20
6' 6'
-10
10" 0
20
-10
Hypothetical 8" 0
case 20
(bonded) -10
6' x 6' 10" 0
20
tress located at the top
stress located at the top


300,000
246.5T
333.8
568.3
223.7T
293.5
531.4
202.0T
256.6
488.6
323.6
286.3
555.1
315.2
243.3
486.9
296.8
209.7
416.2
227.8
253.6
476.0
186.0
200.1
400.4
165.3
158.6
318.4
172.7
197.5
398.8
148.6
156.1
316.8
of the slab


Mid-Edge
700,000
307.7
221.2
441.7
285.4
208.2
433.3
260.5
190.7
412.1
298.9
186.7
435.9
295.8
170.9
405.4
281.6
154.2
362.9


398.2


353.1


290.3


1,100,000 300,000
379.6 259.6
158.7 204.3
366.7 325.8
360.9 251.5
157.9 171.0
371.6 290.9
336.6 248.3
149.9 146.0
361.4 279.2
304.3 331.0
131.1 206.5
364.8 411.7
289.8 340.9
127.7 200.8
352.0 366.5
277.1 340.4
119.7 196.3
324.2 318.7
217.6
182.9
361.4 272.4


323.9


273.9


Where no indicated, the stress is located at the bottom of the slab


Corer
700,000
347.5
134.2
291.0
333.1
119.6
267.8
315.6
104.0
260.9
303.6
152.7
340.0
315.5
160.8
316.7
315.9
161.3
284.5


1,100,000
418.5
95.9
268.0
409.2
89.8
249.9
391.5
80.5
245.9
341.3
125.2
295.9
328.7
135.5
283.6
308.9
138.8
260.2









Table 8-3 Maximum shear stress at the concrete-asphalt interface caused by a 24-kip single axle
load at a temperature differential of +20 F for the bonded slabs.
Shear Stress in the AC Elastic Modulus (psi x 103 )
Interface (psi) Load at the Mid-Edge Load at the Corner
Phase Slab 300 700 1100 300 700 1100
4" 36.0 47.0 53.0 59.3 71.7 78.8
I-a 5" 34.7 44.2 49.6 63.0 75.5 82.7
6" 33.6 41.8 46.6 65.5 77.7 84.8
4" 35.8 46.7 52.3 45.1 56.9 62.8
I-b 5" 34.3 43.5 48.5 46.0 56.8 62.4
6" 33.1 40.9 45.3 46.8 56.5 61.7











Table 8-4 Maximum tensile stresses in the asphalt concrete layer caused by a 24-kip single Axle
load at various critical loading conditions.


Tensile Stress (psi)
Temp
Phase Slab Diff.


I-a
(bonded)
(6' x 6')







I-b
(bonded)
(4' x 4')






II
(partially
bonded)
6' x 6'





Hypothetical
Bonded
6' x 6'


-10
4" 0
20
-10
5" 0
20
-10
6" 0
20
-10
4" 0
20
-10
5" 0
20
-10
6" 0
20
-10
6" 0
20
-10
8" 0
20
-10
10" 0
20
-10
8" 0
20
-10
10" 0
20


AC Elastic Modulus (psi)
Mid-Edge


300,000 700,000 1,100,000 300,000 700,000 1,100,000


97.8
21.8
40.8
100.7
18.3
40.5
101.9
15.7
39.7
91.07
21.3
48.4
92.6
17.3
46.9
91.9
14.4
44.6
107.6
15.7
37.0
110.3
14.1
33.7
104.3
16.2
29.8
100.4
11.8
37.0
95.0
9.2
33.7


172.3
65.6
127.8
179.5
55.95
129.8
182.8
48.2
129.1
161.5
58.7
135.3
165.4
48.8
133.5
165.9
41.2
128.7


121.1


114


105.9


230.9
97.8
202.7
242.6
84.3
208.3
248.0
73.2
208.6
216.4
85.6
206.8
222.9
72.1
206.1
222.9
61.5
206.8


197.9


193.8


188.9


103.6
25.3
45.5
107.5
22.6
44.8
109.8
19.3
43.9
90.0
24.0
57.8
91.6
21.5
55.9
90.9
19.5
63.4
86.7
18.6
42.9


Corer


182.1
79.0
150.8
191.4
74.3
156.8
196.8
66.2
160.0
159.8
68.2
169.4
163.5
63.3
173.2
163.0
58.8
173.0


243.4
122.3
246.7
258.0
116.4
260.7
266.8
104.6
268.8
214.2
101.4
266.7
220.3
95.6
276.4
220.0
90.0
278.5









Table 8-5 Computed stress ratio in the concrete and allowable number of 24-kip single axle
loads under critical loading conditions for the test sections evaluated in this study.
e Slab Stress Stress-strength # of Repetitions of 24-kip
Phase
Thickness (psi) Ratio Axle Loads to Failure
4" 568.3 0.675 3,810
I-a 5" 531.4 0.631 13,231
6" 488.6 0.580 56,178


555.1
486.9
416.2

476.0
400.4
318.4


0.659
0.578
0.494

0.565
0.476
0.378


5,958
59,416
no limit

85,963
no limit
no limit


Table 8-6 Level of tensile
corner


Temp.
Diff

-10
0
+20


12 kips,
linear
analysis
285
198
320


stresses in the 4" slab that failed in Phase I-a. Load applied at the


15 kips,
linear
analysis
289
248
370


18 kips,
linear
analysis
293
297
419


21 kips,
linear
analysis
316
347
468


21 kips,
non-linear
analysis
355


528




























72"


4 "a
48" 48" 48"


Real Situation in the Road


48" 48" 48"


Critical condition in the 3D Model


Figure 8-1 Axle load positioned on slabs with 4-ft joint spacing.


72"


72"


72" 72"


_ 72" 72"


Real Situation in the Road Critical condition in the 3D Model
Figure 8-2 Axle load positioned on slabs with 6-ft joint spacing.


72" 1












Stress Comparison Effect of the AC Elastic Modulus
4"- 6'x6' slab Bonded condition Load at the mid-edge


300,000


700,000

MR (psi)


1,100,000


* Stress is at the Bottom
unless Top (T) is indicated


-4-dT=-10 ---dT=0 -A-dT=+20


Figure 8-3 Effect of AC elastic modulus on maximum tensile stress in concrete caused by a 24-
kip axle load at mid-edge of 4-inch bonded concrete slabs with 6 ft joint spacing.





Stress Comparison Effect of the AC Elastic Modulus
5"- 4'x4' slab Bonded condition Load at the mid-edge


600

500

400

300

200

100


300,000


700,000

MR (psi)


1,100,000


* Stress is at the Bottom
unless Top (T) is indicated


---dT=-10 ---dT=0 -A-dT=+20


Figure 8-4 Effect of AC elastic modulus on maximum tensile stress in concrete caused by a 24-
kip axle load at mid-edge of 5-inch bonded concrete slabs with 4 ft joint spacing.


500
400

300

200

100
0


I


-T- -
--I


---- ------------- --












Stress Comparison Effect of Temperature Differential
Bonded condition 6' x 6' slab Load at the mid-edge


600
500
400
300
200
100
0


Slab Thickness


I-10 F 0 F +20 F

Figure 8-5 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at mid-edge of bonded slabs with 6 ft joint spacing.






Stress Comparison Effect of Temperature Differential
Bonded condition 6' x 6' slab Load at the corner


350
300
S250
200
150
, 100
50
0


Slab Thickness


E -10F 0 F +20 F

Figure 8-6 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at corner of bonded slabs with 6 ft joint spacing.












Stress Comparison Effect of Temperature Differential
Bonded condition 4' x 4' slab Load at the corner


400
-.4U

300
S200
C,
100

0


Slab Thickness


o-10F 0F +20F

Figure 8-7 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at mid-edge of bonded slabs with 4 ft joint spacing.







Stress Comparison Effect of Temperature Differential
Bonded condition 4' x 4' slab Load at the mid-edge


600
500
400
300
200
100
0


Slab Thickness


BE-10F 00F 0+20F

Figure 8-8 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at corner of bonded slabs with 4 ft joint spacing.












Stress Comparison Effect of Temperature Differential
Unbonded condition 6' x 6' slab Load at the mid-edge


500

400
S300

S200
C,
--
100

0


Slab Thickness


E -10 F 2 0 F +20 F

Figure 8-9 Effect of temperature differential on maximum tensile stresses in concrete caused by
a 24-kip single axle load at mid-edge of partially bonded slabs with 6 ft joint spacing.





Stress comparison Effect of Slab Size
Bonded Condition Load at the mid-edge


600

500

S400

S300

200

100


Slab Thickness (in)


* Stress is at the Bottom
unless Top (T) is indicated


-- 6'x6', T=-10
----- 4'x4', T=-10


--- 6'x6', T=0
----- 4'x4', T=0


--6'x6', T=20
---&-- 4'x4', T=20


Figure 8-10 Effect of slab size on the maximum tensile stresses in concrete caused by a 24-kip
single axle load at mid-edge of bonded slabs.


--''-



.T .........-----.---- .
T-T--












Stress comparison Effect of Slab Size
Bonded Condition Load at the corner



------ -



..-------- ----------------
. . . . m . . . . .*


Slab Thickness (in)


* Stress is at the Bottom
unless Top (T) is indicated


-- 6'x6', T=-10
--_-- 4'x4', T=-10


--- 6'x6', T=0
----- 4'x4', T=0


-- 6'x6', T=20
---i-- 4'x4', T=20


Figure 8-11 Effects of slab size on the maximum tensile stresses in concrete caused by a 24-kip
single axle load at corner of bonded slabs.





Stress comparison for the 6" slab
Load applied at the mid-edge


600
500
400
300
200
100


Temperature Differential (F)


* Stress is at the Bottom
unless Top (T) is indicated


- 6" 6' x 6' Bonded E 6" 6' x 6' Unbonded


Figure 8-12 Effects of interface condition on maximum tensile stresses in concrete caused by a
24-kip single axle load at mid-edge of 6-inch slabs with 6 ft joint spacing.


450
400
350
300
250
200
150
100
50
0


...------
T













Stress comparison for the 6" slab
Load applied at the corner


Temperature Differential (F)


* Stress is at the Bottom
unless Top (T) is indicated


- 6" 6' x 6' Bonded - 6" 6' x 6' Unbonded


Figure 8-13 Effects of interface condition on maximum tensile stresses in concrete caused by a
24-kip single axle load at corner of 6-inch slabs with 6 ft joint spacing.





Stress comparison Effect of bond in the interface
Load applied at the mid-edge


600
500
400 -
300 -
200 -
100


0 I
* Stress is at the Bottom
unless Top (T) is indicated


Slab Thickness


S-10 F- Phase II 0 F-F-Phase II
- -A- --10 F Fully bonded -- -OF- Fully bonded


--+20 F Phase II
- +20 F Fully bonded


Figure 8-14 Effects of interface condition on maximum stresses in concrete caused by a 24-kip
single axle load at mid-edge of slab for the test sections in Phase II.


250
200
150
100


... ... -
-


~'"I;'-------__.













Shear Stress in the interface
Bonded Condition 6' x 6' slabs


Slab Thickness (in)


-A-E=300,000 psi, Load at the edge
- -- E=300,000 psi, Load at the corner


-u-E=1,100,000 psi, Load at the edge
------ E=1,100,000 psi, Load at the corner


Figure 8-15 Maximum shear stresses at the interface caused by a 24-kip single load at a
temperature differential of +20 F for the bonded slabs with 6 ft joint spacing.





Shear Stress in the interface
Bonded Condition -4' x 4' slabs


4" 5" 6"
Slab Thickness (in)

-A- E=300,000 psi, Load at the edge -m-E=1,100,000 psi, Load at the edge
---A- E=300,000 psi, Load at the corner ...- -..- E=1,100,000 psi, Load at the corner


Figure 8-16 Maximum shear stresses at the interface caused by a 24-kip single load at a
temperature differential of +20 F for the bonded slabs with 4 ft joint spacing.


90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0


A--- .
S..... I---....--. ---.-.
A ........ ................. ..


--------------------------------------------------


---------------------
A J

















300

250

200

150

100


Tensile Stress in the AC layer
Bonded Condition 6' x 6' slabs +20 F Temp diff.


.-. ..-- -- ..-. . . . . .

------------------ ------------------------------- --








4" 5" 6"
Slab Thickness (in)

-A-E=300,000 psi, Load at the edge -u-E=1,100,000 psi, Load at the edge
- -A- E=300,000 psi, Load at the corner -- E=1,100,000 psi, Load at the corner


Figure 8-17 Maximum tensile stresses in the AC layer caused by a 24-kip single axle load at a
temperature differential of +20 F for the bonded slabs with 6 ft joint spacing.








Tensile Stress in the AC layer
Bonded Condition 4' x 4' slabs +20 F Temp diff.


300

250

u 200
-.
u 150

S100

50

0


Slab Thickness (in)


-A-E=300,000 psi, Load at the edge
- -A- E=300,000 psi, Load at the corner


-u-E=1,100,000 psi, Load at the edge
- ---- E=1,100,000 psi, Load at the corner


Figure 8-18 Maximum tensile stresses in the AC layer caused by a 24-kip single axle load at a
temperature differential of +20 F for the bonded slabs with 4 ft joint spacing.


---------------------------.....................----.......







S............................ ............................ *






















Figure 8-19 4" slab of the non-linear model curled up at corners due to temperature negative
temperature differential


Figure 8-20 4" slab of the non-linear model loaded with 21 kips in the corer and negative
temperature differential









CHAPTER 9
CONCLUSION

9.1 Summary of Findings

A full scale experiment was performed at the APT facility located at the FDOT Material

Research Park to evaluate the feasibility of using whitetopping (WT) pavements in Florida. A

total of nine instrumented WT test sections were constructed and tested using a Heavy Vehicle

Simulator (HVS). Analysis for both strain and temperature from the full scale experiment was

performed to evaluate pavement behavior. Laboratory testing was also performed to characterize

both material properties and pavement response. A 3-D finite element model was developed to

analyze the behavior of the WT pavement test sections. The model was verified and calibrated

using the measured FWD deflections and HVS load-induced strains from the test sections. The

model was then used to evaluate the potential performance of these test sections under a typical

critical temperature-load condition in Florida. A summary of the findings from this study is

presented in the following section.

9.1.1 Bond Strength at the Concrete-Asphalt Interface

In the construction of the test sections with a bonded concrete-asphalt interface (in Phases

I-a and I-b), the asphalt surface was milled, cleaned and sprayed with water before the placement

of concrete. This method was found to produce excellent bonding at the interface. The average

shear strength from the Iowa shear test on the cores from these test sections was 220 psi for

Phase I-a, and 195 psi for Phase I-b. The maximum computed shear stress at the interface under

the critical temperature-load condition for all cases is only 85 psi. No critical combination of

tensile stress and shear stress was found at the interface.

In the construction of the test sections with an unbonded concrete-asphalt interface (in

Phase II), a white-pigmented curing compound was sprayed on the surface of the asphalt to act









as a debonding agent before the placement of concrete. Results of Iowa shear test on the cores

from these test sections indicated an average shear strength of 119 psi before the HVS loading

and 135 psi after the HVS loading. This indicates that partial bonding existed at the interface

though an unbonded condition was intended, and that the bonding improved with additional

loading on the pavement.

Analysis of the strain and temperature data from the test track demonstrated that no curl

occurred in the slabs in either of the phases. The bond strength in the interface was of such

magnitude that the concrete slab was always in contact with asphalt even where there was a

severe temperature differential in the concrete slab.

9.1.2 Development of the 3-D Finite Element Model

A 3-D finite element model, as described in Chapter 6 of this report, was developed for the

analysis of WT pavements. The model was verified and calibrated with the measured FWD

deflections and HVS load-induced strains. It was found that the bonded interface (as existed in

the test sections in Phases I-a and I-b) could be modeled well by modeling the concrete as

perfectly bonded to the asphalt. The partially-bonded interface (as existed in the test sections in

Phase II) could be modeled well by vertical and horizontal springs connecting the concrete layer

with the asphalt layer.

9.1.2.1 Deflection-based calibration

For the fully bonded condition in the interface, the model could satisfactorily replicate the

deflection basin from FWD test by mainly considering material properties obtained from

standard laboratory test. For the partially bonded condition, the model could replicate the

deflection basin by adjusting mainly the value of the vertical springs in the interface.









9.1.2.2 Strain-based calibration

The model could satisfactorily replicate not only the peak strains but also the shape of the

strain profile in all analyzed cases. The model was able to replicate the strains in the fully bonded

condition when the effect of temperature in the AC layer was considered in the model. In the

case of the partially bonded condition, both the AC elastic modulus and the values of the springs

modeling the interface need to be fully considered in the model.

9.1.2.3 Load transfer

For the conditions at the time of the HVS loading on the test sections, it was found that

joints could be modeled well by vertical springs connecting the slabs at the joint. It is believed

that the bond in the interface plays a major role in transferring the load between adjacent slabs.

It is postulated that load transfer at the joints will eventually decrease with time. In the analysis

for long-term performance of the test sections under critical conditions, the worst joint condition

was assumed, and thus springs of zero stiffness were used to model the joint behavior in this

analysis.

9.1.2.4 Interface bond

The concrete slabs in Phase I were effectively well bonded to the asphalt layer, and those

from the Phase II were partially bonded. In both cases, the bond was strong enough to keep the

slab in contact with the asphalt layer, which justifies the use of linear springs in the interface for

the model in Phase II.

9.1.3 Stress analysis of the test sections

9.1.3.1 Effects of elastic modulus of AC

In both the analysis of the measured strain data and the results from the analytical model,

the elastic modulus of the AC layer was found to have great influence on the maximum tensile









stresses in the concrete slab. Thus, for the analysis for the most critical loading condition, the

lowest possible elastic modulus of the AC (at the highest temperature) was used.

9.1.3.2 Effects of concrete panel size

Maximum stresses in the concrete were found to decrease as the joint spacing decreases.

At the most critical loading condition, the concrete slabs with 4 ft joint spacing had lower

maximum stresses than those with 6 ft joint spacing.

9.1.3.3 Effects of bonded versus partially bonded interface

At the condition of negative temperature differentials in the concrete slab, the concrete

slabs with a partially bond interface were found to have higher maximum stresses than those

with a fully bonded interface. However, at the condition of zero or positive temperature

differential in the slab, the maximum stresses in the partially bonded slabs are about the same as

those in the fully bonded slabs.

9.1.4 Performance of the Test Sections

9.1.4.1 Potential performance

The verified and calibrated 3-D finite element model was used to evaluate the potential

performance of the nine test sections under a critical temperature-load condition. Maximum

tensile stresses in the pavement were computed for the critical condition when a 24-kip single

axle load (which is higher than the legal limit of 22 kips in Florida) was placed at the mid-edge

of the slab (which is the most critical loading position) and when the temperature differential in

the concrete slab was +20 F (which is a typical severe temperature condition in the summer

time in Florida.)

The maximum computed stresses in the concrete slabs were all below the flexural strength

of the concrete for all the 9 test sections. Based on the computed maximum stresses in the

concrete, the expected numbers of repetitions of the 24-kip single axle loads at the critical









thermal condition were computed for the nine test sections. The results show that the 4-inch

slabs can be used for heavy (24-kip single axle) load but only for low-volume traffic condition.

The allowable traffic volume increases as the concrete slab thickness increases. In order to be

able to withstand the critical load without fear of fatigue failure (for an infinite number of critical

load repetitions), a minimum slab thickness of 6 inches would be needed for a joint spacing of 4

ft, and a minimum slab thickness of 8 inches would be needed for a joint spacing of 6 ft.

9.1.4.2 Mechanism of failure

The actual performance of the 4" slab that failed during the HVS testing was evaluated

using the 3D analytical model. A variation of the model with nonlinear springs in the interface

was used to model the situation when the slab becomes unbonded due to temperature effect. The

results of the analysis indicated that the slab had become unbonded when it failed.

9.2 Limitations of the Research

Some of the limitations of the research are as follows:

1. During the loading process, a significant portion of the test sections were under the

shade of the HVS. For this reason, the temperature differential measured in the experiment may

not represent the maximum achievable in a real pavement.

2. Only one of the test sections was loaded until failure, and this required a high number of

load repetitions using a very heavy wheel load. The analysis of the performance of the actual WT

pavement was limited to this test section and it may not represent the general case.

3. Only two cases of bond conditions were considered in the full scale experiment, and

even though they were intended to represent the extreme conditions of bond strength (fully

bonded and fully un-bonded), both of them were in the range of partially bonded condition.









4. The structural condition of the asphalt layer under the concrete slabs in the test track can

be considered between fair and good at the time of the HVS loading. This condition might not be

representative of deteriorated asphalt layers considered as candidates for WT resurfacing.

5. The stress estimation using the analytical model should be considered only as

representative of the test sections loaded in the full scale experiment. More field experimentation

needs to be done to extrapolate these results to other field conditions.

6. Due to the high demand on the use of the HVS, it was not possible to load the test

sections for an extended period of time to evaluate the long-term performance of the WT

pavements and the modes of failures under actual traffic and weather conditions.

7. The analysis of the composite pavement did not consider the dynamic effect of the

wheel load traveling at speeds faster than the one use in the HVS testing.

9.3 Recommendations and future work

The developed 3-D finite element model is recommended for use for analysis of WT pave-

ments subjected to load and temperature effects. The model parameters needed in the analysis

include the elastic moduli of the concrete, AC, base and effective subgrade layers. For analysis

of long-term behavior, the joint stiffness can be assumed to be zero. For partially bonded

interface condition, the stiffness values of the springs for modeling the interface are also needed

as model parameters. The elastic moduli of concrete and AC can be determined by testing in the

laboratory, while the other parameters can be determined through the back-calculation method of

matching analytical deflections and strains due to an applied load with measured values.

It is recommended that experimental WT pavement test sections of various designs be

constructed on actual roadways in Florida to evaluate their behavior and performance under

actual environmental and traffic conditions. The experimental pavement sections will be instru-

mented for monitoring of temperature and strains on a long-term basis. This will enable the









monitoring of the behavior of the WT pavements under critical load and temperature conditions,

and the verification of the predicted response from the analytical model. It will also enable the

evaluation of the long-term behavior of the WT pavements under actual traffic and weather

conditions.

In addition, the following recommendations for further research in whitetopping

pavements are offered:

3D Finite element micromodel to study in more detail the interaction between the
concrete slab and the asphalt layer in the interface

Development of laboratory tests to estimate both the shear and tensile strength in the
interface

Full scale experiments that consider many types of interface bond from fully bonded to
fully unbonded condition and also different conditions in the asphalt layer

HVS test long enough to produce cracking in the composite pavement so that design
guidelines can be developed based on observed failure.

A computer model to analyze the dynamic effect of the wheel traveling at typical speeds
in highways.











LIST OF REFERENCES


American Association of State Highway and Transportation Officials, (1993). "Guide for Design
of Pavement Structures," Washington, DC.

American Concrete Pavement Association, (1998). "Whitetopping State of the Practice,"
ACPA Engineering Bulletin EB210P, Skokie, IL.

Armaghani, J.M. and Tu, Diep. (1999). "Rehabilitation of Ellaville Weigh Station with Ultra-
Thin Whitetopping," Transportation Research Record 1654, Transportation Research
Board, National Research Council, Washington, DC, pp. 3-11.

Brown, D. (1995). "Ultra-Thin Whitetopping Emerges as Rehab Technique," Transportation
Builder, V7, No. 1, Jan. 1995, pp 37-41.

Cable, J.K., Grove, J.D., and Heyer, M. (1997). "Ultra-thin Pavements Making the Grade,"
Proceedings, Sixth International Purdue Conference on Concrete Pavement Design and
Materials for High Performance, Volume II, Indianapolis, IN, pp. 245-266.

Cable, J.K, (1998). "Iowa Ultra-thin Whitetopping Research, A Performance Update," Paper
Presented at 1998 Transportation Research Board Annual Meeting, Washington DC.

Cable, J.K. and Ciha, T. (2001). "The Ultra-thin Whitetopping Option," Proceedings of the
Seventh International Conference on Concrete Pavements, Vol. 2, Orlando, Florida,
September, pp. 969-975.

Cole, L.W., and Mohsen, J.P. (1993). "Ultra-Thin Concrete Overlays on Asphalt," Paper
prepared for presentation at the 1993 TAC Annual Conference, Ottawa, Ontario, Canada.

Cole, L.W., Sherwood, J., Qi, X. (1999). "Accelerated Pavement Testing of Ultra-Thin
Whitetopping," Accelerated Pavement Testing International Conference, Reno.

Cown, R.M. (1993). "Experimental Concrete Inlay on Existing Asphalt Pavement, Georgia,"
Department of Transportation, Office of Materials and Research, Concrete Branch, Forest
Park, GA.

Dumitru, N. I., Hossain, M., and Wojakowski, J. (2002). "Construction and Performance of
Ultra-Thin Whitetopping in Kansas," Paper Presented at the 2002 Transportation Research
Board Annual Meeting, Washington, DC.

Edwards, W. F. and Sargand, S. M., (1999). "Response of an Ultra-Thin Whitetopping Pave-
ment to Moving Wheel Loads," Accelerated Pavement Testing International Conference,
Athens, Ohio.

Galal, K. A., Newbolds, S.A., Olek, J., Weiss, W. J., and Nantung, T. (2004). "Stress and Strain
Analysis of Ultra-Thin Whitetopping over Composite Pavement Section using Accelerated
Pavement Testing," Paper Presented at 2004 Transportation Research Board Annual
Meeting, Washington, DC.









Hutchinson, R.L., (1982). "Resurfacing with Portland Cement Concrete," Synthesis of Highway,
Practice, NCHRP 99, Transportation Research Board, Washington, DC.

Khazanovich, Lev, Gotlif, Alex. (2002). "ISLAB2000 Simplified Friction Model." Paper
Presented at 2002 Transportation Research Board Annual Meeting, Washington, DC.

Mack, J.W., Wu, C.L., Tarr, S.M., and Refai, T. (1997). "Model Development and Interim
Design Procedure Guidelines for Ultra-thin Whitetopping Pavements," Proceedings, Sixth
International Purdue Conference on Concrete Pavement Design and Materials for High
Performance, Volume I, Indianapolis, IN, pp. 231-256.

McGhee, Kenneth, H. (1994). "Portland Cement Concrete Resurfacing," Synthesis of Highway
Practice, NCHRP 204, Transportation Research Board, Washington, DC.

Middleton, Brent, Fall, Lynne, Day, Robert. (2005). "Durability and mechanical Properties of
High Performance Concrete for Ultra-Thin Whitetopping Pavements." Paper Presented at
2005 Transportation Research Board Annual Meeting, Washington, DC.

Nelson, Patricia K., Rasmussen, Robert O., (2002). "Delamination Stresses at the Interface of
Bonded Concrete Overlays." Paper Presented at 2002 Transportation Research Board
Annual Meeting, Washington, DC.

Nishiyama, T., Lee, H., Asghar, B.M., (2005). "Investigation of Bonding Condition in Concrete
Overlay by Laboratory Testing, Finite Element Modeling and Field Evaluation,"
Transportation Research Record No. 1933, Washington, DC, pp. 15-23.

Nishizawa, Tasuo, Murata, Yoshiki, Kokubu, Katsuro. (2003). "Mechanical behavior of Ultra-
Thin whitetopping structure under stationary and moving loads," Transportation Research
Record No. 1823, Washington, DC, pp. 102-110.

Portland Cement Association, (1984). "Thickness Design for Concrete Highway and Street
Pavements," Publication No. EB109.01P, Skokie, IL .

Rajan, S., Olex, J., Robertson, T.L., Galal, K., Nantung, T., and Weis, J. (2001). "Analysis of
Performance of Ultra-Thin Whitetopping Subjected to Slow Moving Loads in an
Accelerated Pavement Testing Facility," 7th International Conference on Concrete
Pavements, Orlando, Florida.

Rasmussen, Robert O., Rozycki, Dan K. (2001). "Characterization and Modeling of Axial Slab-
Support Restraint," Paper Presented at 2001 Transportation Research Board Annual
Meeting, Washington, DC.

Risser, R.J., LaHue, S.P., Voigt, G.F., and Mack, J. (1993). "Ultra-Thin Concrete Overlays on
Existing Asphalt Pavement," 5th International Conference on Concrete Pavement Design
and Rehabilitation, Vol. 2, April, Purdue University, Lafayette, IN., pp. 247-254.

Saeed A., Hammons, M.I., and Hall, Jr., J.W. (2001). "Design, Construction, and Performance
Monitoring of Ultra-thin Whitetopping at a General Aviation Airport," proceedings, 2001
ASCE Airfield Pavement Specialty Conference, American Society of Civil Engineers,
Chicago, Illinois.









Saeed, A. and Hall, Jr., J.W. (2001). "Non-destructive Pavement Evaluation and Design of Ultra-
thin Whitetopping at a General Aviation Airport in Tennessee," Proceedings Second
International Conference on Maintenance and Rehabilitation of Pavements and Technical
Control, Auburn University, Alabama.

Speakman, J., and Scott, III, H. (1996). "Ultra-Thin, Fiber-Reinforced Concrete Overlays for
Urban Intersections," Transportation Research Record 1532, Advancements in Concrete
Materials Technology, TRB, National Research Council, Washington, DC.

Sprinkel, M.M., and Ozyildirim C., (1999). "Evaluation of the Installation and Initial Condition
of Hydraulic Cement Concrete Overlays Placed on Three Pavements in Virginia," Interim
Report. Report No. VTRC 99-IR3, April.

Tarr, Scott M., Sheehan, Mattew J., Ardani, Ahmad. (2000). "Mechanistic Design of Thin
Whitetopping Pavements in Colorado," Transportation Research Record No. 1730,
Washington, DC., pp. 64-72.

Tia, M., and Kumara, W. (2003). "Evaluation of performance of Ultra-thin whitetopping by
means of Heavy vehicle simulator (Analysis, Planning and Design phase)." Final Report
UF Project No: 49104504863-12. Gainesville, Florida.

Tia, M., Wu, C.L., Kumara, W. (2002). "Forensic Investigation of the Ellaville Weigh Station
UTW Pavements", UF Project No: 49104504831-12. Gainesville, Florida.

Tritsch, S. (1995). "Whitetopping, Technique Revives Burgeoning Kansas Thoroughfare,"
Roads and Bridges, September, pp. 52-55.

Vandenbossche, J.M. and Fagerness, A.J., (2002). "Performance and Repair of Ultra-Thin
Whitetopping: The Minnesota Experience," Transportation Research Record No. 1809,
Washington, DC., pp.191-198.

Winkelman, Thomas J., (2005). "The Illinois Whitetopping Experience: A Practical Approach,"
Proceeding International Conference on Best Practices for Ultra-thin and Thin
Whitetopping, Denver, Colorado.

Wu, C.L., Tarr, S.M., Refai, T.M., Nagi, M.N., and Sheehan, M.J. (1997). "Development of
Ultra-Thin Whitetopping Design Procedure," Report prepared for Portland Cement
Association (PCA), PCA Serial No. 2124, Skokie, Illinois, January.

Wu, C.L., Tayabji, S.D., Sheehan, M.J., and Sherwood, J. (2001). "Performance and Repair of
UTW Pavements," Proceedings of the Seventh International Conference on Concrete
Pavements, Vol. 2, Orlando, Florida, September, pp. 839-856.

Wu, C.L., and Sheehan, M.J. (2002). "Testing and Performance Evaluation of UTW Pavements
at the Spirit of St. Louis Airport," Transportation Research Record No. 1809, Washington,
DC., pp. 218-227.









BIOGRAPHICAL SKETCH

Patricio Tapia was born in 1965 in Calama, Chile. He is the son of Nelson Tapia and

Carmen Gutierrez. He graduated with a bachelor degree from the Department of Civil

Engineering of the Universidad Catolica del Norte, Chile in 1992. At the same time he received

his PE license. One year after graduation he was recruited as an assistant professor in the

Department of Civil Engineering at the Universidad Catolica del Norte, Chile. In 1992 he

married Juana Mora. In 1993 his son Pablo was born. In 2002, after teaching for 10 years and

attending non-degree program in Japan and USA, he was granted the Fulbright-Laspau

Scholarship to pursuit doctoral studies in the USA. The same year, he enrolled in the Ph.D.

program of the Department of Civil and Coastal Engineering at the University of Florida. In May

2004 he received his master's degree in Engineering from UF. In August 2004 he earned a

teaching assistantship in the material division in the Department of Civil and Coastal

Engineering at UF. In 2006 he received the "Outstanding International Student" award for

academic excellence at UF. The same year he was proposed as a department candidate for the

Teaching Assistant award.





PAGE 1

1 ANALYSIS, TESTING AND VERIFICATION OF THE BEHAVIOR OF COMPOSITE PAVEMENTS UNDER FLORIDA CONDIT IONS USING A HEAVY VEHICLE SIMULATOR By PATRICIO ENRIQUE TAPIA GUTIERREZ A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Patricio Enrique Tapia Gutierrez

PAGE 3

3 To my wife Juana, my son Pablo, Gabr iela, Ana, and my family in Chile

PAGE 4

4 ACKNOWLEDGMENTS I thank m y adviser for his constant support dur ing the course of this investigation. His knowledge and dedication was crucial in the success of this researc h. I also thank to all members of the committee for their valuable contribution to this disserta tion. The Florida Department of Transportation (FDOT) is also gratefully ac knowledged for providing the financial support for this research. The FDOT Materi al Office provided the necessary testing equipment, materials and personnel for this inves tigation. Personnel of the Depart ment of Civil and Coastal Engineering who helped in conducting laboratory tests are also gratef ully acknowledged. The Universidad Catolica del Norte which provided fina ncial support for my staying in the USA and the Fulbright Commission in Chile which granted the Fulbright-Laspau Scholarship, are also gratefully acknowledged. Finally I would like to thank my family here and in Chile and my colleagues in the Department of Civil Engi neering at UCN for their constant support.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES.......................................................................................................................12 ABSTRACT...................................................................................................................................22 CHAP TER 1 INTRODUCTION..................................................................................................................24 1.1 Research Need..................................................................................................................24 1.2 Problem Statement.......................................................................................................... ..26 1.3 Research Hypothesis.........................................................................................................26 1.4 Objectives of Research.....................................................................................................27 1.5 Approach and Scope of Research..................................................................................... 27 1.6 Significance of Research.................................................................................................. 28 2 LITERATURE REVIEW ON WHITETOPPING .................................................................. 30 2.1 General Concepts........................................................................................................... ...30 2.1.1 Ultra-Thin Whitetopping........................................................................................31 2.1.2 Thin Whitetopping..................................................................................................31 2.1.3 Conventional Whitetopping.................................................................................... 32 2.2 Concrete Mixture Pro portions and Properties .................................................................. 32 2.3 Construction Procedures...................................................................................................34 2.4 UTW and TWT Design Considerations........................................................................... 35 2.4.1 Slab Thickness........................................................................................................ 35 2.4.2 Joint Spacing.......................................................................................................... 35 2.4.3 Interface Bonding Strength..................................................................................... 36 2.5 Design Procedure for UTW and TWT Pavements........................................................... 37 2.6 Performance of UTW and TWT Projects.........................................................................40 2.7 Accelerated Pavement Testi ng and Field Testing of UTW .............................................. 42 2.8 Analytical Models.......................................................................................................... ...44 3 INSTRUMENTATION AND CONSTRUC TION OF TEST SECTIONS ............................ 48 3.1 Description of the Testing Phases.....................................................................................48 3.2 Layout of the Test Sections..............................................................................................49 3.2.1 Phase I....................................................................................................................49 3.2.2 Phase II................................................................................................................. ..49 3.3 Layout of the Instrumentation.......................................................................................... 50 3.3.1 Wheatstone Bridge Circuits.................................................................................... 50 3.3.2 Preliminary Stress Analysis.................................................................................... 50

PAGE 6

6 3.3.3 Instrumentation Layout.......................................................................................... 51 3.3.3.1 Phase I..........................................................................................................51 3.3.3.2 Phase II......................................................................................................... 53 3.4 HVS Loading Plan............................................................................................................54 3.5 Data Collection.................................................................................................................54 3.6 Construction of the Test Tracks........................................................................................ 55 3.6.1 Construction of Concrete Test Tracks in Phase I ................................................... 55 3.6.1.1 Asphalt surface preparation and formwork.................................................. 55 3.6.1.2 Concrete mix proportions............................................................................. 55 3.6.1.3 Placement of concrete.................................................................................. 56 3.6.1.4 Placement of strain gages............................................................................. 56 3.6.1.5 Placement of thermocouples........................................................................ 57 3.6.2 Construction of Concrete Test Tracks in Phase II ..................................................57 3.6.2.1 Asphalt surface preparation and formwork.................................................. 57 3.6.2.2 Installation of strain gages and therm ocouples and concrete placement...... 58 4 MATERIALS AND PAVEMENT CHARACTERIZATION ................................................ 74 4.1 Materials Characterization................................................................................................ 74 4.1.1 Interface Bond Strength..........................................................................................74 4.1.1.1 Results from test sections in Phase I-a......................................................... 74 4.1.1.2 Results from test sections in Phase I-b......................................................... 75 4.1.1.3 Results from test sections in Phase II........................................................... 75 4.1.1.4 Comparison before and after HVS loading.................................................. 76 4.1.2 Concrete Properties................................................................................................76 4.1.2.1 Properties of concrete samp led from concrete trucks.................................. 76 4.1.2.2 Properties of concrete from core samples.................................................... 76 4.1.3 Asphalt Concrete Properties................................................................................... 77 4.2 Measurement of Joint Movement..................................................................................... 78 4.3 Measurement of Slab Profile Using a Dipstick................................................................ 78 4.4 FWD Tests.................................................................................................................. ......79 4.4.1 FWD Tests in Phase I-a.......................................................................................... 80 4.4.2 FWD Tests in Phase I-b.......................................................................................... 80 4.4.3 FWD Tests Phase II................................................................................................ 80 5 TESTING OF TEST SECTIONS AND DATA ANALYSIS.............................................. 102 5.1 HVS Loading of Test Sections.......................................................................................102 5.1.1 HVS Loading of Test Sections in Phase I-a......................................................... 102 5.1.2 HVS Loading of Test Sections in Phase I-b......................................................... 103 5.1.3 HVS Loading of Test Sections in Phase II........................................................... 104 5.2 Analysis of Temperature Data........................................................................................104 5.2.1 Temperature Differential......................................................................................104 5.2.2 Temperature in the AC layer................................................................................ 107 5.2.3 Temperature Distribution.....................................................................................108 5.2.4 Summary of temperature analysis........................................................................108 5.3 Analysis of Strain Data................................................................................................... 109

PAGE 7

7 5.3.1 Dynamic Strain versus Static Strain..................................................................... 109 5.3.2 Measured strain an d calculated strain ................................................................... 110 5.3.3 Effect of Temperature on the Strain..................................................................... 112 5.3.4 Effect of the load magnitude on the strain............................................................113 5.3.5 Effect of the loadin g period on the strain ............................................................. 114 5.3.6 Evaluation of the bond condition using strain ratios............................................116 6 DEVELOPMENT OF A 3-D FINITE ELEMENT MODEL............................................... 158 6.1 Finite Element Program.................................................................................................. 158 6.2 Six-Slab and Twelve-Slab 3-D Finite Element Models................................................. 158 6.3 Solid 20-Node Finite Element........................................................................................159 6.4 Modeling of Concrete Slab Joints................................................................................... 160 6.5 Modeling of Materials....................................................................................................160 6.6 Modeling of Concre te-Asphalt Interface ........................................................................ 160 6.7 Modeling of Loads and Temperature Effects................................................................. 162 7 MODEL CALIBRATION AND VERIFICATION............................................................. 169 7.1 Overview of Model Calibration...................................................................................... 169 7.2 Deflection-Based Calibration of Model Parameters....................................................... 170 7.2.1 Phases I-a and I-b................................................................................................. 170 7.2.2 Phase II................................................................................................................. 172 7.3 Strain-Based Calibration of Model Parameters.............................................................. 174 7.3.1 General Approach................................................................................................. 174 7.3.2 Phase I-a............................................................................................................... 175 7.3.3 Phase I-b............................................................................................................... 177 7.3.4 Phase II................................................................................................................. 179 7.4 Summary of Calibration Results.....................................................................................180 8 EVALUATION OF POTENTIAL PERFOR M ANCE OF THE WT DESIGNS................. 211 8.1 Overview.........................................................................................................................211 8.2 Assumptions for the evaluation of the pot ential perform ance of the test sections......... 211 8.2.1 Critical loading conditions................................................................................... 211 8.2.2 Model parameters................................................................................................. 212 8.3 Results of Critical Stress Analysis.................................................................................. 212 8.3.1 Maximum stresses in the concrete slabs............................................................... 212 8.3.1.1 Effects of elastic m odulus of AC layer ...................................................... 213 8.3.1.2 Effects of temperature differential.............................................................213 8.3.1.3 Effects of panel size................................................................................... 214 8.3.1.4 Effects of bonded versus partially bonded interface.................................. 215 8.3.2 Maximum shear stresses at the interface.............................................................. 216 8.3.3 Maximum stresses in the AC layer....................................................................... 217 8.4 Potential performance of the test sections...................................................................... 218 8.5 Evaluation of the actual performance of the test sections.............................................. 219

PAGE 8

8 9 CONCLUSION..................................................................................................................... 237 9.1 Summary of Findings.....................................................................................................237 9.1.1 Bond Strength at the C oncrete-Asphalt Interface ................................................. 237 9.1.2 Development of the 3-D Finite Element Model................................................... 238 9.1.2.1 Deflection-based calibration......................................................................238 9.1.2.2 Strain-based calibration.............................................................................. 239 9.1.2.3 Load transfer.............................................................................................. 239 9.1.2.4 Interface bond.............................................................................................239 9.1.3 Stress analysis of the test sections........................................................................ 239 9.1.3.1 Effects of elastic m odulus of AC............................................................... 239 9.1.3.2 Effects of concrete panel size..................................................................... 240 9.1.3.3 Effects of bonded versus partially bonded interface.................................. 240 9.1.4 Performance of the Test Sections......................................................................... 240 9.1.4.1 Potential performance................................................................................ 240 9.1.4.2 Mechanism of failure................................................................................. 241 9.2 Limitations of the Research............................................................................................ 241 9.3 Recommendations and future work................................................................................242 LIST OF REFERENCES.............................................................................................................244 BIOGRAPHICAL SKETCH.......................................................................................................247

PAGE 9

9 LIST OF TABLES Table page 2-1 Concrete Mix Proportions Used in Louisvil le Experim ental Project (Riser et al 1993)..................................................................................................................................46 2-2 Concrete Mix Proportions Used in Leawood, Kansas (Wu et al. 1997)............................ 46 2-3 Concrete Mix Proportions Used in SNH, Tennessee (Saeed et al. 2002).......................... 46 2-4 Mix proportions for eight white topping projects in Illinois...............................................47 3-1 Mix Designs of Concrete Used in Phases I and II............................................................. 59 4-1 Results of the Iowa Shear Te sts on the cored samples from test sections in Phase I-a after HVS loading..............................................................................................................82 4-2 Results of Iowa Shear Tests on the cored sam ples from te st sections in Phase I-b after HVS loading..............................................................................................................83 4-3 Results of Iowa Shear Tests on cores fr om test sections in Phase II after HVS loading................................................................................................................................84 4-4 Summary of the interface bond streng th before and after HVS loading............................ 84 4-5 Properties of fresh concrete used....................................................................................... 85 4-6 Properties of hardened concrete sampled from truck in Phase I-a..................................... 85 4-7 Properties of the hardened concre te sam pled from truck in Phase II................................ 85 4-8 Results of Indirect Tensile Strength Test on the concrete sam ples taken from the test sections in Phase I-a after HVS loading............................................................................ 86 4-9 Results of Resilient Mod ulus and Indire ct Tensile Strength Tests on the asphalt concrete samples obtained from test sec tions in all phases after HVS loading................. 86 4-10 Results of Penetration and Absolute Visc osity T ests on the recovered asphalt binders from cores from Phase I-a after HVS loading................................................................... 87 5-1 HVS loading period and num ber of 12-kip wh eel passes on the test sections in Phase I-b. .......................................................................................................................... ..120 5-2 Summary of HVS loading on test sections in Phase II....................................................120 5-3 Number of hours in a day when the temper ature differential was in certain ranges for the 4-inch slab in Phase I-a.............................................................................................. 120

PAGE 10

10 5-4 Number of hours in a day when the temper ature differential was in certain ranges for the 5-inch slab in Phase I-a.............................................................................................. 120 5-5 Extrem e values for temperature different ial and temperature in the AC Layer.............. 121 5-6 Measured static and dynam ic strains for gage s 1, 2 and 5 in the 4-inch slab in Phase I-a caused by 9 and 12-kip loads......................................................................................121 5-7 Measured static and dynam ic strains for gage s 2 and 4 in the 5-inch slab in Phase I-a caused by 9 and 12-kip loads........................................................................................... 122 5-8 Measured static and dynam ic strains for gage s 2 and 4 in the 5-inch slab in Phase I-a caused by 15and 18-kip loads........................................................................................ 122 5-9 Verified depths of st rain gages in Phase I-b. ................................................................... 123 5-10 Verified depths of st rain gages in Phase II. ..................................................................... 123 5-11 Strain Ratios to evaluate th e degree of bond in the interface .......................................... 123 7-1 Extrem e values of the AC resilient modul us based on extreme temperature during the HVS test, using Eq. 7.1....................................................................................................182 7-2 Elastic m odulus and Poissons ratio of the pa vement materials used in the 3-D finite element model.................................................................................................................. 182 7-3 HVS loading periods for Phase I-a. .................................................................................182 7-4 HVS loading periods for Phase I-b. .................................................................................182 7-5 HVS loading periods for Phase II.................................................................................... 183 7-6 Summary of the best estimated parameters of the 3-D m odel for all test sections.......... 184 8-1 Model parameters of the 3-D model for each test section used in the analysis............... 222 8-2 Maxim um tensile stresses in the concrete sl abs caused by a 24-kip single axle Load at various critical loading conditions...............................................................................223 8-3 Maxim um shear stress at the concrete-a sphalt interface caused by a 24-kip single axle load at a temperature differen tial of +20 F for the bonded slabs............................ 224 8-4 Maxim um tensile stresses in the aspha lt concrete layer caused by a 24-kip single Axle load at various critical loading conditions.............................................................. 225 8-5 Computed stress ratio in the concrete a nd allowable number of 24-kip single axle loads under critical loading c onditions for the test sections evaluated in this study.......226

PAGE 11

11 8-6 Level of tensile stresses in the 4 slab th at failed in Phase I-a. Load applied at the corner...............................................................................................................................226

PAGE 12

12 LIST OF FIGURES Figure page 2-1 Transition between UTW and Adjoining Asphalt Pavem ent (ACPA, 1998).................... 47 3-1 Layout of the Test Sections for Phase I............................................................................. 60 3-2 Layout of the Test Sections for Phase II............................................................................ 60 3-3 Strain Gage Arrangements in a Half Bridge Circuit. ......................................................... 61 3-4 Connection of the Active and Dummy Stra in Gages in the Half Bridge Circuit. .............. 61 3-5 Instrumentation Layout for the Test Slabs in Phase I-a..................................................... 62 3-6 Vertical Positions of the Strain Gages in Phase I-a. ..........................................................63 3-7 Vertical Positions of th e Therm ocouples for the 4, 5 and 6 Slabs in Phase I............... 64 3-8 Instrumentation Layout for the Test Slabs in Phase I-b..................................................... 65 3-9 Vertical Positions of the Strain Gages in Phase I-b. ..........................................................66 3-10 Instrumentation layout for Phase II.................................................................................... 66 3-11 Vertical positions of the strain gages in Phase II. ..............................................................67 3-12 Vertical positions of thermocouples in Phase II................................................................ 67 3-13 Milled surface before concrete placem ent on Lane 6 in Phase I-a.................................... 68 3-14 Formwork prepared for Lane 7 in Phase I-b...................................................................... 68 3-15 After placement of concrete on Lane 6 in Phase I-a.......................................................... 69 3-16 Placement of top and bottom strain gages on Lane 7 in Phase I-b.................................... 69 3-17 Placement of surface strain ga ges at a join t in Phase I-a................................................... 70 3-18 Strain gages in a protective PVC pipe before placing concrete. ........................................70 3-19 Removal of concrete slabs from Lane 6 in Phase I-a......................................................... 71 3-20 Formwork for test slabs in Phase II................................................................................... 71 3-21 Grooves on asphalt surface for placem ent of strain gages and thermocouples cables in Phase II..........................................................................................................................72

PAGE 13

13 3-22 Asphalt surf ace with white curing compound before concrete placement in Phase II...... 72 3-23 Finishing of the concrete fo r the test track in Phase II. ..................................................... 73 3-24 Curing of concrete by sprinkling with water.....................................................................73 4-1 Cores samples from Lane 6 in Phase I-a after HVS loading............................................. 88 4-2 Location of the cores taken after loading in Phase I-a. ......................................................89 4-3 Location of the cores taken after load ing for each test section in Phase I-b. ..................... 90 4-4 Relationship between Temperature and Res ilient Modulus for the AC layer in the com posite pavement........................................................................................................... 90 4-5 The Whitmore gage with invar bar.................................................................................... 91 4-6 Measured gage spacing from a 4-inch slab in Lane 6........................................................ 92 4-7 Measured gage spacing from a 5-inch slab in Lane 6........................................................ 92 4-8 Measured gage spacing from a 6-inch slab in Lane 6........................................................ 93 4-9 Changes of joint spacing on a selected day....................................................................... 93 4-10 Grid marked on slabs for the Dipstick measurement......................................................... 94 4-11 The Dipstick instrument................................................................................................... ..94 4-12 Dipstick measurements at two critical temperatures......................................................... 95 4-13 FWD load and sensor locations for FW D Test at slab center in Phase I-a. ....................... 96 4-14 FWD load and sensor locations for FW D Test at slab corner in Phase I-a. ...................... 96 4-15 FWD load and sensor locations for FWD Te st at the slab corner in Phase I-b................. 97 4-16 FWD load and sensor locations for FWD Test at the slab edge in Phase I-b.................... 97 4-17 FWD load and sensor locations for FWD Test at the slab center in Phase I-b.................. 98 4-18 FWD Test at the slab corner and meas uring deflections on the opposite slab.................. 98 4-19 FWD Test at the mid-edge and measuri ng deflections in the loaded slab......................... 99 4-20 FWD Test at the center and measuring de flections along the tran sverse center line. ....... 99 4-21 FWD Testing Plan for the mid-edge and corner load in Phase II.................................... 100 4-22 FWD Testing Plan for the cen ter load at the two ends of the test track in P hase II........ 100

PAGE 14

14 4-23 Comparative analysis of load transfer factor for the test sections in Phase I-b and II. ....101 5-1 Corner cracks on 4-inch slabs in Phase I-a after 21-kip wheel loads. ............................. 124 5-2 Shrinkage cracks on a 4-inch test slab in Phase I-a.........................................................124 5-3 Shrinkage cracks on a 5-inch test slab in Phase I-a.........................................................125 5-4 Shrinkage cracks on a 6-inch concrete slab in Phase I-a................................................. 125 5-5 Temperature differential variation in the 4-inch slab in Phase I-a. ..................................126 5-6 Temperature differential variation in the 5-inch slab in Phase I-a. ..................................126 5-7 Temperature differential variation in the 6-inch slab in Phase I-a. ..................................127 5-8 Temperature differential variation in the 6-inch slabs in Phase I-b. ................................ 127 5-9 Temperature differential variation in the 5-inch slabs in Phase I-b. ................................ 128 5-10 Temperature differential variation in the 4-inch slabs in Phase I-b. ................................ 128 5-11 Temperature differential variation in the 10-inch slab in Phase II. ................................. 129 5-12 Temperature differential variation in the 8-inch slab in Phase II ....................................129 5-13 Temperature differential variation in the 6-inch slab in Phase II. ...................................130 5-14 Temperature variation on the surface of the as phalt layer for the 4 slab in Phase I-a... 130 5-15 Temperature variation on the surface of the as phalt layer for the 5 slab in Phase I-a... 131 5-16 Temperature variation on the surface of the as phalt layer for the 6 slab in Phase I-a... 131 5-17 Temperature on the surface of the AC la yer for the 6-inch slab in Phase I-b. ................. 132 5-18 Temperature on the surface of the AC layer for the 5 slab in Phase I-b. ....................... 132 5-19 Temperature on the surface of the AC layer for the 4 slab in Phase I-b. ....................... 133 5-20 Temperature on the surface of the AC la yer in the 10-inch slab in Phase II. .................. 133 5-21 Temperature on the surface of the AC layer in the 8-inch slab in Phase II. .................... 134 5-22 Temperature on the surface of the AC layer in the 6-inch slab in Phase II. .................... 134 5-23 Temperature distribution along th e depth of the 4 in concrete slab in Phase I-a at maximum positive temperature differential.....................................................................135

PAGE 15

15 5-24 Temperature distribution along the depth of the 5in concre te slab in Phase I-a at maximum positive temperature differential.....................................................................135 5-25 Temperature distribution along the depth of the 6in concre te slab in Phase I-a at maximum positive temperature differential.....................................................................136 5-26 Temperature distribution along the depth of the 4in concre te slab in Phase I-b at maximum temperature differentials................................................................................. 136 5-27 Temperature distribution along the depth of the 5in concre te slab in Phase I-b at maximum temperature differentials................................................................................. 137 5-28 Temperature distribution along the depth of the 6in concre te slab in Phase I-b at maximum temperature differentials................................................................................. 137 5-29 Temperature distribution along the depth of the 6in concre te slab in Phase II at maximum temperature differentials................................................................................. 138 5-30 Temperature distribution along the depth of the 8in concre te slab in Phase II at maximum temperature differentials................................................................................. 138 5-31 Temperature distribution along the depth of the 10in conc rete slab in Phase II at maximum temperature differentials................................................................................. 139 5-32 Comparison of dynamic and static strain for ga ge 1 in the 4-inch slab in Phase I-a....... 139 5-33 Comparison between static and dynamic strain for Gages 2 and 5 in the 4-inch slab in Phase I-a................................................................................................................... ....140 5-34 Measured dynamic and static strains at gage 2 in the 5-inch slab in Phase I-a ............... 140 5-35 Measured dynamic and static strains at gage 4 in the 5-inch slab in Phase I-a ............... 141 5-36 Measured strains at two differe nt depths at the m id-edge of the 6-inch slab in Phase I-a. ............................................................................................................................141 5-37 Zeroed strains profile at two different depths at the m id-edge of the 6-inch slab in Phase I-a...................................................................................................................... .....142 5-38 Strain in the com posite pavement as a function of time.................................................. 142 5-39 Effect of temperature differential on the p eak strain for the 6 slab in Phase I-a. ........... 143 5-40 Effect of temperature differential on the p eak strain for the 4 slab in Phase I-b. ..........143 5-41 Effect of temperature differential on the p eak strain for the 5 slab in Phase I-b. ..........144 5-42 Effect of the temperature differential on th e peak strain for the 6 slab in Phase I-b ..... 144

PAGE 16

16 5-43 Effect of the tem perature differential on the peak strain for the 6 slab in Phase II....... 145 5-44 Effect of the tem perature differential on the peak strain for the 8 slab in Phase II....... 145 5-45 Effect of the tem perature differential on the peak strain for the 8 slab in Phase II....... 146 5-46. Effect of AC tem perature on the peak st rain for the 6 slab in Phase I-a........................ 146 5-47 Effect of AC tem perature on the peak st rain for the 4 slab in Phase I-b........................ 147 5-48 Effect of AC tem perature on the peak st rain for the 5 slab in Phase I-b........................ 147 5-49 Effect of AC tem perature on the peak st rain for the 6 slab in Phase I-b........................ 148 5-50 Effect of AC tem perature on the peak strain for the 6 slab in Phase II.......................... 148 5-51 Effect of AC tem perature on the peak strain for the 8 slab in Phase II.......................... 149 5-52 Effect of AC tem perature on the peak st rain for the 10 slab in Phase II........................ 149 5-53 Relationship between strain and load in Phase II. Gages at the Mid-Edge. .................... 150 5-54 Relationship between strain and load in Phase II. Gages at the corner. ..........................150 5-55 Relationship between strain and load in Phase II. Gages in the AC layer. ...................... 151 5-56 Variation of peak strains during the HVS test in the 6-inch slab in Phase I-a. ................ 151 5-57 Variation of m aximum strains during the HVS test in the 4-inch slab in Phase I-a........ 152 5-58 Variation of m aximum strains in the 6-inch concrete slab and on the surface of the asphalt layer at Location 1 (mid edge of the slab) during HVS test in Phase I-b............ 152 5-59 Variation of peak strains in the 6-inch conc rete slab and on the surface of the asphalt layer at Location 2 (corner of the sl ab) during H VS test in Phase I-b............................. 153 5-60 Variation of m aximum strains in the 5-inch concrete slab and on the surface of the asphalt layer at Location 1 (mid-edge of the slab) during HVS te st in Phase I-b............ 153 5-61 Variation of peak strains in the 4-inch conc rete slab and on the surface of the asphalt layer at Location 2 (corner of the sl ab) during H VS test in Phase I-b............................. 154 5-62 Variation of peak strains in the 6-inch sl ab at Location 1 during HVS test in Phase II. .154 5-63 Variation of peak strains in the 6 slab at Location 5 during HVS test in P hase II......... 155 5-64 Variation of peak strains in the 8 -inch slab at Location 1 during HVS test in Phase II..155 5-65 Variation in peak str ains in the 8-inch slab at Location 5 during HVS test in Phase II..156

PAGE 17

17 5-66 Variation of peak strains in the 10-inch slab at Location 1 during HVS test in Phase II. ........................................................................................................................... .156 5-67 Variation of peak strains in the 10-inch Slab at Locati on 2 during HVS test in Phase II. ........................................................................................................................... .157 6-1 Six-slab 3-D finite element model................................................................................... 164 6-2 Twelve-slab 3-D finite element model............................................................................ 164 6-3 Mesh pattern in the XY Plane for the 6-slab model......................................................... 165 6-4 Mesh pattern in the XY plane for the 12-slab model....................................................... 166 6-5 Twenty-node 3D solid element used in the analytical model.......................................... 166 6-6 Springs to model load transf er at concrete slab joints. .................................................... 167 6-7 Springs in th e concrete-a sphalt interface to model th e partial bond condition................ 167 6-8 Non-linear s prings to model the full y un-bonded condition in the interface................... 168 7-1 Matching of deflection basin in the long itudinal direction caused by a 12-kip FWD load applied to the center of a 4 slab in Phase I-a. .........................................................185 7-2 Matching of deflection basin in the long itudinal direction caused by a 12-kip FWD load applied to the center of a 6 slab in Phase I-b. .........................................................185 7-3 Matching of deflection basin in the transverse direction caused by a 12-kip FW D load applied to the center of a 6 slab in Phase I-b..........................................................186 7-4 Matching of deflection basin in the long itudinal direction caused by a 12-kip FWD load applied to the center of a 4 slab in Phase I-b. .........................................................186 7-5 Matching of deflection basin in the transverse direction caused by a 12-kip FW D load applied to the center of a 4 slab in Phase I-b..........................................................187 7-6 Matching of deflection basin along the edge of loaded slab caus ed by a 12-kip FWD load applied to the corner of a 4 slab in Phase I-a. ........................................................187 7-7 Matching of deflection basin along the e dge of unloaded slab caused by a 12-kip FW D load applied to the corner of a 4 slab in Phase I-a................................................ 188 7-8 Matching of deflection basin along the edge of loaded slab caus ed by a 12-kip FWD load applied to the corner of a 4 slab in Phase I-b. ........................................................188 7-9 Matching of deflection basin along the edge of loaded slab caus ed by a 12-kip FWD load applied to the corner of a 5 slab in Phase I-b. ........................................................189

PAGE 18

18 7-10 Matching of deflection basin along the edge of loaded slab ca used by a 12-kip FWD load applied to the corner of a 6 slab in Phase II. ..........................................................189 7-11 Matching of deflection basin along the edge of loaded slab caus ed by a 12-kip FWD load applied to the m id-edge of a 6 slab in Phase II...................................................... 190 7-12 Matching of deflection basin along the e dge of unloaded slab caused by a 12-kip FW D load applied to the mid-edge of a 6 slab in Phase II............................................ 190 7-13 Matching of deflection basin along the edge of loaded slab caus ed by a 12-kip FWD load applied to the corner of an 8 slab in Phase II. ........................................................ 191 7-14 Matching of deflection basin along the e dge of unloaded slab caused by a 12-kip FW D load applied to the corner of an 8 slab in Phase II............................................... 191 7-15 Matching of deflection basin along the edge of loaded slab caus ed by a 12-kip FWD load applied to the m id-edge of an 8 slab in Phase II.................................................... 192 7-16 Matching of deflection basin along the e dge of unloaded slab caused by a 12-kip FW D load applied to the mid-edge of an 8 slab in Phase II.......................................... 192 7-17 Matching of deflection basin along the edge of loaded slab caus ed by a 12-kip FWD load applied to the corner of a 10 slab in Phase II. ........................................................ 193 7-18 Matching of deflection basin along the edge of loaded slab ca used by a 12-kip FWD load applied to the m id-edge of a 10 slab in Phase II.................................................... 193 7-19 Matching of deflection basin along the edge of unloaded slab caused by a 12-kip FW D load applied to the mid-edge of a 10 slab in Phase II.......................................... 194 7-20 Strain comparison at Gage 1 in the 6 slab in Phase I-a. ................................................. 194 7-21 Strain comparison at Gage 2 in the 6 slab in Phase I-a. ................................................. 195 7-22 Strain comparison at Gage 3 in the 6 slab in Phase I-a. ................................................. 195 7-23 Strain comparison at Gage 2 in the 5 slab in Phase I-a. ................................................. 196 7-24 Strain comparison at Gage 1 in the 4 slab in Phase I-a. ................................................. 196 7-25 Strain comparison at Gage 2 in the 4 slab in Phase I-a. ................................................. 197 7-26 Strain comparison at Gage 3 in the 4 slab in Phase I-a. ................................................. 197 7-27 Strain comparison at Gage 1 in the 6-inch slab in Phase I-b........................................... 198 7-28 Strain comparison at Gage 2 in the 6-inch slab in Phase I-b........................................... 198 7-29 Strain comparison at Gage 3 in the 6-inch slab in Phase I-b........................................... 199

PAGE 19

19 7-30 Strain comparison at Gage 1 in the 5-inch slab in Phase I-b........................................... 199 7-31 Strain comparison at Gage 3 in the 5-inch slab in Phase I-b........................................... 200 7-32 Strain comparison at Gage 2 in the 5-inch slab in Phase I-b........................................... 200 7-33 Strain comparison at Gage 2 in the 4-inch slab in Phase I-b........................................... 201 7-34 Strain comparison at Gage 3 in the 4-inch slab in Phase I-b........................................... 201 7-35 Strain comparison at top of Location 1 (m id edge) of the 10-inch slab in Phase II. ....... 202 7-36 Strain comparison at bottom of Location 1 (m id edge) of the 10-inch slab in Phase II. 202 7-37 Strain comparison on the surface of the AC layer at Location 1 (m id edge) of the 10inch slab in Phase II......................................................................................................... 203 7-38 Strain comparison at top of Location 5 (sla b corner) of the 10-inc h slab in Phase II. .... 203 7-39 Strain comparison at bottom of Location 5 (slab corner) of the 10inch s lab in Phase II.......................................................................................................................................204 7-40 Strain comparison on the surface of the AC layer at Location 5 (slab co rner) of the 10-inch slab in Phase II....................................................................................................204 7-41 Strain comparison at top of Location 1 (m id edge) of the 8-inch slab in Phase II. ......... 205 7-42 Strain comparison at bottom of Location 1 (m id edge) of the 8-inch slab in Phase II. ... 205 7-43 Strain comparison on the surface of the AC layer at Location 1 (m id edge) of the 8inch slab in Phase II......................................................................................................... 206 7-44 Strain comparison at top of Location 5 (sla b corner) of the 8-inch slab in Phase II. ...... 206 7-45 Strain comparison at bottom of Location 5 (s lab corner) of the 8-inch slab in Phase II.......................................................................................................................................207 7-46 Strain comparison on the surface of the AC layer at Location 5 (slab co rner) of the 8inch slab in Phase II......................................................................................................... 207 7-47 Strain comparison at top of Location 1 (m id edge) of the 6-inch slab in Phase II. ......... 208 7-48 Strain comparison at bottom of Location 1 (m id edge) of the 6-inch slab in Phase II. ... 208 7-49 Strain comparison on the surface of the AC layer at Location 1 (m id edge) of the 6inch slab in Phase II......................................................................................................... 209 7-50 Strain comparison at top of Location 5 (sla b corner) of the 6-inch slab in Phase II. ...... 209

PAGE 20

20 7-51 Strain comparison at bottom of Location 5 (s lab corner) of the 6-inch slab in Phase II.......................................................................................................................................210 7-52 Strain comparison on the surface of the AC layer at Location 5 (slab co rner) of the 6inch slab in Phase II......................................................................................................... 210 8-1 Axle load positioned on slabs with 4-ft joint spacing...................................................... 227 8-2 Axle load positioned on slabs with 6-ft joint spacing...................................................... 227 8-3 Effect of AC elastic m odulus on maximum tensile stress in concrete caused by a 24kip axle load at mid-edge of 4-inch bonde d concrete slabs with 6 ft joint spacing......... 228 8-4 Effect of AC elastic m odulus on maximum tensile stress in concrete caused by a 24kip axle load at mid-edge of 5-inch bonde d concrete slabs with 4 ft joint spacing......... 228 8-5 Effect of temperature differential on m aximu m tensile stresses in concrete caused by a 24-kip single axle load at mid-edge of bonded slabs with 6 ft joint spacing................ 229 8-6 Effect of temperature differential on m aximu m tensile stresses in concrete caused by a 24-kip single axle load at corner of bonded slabs with 6 ft joint spacing..................... 229 8-7 Effect of temperature differential on m aximu m tensile stresses in concrete caused by a 24-kip single axle load at mid-edge of bonded slabs with 4 ft joint spacing................ 230 8-8 Effect of temperature differential on m aximu m tensile stresses in concrete caused by a 24-kip single axle load at corner of bonded slabs with 4 ft joint spacing..................... 230 8-9 Effect of temperature differential on m aximu m tensile stresses in concrete caused by a 24-kip single axle load at mid-edge of partially bonded slab s with 6 ft joint spacing.............................................................................................................................231 8-10 Effect of slab size on the maximum tens ile stresses in concrete caused by a 24-kip single axle load at m i d-edge of bonded slabs.................................................................. 231 8-11 Effects of slab size on the maximum tens ile stresses in concrete caused by a 24-kip single axle load at corner of bonded slabs. ...................................................................... 232 8-12 Effects of interface condition on maximum tensile stre sses in concrete caused by a 24-kip single axle load at midedge of 6-inch slabs with 6 ft joint spacing.................... 232 8-13 Effects of interface condition on maximum tensile stre sses in concrete caused by a 24-kip single axle load at corner of 6-inch slabs with 6 ft joint spacing......................... 233 8-14 Effects of interface condition on maximum stress es in concrete caused by a 24-kip single axle load at mid-edge of slab for the test sec tions in Phase II............................... 233

PAGE 21

21 8-15 Maximum shear stresses at the inte rface caus ed by a 24-kip single load at a temperature differential of +20 F for th e bonded slabs with 6 ft joint spacing.............. 234 8-16 Maxim um shear stresses at the interf ace caused by a 24-kip single load at a temperature differential of +20 F for th e bonded slabs with 4 ft joint spacing.............. 234 8-17 Maxim um tensile stresses in the AC laye r caused by a 24-kip singl e axle load at a temperature differential of +20 F for th e bonded slabs with 6 ft joint spacing.............. 235 8-18 Maxim um tensile stresses in the AC laye r caused by a 24-kip singl e axle load at a temperature differential of +20 F for th e bonded slabs with 4 ft joint spacing.............. 235 8-19 4 slab of the non-linear model curled up at corners due to temperature negative tem perature differential.................................................................................................... 236 8-20 4 slab of the non-linear model loaded with 21 kips in the corner and negative tem perature differential.................................................................................................... 236

PAGE 22

22 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ANALYSIS, TESTING AND VERIFICATION OF THE BEHAVIOR OF COMPOSITE PAVEMENTS UNDER FLORIDA CONDIT IONS USING A HEAVY VEHICLE SIMULATOR By Patricio Enrique Tapia Gutierrez December 2007 Chair: Mang Tia Cochair: Fazil Najafi Major: Civil Engineering Whitetopping (WT) is a rehabilitation method to resurface deteriorated asphalt pavements. While some of these composite pavements have performed very well carrying heavy load, other have shown poor performance with early cr acking. With the objectiv e of analyzing the applicability of WT pavements under Florida co nditions, a total of nine full-scale WT test sections were constructed and tested using a He avy Vehicle Simulator (HVS) in the APT facility at the FDOT Material Research Park. The test se ctions were instrumented to monitor both strain and temperature. A 3-D finite element model wa s developed to analyze the WT test sections. The model was calibrated and verified using m easured FWD deflections and HVS load-induced strains from the test sect ions. The model was then used to ev aluate the potential performance of these test sections under critical temperature-load condition in Florida. Six of the WT pavement test sections had a bonded concrete-as phalt interface by milling, cleaning and spraying with water the asphalt surf ace. This method produced excellent bonding at the interface, with shear strength of 195 to 220 psi. Three of the test sections were intended to have an unbonded concrete-asphalt interf ace by applying a debonding agent in the asphalt

PAGE 23

23 surface. However, shear strengths between 119 a nd 135 psi and a careful an alysis of the strain and the temperature data indi cated a partial bond condition. The computer model was able to satisfactor ily model the behavior of the composite pavement by mainly considering material pr operties from standard laboratory tests and calibrating the spring elements used to model the interface. Reasonable matches between the measured and the calculated strains were achieved when a temperature-dependent AC elastic modulus was included in the analytical model. The expected numbers of repetitions of the 24-kip single axle loads at critical thermal condition were computed for the nine test se ctions based on maximum tensile stresses and fatigue theory. The results showed that 4 slabs can be used for heavy loads only for lowvolume traffic. To withstand the critical load wit hout fear of fatigue failure, 6 slabs and 8 slabs would be needed for joint spaci ngs of 4 and 6, respectively.

PAGE 24

24 CHAPTER 1 INTRODUCTION 1.1 Research Need The increasing truck weights and tire pressu res o n our pavements in recent years have pushed the demand on the performance of our pa vements to a higher level. Many asphalt pavements have experienced rutting while many ot hers have experienced longitudinal cracking. One of the possible solutions to this problem is the use of white topping (WT), which is placing a concrete layer over an existing asphalt paveme nt. Whitetopping has an advantage over an asphalt overlay in that the concrete surface is st ronger and thus is more resistant to rutting and surface-initiated cracking. The better durability and long-term performance characteristics of concrete pavement surfaces can significantly reduce traffic delays associated with the frequent maintenance of asphalt pavements. In addition, wh en concrete surfaces are used, skid resistance and safety can be substantially improved, especia lly under wet conditions. In recent years, with the sky-rocketing price of asphalt, concrete is becoming more competitive in cost with that of asphalt. This makes the use of whitetopping a more economically viable alternative for rehabilitation of asphalt pavements. There are three types of WT pavements based on the thickness of the concrete slab. Ultra-Thin Whitetopping (UTW) is a relative ly new technique for re surfacing deteriorated asphalt pavements. It involves pl acing very thin concrete slabs, 2 to 4 inches thick, on an old asphalt pavement to create a bonded (or partially bonded) composite pavement. The reduction of thickness is justified by the use of a high qual ity concrete, shorter joint spacing, and good bond between the concrete and the existing asphalt pavement. Thin Whitetopping (TWT) involves placing relatively thicker concrete slabs, normally 5 to 8 inches thick, bonded (or partia lly bonded) over an existing asphalt pavement.

PAGE 25

25 Similar to UTW pavements, TWT pavements us e short joint spacing and good bond between the concrete and the asphalt layer. Conventional whitetopping (CWT) involves plac ing concrete slabs which are typically greater than 8 inches in thickness. The concre te slabs are typically not bonded to the underlying asphalt layer. Experimental UTW pavements have been c onstructed in many states, including Colorado, Georgia, Iowa, Kansas, Kentucky, Missouri, Ne w Jersey, North Carolina, Pennsylvania and Tennessee. Preliminary evaluations of these recently constructed UTW projects have shown that UTW is a viable rehabilitation method for asphalt pavements. The Florida Department of Transportation (FDOT) has also experimented w ith UTW in recent years. Three UTW test tracks were constructed behind the FDOT State Materials Office in Gainesville in 1996. An experimental UTW project was also constructed at the Ellaville Truck Weigh Station on I-10 in Northwest Florida in 1997. However, the performan ce of these test sections were less than ideal, with the observation of some early cracking on the concrete surface. These problems were attributed mainly to the fact that all of the UTW test sections were inadequately designed for the traffic at the Ellaville Weigh Station (Tia et al, 2002). While the UTW technique may provide durable wearing surface for normal traffic loads on residential and city streets, low-volume roads, street intersections, ge neral aviation airports, and park ing areas, the UTW technique was probably not an appropriate rehabilitation altern ative for weigh stations subjected to frequent applications of heavy truck traffic. The use of TWT or CWT might have been a more appropriate choice in su ch an application. With the potential economical and technical benefits of WT pavements, there is a need to effectively evaluate the feasibility and proper application of UTW, TWT and CWT pavements in

PAGE 26

26 Florida, so that the WT techniques can be pr operly and effectively u tilized to achieve the maximum benefits to the traveling public. 1.2 Problem Statement To effectively evaluate WT pave ments, it is necessary to have a reliable model or methodology to analyze their behavi or under the effects of loadi ng, geometric and environmental conditions. This model is to be developed based on existing knowledge on modeling and behavior of WT pavements and ta king advantage of the state-of -the-art tools in 3-D finite elements method. The model will consider the effects of joints, inte rface bond, temperature, and other pertinent material propertie s and pavement parameters. The model is to be calibrated and verified using experimental resu lts from full scale experiments cons idering most of the variables affecting the behavior of WT pavements. The main model characteristic should be its capability of modeling different interface conditions. In addition to providing the necessary data for the model calibration, the full scale experiments will allow for observation of relationships between pavement performance and pavement response (measured strains and calcu lated stresses). In this way, a better understanding of the behavior of the WT pavements can be accomplished. 1.3 Research Hypothesis The following hypotheses were inve stigated in this research: 1. Conventional proced ures for WT interface treatments provide enough bond strength for the composite pavement to become bonded or pa rtially bonded, and such interface condition can be adequately modeled using linear springs connecting the layers. 2. Load transfer at the concrete slab join ts due to the interloc king mechanism can be modeled using vertical linear spri ngs connecting the adjacent slabs.

PAGE 27

27 3. Temperature is a critical factor affecting the behavior of WT pavements. It affects the AC as a supporting layer and severe temperature conditions may cause a composite pavement to become de-bonded, and to have lowe r load carrying capabilities. 1.4 Objectives of Research The m ain objectives of this research are as follows: (1) To develop analytical models for analys is of the behavior of UTW, TWT and CWT pavements. These models were to be calibra ted and verified using experimental results. (2) To evaluate the effects of temperature and other variables in the behavior of WT pavement and their role in the failure mechanism. (3) To evaluate the applicability of UTW, TWT and CWT techniques for rehabilitation of deteriorated asphalt pave ments in Florida. 1.5 Approach and Scope of Research The objectives of this research study were achieved through the following main tasks: 1. A literature review on the state-of-the-art of WT pavements. 2. Development of an experimental design a nd instrumentation plan for evaluation of several UTW, TWT and CTW pavement test secti ons by means of accelerated pavement testing. 3. Construction and testing of the UTW, TWT a nd CWT test sections located in the Florida DOT Research Park by means of the HVS. 4. Characterization of the test sections by laboratory testi ng of cored samples and FWD testing to obtain pavement parameters. 5. Development of analytical models for analysis of UTW, TWT and CWT pavements, and calibration and verification of the models by compar ing the analytical resu lts with experimental results from the HVS test sections.

PAGE 28

28 6. Use of the developed analytical model to estimate the stresses in the WT pavements sections under critical loading conditions to study the applicability of these composite pavements under Florida conditions. 1.6 Significance of Research In the past, there have been m any studies where WT pavements were constructed and their performance observed. There have also been so me studies where WT pavements were modeled and analyzed with respect to the various factors which may affect their performance. However, there has been little work done where the WT pa vements were instrumented and the measured responses were compared to the analytical results to validate the models used. This research evaluated WT pavements by ta king into consideration all the pertinent factors affecting their behavior. All the variables that have been previously shown as important in the performance of WT paveme nts, such as thicknesses of the concrete and AC layers, joint spacing, temperature and level of bond in the in terface were assessed in this research. A 3-D finite element method was used to model the be havior of the WT pavements, and a full-scale experiment using full-scale instrumented pavement test sections and full-scale wheel loads was used to evaluate the behavior and performance of various WT pavement designs. The measured responses from the test sections were used to va lidate and fine-tune the analytical model. The validated and fine-tuned model was used to predict the level of stresses and performance of the WT pavement sections. The significance of this research work is that the analytical model developed was validated and fine-tuned using measured responses from full-scale and instrumented WT pavements, which has not been done before. The 3-D finite element model has also further refinements from previous work in this area. In the past rese arch work, 2-D models using 4-node elements and 3D models using 8-node elements have been used to analyze this type of pavements. In this

PAGE 29

29 research, a 3-D finite element model with 20-node 3D solid elem ents was used to model the pavement structure. Previous models have used a 4-slab system to model the WT pavements. In this research, a 12-slab system was used. Us ing a 12-slab system can possibly give better modeling of the effects of the adjacent slabs. The bonded interface betw een the concrete slab and the AC layer was also modeled with special elements to cover the range from a fully bonded to fully un-bonded condition.

PAGE 30

30 CHAPTER 2 LITERATURE REVIEW ON WHITETOPPING 2.1 General Concepts The concept of resurfacing existing asphalt pa vem ent using Portland cement concrete (whitetopping) is not new. In fact, the first reported use of whitet opping dates back to 1918 (Hutchinson, 1982). However, this technology ha s improved over the years as the concrete paving technology has improved. Plain concrete, reinforced concrete and fibrous (fiber reinforced) concrete have been used to re surface flexible pavements (Hutchinson, 1982 and McGhee, 1994). In the 1940s and 1950s, plain concrete was mainly used at civil and military airports. Concrete thickness used in these pr ojects ranged from 8 to 18 in. (200 to 460 mm). Since 1960, plain concrete has been extensively used to resurface existing highway pavements in states such as California, Utah, and Iowa. Concrete thickness of these resurfacing projects ranged from 7 to 10 in. (175-250 mm). Continuously reinforced concrete and fiber-reinforced concrete were also used on a limited number of projects. NCHRP synthesis 204 listed 189 whitetopping projects construc ted in the United States between 1918 and 1992. This list included streets, highways, and airfield projects. There are several advantages of using whitetopping for reha bilitating asphalt pavements. Whitetopping provides long-term benefits to the traveling public, and to roadway or airport agencies. Concrete durability and long-term pe rformance characteristics decrease the maintenance required and life cycle costs of pavement s. As a result, concrete surfaces significantly reduce traffic delays associated with the frequent maintenance of asphalt pavements. In addition, when concrete surfaces are used, skid resistance and safety are substantially improved, especially under wet conditions. These advantages promote and contribute to the use of concrete pavements over asphalt surfaces.

PAGE 31

31 2.1.1 Ultra-Thin Whitetopping Ultra-thin w hitetopping (UTW) is a relatively new technique fo r resurfacing deteriorated asphalt pavements. It involves pl acing very thin concrete slabs (2 to 4 in. thick) on top of an asphalt pavement to form a bonded (or partially bonded) composite pavement. The reduction of thickness is justified by the use of high quality concrete with relativel y high strength, shorter joint spacing, and bond between the concre te and the existing asphalt pavement. The first UTW experimental project was constr ucted on the access road to a waste disposal landfill in Louisville, Kentuc ky in September 1991(Cole and Mohsen, 1993, Brown, 1995, and Risser et al. 1993). The concrete mixture wa s designed to provide relatively high early compressive strength, 3,500 psi at 24 hours. A low water-cement ratio of 0.33 was selected to achieve higher strength and redu ce drying shrinkage. Two concre te slab thicknesses, 2 in. and 3.5 in., and two joint spacings, 2 ft and 6 ft were used. The Louisville UTW pavement has performed well, carrying many more traffic loads th an predicted by design pr ocedures available at that time. Following the success of the Louisville UTW project, many other states, including Tennessee, Georgia, North Carolina, Kansas Iowa, Pennsylvania, New Jersey, Colorado, Missouri, Mississippi, Virginia and Florida, ha ve constructed and are currently evaluating UTW projects. Over 200 UTW pavements have been bu ilt in the last decade. The development of a mechanistic design procedure for UTW pavement s in 1997 represented another major step in advancing this promising technique (Wu et al. 1997, Mack et al. 1997, and ACPA, 1997). 2.1.2 Thin Whitetopping Thin W hitetopping (TWT) is a variation of the UTW where thicker concrete slabs are used. Slab thicknesses in the range of 5 to 7 inches are normal for this type of pavements. TWT may have a bonded or an unbonded interface between th e concrete slab and the AC layer. While

PAGE 32

32 more attention has been paid to investigate th e behavior of UTW paveme nts, few studies have focused in the alternative of using TWT when th e conditions for the thinner slabs cannot be met. Among the states that have undertaking projects involving TWT are Colorado (Tarr et al. 1997), Minnesota (Vandenbossche et al. 2002), and Mississippi. 2.1.3 Conventional Whitetopping Whitetopping pavem ents with a slab thickness greater than 8 inches are commonly known as Conventional Whitetopping (CWT). CWT pavements are generally used for pavements subjected to heavier traffic lo ads, and have been designed based on the assumption that the existing asphalt concrete (AC) layer does not contri bute directly to the lo ad-carrying capacity of the pavement structure. Rather, the AC layer is considered to se rve as a base layer for the new concrete overlay, and no bond is cons idered to exist between the ove rlay and the existing asphalt. Longer joint spacing (comparable to those of conventional jointed concrete pavements (JCP)) is generally incorporated in CWT. 2.2 Concrete Mixture Proportions and Properties The concrete m ix for a particular UTW and TWT project is often selected based on the requirements for early opening to traffic. A nor mal mix design includes cement, coarse and fine aggregates, air-entraining agent, admixtures, a nd a lower water-cement ratio. Fibers have been used in many UTW projects; however, the effect s of fiber have not been well documented. Compared to aggregate used for thicker concrete pavements, the top-size of coarse aggregate for UTW and TWT is reduced. Materials and mix pr oportions selected for th e first experimental project in Louisville, Kentuc ky are shown in Table 2-1. The concrete mixture was designed to provide re latively high strength at early ages (3500 psi or 24.2 MPa at 36 hours). A lo w water-to-cement ratio (0.33) wa s selected to achieve higher strength and reduce drying shrinkage. Polypropylen e fibers were used to enhance the flexural

PAGE 33

33 strength, increase impact and freeze-thaw resi stance, and further reduce drying and plastic shrinkage cracking. Similar mix proportions were used in the Tennessee (Speakman et al. 1996) and Georgia (Cown, 1993) projects. In the Leawood, Kansas s ite, the mixture proportions were slightly different; cement content was less (611 lb/yd4), but a setting-accelerating admixture was used (Dumitru et al. 2002). Compressive strength of 3000 psi (20.7 MPa) at 24 hours was achieved. Three pounds of polypropylene fibers were also used in this mixture. The mix design used in the Kansas UTW project is presented in Table 2-2. In April 2000, an old asphalt runway was re habilitated using the UTW technique at the Savannah-Hardin County Airport (SNH) in Tenne ssee (Saeed et al. 2001 ). The runway was original constructed in 1962 and was subsequen tly overlaid and extended in 1975. The original pavement consisted of an AC surface of 3.5 in. a nd a crushed aggregate base of 6.5 in. A 3-in. AC overlay was added during the 1975 rehabilitation. At the time of UTW construction, the AC su rface had exhibited significant fatigue and thermo cracking. Design UTW thickness was 4 in. with a joint spacing of 48 in. Design concrete flexural strength was 700 psi. Th e concrete mix design is shown in Table 2-3. In 1997, the Illinois Department of Transporta tion (IDOT) began investigating the use of whitetopping as an intersection repair method (Winkelman, 2003). Eight projects, including main lines and intersections, were selected to be analy zed in this research. The projects are identified in Table 2-4 along with the mixture design used in each case. All of the mixture designs contained an air entraining admixture. The mixt ure designs also included a water reducer except for the project on Clay County Highway 3. The project listed in Decatur contained a super-

PAGE 34

34 plasticizer. The water to cement ratios for all of these projects ranged from 0.34 to 0.36 except for the Clay County project, which was 0.46. The performances of the thin whitetoppi ng on the four mainline pavements and the ultrathin whitetopping sections on the intersections have been reported as excellent. Middleton et al. 2005, investigated the impact of different cement materials, synthetic fiber types, and curing procedures on compressive strength, flexural strength, shrinkage and scaling durability of concrete. According to the results, high early strength concrete containing ordinary or rapid-hardening cement along with a low wate r-cement ratio gave excellent compressive and flexural strength at the age of one day. It was also noted that it provided high resistance to scaling from de-icing agent. The use of Class C fly ash gave acceptable final strength, but the early strength was lower than the one obtained in mixes without fly ash. On the other hand the fly ash concrete showed poor scaling resistance. 2.3 Construction Procedures UTW pavements are constructed with slipform or fixed form pavers in essentially the same way as conventional concrete pavements, with some special provisions. The construction procedures consist of the following steps: preparing asphalt surface, placing the concrete, finishing, surface texturing, curing, and sawing the joints. Asphalt pavement surface preparation prior to concrete placement is a very important procedure to achieve better bond and good performance of UTW. Milling, followed by cleaning with compressed air to remove all laitance, dust, gr it and all foreign materials, is the best way to prepare the asphalt surface. It is recommended that an adequate asphalt thickness of a minimum of 3 in. (75 mm) after milling, be used, if possible (Mack et al. 1997). Concrete used in UTW can be produced at a ready-mix plant and delivered to the site by ready-mix trucks. Normal slipform pavers can be used to spread, scre ed, and consolidate the

PAGE 35

35 concrete in an efficient manner. After the su rface is finished and textured, a curing compound is immediately sprayed on the entire surface to ach ieve adequate curing. The curing compound is generally applied at a rate twice the normal a pplication rate for thicker concrete pavements because thin concrete slabs can lose water rapidly (ACPA, 1998). Joint sawing must be performed as soon as surface conditions permit, or when the concrete is able to support the equipment and the operator. Usually joints are not sealed because joint openings are generally narrow due to the short joint spacing. 2.4 UTW and TWT Design Considerations A param etric investigation of the variables affecting the performance of whitetopping pavements was performed in Gainesville, Florida. Confirming the results from previous studies, Tia el al. 2003, concluded that th e main factors affecting the be havior of UTW and TWT are the thickness of the concrete slab, the joint spacing, the bond in the interface and the thickness of the asphalt layer. 2.4.1 Slab Thickness A m ajor benefit of UTW is the reduction in concrete thickness. As defined above, the recommended thickness for ultra-thin whitetopping is not more than 4 in. (100 mm). For most of the experimental projects, thickn ess ranges from 2 in. (50 mm) to 4 in. (100 mm). The reduction in thickness is justified by the use of high qual ity concrete with relatively high compressive and flexural strength, closely-spaced joints, a nd good bonding between UTW and the existing asphalt base. 2.4.2 Joint Spacing Spacing between joints is an im portant factor controlling the performance of UTW. In the Louisville experimental project, 6-ft transverse and longitudinal joint spacings (6-ft panels) were compared with 2-ft panels. Although the 6-ft pa nels exhibited cracking, no signs of distress were

PAGE 36

36 observed on the 2-ft panels. This performance was attributed to the smaller panels (2-ft) transferring the load completely to the flexib le base and the concre te being mostly under compression. Comparatively, when la rger panels are used, some of the load is absorbed by slab bending. The contribution of the existing asphalt is based on the assumption that the overlay is bonded to the flexible base. The American Concrete Pavement Association (ACPA) has recommended that joint spacing be about 12 to 15 times of the sl ab thickness. For example, spacing for UTW of 2 in. (50 mm) of thickness s hould be between 2 and 3 ft (0.6 and 0.9 m). Spacing at a Georgia site wa s 2-ft (0.6 m) and in Kansas 3-ft (0.9 m) in one section, and 4-ft (1.2 m) in the other. It is not practical to install dowel bars, tie bars, or keyway in UTW pavements because of the very thin slabs. Field ev aluation has indicated that load transfer provided by aggregate interlock is generally high becau se of the short joint spacing and the support provided by the asphalt layer. For UTW pavements, field performance demonstrated the need for thicker slabs at the transition areas between the UTW and the asphalt roadways. Figure 2-1 shows transition details for UTW pavements, as recommended by ACPA. 2.4.3 Interface Bonding Strength Bond between UTW and existing asphalt pavement is a key factor c ontrolling the performance of the UTW composite pavement. The ex istence of bond strength not only significantly reduces the stresses in the conc rete section, but also allows the section to perform and be analyzed as a composite section. A wide range of bond strength (shear st rength between the two layers) was measured at some of the experimental sites; from the average strength of 50 psi (0.34 MPa) measured at the Swedish site to the ove r 200 psi (1.38 MPa) measured at some Florida UTW test sections. Bond strength can be improved by milling the old asphalt pavement; the

PAGE 37

37 roughness in the surface and the exposed aggregates lock the layers together thus increasing the bond. An experiment intended to evaluate the bonded condition in the in terface of concrete overlay was performed in Iowa by Nishiyama et al. 2005. This experiment included field and laboratory testing to evaluate th e bond at different times after c oncrete placement. The project concluded that bond strength betwee n an overlay and the existing pavement gradually increases over the time regardless of the initial bond strength. To evaluate the interface condition, other authors have propos ed the use of the vertical tensile strength in the interface as an indicato r of the bond, in addition to the shear strength. By performing pull-off tests on the composite samples, it is also possible to characterize the vertical interaction between layers. Rasmussen et al., 2000, presented a method to ch aracterize the axial sl ab support restraint by running full-scale push-off test to obtain the stiffness and strength of the interface bond. These parameters can be used in analytical models to better represent the interface condition in whitetopping pavements. 2.5 Design Procedure for UTW and TWT Pavements Because of the thin s labs, the Portland Ce ment Associations (PCA) thickness design procedure (PCA, 1984) does not completely apply to UTW. Furthermore, the AASHTO thickness design method (AASHTO, 1993) does not account for the bond between the two layers. As the development of the UTW technique continued, it became apparent that a design procedure for selecting the optimum UTW thickness and joint spacing subjected to anticipated traffic and environmental loading was needed. In 1994 the PCA sponsored a comprehensive research effort aimed at developing a mechanistic based design procedure for UTW (W u et al. 1997 and Mack et al. 1997). The

PAGE 38

38 PCAs UTW pavement design proc edure was developed in 1997. Wo rk conducted in developing the design procedure included a thorough literature review of past and current work; condition surveys of UTW sites in Georgia and Tennessee; instrumentation and load testing of several test sections; development of a 3-dimensional fini te element model for UTW analysis; and development of a design procedure a nd construction guidelines. Development of this design procedur e included the following elements: Verification of the 3-dimens ional finite element model Strain (stress) data collected from the Spirit of St. Louis Airport UTW test s ections were used to calibrate and verify the three-dimensional model that was developed in this study and was used as an analytical tool to analyze the UTW pavement behavior and to deve lop the design procedure. To calibrate and verify the model, stresses were computed under each loading and temp erature condition, and were compared with the load testing results obtained from the test pavements. Identification of degree of bonding existed in the field. It has been established from the results of field testing that UTW pavements behaved as partially bonded composite pavements. Using field data and the 3-D mode l, an effort was made to quantify the additional structural capacity (or load carrying capaci ty) that could be offered by the asphalt layer to the UTW pavements. Correlation between stresses calculated from ILSL2 and the 3-dimensional model Since many computer runs would be required to develop the design guidelines, the use of the 3dimensional model was not feasible due to the ti me needed for each run. Instead, correlations were developed between stresses computed by these two mode ls, and were used in the development of the design guidelines.

PAGE 39

39 Development of design guidelines. Following are the processes involved in the development of the UTW design guidelines: Stresses induced by loads and temperature were separately computed for fully bonded UTW pavements using the 2-dimensional mode l, ILSL2. A wide range of pavement parameters and material properties were covered. The 2-D model stresses were converted to 3-D model stresses using the conversion equations derived in the previous step. The converted 3-D model stresses were in creased by 36% to account for the partially bonded condition, as observed in the field testing. Equations were developed to correlate the c onverted and adjusted st resses to different pavement parameters. Stresses and strain s were then calculated for typical parameters for UTW pavements under different lo ading and temperature conditions, and were tabulated. UTW pavement thickness design was accomplished by limiting both the concrete and asphalt strains within safe limits under anti cipated traffic and environmental loadings in the pavement's design life. A mechanistic design procedure for TWT was de veloped by the state of Colorado (Tarr et al. 1998). The procedure followed to develop this design method was very similar to the one used in the PCA method. Experiments using slab thicknesses betw een 5 and 7 inches were performed and the measured strain values were used to calibrate and verify the computer model. Theoretical design equations to predict critical stresses and strain were identified. Correction factors for stresses due to load position and differences between th e calculated and the measured stresses were included in the model. Also correction factors for the strains at the interface between the concrete slab and the asphalt layer were evaluated to account for the bond condition. An additional factor to consider the effect of temperature was added to the model, which modifies the stress in the concrete. The predic tion equations were developed for load varying from 20 kips to 40 kips single-axle load. Failure criteria for both materials were assumed from fatigue relationships. As considered in the PCA design procedure, the number of load repetitions

PAGE 40

40 for the concrete slab is a function of the flexural stress-to-strength ratio. The failure criterion for the asphalt concrete is based on the allowable numb er of repetitions, which was considered as a function of the asphalt elastic modulus and the volume of binder and voids. In this method the number of load repetitions that the asphalt concrete has alrea dy carried is also considered. 2.6 Performance of UTW and TWT Projects Most of the experim ental UTW and TWT project sites have performed very satisfactorily. At the Louisville site, two overlay thicknesses we re evaluated (2 and 3.5 in.), and joint spacings of 6 ft and 2 ft. were used. UTW test sections with 2-ft joint spacing showed much less cracking than those with 6-ft joint spacing and 2-in. thick slabs. Evaluations in Georgia, c onducted two years after construction, indicated good performance (Wu et al. 2001). The UTW test sections we re located in a truck weigh station on I-85. The test sections had a design UTW thickness of 2.5 in. (64 mm) and the existing asphalt thickness was about 11 in. (279 mm). It was reported that the test sections had been subjected to 351,000 18-kip (80-kN) Equivalent Si ngle Axle Loads (ESAL) in two years. Only 2% of the slabs had cracks for the test sect ions with fiber reinforced conc rete and 5% of slabs had cracks for the non-fiber sections. A design evaluation us ing the PCAs design procedure also indicated that the UTW test sections were designed adequately. A wealth of information on the behavior of UTW was generated from the Tennessee experience (Wu et al. 2001). After the first experimental project wa s constructed in Nashville in May of 1992, six more sites were built in Maryville, Chattanooga, McMinnville and Athens. Cracks observed on the first site (Nashville) were mainly attributed to the asphalt base. It was observed that during milling, the asphalt was comp letely removed and the concrete overlay was supported by a cobblestone base. A design evaluation also indicated that the test section was severely under-designed.

PAGE 41

41 A UTW pavement project was cons tructed at the Spirit of St. Louis Airport to carry lightload aircraft (gross weight of 12,500 lb or 5,670 kg ) traffic in 1995. The design thickness of the UTW pavements was 3 in. (89 mm) with a jo int spacing of 50 in. (1.3 m). The existing asphalt pavement was milled before concrete pl acement to create a rough surface. The asphalt thickness after milling was about 3.1 in. (79 mm). A visual condition survey of the UTW pavements was performed on July 2001 (Wu et al. 2002). It was observed that, after over six years of service, the UTW pavements performed extremely well, with very little distresses observed on the entire site. Out of the more than 7,200 panels, only 18 panels (0.25%) have exhibited distresses, with majority of them in the form of corner cracking. The Iowa and Minnesota Departments of Tran sportation (DOT) have also been actively involved in the UTW technique development and evaluation. In 1994, Iowa constructed 7.2 miles (11.6 km) of UTW pavements on a segment of Highway 21 (Cable et al. 1997 and Cable et al. 2001). The research was designed to evaluate the long-term performance of UTW pavements and their applicability in Iowa. Four major va riables were included for evaluation, resulting in a total of 41 test sections. The four variables were overlay th ickness, joint sp acing, the use of fiber, and the asphalt surface preparation. The pavements were subjected to an estimated Average Daily Traffic (ADT) of 1,350 (or 40 ESAL s per day). Through continuous monitoring of the test sections, after seven years in service, the test secti ons have performed well and have exhibited minimal distresses. In October 1997, the Minnesota DOT constructe d several thin and ultra-thin whitetopping pavements on I-94 at the Minneso ta Road Research (Mn/Road) (Vandenbossche et al. 2002). The existing asphalt pavement was in fairly good condition, with minor cracking and rutting. The asphalt pavements were milled to the depth of the overlay thickness to maintain the original

PAGE 42

42 pavement surface elevation. The UTW test sections had two different thicknesses, 3 in. and 4 in. (75 mm and 100 mm), with two differe nt joint patterns, 4 ft by 4 ft and 5 ft by 6 ft (1.2 m by 1.2 m and 1.5 m by 1.8 m). After 3 years and over 4.7 million ESALs, cracking (transverse and corner cracking) was observed in the UTW test sections. The majo rity of the cracking was in the truck lane. It was indicated in the research st udy that most of the corner cracking occurred along the inside longitudinal joint due to its location di rectly in the inside wheel path. Transverse cracking often occurred in the outside wheel path near a transverse joint. This indicated that using a joint layout that keeps the longitudinal joints outside the wheel paths could improve UTW pavement performance. A forensic investigation of the UTW constructe d in the Ellaville Weigh Station in Florida (Tia et al. 2002) showed that the poor performanc e of most of the six te st sections was mainly due to an inadequate design of the overlay. The thickness of the test slabs were 3 to 4 inches, with joint spacing of 4 a nd 6 feet. The premature cracking, the extensiveness and the severity of the cracking and the rapid progress of the cracking ar e also attributed to th e lack of control on the layer thickness especially in the AC, which pres ented no thickness at all in some sections. The problem was aggravated by the loss of bond between the concrete slab and the asphalt layer, which had a great effect on the ra pid progress of cracking and the large percentage of shattered slabs. 2.7 Accelerated Pavement Testin g and Field Testing of U TW In the spring of 1998, the Federal Highway Ad ministration (FHWA), in partnership with concrete industry groups, undertook a study to ev aluate the performance of ultra-thin whitetopping under accelerated traffic load. Eight (8 ) sections of existing asphalt pavements were whitetopped and subjected to FHWAs accelerated lo ading facility (ALF) in McLean, Virginia. The experimental results indi cate the bond between AC and concrete decreases the critical

PAGE 43

43 tensile stresses in the concrete overlay as th e UTW section acts in a composite manner. Dynamic strain measurement of longitudinal stra in indicates the concrete overlay experiences significant stress reversal as th e wheel rolls over the pavements (Cole et al. 1999). The Indiana Department of Transportation (INDOT) conducted an experimental study at the APT facility to invest igate the performance of UTW pavement in Fall 1999 (Rajan et al. 2001). Four concrete mixtures were tested under slow moving load s. The results indicated the joint spacing of 1.2 m (4) in all the lanes was su fficient. The measured strains in the overlay were proportional to the a pplied load. The results showed that the pavement response was linear within the overlay. However the increase of temper ature in one of the lane s affected the linearity of the pavement response. Strains at the aspha lt surface increased consid erably because of the temperature gradient. The researchers have found that the concrete overlay experienced significant stress reversal as the wheel rolled over the pavement. It was also observed that the overlay thickness and the asphalt stiffness signi ficantly affect the stra ins because of their influence on the location of the neutral axis (Rajan et al. 2001). An UTW pavement with 3-inch thick concrete on 6 inch asphalt wa s constructed in the APT facility in Lancaster, Ohio for the purpose of measuring response under controlled loading and environmental conditions (Edwards et al. 1999). The meas ured strain was relatively proportional to the magnitude of the applied lo ad over the range of 6 to10 kips at 50 F and 5 mph. The measured tensile strains were higher on the AC surface than would be expected by projecting strains measured in the PCC layer down to the AC/PCC interface. The placement of the 3-inch thick PCC layer on 6-inch thick layer of AC lo wered the neutral axis of this UTW pavement structure to the lower portion of the PCC where tensile strains were minimal.

PAGE 44

44 The researchers at Purdue University a nd INDOT conducted an UTW pavement experiment to investigate the applicability of PCA design guidelines for UTW designs (Galal et al. 2004). Preliminary results indicate that the PCA design equations may be able to be used if the concept of an equivalent thickness is employed. Equivalent section was hypothesized to take into account the additional laye r in the existing composite pavement structure. There was a good agreement between the measured strain in the composite UTW section a nd the computed strains from the ESLYM5 program. Nishizama et al. 2003 performed full-scale e xperiments to evaluate the mechanical behavior of UTW pavements in Japan. The expe riment included the appropriate instrumentation with strain gages embedded in the concrete slab s and thermocouples to monitor strain and temperature respectively. Two joint spacings were investigated (4 and 6). The loading periods included summer and winter time. The load was a pplied in two ways: stationary and moving. In the stationary load test, a load was applied on th e edge of the slab through a 30 cm diameter plate by a jack and a crane truck. The load was applied at increments of 9.8 kN up to 49 kN and strains were measured at each load increment. In the moving load test, the crane truck traveled on the test pavement at a low speed and dynamic stra ins were measured every 0.1 sec. A single thickness of 4 inches for both the concrete slab an d AC layer was used in the test. The concrete slab was placed bonded to the AC layer. The m easured strains were co mpared with the calculated ones obtained from a 3D FE m odel and the comparison showed a good match. 2.8 Analytical Models Many efforts have been m ade to model white topping pavements. Most of them have focused on modeling the interface interaction between the concrete slab and the AC layer. In addition to the bond interface evaluation, the analyti cal models have to consider also the load

PAGE 45

45 transfer at joints. The general a pproach to these two types of inte raction is the use of springs to represent the stiffness in different directions. A 2D finite element model, NSLIP (Nels on et al. 2002) was developed to model the interface condition in composite pavements to study the delamination. In this model a 4-node slip finite element with displacement in the norma l and shear direction was used. The stiffness in the normal direction was considered as infinitely large when the interface was in compression. It was hypothesized that when the interfacial strength is exceeded, delamination occurs in the interface. The concept of friction factor was used in combination with the ISLAB2000 FE program (Khazanovich at al. 2002) to characterize the bond in the interface of com posite pavements. The ISLAB2000 program contains two models to analyze the interface: the modified Coulomb friction model and the simplified friction mode l. For the analysis, Khazanovich utilized the simplified friction approach in conjunction with the transformed section concept. The simplified friction model allowed for modeling of intermediate degrees of interaction between the fully bonded and un-bonded and the friction factor was shown to be a good indicator of level of bond. Nishizawa et al. 2003, developed PAVED3D, a 3D FE model using 3D solid elements with 8 nodes per element. This model has the capabilit y to model several slab s with the respective joints. The interaction between c oncrete slabs is modeled through spring elements that represent the stiffness in the plane of the joint (Ks and Kt), and in the normal direction (Kn). Springs were also used to model the interface in teraction between the concrete slab and the AC layer. Similar to the joint case, three springs were utilized. The findings of this research were that by varying the stiffness of the springs used in the interface it is possible to model different levels of bond.

PAGE 46

46 Table 2-1 Concrete Mix Proportions Used in Louisville Experiment al Project (Riser et al 1993). Constituents Quantity Cement (ASTM C 150 Type I) 800 lb/cy Coarse Aggregate 1800 lb/cy Fine Aggregate 1150 lb/cy Water 260 lb/cy Polypropylene Fibers 3 lb/cy High Range Water Reducer 14 oz/100 lb cement Table 2-2 Concrete Mix Proportions Used in Leawood, Kansas (Wu et al. 1997). Constituents Quantity Cement (ASTM C 150 Type I) 611 lb/cy Coarse Aggregate SSD (Crushed Limestone) 1730 lb/cy Fine Aggregate SSD (Natural Sand) 1345 lb/cy Total Water 225 lb/cy Pave Air 5 oz/cy Pozzutec 65 oz/cy Rheobuild 43 oz/cy Table 2-3 Concrete Mix Proportions Used in SNH, Tennessee (Saeed et al. 2002). Constituents Attributes Cement Type I, 573 lb/cy Fly Ash Maximum 15% by weight of cementitious materials Strength 700 psi flexural Slump Between 1/2 and 2 in. Water/Cement Ratio 0.35 Air Content 6% by volume Synthetic Fiber 3 lb/cy

PAGE 47

47 Table 2-4 Mix proportions for eight whitetopping projects in Illinois. Project-Location Coarse Aggregate (lb) Fine Aggregate (lb) Cement (lb) Water (lb) Polypropylene Fibers (lb) Decatur Intersection of US-36 and Oakland Avenue 17131210705239 N/A Carbondale Intersection of US-51 and Pleasant Hill Road 18051008755273 3.0 Harrisburg Intersection of US-45 and Illinois Rt. 13 1811975755302 3.0 Anna Illinois Rt. 146 (Intersection of Vienna and Main Streets) 1811975755302 3.0 Tuscola US-36 17041035755255 N/A Clay County Highway 3 18141286534244 N/A Piatt County Highway 4 19571220534179 N/A Cumberland County Highway 2 18361256575197 N/A Figure 2-1 Transition between UTW and Ad joining Asphalt Pavement (ACPA, 1998)

PAGE 48

48 CHAPTER 3 INSTRUMENTATION AND CONSTRUC TION OF TEST S ECTIONS 3.1 Description of the Testing Phases The f irst step in investigating the applicability of WT pavement in Florida was the design and construction of a full scale experiment for Accelerated Pavement Testing (APT). The FDOT Materials Office has a Heavy Vehicle Simulator (HVS) and an Accelerated Pavement Testing (APT) facility for the operation of this HVS. The HVS can appl y realistic full-size wheel loads to full-size pavements to assess their behavior and performance direct ly. The HVS has the capability to simulate 20 years of interstate traffi c on a pavement test section within a period of 1 to 4 months. This accelerated pavement testing facility provided an excellent opportunity for evaluating the long-term performa nce of WT pavements in Florida in a direct and effective manner. The HVS testing of the test sections in th is study were divided into two main phases, namely Phase I using bonded composite pavements, and Phase II using un-bonded composite pavements. Phase I was divided in two sub pha ses by using two different joint spacings. The description of the pha ses is as follows: (1) Phase I-a It involved three test sections on Lane 6 of the APT test area at the FDOT State Materials Research Park. The concrete sl abs were placed bonded to the top of an asphalt concrete layer, and had a panel size of 6 feet by 6 feet. The three test sections had concrete slab thicknesses of 4, 5 and 6 inches. The thickness of the underlying AC varied from 4 to 5 inches. (2) Phase I-b It involved three test sections on Lane 7 of the APT test area. Similar to Lane 6 in Phase I-a, Lane 7 also had three test sections with concrete sl ab thickness of 4, 5 and 6 inches placed bonded to the top of a similar AC laye r. The panel size in this case was 4 feet by 4 feet.

PAGE 49

49 (3) Phase II It involved three test se ctions on Lane 6 constructe d after the test sections from Phase I-a were tested and removed. In this case, the concrete slabs were 6, 8 and 10 inches thick, placed un-bonded to the top of the asphalt la yer. The concrete panel size was 6 feet by 6 feet. Although there was some variati on in the asphalt thickness in the test track, the thickness of the asphalt layer was not considered as a vari able in this experiment and an average thickness of 4.5 inches was used in the analysis. 3.2 Layout of the Test Sections 3.2.1 Phase I The test track in Phase I-a (on Lane 6) consiste d of three test sections of 4, 5 and 6 inches of concrete placed bond ed to the existing AC layer, with 6 ft by 6 ft joint spacing. The test track in Phase I-b (on Lane 7) consiste d of three test sections with th e same thicknesses as those used on Lane 6, but with 4 ft by 4 ft joint spacing. The concrete overlay in Lane 7 was also bonded to the existing AC layer. While the 4-inch concrete slabs are considered UTW, the 5and 6-inch slabs fall in the category of TWT. Figure 3-1 shows the layout of the test sections in Lanes 6 and 7. The test sections are confined by two ends and transition conc rete slabs constructed to support the HVS. To ensure a bonded condition in the interf ace in these test sections, the AC surface was milled and cleaned prior to the concrete placement. 3.2.2 Phase II Af ter the removal of the test sections for Ph ase I-a, Lane 6 was overlaid with 6, 8 and 10 inches of concrete placed un-bonded to the existi ng AC layer, with 6 ft by 6 ft joint spacing. Because it was not possible to remove the concre te slab without damaging the AC layer (a very strong bond was observed), the existi ng asphalt layer was also removed and replaced with a new one with the same thickness and properties. To ensure an un-bonded condition in the interface in

PAGE 50

50 these test sections, a white pigmented curi ng compound was sprayed on the asphalt surface prior to the concrete placement. Figure 3-2 s hows the test section layout for Phase II. 3.3 Layout of the Instrumentation 3.3.1 Wheatstone Bridge Circuits To m onitor the strains in the test track, Wheatst one half-bridge circuits were used. In this configuration one strain gage wa s used as an active gage to monitor the load-induced strain, while another one was used as a dummy gage for temperature compensation. The Wheatstone half-bridge circuit used is shown in Figures 3-3 and 3-4. Th e active gage with a resistance of RA is subjected to a temperature-induced strain (y ) and a load-induced stra in (x) simultaneously. The dummy gage with a resistance of RD, is subjected only to a temper ature-induced strain (y). The effect of the temperature-induced strain (1+y) is canceled out in this half bridge circuit, and only the load-induced strain is measured. 3.3.2 Preliminary Stress Analysis To determ ine the instrumentati on layout, a stress analysis was performed to estimate the maximum stresses. The capability of the ADI NA program to consider a bonded condition between layers was used to model the composite pavement for the preliminary stress analysis in Phase I. A 3D model considering four slabs was built to evaluate th e stresses under critical combinations of load and temperature. Two critic al load conditions were considered in the stress analysis: at the mid edge and at the corner of the slab. Also three case s of temperature differential were applied to th e model: -10, 0, and 10 C. The temperat ure differential is defined as the difference between the temperature at the top of the concrete slab and the temperature at the bottom. The FEACONS IV (Finite Element Analysis of CONcrete Slabs version IV) program was used to calculate the an ticipated stresses on the test slabs for the un-bonded condition in Phase II.

PAGE 51

51 The FEACONS program was developed at the Univer sity of Florida for the FDOT for analysis of concrete pavements subject to load and ther mal effects. This program was chosen for use since both the University of Fl orida and FDOT have extensive experience with this program and its reliability has been demonstr ated in previous studies. The FEACONS program was used to analyze the stresses in the test slabs when subjected to a 12-kip (53-kN) single wheel load with a tire pressure of 120 psi (827 kP a) and a contact area of 100 in2 (645 cm2 ), and applied along the edge of the slab, which represents the most critical loading location. Similar to the case of bonded interface (Phase I), the analysis was performed for two different load positions, at the corner of the slab and at the middle of the edge, for the same temperature differentials in the concrete slab s. No load transfer at the joints was assumed in the analysis, which repr esents the worst condition. 3.3.3 Instrumentation Layout W ith the results from the stress analysis, it was possible to identify the locations where the maximum stresses and strains in the test slab would occur so that strain gages could be placed to monitor these maximum induced strains. The foll owing sections describe the instrumentation layout for Phases I and II. 3.3.3.1 Phase I Figure 3-5 shows the instrum entation layout for the 6 ft by 6 ft test section (Phase I-a), which were placed on Lane 6 of the APT test area Three locations (Location 1, 2 and 3) were identified to have the maximum anticipated strain s due to the HVS load. Thus, strain gages were placed at these three locations. Figure 3-6 shows the vertical positions of the gages at these three locations along with the gage id entification. While Location 1 had two gages, the other two locations had only one. At Location 1, one embedded strain gage was placed at a depth of 1 inch from the concrete surface, while the other embe dded strain gage was placed 0.5 inch from the

PAGE 52

52 bottom of the concrete layer. Location 2 had a st rain gage embedded 1 inch from the surface of the concrete slab. Location 3 had a strain ga ge embedded 0.5 inch from the bottom of the concrete slab. Two surface gages were also used to monitor any micro cracks that may occur in the concrete surface. These two gages were located next to the transversal join t at the middle of the slab, in the adjacent panels. These surface gages, lo cated in the adjacent slabs, were also used to evaluate the load tran sfer at the joints. Figure 3-5 also shows the locations of thermocouples to monitor the temperature in the slab. Two positions were considered for the thermo couples, one at the center of the slab and the other in the corner. Figure 3-7 depicts the ve rtical position of the thermocouples. They were placed 1 apart along the depth of the slab with the first starting at 1 from the surface. An additional thermocouple was placed on the surface of the AC layer to monitor daily variation of temperature in the asphalt layer. Figure 3-8 shows the instrumentation layout for the 4 ft x 4 ft slabs in Phase I-b. In this case only two locations (namely Loc1 and Loc2) were identified as locations with maximum stresses and for placement of embedded gages. Four strain gages were placed on the surface of the concrete slab to monitor any micro cracks oc curring in the slabs and load transfer at the joints. The vertical po sitions for the gages are indicated in Figure 3-9 al ong with the gage identification. At Location 1, three gage s were used -one at 1 inch from the surface of the concrete slab, the second at 0.5 inch from the bottom of the concrete slab, and the third one at 0.5 inch below the surface of the AC layer. The placemen t of these three gages allowed for not only the monitoring of the maximum strain at the top and bottom of the c oncrete slab, but also for the comparison of the strain values near the interface of the layers. By comparing the strain values

PAGE 53

53 near the interface, it would be possible to know how well the bonding condition would be in the composite pavement. Location 2 had two gages -one at 0.5 inch a bove the bottom of the concrete slab, and the other at 0.5 inch below the surface of the AC layer. Figure 3-8 also shows the two locations for th e thermocouples in Phase I-b. The vertical positions of the thermocouples in Phase I-b were the same as those used in Phase I-a, which are shown in Figure 3-7. 3.3.3.2 Phase II Sim ilar to Phase I, the strain gages were pl aced at the locations of maximum anticipated stresses due to the HVS loads. Figure 3-10 s hows the instrumentation plan adopted. Taking advantage of the upgrade to the data collection equipment that allowed for more channels for data acquisition, five locations were selected for placement of strain gages. For each of these locations, a set of three strain gages were installed to monitor maximum strains in the concrete slab and the strain at the surface of the AC layer. One strain ga ge was placed at one inch under the surface of the concrete slab. A second strain gage was placed 1 inch from the bottom of the concrete slab. The third strain gage was located 1/2 inch below th e top of the asphalt layer. The vertical positions of th e strain gages are shown in Figure 3-11. Unlike the previous phase, surface gages were not used in Phase II. The locations for the thermocouples are shown in Figure 3-10. In Phas e II, three locations were used to monitor the temperature in the slabs. For each of the three thermocouple locations, a set of thermocouples were plac ed at depths of 1, 3, 5, 7 & 9 in ches for the 10-inch slabs, at depths of 1, 3, 5 & 7 inches for the 8-inch slabs, and at depths of 1, 3 & 5 inches for the 6-inch slabs. At each of the locations, a thermocouple wa s also placed in the asph alt layer at a depth of

PAGE 54

54 1/2 inch from the top of the asphalt layer. The ve rtical positions of the thermocouples are shown in Figure 3-12. 3.4 HVS Loading Plan Testing of the com posite pavements was performed using a Heavy Vehicle Simulator (HVS), Mark IV model. HVS loading was schedu led to start 28 days after concrete placement with an initial load of 9,000 lb, super single tire, with a contact pressure of 120 psi. The wheel load traveled at a speed of 8 mph, in a uni-d irectional mode with no wander, and along the longitudinal edge of the test slab. Loading along the edge was chosen because it represents the most critical loading cond ition for a concrete slab. If the composite pavement test sections coul d withstand the 9-kip lo ad with no visible or detectable cracks for a certain pe riod of time, the load would be increased to 12 kips, 15 kips, 18 kips, and 21 kips, to observe the behavior of the test sections under different loads, and the load at which cracking would occur. 3.5 Data Collection For each test slab, th e strain gages were connect ed to a strain indicator unit, Vishay System 6000 (Model 6100) for strain reading and data acquisition. This system has the ability to take individual strain readings at a very high fr equency. This enabled the recording of dynamic strains as the wheel passed ove r the pavement. Data collection for load-induced strain was started immediately after the star t of HVS loading. Strain data were collected for 30 seconds at one-hour intervals. The rate of data collecti on was 100 strain values per second. This rate allowed for the capture of the progression of the strain and to especially observe the strain reversal phenomenon. Strain gage readings due to a static wheel load were also taken for two wheel loading positions, namely corner (pt 1), and mid-edge (pt 2). Static readings were recorded while the wheel was traveling at slow speed towards the static loading position and

PAGE 55

55 while it stayed at the two load positions (pt 1 an d pt 2) for 20 seconds each. Static strains were measured only for Phase I-a for comparison purpos es with the dynamic load application. Phase I-b and Phase II included only dynamic strain data co llection. While in Phase I, strain data were collected only for the loading pe riod, strain data co llection in Phase II wa s started a few days after concrete placement to monitor st rain due to temperature changes. All the thermocouples were connected to th e same data acquisition system. Temperature data were collected during the en tire day at 5-minute intervals. In both phases, data collection for temperature was started before the loading pe riod, especially in Phase II where strain due to temperature changes was monitored. 3.6 Construction of the Test Tracks 3.6.1 Construction of Concrete Test Tracks in Phase I 3.6.1.1 Asphalt surface preparation and formwork The concrete test track for Phase I-a was constructed on Lane 6 on June 10, 2004, and that for Phase I-b was constructed on Lane 7 on A ugust 10, 2004. These two concrete test tracks were constructed over an existing four-inch thick as phalt surface (two 2 in lifts of asp halt) at the APT test area at the FDOT State Materials Research Park. The asphalt surface was milled and cleaned prior to the placement of formworks fo r the test track. Figure 3-13 shows the milled asphalt surface in Lane 6. In the construction of Lane 7, a tapered formwork was used at the transition from one thickness to the other to make the placement of the concrete easier. Figure 314 shows the formwork used for Lane 7. 3.6.1.2 Concrete mix proportions A m inimum 24-hour compressive strength of 2500 psi and minimum 28-day strength of 5800 psi were specified for the concrete for the te st tracks. The mix designs used for Lanes 6 and 7 in Phases I-a and I-b are shown in Table 3-1.

PAGE 56

56 3.6.1.3 Placement of concrete Before placing the concrete on top of the as phalt layer, water was sprayed on the asphalt surface to promote a good bond between the concrete and the asphalt and to prevent the reduction of water from the concrete. Samples of concrete were taken from a selected truck during concrete placement. The slump, air content and temperature of the fresh concrete were measured. Samples were fabricated for comp ressive strength and el astic modulus, maturity test, and flexural strength tests. Figure 3-15 show s the concrete test tr ack after placement of the concrete. After placement and finishing of the concrete on the test track, saw cuts were made to a one third (1/3) of the thickness to form the join ts for the slabs. A diamond-bladed saw was used for these cuts to ensure a smooth, straight vertical surface. 3.6.1.4 Placement of strain gages Em bedded strain gages were inst alled in the test slabs at th e location described in Section 3.3.3. Each strain gage was fixed between two steel rods fixed to the base layer. At the locations where both the top and bottom embedded gages need ed to be placed, one gage was fixed at the top of two rods and the other ga ge was fixed at the bottom of th e two rods using nuts and bolts. Figure 3-16 shows the placement of the top and bottom strain gages in a 4-inch test slab in Lane 7. The strain gages were placed in the concrete with a distance of 1 inch from the top and 0.5 inch from the bottom of the c oncrete. Surface gages were placed before loading was started on the test slab. The surface of the concrete wher e surface gages were placed was cleaned using a sand paper and applied with the recommended glue to bond the st rain gages to the concrete. Figure 3-17 shows a picture of th e surface gages at a joint on the 4-inch test slab in Lane 6.

PAGE 57

57 A PVC cylinder was placed around the embedded strain gages during placement of concrete, as shown in Figure 3-18. The concrete was placed in the cylinder manually to prevent disturbance from the concrete handling instruments. After the concrete was placed to the same thickness both inside and outside the PVC pipe, the PVC pipe wa s then removed by pulling it out vertically. 3.6.1.5 Placement of thermocouples Therm ocouples were placed at various depths in the test slabs to monitor temperature variation. This was achieved by fixing the thermoc ouples to a PVC rod at different heights. The thermocouple-attached rods were fixed to the asphalt layer. Similar to the case of the strain gages, a PVC cylinder was placed around each of th e rods to protect it from the concrete handling instruments during the concrete placement. The concrete was placed manually in the PVC pipe to prevent any disturbance from the concrete handling instrument. The PVC pipe was removed after the concrete was pl aced to the same thickness inside and outside of the PVC pipe. This procedure ensured the proper position of the thermocouples by preventing any disturbance from the concrete placement. 3.6.2 Construction of Concrete Test Tracks in Phase II 3.6.2.1 Asphalt surface preparation and formwork Af ter testing on the concrete slabs in Phase I-a was finished, the concrete slabs were removed from Lane 6, and the test sections in Phase II were placed on Lane 6. When attempts were made to remove the concrete slabs wit hout removing the underlying asphalt layer, it was found that the concrete slabs were bonded so well to the asphalt la yer that a lot of the underlying asphalt concrete was also removed at the same time. Figure 3-19 shows the condition of Lane 6 during the slab removal process. Due to this situation, a new asphalt la yer was placed on Lane 6 before the concrete slabs in Phase II were placed.

PAGE 58

58 Figure 3-20 shows the formwork for the concrete slabs in Phase II. Each test section was 18 feet long, with transition zones separating one test section from another. A tapered formwork was used at the transition from one thickness to the other to make the placement of the concrete easier. 3.6.2.2 Installation of strain gages and thermocouples and concrete placement The m ethod of installation of strain gages a nd thermocouples was similar to that used in Phase I. Grooves were cut with a diamond saw on the surface of the as phalt surface for placement of the strain gage and thermocouple wires (Figure 3-21). Thermocouples and strain gages were installed at specified locations on the test slabs and covered by PVC pipes before placement of concrete, as described in Section 3.6.1. Before the concrete was placed, a white pi gmented curing compound was applied to the asphalt surface to act as a debonding agent between th e asphalt and the co ncrete slab. The prepared asphalt surface is shown in Figure 3-22. The concrete in Phase II was placed on Oct ober 11, 2005. The mix design of the concrete used is shown in Table 3.1. The same procedure used in Phase I to protect the instrumentation during concrete placement was followed in this case. Figure 3-23 shows the finishing of the concrete test track in Phase II. The joints for the concrete slabs were sawed the following day. Each test section was sawed into six 6 ft x 6 ft panels. The concrete slabs were kept moist by sprinkling with water for at least 3 days to ensure adequate curing (Figure 3-24).

PAGE 59

59 Table 3-1 Mix Designs of Concre te Used in Phases I and II. Lane No. Material Target Actual Moist, % Remarks Cement 508 lb 506 lb D57 Stone 1801 lb 1798 lb 1.5 Pit # 08-012 DOT Sand 1328 lb 1316 lb 5.0 Pit # 76-349 Air entrainment, MBAE 90 1 oz 1.1 oz Admixture, MBL 80 45 oz 40.2 oz Water 15.6 Gal 17.2 Gal Lane 6 PHASE I-a (06/08/2004) W/C Cement 508 lb 504 lb D57 Stone 1801 lb 1810 lb 1.8 Pit # 08-012 DOT Sand 1328 lb 1346 lb 5.0 Pit # 76-349 Air entrainment, MBAE 90 1 oz 1.1 oz Admixture, MBL 80 45 oz 45 oz Water 15.6 Gal 16.5 Gal Lane 7 PHASE I-b (10/08/2004) W/C Cement 508 lb D57 Stone 1750 lb 6.99% Pit # 08-004 Silica Sand 1265 lb 5.84% Pit # 76-349 Air entrainment, MBAE 90 2 oz Admixture, MBL 80 65.0 oz Water 255 lb Lane 6 PHASE II W/C 0.502

PAGE 60

60 Figure 3-1 Layout of the Te st Sections for Phase I. Figure 3-2 Layout of the Test Sections for Phase II.

PAGE 61

61 Figure 3-3 Strain Gage Arrangeme nts in a Half Bridge Circuit. Figure 3-4 Connection of the Ac tive and Dummy Strain Gages in the Half Bridge Circuit.

PAGE 62

62 Loc2 Loc1 27" 12" 9" 6" 3" Embedded Gauge 72" Direction Wheel Path Thermocouple Surface Gauge 36" 24" 6" Loc3 72" 72" Figure 3-5 Instrumentation Layout for the Test Slabs in Phase I-a.

PAGE 63

63 Location 2Gauge 4Location 3 0.5" Concrete AsphaltGauge 3 1" Location 1Gauge 1 Gauge 2 1" 0.5" Concrete Asphalt Asphalt Concrete Figure 3-6 Vertical Positions of the Strain Gages in Phase I-a.

PAGE 64

64 1" 1" 4" 5" 1" 6" Asphalt Concrete 1" 1" 1" 1" Concrete Asphalt 1" 1" 1" 1" Concrete Asphalt 1" 1" 1" 1" Figure 3-7 Vertical Positions of the Thermocouples for the 4, 5 and 6 Slabs in Phase I.

PAGE 65

65 Embedded Gauge 48" Thermocouple Surface Gauge Wheel Path Direction 6" 12" 24" Loc1 Loc2 22" 14.5" 12" 6" 3" 48" 48" Figure 3-8 Instrumentation Layout for the Test Slabs in Phase I-b.

PAGE 66

66 0.5" Gauge 3 Gauge 5 0.5" Location 2 Concrete Asphalt 1" Gauge 1 Gauge 4 Gauge 2Location 1 0.5" 0.5" Concrete Asphalt Figure 3-9 Vertical Positions of the Strain Gages in Phase I-b. 6" 32" 36" 7" 5 4" 15" 1 3" 2 4 3 36" Wheel Path Direction 25" 6" Thermocouple Gauge Figure 3-10 Instrumentation layout for Phase II.

PAGE 67

67 Figure 3-11 Vertical positions of the strain gages in Phase II. 1" 10" 2" 6" Asphalt 2" 2" 1" Concrete 2" 2" 2" 8" Asphalt Concrete Concrete Asphalt 2" 2" 2" 1" Figure 3-12 Vertical positions of thermocouples in Phase II.

PAGE 68

68 Figure 3-13 Milled surface before concre te placement on Lane 6 in Phase I-a. Figure 3-14 Formwork prepared for Lane 7 in Phase I-b.

PAGE 69

69 Figure 3-15 After placement of co ncrete on Lane 6 in Phase I-a. Figure 3-16 Placement of top and bottom strain gages on Lane 7 in Phase I-b.

PAGE 70

70 Figure 3-17 Placement of surface strain gages at a joint in Phase I-a. Figure 3-18 Strain gages in a protec tive PVC pipe before placing concrete.

PAGE 71

71 Figure 3-19 Removal of concrete slabs from Lane 6 in Phase I-a. Figure 3-20 Formwork for test slabs in Phase II.

PAGE 72

72 Figure 3-21 Grooves on asphalt surface for pl acement of strain gages and thermocouples cables in Phase II. Figure 3-22 Asphalt surface with white curing compound before concrete placement in Phase II.

PAGE 73

73 Figure 3-23 Finishing of the concrete for the test track in Phase II. Figure 3-24 Curing of concrete by sprinkling with water

PAGE 74

74 CHAPTER 4 MATERIALS AND PAVEMENT C HARACTERIZATION 4.1 Materials Characterization Tests were p erformed to characterize the paveme nt materials used in the test sections in this study. The properties measured include the concrete-asphalt in terface bond strength, compressive strength, splitting tensil e strength and elastic modulus of the concrete, resilient modulus and indirect tensile strength of the asphalt concre te, and the penetration a nd absolute viscosity of the recovered asphalt from the asphalt cores. 4.1.1 Interface Bond Strength 4.1.1.1 Results from test sections in Phase I-a Iowa shear tests were perform ed on both the 4inch and the 6-inch diameter core samples extracted from the 4-inch slabs in Phase I-a befo re loading started (at 28 days or later). The average shear strength for two 6-inch diameter core samples was 207.5 psi, while the shear strength for one 4-inch diameter sample was 165 psi. Six 6-inch diameter cores of the concrete/as phalt composite layer were extracted from each of the test sections at the end of the HVS testing in Phase I-a The locations of the cores were selected so that the bond strength for different conditions could be evaluated. For each test section, two cores were obtained from the wheel pa th (loaded area) one from the corner and one from the mid edge of the slab. Four cores were obtained outside the wheel path one from the center of an unloaded slab, one fr om the center of the loaded sla b, one from the mid edge of the longitudinal joint, and one from th e mid edge of the transverse joint. Figure 4.1 shows the core samples from the 6-inch slabs. Figure 4.2 shows the locations of the cores with the measured bond strengths displayed next to th em. Table 4-1 shows the results of the Iowa tests on the cores extracted from the test lane in Ph ase I-a (Lane 6) after HVS loading.

PAGE 75

75 An examination of the data shows that the measured bond strength was not affected by their locations on the pavement. No loss of bonding due to repeat ed loading was observed. The loaded area had equally high bond strength as the unloaded area. The average bond strength was about 220 psi. 4.1.1.2 Results from test s ections in Phase I-b For the test sections in P hase I-b, a total of 7 cores of the concrete/asphalt composite layer were extracted from each test section. Six of these samples were tested for interface bond condition after loading. Table 42 summarizes the results of the Iowa Shear Test run in these samples. It can be observed that the average shear strength was 195 psi. Similar to the case for Phase I-a, no significant difference in the shea r strength was observed among the cores taken from the corner, mid edge and center of the slabs both in and out the wheel path. This means that the area of the slab interface load ed for a short period of time did not experience more deterioration than that out of the wheel path. Figure 4-3 shows the locati ons of the cores taken after HVS loading for each test section in Phase I-b. 4.1.1.3 Results from test sections in Phase II Four cores were taken f rom the 10-inch sl abs in Phase II before the HVS loading was started, and Iowa shear tests were run on these cores to determine the interface shear strength. Though the concrete-asphalt interf ace was intended to be unbonded, the cores indicated that the concrete was partially bonded to the asphalt layer. The average interface shear strength from the four cores was 118.6 psi. After HVS loading, six cores were taken from each test slab, and Iowa shear tests were run on these cores for comparison purpose. The locations of the cores in the 6 x 6 slabs were the same as the one used for the Ph ase I-a, which is shown in Figure 4-2. Table 4-3 displays the

PAGE 76

76 results of the Iowa shear tests on these cores. The average shear strengt h for these cores was 135 psi. 4.1.1.4 Comparison before and after HVS loading Table 4-4 su mmarizes the measured interface bond strength before and after HVS loading for both Phase I and Phase II. It can be noted th at for the test sections in Phases I-a and I-b, where a bonded condition was intended, there wa s a small increase in bon d strength after HVS loading. For the test sections in Phase II, where an unbonded c ondition was originally intended, there was a larger gain in bond strength after HVS loading. 4.1.2 Concrete Properties 4.1.2.1 Properties of concrete sampled from concrete trucks Sa mples of concrete were taken from a sel ected truck during the pl acement of the test slabs. The slump, air content and temperature of the fresh concrete were measured. Samples were fabricated for compressive strength and elas tic modulus, and flexural strength tests. The properties of fresh concrete used in the construction of the test sec tions in Phases I-a, I-b and II are shown in Table 4-5. Additional concrete samples from Phase II were prepared to evaluate its coefficient of thermal expansion. The average value of this parameter was determined to be 6.5 X 10-6 1/F. Tables 4-6 and 4-7 show the comp ressive strength, elastic modulus and flexural strength at various curing times of the c oncrete sampled from the truck in Phase I-a and Phase II, respectively. All th ese tests were performed by FDOT personnel at the FDOT facility. 4.1.2.2 Properties of concrete from core samples Splitting ten sile strength test wa s run on the concrete portion of the core samples after the Iowa bond strength test. From the core samples obtained from the test sections in Phase I-a before the start of the HVS loading, the average indirect tensile strength of concrete from three

PAGE 77

77 samples was 610 psi. As described earlier, 18 core samples (6 from each test section) were taken from the test sections in Phase I-a after the HVS loading. Table 4-8 displays the results of the indirect tensile strength test on the concrete portion of these 18 core samples. The average indirect tensile strength ranged from 473 psi for the 4-inch concrete sl abs to 509 psi for the 5inch slabs. 4.1.3 Asphalt Concrete Properties Resilient m odulus test was performed on the as phalt portion of a core that was obtained from a 4-inch slab in Phase I-a before HVS load ing. This test was run by FDOT personnel at the FDOT facility. The resilient modulus at 10 C was determined to be 1,263 ksi. Resilient modulus and indirect tensile strength tests were also run on the asphalt portion of the cores taken from the test sections in Phase Ia, Phase I-b and Phase II at the end of each HVS loading period. These tests were performed in the Asphalt Lab of the Department of Civil and Coastal Engineering at UF. The results of these tests are summarized in Table 4-9. Figure 4-4 shows the effect of the temperature in the resilient modulus of the AC layer. From this graph it can be observed that the MR of the AC layer can be drastically reduce from 1800 ksi at 5 C to a value lower than 400 psi at 40 C. The adjusted curve and the equation shown in this Figure 4-4 will be used in the ne xt chapters for both calibrating the analytical model and estimating the level of stresses. The asphalt binders were extract ed and recovered from these asphalt concrete samples. Penetration tests at 25 C and the absolute visc osity test at 60 C we re run on the recovered asphalt binders. These tests we re performed by FDOT personnel at the FDOT facility. The test results are shown in Table 4-10. The recovere d asphalt binders were shown to be fairly consistent in properties with th e penetration ranging from 20 to 26, and the viscosity at 60 C varying from 61,000 to 73,000 Poises. The viscosity at 60 C of a recovered asphalt binder from

PAGE 78

78 a new pavement in Florida is generally in the range of 6,000 to 10,000 Poises. Thus, the recovered asphalt represented an asphalt bi nder which had been substantially aged. 4.2 Measurement of Joint Movement Two pairs of W hitmore plugs were placed at the joints of each test sl ab to monitor joint movement. Each pair of Whitmore plugs were placed at a distance of 6 inches apart from one another, and one on each side of the joint. Thes e plugs were fixed to concrete before the fresh concrete stiffened during placement. Figure 4-5 shows the Whitmore plugs fixed at a joint. The Whitmore gage with the standard Invar bar is shown in Figure 46. The invar bar is a reference bar which was used to calibrate the Whitmore ga ge. The distance between the gage points was measured in early morning before 7 AM and in mid afternoon around 3 PM, which represent the two extreme temperature conditions in a day. In addition, joint movements were measured every two hours from 6 AM to 5 PM on some selected days to monitor the slab movement throughout the day. Figures 4-7 through 4.9 show the measured Whitmore plug spacing from a 4-inch, 5inch and 6-inch slab in Lane 6 (with 6 ft by 6 ft joint spacing) at 7 AM and 3 PM in some selected days. Significant joint movement was ob served at the joints with extended cracks. The joint movement was minimal until the extended cracks were formed at the joints. Figure 4-10 shows the change of gage spacing for all three sl abs on a selected day. He re, the gage spacing is shifted to an original gage length of 10 inch es for comparison purpose. Thus, only the changes in gage spacing, rather than the absolute gage spacing, are to be read from this plot. 4.3 Measurement of Slab Pr ofile Using a Dipstick A grid was drawn on four test slabs with grid po ints spaced 12 inches apart. A Dipstick profiler was used to measure the elevation of the grid points with respect to a reference point established close to the test sl ab. Figure 4-10 shows the grid that covers four slabs. The coordinates (1, 0) represent the reference point. Figur e 4-11 shows the dipstick profiler with one

PAGE 79

79 leg of it placed on the reference point and the ot her on a point in concre te slab. The elevation collected at 7 AM and 3 PM were plotted as show n in Figure 4-12. It can be observed that this 4-slab unit curled up at the middle of the slab in the afternoon and curled up at the edges in the morning. The curling of these 4 slabs together agrees w ith the typical rigid pa vement behavior at positive and negative temperature differentials. This observation indicates that the 4 slabs act as one continuous slab before cr acking occurred at the joints. 4.4 FWD Tests Falling W eight Deflectometer (FWD) tests we re performed on the composite pavement test sections in all phases of the study. The measured FWD deflection basins were used to estimate the elastic modulus of the pavement mate rials and the stiffness of the springs used to model the load transfer at the joints and conc rete-asphalt interface through a back-calculation process. This back-calculation pr ocess also allowed for the verifi cation of the elas tic modulus of the concrete and the asphalt layer, previ ously evaluated from laboratory testing. FWD tests were run at midday between 1:30 PM and 3:30 PM and at early morning between 7 AM and 8 AM. At mid day, the temperature differential tends to be positive and slab tends to curl down at the edges and joints. This is the best time to run the FWD test for evaluation of joints because the slab is more likely to be in full contact with the layer underneath at both the edges and joints. From midnight to early morning, the temperature differential tends to be negative and the slab tends to curl down at the center of the slab. This is an ideal time to run the FWD test at the center of the slab for ev aluation of the condition of the concrete slab and the layer underneath. In order to reduce the effect s of the joints, the FWD test run in the morning was performed on a transition slab with the same slab thickness but with a larger panel size.

PAGE 80

80 4.4.1 FWD Tests in Phase I-a FW D tests were run on the 4 slab of the te st section in Phase I-a Figures 4-13 and 4.14 show the schemes of the morning and afternoon tests, respectivel y. The morning test was run at 7AM with an average pavement temperature of 65 F. The afternoon test was run between 2 PM and 4 PM with an average pavement temperature of 99 F. In the case of the afternoon test, a load of 12 kips was applied to both the corner an d at the mid-edge of the slab. In all cases the distance between sensors was 12 inches. 4.4.2 FWD Tests in Phase I-b FW D tests were run on the test sections in Phas e I-b. Three sets of reading were taken for three different loads (at around 9, 12, and 15 kips). A replicate test was ru n right after each test was completed to check for consistency. Figures 4-15 and 4-16 show the FWD load and sensor positions used for the FWD test at the slab corner and slab edge, respectively. These schemes correspond to those applied in the 4 x 4 slabs (in Phase I-b). The average pavement temperatures were 56 and 75 F for the morning and afternoon tests, re spectively. Figure 4-17 shows the FWD load and sensor positions used for the FWD test at the slab center. Figures 4-18 through 4-20 show pictures of the FWD tests run at the slab corner, slab edge and slab center, respectively. 4.4.3 FWD Tests Phase II Figures 4-21 and 4-22 show the FWD testing plan for the tes t sections in Phase II. This plan is very similar to the one used for Phase I-a since both used 6 x 6 slabs. Morning and afternoon tests were performed to evaluate the el astic modulus of the layers and load transfer characteristics at the joints, respectively. Three sets of reading were taken for three different loads (at around 9, 12, and 15 kips). The average pavement temperatur es were 78 and 108 F during the morning and afternoon tests, respectively.

PAGE 81

81 Using the deflections caused by the FWD load, an analysis of the load transfer was undertaken. The following load transf er factors where defined to re present the load transfer at both longitudinal and transversal joints: FLon = (D0 U)ME/(D0 L)ME x 100 (4-1) FTra = (D-12 U)C/(D0 L)C x 100 (4-2) Where: FLon = Longitudinal load transfer factor FTra = Transversal load transfer factor (D0 L)ME = Deflection under the FWD load on the loaded slab, when the test was run at the mid-edge (D0 U)ME = Deflection across the longitudinal joint, in front of D0 L, on the unloaded slab when the test was at the mid-edge (D0 L)C = Deflection under the FWD load on the loaded slab, when the test was run at the corner (D-12 U)C = Deflection across the transversal joint, 12 inches apart of D0, in the direction of the sensors, on the unloaded slab wh en the test was run at the corner. Figure 4-23 shows a comparative analysis of th e load transfer using the factors defined above. According with the definition as the factor approach to 1.0, more load is transferred from one slab to the adjacent slab. While these factors cannot be used to determin e the degree of load transfer for composite pavements, they can be used to compare th e load transfer based on different pavement characteristics. It is clear from the graph that the load transfer at bot h the longitudinal and the transversal joint increases as the thickness of the slab increases. Similarly, there is more load transfer at the transversal joints than at the longitudinal joints. By comparing the 6 slab in both phases, it seems that the shorter slabs can tran sfer more load. However this result might be influenced by the fact that Phase I-b consider ed a bonded condition in the interface while Phase II was intended to be unbonded. It may be possibl e that the bond in the interface co llaborated with the load transfer sin ce the AC layer is continuous.

PAGE 82

82 Table 4-1 Results of the Iowa Sh ear Tests on the cored samples from test sections in Phase I-a after HVS loading. Testing Date Sample # Diameter (mm) Diameter (in) Area (in2 ) Load (lbs.) Shear Strength (psi) 6/28/2005 1-4 151.02 5.946 27.76 5860 211.1 6/28/2005 2-4 151.02 5.946 27.76 5800 208.9 6/28/2005 3-4 151.33 5.958 27.88 6000 215.2 6/28/2005 4-4 151.00 5.945 27.76 5720 206.1 6/28/2005 5-4 151.24 5.954 27.85 5660 203.3 6/28/2005 6-4 150.98 5.944 27.75 6440 232.1 6/28/2005 1-5 151.24 5.954 27.85 5680 204.0 6/28/2005 2-5 151.21 5.953 27.83 5980 214.8 6/30/2005 3-5 151.26 5.955 27.85 5940 213.3 6/28/2005 4-5 151.14 5.950 27.81 6860 246.7 6/30/2005 5-5 151.27 5.955 27.86 7360 264.2 6/30/2005 6-5 150.95 5.943 27.74 5992 216.0 6/30/2005 1-6 151.19 5.952 27.83 6000 215.6 6/30/2005 2-6 150.99 5.944 27.75 6140 221.2 6/30/2005 3-6 151.12 5.950 27.80 6040 217.3 6/30/2005 4-6 151.23 5.954 27.84 5020 180.3 6/30/2005 5-6 151.05 5.947 27.78 3900 140.4* 6/30/2005 6-6 151.22 5.954 27.84 7520 270.1 Average5995 215.6 St. Dev.799 28.7 Average without specimen 5-66118 220.0 St. Dev. without specimen 5-6622 22.3 Outlier, excluded in the computation of average and standard deviation.

PAGE 83

83 Table 4-2 Results of Iowa Shear Tests on the cored samples from te st sections in Phase I-b after HVS loading. Core ID Diameter (in.) Cross Sectional Area (in2 ) Load (lb.) Shear Stress (psi) L7-1A 5.99 28.185340 189.5 L7-2A 5.98 28.095230 186.2 L7-3A 5.98 28.095450 194.0 L7-4A 5.98 28.095030 179.1 L7-5A 5.97 27.995500 196.5 L7-6A 5.95 27.817040 251.5 L7-1B 5.97 27.997040 251.5 L7-2B 5.98 28.095590 199.0 L7-3B 5.97 27.995650 201.8 L7-4B 5.97 27.996270 224.0 L7-5B 5.97 27.994520 161.5 L7-6B 5.97 27.995050 180.4 L7-1C 5.96 27.907080 253.8 L7-2C 5.97 27.995770 206.1 L7-3C 5.99 28.185680 201.6 L7-4C 5.98 28.094500 160.2 L7-5C 5.97 27.995130 183.3 L7-6C 5.97 27.994780 170.8 Average: 194.9 Standard Deviation: 26.6 Minimum: 160.2 Maximum: 253.8

PAGE 84

84 Table 4-3 Results of Iowa Shear Tests on cores fr om test sections in Ph ase II after HVS loading. Core ID Diameter (in.) Cross Sectional Area (in2 ) Load (lb.) Shear Stress (psi) L6-1A 5.94 27.714720 170.3 L6-2A 6.00 28.273980 140.8 L6-3A 5.97 27.994140 147.9 L6-4A 5.98 28.09 n/a n/a L6-5A 5.97 27.993800 135.8 L6-6A 5.97 27.994250 151.8 L6-1B 5.99 28.184300 152.6 L6-2B 5.95 27.81 n/a n/a L6-3B 5.91 27.432140 78.0 L6-4B 5.94 27.713550 128.1 L6-5B 5.94 27.713190 115.1 L6-6B 5.96 27.904200 150.5 L6-1C 5.95 27.814990 179.5 L6-2C 5.94 27.714280 154.4 L6-3C 5.94 27.711730 62.4 L6-4C 5.94 27.714270 154.1 L6-5C 5.95 27.812640 94.9 L6-6C 5.95 27.81 n/a n/a Average: 134.4 Standard Deviation: 33.3 Minimum: 62.4 Maximum: 179.5 Table 4-4 Summary of the interface bond st rength before and after HVS loading. Intended Bond Condition Slab Size Shear Strength Before Loading (psi) Shear Strength After Loading (psi) Bonded (Phase I-a) 6 6 207.5 220 Bonded (Phase I-b) 4 4 -194.5 Un-bonded (Phase II) 6 6 118.6 134.4

PAGE 85

85 Table 4-5 Properties of fresh concrete used. Properties Phase I-a Phase I-b Phase II Slump, inch 3.5 5.75 3.0 Air, % 2.50 2.00 2.4 Unit Weight, pcf 145.8 142.9 143.2 Temperature, F 93 92 78 Table 4-6 Properties of hardened conc rete sampled from truck in Phase I-a. Curing Time, Days Compressive Strength, psi Elastic Modulus, ksi Flexural Strength, psi 1 1,690 3 2,940 7 3,930 3,440 14 4,750 3,737 732 28 5,980 3,940 772 56 6,750 4,380 847 Table 4-7 Properties of the hardened c oncrete sampled from truck in Phase II. Curing Time, days Compressive Strength, psi Elastic Modulus, ksi Flexural Strength, psi 1 1,933 3 3,608 7 4,651 3,307 14 3,875 808 28 6,083 4,004 855 56 6,612 4,272

PAGE 86

86 Table 4-8 Results of Indirect Te nsile Strength Test on the concrete samples taken from the test sections in Phase I-a after HVS loading. Diam L1 L2 L3 L4 L Load Strength Sample in in in in in in lb psi 4-1 6 3.81 3.863.663.783.7814610 410.4 4-2 6 3.94 4.043.924.204.0314900 392.8 4-3 6 4.11 4.184.144.044.1217199 440.6 4-4 6 4.02 4.094.104.144.0924220 628.7 4-5 6 4.43 4.264.304.324.332510* 4-6 6 4.44 4.494.554.524.5020880 492.3 Average 473.0 5-1 6 5.40 5.255.385.345.3419121 379.7 5-2 6 5.29 5.185.365.185.2522270 449.9 5-3 6 5.29 5.315.235.235.2734690 699.1 5-4 6 5.40 5.465.455.495.4530050 585.0 5-5 6 5.45 5.475.475.525.4825720 498.2 5-6 6 5.67 5.445.895.475.6223320 440.5 Average 508.7 6-1 6 6.31 6.236.256.256.2625440 431.2 6-2 6 6.18 6.206.156.166.1726840 461.4 6-3 6 6.12 6.186.006.006.0827340 477.5 6-4 6 6.13 6.206.106.116.1431150 538.7 6-5 6 6.00 6.006.016.036.0130640 540.9 6-6 6 5.94 5.915.946.005.9527570 491.8 Average 490.3 Outlier Table 4-9 Results of Resilient Modulus and Indirect Tensile Stre ngth Tests on the asphalt concrete samples obtained from test se ctions in all phases after HVS loading. T ( C) MR (ksi) Strength (psi) 5 1,780 10 1,390 15 1,170 272 20 916 25 750 30 590 40 325 78

PAGE 87

87 Table 4-10 Results of Penetration and Absolute Viscosity Tests on the recovered asphalt binders from cores from Phase I-a after HVS loading. Sample Needle ID Pen (@ 25 C) Avg. Pen Tube Bulb ConstantSeconds Viscosity (Poises) (@ 60 C) A-498 20 B-171 21 Core 4 B-464 22 21.00 400/R164 D 810.0090.30 73,143 A-498 23 B-171 23 Core 5 B-464 22 23.00 400/E171 D 684.0096.30 65,869 A-498 23 B-171 25 Core 6 B-464 26 25.00 400R/E357D 872.0069.60 60,691

PAGE 88

88 Figure 4-1 Cores samples from Lane 6 in Phase I-a after HVS loading.

PAGE 89

89 270.1 232.1 216 3 1 5 221.2 215.6 1 217.3 2 3 180.3 4 140.46" 6'x6' Shear Strength (psi)214.8 204.0 2 213.3 246.7 4 Shear Strength (psi) 5" 6'x6'208.9 211.1 215.2 2 264.2 5 4 206.14" 6'x6' Shear Strength (psi)1 3 203.3 5 6 6 6 Figure 4-2 Location of the cores taken after loading in Phase I-a.

PAGE 90

90 Location of cores (4'x4' Slabs) 7 2 1 3 4 6 5 Figure 4-3 Location of the cores taken after loading for each test section in Phase I-b. Effect of the Temperature in the AC Elastic ModulusMR = 2321.9e-0.0473TR2 = 0.9929 0 200 400 600 800 1000 1200 1400 1600 1800 2000 01 02 03 04 05 0 Temperature (C)Resilient M odulus (ksi ) Figure 4-4 Relationship between Temperature and Resilient Modul us for the AC layer in the composite pavement.

PAGE 91

91 Figure 5.4. A Pair of Whitmo re Plugs Fixed at a Joint. Figure 4-5 The Whitmore gage with invar bar.

PAGE 92

92 4 Inch slab-0.04 0.01 0.06 0.11 0.16 0.21 0.266/66/86/106/126/146/166/186/206/226/246/266/286/307/27/4Measured Gauge Spacing, in North-7.00AM South-7.00AM North-3.00PM South-3.00PM Figure 4-6 Measured gage spacing from a 4-inch slab in Lane 6. -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.166/66/86/106/126/146/166/186/206/226/246/266/286/307/27/4Measured Gauge Spacing, in North-7.00AM South-7.00AM North-3.00PM South-3.00PM Figure 4-7 Measured gage spacing from a 5-inch slab in Lane 6.

PAGE 93

93 -0.03 -0.01 0.01 0.03 0.05 0.07 0.096/66/86/106/126/146/166/186/206/226/246/266/286/307/27/4Measured Gauge Spacing, in North-7.00AM South-7.00AM North-3.00PM South-3.00PM Figure 4-8 Measured gage spacing from a 6-inch slab in Lane 6. 9.99 10 10.01 10.02 10.03 10.04 10.05 10.06 6:00 AM8:00 AM10:00 AM12:00 PM2:00 PM4:00 PM6:00 PMJoint space, inch 6-inch 5-inch 4-inch Figure 4-9 Changes of joint spacing on a selected day.

PAGE 94

94 Figure 4-10 Grid marked on slabs for the Dipstick measurement. Figure 4-11 The Dipstick instrument. Slab Joints

PAGE 95

95 Figure 4-12 Dipstick measurements at two critical temperatures.

PAGE 96

96 Figure 4-13 FWD load and sensor locations fo r FWD Test at slab center in Phase I-a. Figure 4-14 FWD load and sensor locations fo r FWD Test at slab corner in Phase I-a.

PAGE 97

97 71 4 ft 4 ft 7 7 7 1 2 34 5 4 2 3 3 2 4 6 5 6 5 6 CORNER LOADING2 3 4 5 6 Figure 4-15 FWD load and sensor locations for FWD Test at the slab corner in Phase I-b. CENTER EDGE LOADING 7 1 23 3 2 7 7 1 2 3 4 5 6 4 ft 4 ft 7 2 3 4 5 6 6 5 4 5 4 6 Figure 4-16 FWD load and sensor locations for FWD Test at the slab edge in Phase I-b.

PAGE 98

98 6 ft 6 ft 6 547 2 3 75 34 2 6 10' 2" 10'2" Figure 4-17 FWD load and sensor locations for FWD Test at the slab center in Phase I-b. Figure 4-18 FWD Test at the slab corner and measuring deflections on the opposite slab.

PAGE 99

99 Figure 4-19 FWD Test at the mid-edge and measuring deflections in the loaded slab. Figure 4-20 FWD Test at the cent er and measuring deflections al ong the transverse center line.

PAGE 100

100 CENTER EDGE LOADING (PM TESTING PLAN) D1 positioned beneath applied load 6 ft 6 ft 30" 7 7 4 4 1 2 3 2 3 5 6 5 6 CORNER LOADING (PM TESTING PLAN) 6 ft 7 1 7 6 ft 32 4 3 2 4 5 6 65 D1 positioned beneath applied load Figure 4-21 FWD Testing Plan for the mi d-edge and corner load in Phase II. 10'9" D1 positioned beneath applied load CENTER LOADING (AM TESTING PLAN) 6 ft 6" Slab End 6 ft 3.08' 2 7 7 2 3 45 645 3 6 6' 10" 6 ft 6 ft 10" Slab End Load 3 7 3 2 7 2 4 5 6 64 5 Load 10' 3" 6' 3" Figure 4-22 FWD Testing Plan for the center load at the two ends of the test track in Phase II.

PAGE 101

101 Load Transfer at Joints65 70 75 80 85 904 (Phase I-b)5 (Phase I-b)6 (Phase I-b)6 (Phase II)8 (Phase II)10 (Phase II)Thickness (in)Deflection Ratio (R) Transversal Longitudinal Figure 4-23 Comparative analysis of load transfer factor for the test sections in Phase I-b and II.

PAGE 102

102 CHAPTER 5 TESTING OF TEST SECTIONS AND DATA ANALYSIS 5.1 HVS Loading of Test Sections The HVS testing has the primarily objectiv e of providing the nece ssary data for the calibration of the computer model. HVS testing of the test sections was performed to measure strain, evaluate temperature effect and even tually observe perform ance of the composite pavement under load repetition. The following paragraphs are a description of the HVS test for each phase. 5.1.1 HVS Loading of Test Sections in Phase I-a The three test sections on Lane 6 (in Phase I-a), with a panel si ze of 6 ft x 6 ft and concrete slab th ickness of 4, 5 and 6 inches, were loaded by the HVS according to the plan as described in Section 3.4. HVS loading of the test section with 4-inch slabs was started on July 11, 2004. This test section performed well with no observed cracking after 3 days of loading at 9 kips with a total of 36,407 passes, followed by 12 days of loading at 12 kips with a total of 146,748 passes. Then the load was increased to 15 kips for 3 days with 35,9 18 passes and then to 18 kips for two days with 21,727 passes. Finally, the load was increased to 21 kips, and corner cracks developed after 12,187 passes, as shown in Figure 5-1. This wa s the only load-induced crack observed during this phase and in the entire experiment. It is to be noted that the HVS loading had to be suspended for two times during this entire te st duration due to mechanical problems. As indicated, the application of a very large wheel load and a hi gh number of load repetitions were necessary to break the thinnest slab in the full sc ale experiment. At that time, it was realized that for the thicker slabs it would take much more time to break them and the experiment was

PAGE 103

103 confined to study the level of st rains and stresses due to normal wh eel load, which is 12 kips per wheel (24 kips per axle). HVS loading of the test section with 5-in ch slabs was performed from November 1, 2004 through January 14, 2005. This test section was loaded with 79,014 passes of 9-kip, 130,186 passes of 12-kip, 25,638 passes of 15-kip and finally 128,817 passes of 18-kip HVS wheel load. No load-induced cracks were obs erved during and at the end of th is testing period. Only a few small shrinkage cracks were obser ved on the surface of the concrete. HVS testing of the test section with 6inch slabs was performed from May 23, 2005 through May 26, 2005. This test section was loaded for three days with a total 41,239 passes of 12-kip wheel load. No load-induced cracks were observed, and only a few small shrinkage cracks were observed on the surface of the concrete. Shrinkage cracks were observed in most of the slabs in Phase I-a before the loading period. Shrinkage cracks on the test slabs of the 4, 5 and 6-inch test sections are shown in Figures 5-2 through 5-4, respectively. In addition, a lot of sh rinkage cracks in the tr ansverse direction were observed in the transition zones wh ere the thickness changed from 4 inches to 5 inches and from 5 inches to 6 inches. 5.1.2 HVS Loading of Test Sections in Phase I-b The three test sections on Lane 7 (in Phase I-b), with a panel si ze of 4 ft x 4 ft and concrete slab thickness of 4, 5 and 6 inches, were load ed with 12-kip HVS wheel loads at 120 psi tire pressure. Table 5.1 shows the dates of the test and the number of wheel passes applied to each of the three test sections. No lo ad-induced or shrinkage crack was observed in any of the three test sections at the end of each loading period.

PAGE 104

104 5.1.3 HVS Loading of Test Sections in Phase II The test sections in Phas e II with a panel size of 6 ft x 6 ft, concrete slab thickness of 6, 8 and 10 inches, and un-bonded concrete-asphalt interface were loaded by three levels of HVS wheel loads to study the linearity of the load-strain relationship a nd eventually observe any loadinduced crack. Table 5.2 summarizes the HVS loads, numbers of wheel passes and loading periods for the test sections in Phase II. Unfortunately no load-induced or shrinkage cracks were observed in any of the slabs at the end of the tests. 5.2 Analysis of Temperature Data Temperature data were analyzed in three ways. First the temperature differential was estimated based on the values from the thermo couple located near or on the surface of the concrete slab and the thermocoupl e located near the bottom of the slab or the one located in the surface of the asphalt layer. This value was to be used to estimate the level of stresses under the combination of wheel load and temperature effect using the analytical model in later chapters. Secondly, the absolute temperat ure in the AC layer was es timated directly from the thermocouple located in the surface of the layer. This value was used to estimate the effect of the temperature on the ACs elastic modulus and thus its effect on the composite pavement as a supporting layer. Finally the distri bution of the temperature along th e depth of the concrete slab was analyzed to better address its effects on the stresses. An investigation of the effect of both the temperature differential and the temperature in the AC layer on the peak strain was also undertaken, which is shown in the next section in this chapter. 5.2.1 Temperature Differential In Phase I-a, two sets of therm ocouple wires were used to monitor the temperature in the slabs. Thermocouples were installed at various depths, at one-inch increments, in the test sections to monitor the temperature distribution. One set of thermocouples was installed at the

PAGE 105

105 slab corner at the side that would be loaded by the HVS wheel and the other set of thermocouples was installed at the slab center. Temperature differentials between the top and the bottom of the slab were computed and plotted against time for th e 4-inch, 5-inch, and 6-inch slabs in Figures 55, 5-6 and 5-7, respectively. It was observed that for the 4-inch slab, the te mperature differential at the slab corner was relatively low compared with that at the slab cente r. In all cases, the slab corner was under the shade of the HVS during the day time, and thus di d not get as much heating from the sun as the slab center did. This explains why the slab corner had lower positive temperature differentials than the slab center, as the slab corner was less heated by direct sunlight in the day time. However, the temperature differentials computed for the 5-inch slab did not show a significant difference between slab center and slab corner. It is to be note d that the temperature readings from the 5-inch slab were taken during the winter, while those from the 4-inch slab were taken in the summer. For the 6-inch concrete slabs, temperature was recorded between May 20 and May 25, 2005. In this case, temperature data were reco rded at one minute intervals for the first three days and at 5-minute interv als for the other days. For the 6-inch slabs, the maximum positive temp erature differential at the slab corner was +18.5 F (+10.28 C), while the ma ximum negative at the slab corn er was -8.6 F (-4.8 C) during the HVS loading of these slabs. For the slab center, the maximum positive and negative values were +25 F (13.9 C) and -5.2 F (-2.9 C), respectively The temperature differentials for the 4-inch and 5-inch slabs, which were monitored and loaded a longer time compared with the 6 slab, were analyzed. Stat istical analyses were performed to determine the number of hours in a day when the temperature differential was in a certain range. Table 5.3 shows the results of th e statistical analysis for the 4-inch slab in

PAGE 106

106 different months during the tes ting period. The 4-inch slab was tested during the summer of 2004 and the thermocouple used to do th is analysis was located under th e shade of the HVS. That is the reason why most of the time the temperature differential was between -5 C and +5C. Table 5.4 shows the results of the statistic al analysis for the 5-inch slab. The 5-inch slab was tested during the winter of 2004 and it can be observe d that most of the time the temperature differential was between -5 C and +5 C Similar to the case in Phase I-a, temperature in the 4x 4 concrete slabs in Phase I-b was measured by means of thermocouples installed in the slabs. They were placed at one inch apart vertically in the slab at two diffe rent locations, namely at the center and at the corner of the slab. The temperature data were collected (1) betwee n June 17 and June 22, 2005 for the 6-inch slabs, (2) between June 13 to June 17, 2005 for the 5-inch slabs, and (3) between June 9 and June 12, 2005 for the 4-inch slabs. Figures 5-8 through 5-10 show the plots of te mperature differentials versus time in Phase I-b, as measured at the slab edge and at the slab center of the 6-, 5and 4-inch slabs, respectively. The maximum positive and negative temperature differentials were (1) +28.8 F (+16 C) and -6.3 F (-3.5 C) for the 6-inch slabs, (2) +27 F (+15 C) and -5.4 F (-3 C) for the 5-inch slabs, and (3) +18 F (+10 C) and -2.7 F (-1.5 C) for the 4-inch slabs during their respective test periods. Temperature data in Phase II were collected every 5 minutes during the loading period. Thermocouples were installed in three differe nt locations in the slab. At each location, thermocouples were installed at 1 inch from the top of the slabs and then at 2-inch increments along the thickness of the slab. One thermocoupl e was installed in the interface between the concrete slab and the asphalt surface. Figures 5-11 thr ough 5-13 show the temperature

PAGE 107

107 differential for the 10-, 8and 6-inch slabs, re spectively. The slabs in Phase II were loaded during winter time, and for this reason the maximum positive temperature differential was less than 15 F. The maximum negative temperature diffe rential reached -10 F in the 10-inch slab. The actual temperature differential should be sli ghtly higher since the thermocouple used to estimate the temperature at the top was located at 1 inch below the surface of the slab. 5.2.2 Temperature in the AC layer Figures 5-14 through 5-16 show the variation of the temperature on the surface of the asphalt concrete layer du ring the loading period in Phase I-a. This is an important variable to be considered in the analysis since the elastic modulus of the asphalt layer varies substantially with the changes in temperature as show n in previous chapter, affecti ng the behavior of the composite pavement. In the cases of 4and 5-in te st sections, which were loaded and monitored for a longer time compared with the 6 slab, a s easonal variation of the temperature in the AC layer could be observed. While the 4 slab kept a temperatur e between 25 and 35 C during the whole summer of 2004, the 5 slab varied from 30 C in the Fall of 2004 to 10 C in the Winter of 2004-2005. The 6 slab in Phase I-a was monitored for onl y 4 days and a variation of no more than 12 C was observed. Figures 5-17 through 5-19 show the temperat ure on the surface of the asphalt layer in Phase I-b. All these temperatures were measured during the very hot summer of 2005, and temperatures in the range of 78-106 F (25 41 C ) were observed. The thinnest test section in this phase (4) was the one that showed less vari ation in the AC temperature. On the other hand, the thicker slab (6) showed the gr eatest AC temperature variation.

PAGE 108

108 Figures 5-20 to 5-22 show the temperature in the AC layer for Phase II. Since the loading period was during the winter time, the temperature in the asphalt layer was in the range of 50 to 73 F (10 to 23 C). Unlike the test sections in Ph ase I-b, all test sections in Phase II showed a similar variation in the AC temperature. 5.2.3 Temperature Distribution Figures 5-23 through 5-25 show the tem per ature distribution al ong the depth of the concrete slab for the cases of maximum positive and maximum negative temperature differential. A significant difference can be observed in the temperature distribution when comparing the cases of positive and negative temperature differe ntials. The distribution tends to be non-linear under a negative temperature differe ntial condition. Also the distri bution tends to be non-linear for thicker slabs and the non-linearity is more severe when considering the temperature measured in the center of the slab compared with the one measured in the corner. It seems that the shade provided by the HVS has also an eff ect on the temperature distribution. Figures 5-26 through 5-28 show the temper ature distribution al ong the depth of the concrete slab in Phase I-b. Similar to the previo us phase, some differences can be observed from the temperature values measured by thermocouples at different locations. Again the nonlinearity tends to increase as the thickness of the slab increases and that is also affected by the shade of the HVS. Figures 5-29 through 5-31 show the temperat ure distribution along th e slabs for the tests sections in Phase II, which show the same trends as in the previous phases. 5.2.4 Summary of temperature analysis Table 6-5 sh ows a summary of the extreme values for temperature differential and the absolute temperature on the AC layer for all the sl abs in the different phases. It can be noticed

PAGE 109

109 that a maximum temperature of 106 F (41 C) a nd a minimum of 48 F (8.9 C) were reached on the surface of the asphalt concrete layer depending on the season. After analyzing all the cases of temperature distribution along the depth of the concrete slab, the following can be stated: (1) Temperature distribution tends to be linea r when the slab is directly heated by the sunlight. During the day, light sun heats the slab uniformly so that the heat flows rapidly along the depth of the slab allowing th e temperature distributi on to be linear for positive temperature differential (2) Non-linearity of the temp erature distribution is commonl y observed when the slab is affected by negative temperature differentials. It seems that the surface of the concrete slab cools faster than the rest of the slab (3) Non-linearity of the temper ature distribution increases as the thickness of the slab increases 5.3 Analysis of Strain Data 5.3.1 Dynamic Strain versus Static Strain The first type of analysis using the m easured strains collected from the test sections was a comparison between maximum strains caused by dyna mic and static loads. The maximum strains (compression or tension) measured when the slab was loaded by a moving HVS wheel load were compared to the corresponding maximum stra ins when the same HVS load was applied statically. Table 5.6 shows the maximum static and dynamic strains as measured by Gages 1, 2 and 5 of the 4-inch slab in Phase I-a, when the wheel load of various magnitudes was applied at the mid-edge or corner of the slab. Figure 5-32 shows the comparison of static and dynamic strain for Gage 1. Comparison of measured st rains caused by dynamic loads and static loads indicated that they were very close to one another for this gage, which was in compression. However, the difference between static strain and dynamic strain increased with time due possibly to the micro cracks induced in the conc rete. There were significant differences between the strains measured for dynamic a nd static loads at Gages 2 and 5, which were in tension. It

PAGE 110

110 appears that the effects of micro cracks were mo re significant when the concrete was in tension than when it was in compression. Figure 5-33 shows the comparison of static and dynamic strains for Gages 2 and 5. Comparisons were also made for the static a nd dynamic strains for the strain gages in the 5-inch slab in Phase I-a. Dynamic strains cause d by slow moving HVS loads at 1 mph were also measured to detect any positioning errors and to determine the strains measured at slow speed. Table 5.7 shows the static and dynamic strains for Gages 2 and 4 caused by 9 and 12 kip loads. Dynamic strain measurement at slow speed was ma de at the end of the testing. Table 5.8 shows the strains data obtained with slow-speed loads at 1 mph, along with the strain s obtained with static loads and loads at 8 mph, caused by 15 and 18 kips wheel loads. The plots of dynamic and static strains for Gage 2 and 4 are shown in Figures 5-34 and 5-35, respectively. Static and low-speed load produce higher stra ins compared with dynamic 8mph load since in the first two cases the loading time period a llows the pavement to completely develop the strain. During the dynamic load, a ny point of the pavement is load ed for a very short period of time. 5.3.2 Measured strain and calculated strain Dyna mic strain data were collected at each hour for 30 seconds during the loading period. Figure 5-36 shows a plot of typical strain data for two gages at the mid-edge of the slab. One gage was located at a depth of 1 inch from the t op of the slab, and the other gage was located 1 or 0.5 above the bottom of the concrete slab. Th e positions of the strain gages for the different phases have been previously shown in Figures 3-5 and 3-6, Chapter 3. Figure 5-36 is an example taken from Phase I-a and it clearly depicts the strain reversal that was observed during the passing of the wheel load in all test sections. For the gage located near the top of the slab in the mid edge, as the wheel approached the gage location, tensile strain

PAGE 111

111 was measured by the gage. When the wheel was directly above the gage location, the strain reversed to a high value of compressive strain, and when the wheel moved away from the gage location, the strain again reversed to a tensile st rain. Strains of opposite signs can be observed for the strain gage located near the bottom of the slab. The peak strain reversal happens at the time the wheel load is positioned at the corner of the slab. To isolate the effect of the wheel load and to overcome the fact that the strain gages cumulated a lot of strain at each reading, the st rain profile shown in Figure 5-36 was zeroed so that it looked like the one depicted in Figure 5-37. Then the peak load-induced strain in compression (at the top gage) and tension (at th e bottom gage) were extracted from each 30-sec data file. Where a strain gage was installed, the peak strain on the surface of the AC layer was also extracted. Peak load-induced strain here refers to the highest value of strain observed in the 30-sec strain data as shown in Figure 5-37, which happened when the wheel load passed directly above the strain gage location. Maximum strains correspond to th e strains calculated at the very top and at the very bottom of the concrete slab. These strains were ca lculated based on the peak strain as measured by gages near the top and near th e bottom of the slab and assuming a linear variation of the strain along the depth of the slab. Verification of the vertical position of the embedded stra in gages was undertaken to improve the accuracy in the calculation of maximu m strains. Small variation in the position of the gages in thin slabs in Phase I can drastically affect the resu lt of the strain analysis. Some vertical positions of the gages were verified through cores taken at the locations of the strain gages after HVS testing. A total of 7 gage positi ons were verified in Phase I and 9 were verified in Phase II.

PAGE 112

112 Knowing the exact position of the strain ga ges allows for a better estimation of the maximum strain at the top and at the bottom of the slab. Most of the analysis involving strain was done with the data collected for the smaller load in the experiment (12 kips) so that the concrete slab behaved as a linear elastic material. In the cases where th e exact position of the strain gages was not known, the targ et position was used. In Phase I the top gage was planned to be 1 below the top surface of the slab and the bottom gage was intended to be 0.5 above the bottom surface. In Phase II each gage was 1 away from the respective concrete surface. 5.3.3 Effect of Tempera ture on the Strain The peak strain extracted from the data file as explained in the pr evious section did not completely consider the effect of the temperatur e differential. As indicated before the strain profile for each 30-sec data set wa s zeroed not only to isolate the load-induced strain but also to delete the strain cumulated in the gages after each reading. Each strain data set was 30 sec long with 4 or 5 peaks that lasted only 0.6 sec each (F igure 5-36). The fastest temperature differential change measured during the HVS test was 4 F in 10 minutes. Because both the temperature differential and the correspondent temperature-induced strain remained almost constant within the 30-sec loading period, the effect of the temperat ure was lost at the end of the zeroing process. The wheel load-induced strain can be a functi on of the temperature, since its magnitude can be affected by the slab curling due to a te mperature differential. When the slab curls the wheel load might induce more strain in the slab. In the composite pavement, if the bond between th e layers is perfect, the slab is completely restricted to curl, then the load-induced strain is no longe r a function of temperature. The effect of the temperature on the load-induc ed strain was investigated using the data collected during the HVS test. Plot s of strain versus both temperat ure differential in the concrete slab and temperature in the AC layer were developed. The temperature measured for each

PAGE 113

113 thermocouple was obtained at the time when the strain data was collected. As mentioned in Chapter 3, temperatures values were measured at 5 minutes intervals and strain values were recorded at 60 minutes intervals. The strain extracted from each data file (Figure 5-37) is the load-induced strain. Figures 5-39 through 5-45 show plots of measured strain versus temperature differential in the concrete slab. It can be noticed that the meas ured load-induced strain did not vary too much with temperature differential. Th is can be interpreted as the slabs were bonded well enough to the asphalt layer so that very little or no curling occurred. The slabs in Phase II were intended to be fully un-bonded, however the resu lt from this analysis showed that they have a considerable degree of bond. Temperature in the AC layer was found to have an effect on the load-induced strain, as it can be seen from Figures 5-46 through 5-52. This re sult demonstrates the f act that temperature in the asphalt layer affects the ACs elastic modulus, and therefore the AC as a supporting layer. As the temperature increases, the elastic modulus of the AC layer decrease s and therefore both the compressive strain and the tensile strain in the concrete slab increases. These figures also show the variation of the strain in the surface of the AC layer with respect to the temperature in that layer. The ACs elastic modulus tends to decrease when temperat ure increases. At the same time stresses in the AC layer tend to decrease since th e higher stiffness of the concrete makes the slab to take more stresses compared to the AC layer. Therefore the strain in the AC layer can increase or decrease depending on how small the new stre ss in the AC layer is. That is probably the reason why a trend cannot be observed in these plots for the tensile strain in the asphalt layer. 5.3.4 Effect of the load m agnitude on the strain Test section s in Phase II were also loaded with higher wheel loads to investigate both the performance of the composite pavement and the relationship between strain and load magnitude. In addition to the 12 kips wheel load, the test sec tions in Phase II were loaded with 15 kips for 5

PAGE 114

114 days followed by another five days with a 18-ki ps wheel load. As mentioned earlier in this chapter, no load-induced cracks were observed in any of the loading period in any of the test slabs. In Figures 5.53, 5.54 and 5.55, the average strains were calculated fo r each loading period and plotted versus load to s how the strain-load relationship. Figures 5-53 and 5-54 show a linear relationshi p between strain and lo ad in both locations at the top and at the bott om of the concrete slab. The strain ga ge at the bottom of the 6 slab in the corner did not work properly and thus was not included in the plots. Figure 5-55 shows the case of the tensile strain in the AC layer as a function of the load. Though a fairly linear relationship between load and strain can be obs erved for the gages located at the mid-edge, a certain degree of non-linearity can be observed in the gages located at the corner. In Figure 5-55, it can also be observed that as the slab becomes th icker and the load increases, the strain tends to remain constant or decrease. By inspecting the plot s involving the strain in the AC layer, it seems that this layer become more rigi d as the load increases. 5.3.5 Effect of the loading period on the strain Plots of strain versus number of passes of the wheel load were developed to show the effect of the loading period. The strains in thes e plots (peaks and m aximum) were extracted from the strain data files in the same way as it was done in the previous section. These strains can be considered as pure load-induce st rains since they do not include the temperature-induced strain component ( eT) and very little cu rling occurred in the slabs as shown in previous sections. Figure 5-56 shows a plot of the peak load-i nduced strains measured at different gage locations as a function of the number of passes of the wheel load for the 6-inch slabs in Phase Ia. Gages 1 and 2 were located at the mid edge (Location 1) at the top and bottom of the slab, respectively. Gage 3 was located at the top of the concrete slab in Location 2. Gage 4 was

PAGE 115

115 located at the bottom of Location 3, in the corner of the slab. Similarly, Figure 5-57 shows a plot of the maximum strains at Location 1 for the 4-inch slab in Phase I-a. For the 5-inch slabs in Phase Ia, the strain gage at the top of the slab at Location 1 (Gage 1) was out of order from the beginning of th e test, and thus maximu m strains could not be estimated from the peak strains. It can be observed from Figures 5-56 and 5-57 that no appreciable change in the load -induced strains occurred over the testing period. This may indicate that no crack had occurred near the lo cations of the strain gauges during the testing period for the specific load shown in the plot. Similar to the procedure used in Phase I-a, st rain data in Phase I-b were collected every hour during HVS loading of each test section, at the rate of 100 measurements per second for 30 seconds. In addition to the peak strain at the top and at the bottom of the concrete slab as indicated in Phase I-a, the strain s in the AC surface were also extracted from each strain record. Figure 5-58 depicts the maximum load-induced strains at the t op and at the bottom of the concrete at Location 1 in the 6-in ch slabs as a function of the nu mber of HVS wheel passes. This figure also shows the strains at the surface of th e asphalt layer. Figure 5-59 shows the strain measured by the gage located at the bottom of the concrete slab and the gage located on the surface of the asphalt, both of them at the corner of the 6 slab. Figure 5-60 shows similar plots of load-induced strains versus number of HVS wh eel passes for the 5-in ch concrete slabs in Phase I-b. Figure 5-61 shows the peak strain s at the bottom of the concrete slab and on the surface of the AC layer for the 4 slab in Phase I-b. In th is case, the analysis of maximum strain was not possible to perform due to the failu re of the gage located at the t op of the slab. The values shown

PAGE 116

116 in this figure cannot be used to check the bonded condition in the inte rface, since the strains plotted here were measured by gages close to the interface but at different depths. From all the figures depicting load-induced st rains in Phase I-b, no si gnificant variation of the strain can be observed as the number of load repetitions incr eases. The only exception to this is perhaps the gage located at the corner of the 6 slab, where there is a noticeable increase in the strain at the end of th e loading period. This might be the e ffect of micro crack s developed near the location of the strain gages. In Phase II, a total of 15 strain gages were installed in the sl abs to record the strain at different locations. Figures 5-62 through 5-67 sh ow the maximum measured HVS load-induced strains in Location 1 (at mid-edge ) and Location 5 (at slab corner) as a function of the number of passes of the 12-kip wheel load. The measured st rains shown include (1) the strain near the top of the concrete layer, (2) the stra in near the bottom of the concrete layer, and (3) the strain on top of the asphalt layer. Using the measured strain near the top of the concrete layer (Gage 1 and 13) and that near the bottom of the concrete laye r (Gage 2 and 14), the maximum strain at the top and the bottom of the concrete layer were computed by linear extrapolation. These maxi mum calculated strains are also shown in Figures 5-62 through 5-67 along with the gage measurements. Similar to the results from Phase I-a and Phase I-b, no significant vari ation in the strain was observed as a function of the number of passes for the 8and 6-inch slabs. However for the 10-inch slab, a slight increase in strain was obser ved at the end of the 12-kips loading period at the corner of the concrete slab. 5.3.6 Evaluation of the bond condition using strain ratios To better evaluate the bond condition using the strain data collected during the HVS loading, strain ratios were developed using av eraged values from plot s in section 5.3.5. Two

PAGE 117

117 types of strain ratios were used to quantitative ly evaluate the degree of bond in the interface of the composite pavement: R1 = et/ ec = ratio between the maximum tensile strain and the maximum compressive strain in the concrete slab R2 = eAC/ et = ratio between the tensile strain in the surface of the AC layer and the maximum tensile strain at the bottom of the concrete slab These ratios were calculated usi ng the measured peak strains in both the concrete slab and the AC layer and the maximum calculated strains in the slab. Theoretically, R1 should be close to 1.0 when the bond in the interface is very weak, and the composite pavement can be considered as fully un-bonded. The value is also close to 1.0 when the composite pavement is bonded but the asphalt layer is very thin compared to the concrete slab. In both cases there is no collaboration of the asphalt layer in reducing the tensile strain in the concrete. Theoretically, if the concrete remain linearly elastic, the compre ssive strain should be th e same as the tensile strain. In general, it can be said that the lower the value of R1, the better the bond in the interface. Strain ratio R2 should be close to 1.0 if the composite pavement is fully bonded. In this case there is full bonding between the concrete and the asphalt laye r, and therefore the tensile strain in the interface sh ould be the same for both materials. Values of R2 far from 1.0 mean that the composite pavement has a par tial bond condition in the interface. Table 5-11 summarizes the strain ratio va lues for the different phases in the HVS experiment. In some cases, R1 could not be calculated because th e gage located either at the top or at the bottom of the concrete slab failed, ma king impossible to estimate the maximum strains. In other cases, R2 could not be calculated because there wa s no gage in the asphalt layer as it was the case in Phase I-a.

PAGE 118

118 Few things could can be inferred with respec t to the interface bond condition in Phase I-a since few strain gages were placed to monitor the strain and no gage was installed on the asphalt surface. However from the strain ratio at the mi d-edge of the slab, it can be seen that the composite pavement behaved monolithically since the strain at the bottom was lower than the strain at the top (R1<1.0). The low value of R1 for the thinnest slab is a consequence of the comparable thicknesses between the c oncrete slab and the AC layer. For the 5 slab in Phase I-b, the R1 value was lower than 1.0, what agrees with the predicted behavior. This value indicates that a good bond was achie ved between the concrete and the asphalt layer, which resulted in the reductio n of the tensile strains at the bottom of the concrete layer. However, for the 6 slab in Phase I-b, the maximum compressive strains were only slightly higher than the maximum tensile strains (R1 very close to 1.0). This might be caused by either (1) the loss of bonding between th e concrete and the asphalt layer, or (2) inappropriate placement or functioning of the strain gages. By inspecting the R2 values that also indicates the degree of bond in the interface, both the 5 and the 6 slabs seem not to have a fully bond condition in the interface (R2<1.0). One possible explanation for this is that the gage located at the bottom of the 5 slab could have been displaced up during concrete placement, and thus the measured strain was not only lower than what was expected, but al so different compared with the one measured in the AC layer. Unfortunately, the exact pos ition of gages in the 5 slab was not possible to check. The 10 slab in Phase II shows characteri stics of an un-bonded composite pavement. The R1 value at both the mid-edge and the corner of the slab are very close to 1.0 and the R2 value is far from 1.0 (~0.6). For the 8 and 6 slabs the R1 and R2 values at the mid-e dge of the slab are lower than 1.0. Because the R1 values are still high for these two slabs, the in terface condition in

PAGE 119

119 these slabs can be considered as a par tially bonded. However, the high value of R1 in Phase II can also be due to the difference in thickness between the two layers and th erefore there is little collaboration of the AC layer in taking the bending. A difference of more than 15 micro-strains was observed between the two laye rs at the interface in the 10-inch slab, while the difference was much lower for the 8and 6-inch slabs. The fact that the 10-inch slab had a relatively poorer bond at the interface as compared with the 8and 6-inch slabs can be partially explained in terms of the loading period. The 10inch slab was the first slab to be loaded in Ph ase II, at 28 days after th e concrete was placed. It has been reported in other studies that the bond in the in terface tends to increase with time due to mainly the effect of the slab weight. The 8-inch and the 6-inch slabs were loaded at 2 and 3 months after concrete placement. In these tw o cases, the bond at the inte rface might have gained somewhat during this time. A summary review of Table 5.11 indicates that, in genera l, the bond condition in the interface ranged from fully bonded to partially bonded. By looking at the R1 value of the test sections in Phase I, a high level of bond in th e interface was achieved, with the exception of the 6 slab with a joint spacing of 4 ft (Phase I-b). The bond in Phase II seems to be lower than Phase I as it was meant to be. By looking at the value of R1 and R2 in the corner of the slab, it looks like the composite pavement in Phase II de veloped a low level of bond in that particular location.

PAGE 120

120 Table 5-1 HVS loading period and number of 12-kip wheel passes on the test sections in Phase I-b. Slab thickness Starting date Endi ng date # of 12-kip wheel passes 4 June 9, 2005 June 12, 2005 40,650 5 June 14, 2005 June 17, 2005 38,800 6 June 20, 2005 June 22, 2005 26,040 Table 5-2 Summary of HVS loadi ng on test sections in Phase II. Slab Load Starting date Ending date # of passes 12 kips November 14, 2005 December 4, 2005 87,508 15 kips December 4, 2005 December 11, 2005 86,954 10 18 kips December 11, 2005 December 16, 2005 72,554 12 kips January 9, 2006 January 15, 2006 73,662 15 kips January 15, 2006 January 21, 2006 60,923 8 18 kips January 21, 2006 January 28, 2006 67,015 12 kips January 30, 2006 February 5, 2006 73,108 15 kips February 5, 2006 February 10, 2006 67,015 6 18 kips February 10, 2006 February 15, 2006 65,908 Table 5-3 Number of hours in a day when the te mperature differential was in certain ranges for the 4-inch slab in Phase I-a. T<-5F -5FT<+5F 5F>T<10F 10F>T<15F Month Ave Max Min Ave MaxMin Ave MaxMin Ave MaxMin Ave MaxMin July 0.0 0.0 0.0 15.2 21.211.86.810.02.71.32.50.0 0.6 20.0 Aug. 0.4 1.2 0.0 14.7 16.712.87.59.83.81.12.30.0 0.3 1.20.0 Sep. 0.0 0.0 0.0 15.1 16.813.06.87.84.81.32.30.7 0.8 1.70.0 Oct. 0.1 0.7 0.0 15.3 17.713.26.08.23.71.42.30.8 1.1 1.80.0 Season 0.1 1.2 0.0 15.1 21.211.86.710.02.71.32.50.0 0.7 2.00.0 Time in hours Table 5-4 Number of hours in a day when the temperature differential was in certain ranges for the 5-inch slab in Phase I-a. T<-5F -5FT<+5F 5F>T<10F 10F>T<15F Month Ave Max Min Ave MaxMinAveMaxMin AveMaxMin Ave Max Min Nov. 0.8 7.7 0.0 15.3 24.08.37.011.80.00.91.70.0 0.0 0.00.0 Jan. 0.0 0.0 0.0 13.8 15.812.09.711.08.20.51.30.0 0.0 0.00.0 Season 0.6 7.7 0.0 14.9 24.08.37.711.80.00.92.00.0 0.0 0.00.0 Time in hours

PAGE 121

121 Table 5-5 Extreme values for temperature diffe rential and temperature in the AC Layer. Temperature Differential F (C) Temperature in AC Layer F (C) Phase Slab Min Max Min Max 4 -4.6 (-2.5)15.1 (8.4)78.1 (25.6) 97.7 (36.5) 5 -5.7 (-3.2)9.7 (5.4)50.7 (10.37) 86.2 (30.1) I-a 6 -8.4 (-4.7)24.7 (13.7)78.5 (23.6) 89.2 (31.8) 4 -3.1 (-1.7)15.5 (8.6)81.1 (27.3) 93.0 (33.9) 5 -5.6 (-3.1)26.8 (14.9)78.8 (26.0) 99.9 (37.7) I-b 6 -5.2 (-2.9)29.9 (16.6)78.3 (25.7) 105.6 (40.9) 6 -6.9 (-3.8)15.0 (8.3)48 (8.9) 71.6 (22.0) 8 -7.8 (-4.3)14.0 (7.8)53.6 (12.0) 72.9 (22.7) II 10 -9.3 (-5.2)12.2 (6.8)56.7 (13.7) 69.3 (20.7) Table 5-6 Measured static and dyn amic strains for gages 1, 2 and 5 in the 4-inch slab in Phase Ia caused by 9 and 12-kip loads. Measured Strain, 10-6 in/in Gage 1 Gage 2 Gage 5 Date Load Static Dynamic Static Dynamic Static Dynamic 7/8/2004 9 2224302521 14 7/9/2004 9 2222202218 15 7/10/2004 9 2328352725 15 7/12/2004 12 3036302924 16 8/4/2004 12 3932503522 16 9/23/2004 12 3532473124 15 9/24/2004 12 3637473429 18 10/1/2004 12 4338533731 18 10/2/2004 12 573834 17 10/4/2004 12 503428 19 10/5/2004 15 534232 24 10/6/2004 15 544134 23 10/7/2004 15 634735 28

PAGE 122

122 Table 5-7 Measured static and dyn amic strains for gages 2 and 4 in the 5-inch slab in Phase I-a caused by 9 and 12-kip loads. Dynamic Strain (x10-6) Static Strain (x10-6) Date Load Gage No 2 Gage No 4 Gage No 2 Gage No 4 11/1/2004 9 21-6-421 11/2/2004 9 21-5-320 11/3/2004 9 21-5-417 11/4/2004 9 20-6-425 11/6/2004 9 17-5-4 11/7/2004 9 18-7-4.530 11/8/2004 12 21-6-427 3 11/9/2004 12 20-8-522 11/10/2004 12 20-7-735 2 11/15/2004 12 20-6-733 2 11/16/2004 12 18-5-623.5 3 11/17/2004 12 20-7-731 2 11/18/2004 12 20-7-730 11/19/2004 12 21-7-632 11/22/2004 12 21-8-634 11/23/2004 12 23-8-731 Table 5-8 Measured static and dyn amic strains for gages 2 and 4 in the 5-inch slab in Phase I-a caused by 15and 18-kip loads Dynamic 8 mph Static Dynamic-1 mph Date Load Gage No 2 Gage No 4 Gage No 2 Gage No 4 Gage No 2 Gage No 4 12/28/2004 15 18-4-526 23 -4-5 12/29/2004 15 19-6-637 25 -7-7 12/30/2004 15 20-7-636 28 -8-8 1/3/2005 18 25-9-729-232 -11-9 1/4/2005 18 25-10-943 33 -13-11 1/5/2005 18 27-13-939-435 -13-11 1/6/2005 18 27-11-940-436 -14-11 1/7/2005 18 28-9-946-437 -14-12 1/9/2005 18 33-12-1145-740 -14-9 1/10/2005 18 30-11-938-338 -16-10 1/11/2005 18 30-11-1041-439 -15-11 1/13/2005 18 31-11-950-640 -14-13 1/14/2005 18 32-10-951-639 -13-12

PAGE 123

123 Table 5-9 Verified depths of strain gages in Phase I-b. Location 1 (mid-edge) Location 2 (corner) Intended Position 6 5 4 6 5 4 Top Gage 1 from top 1.25-0.75N/A N/AN/A Bottom Gage 0.5 from bottom 5.5-3.25.3 4.33.3 Table 5-10 Verified depths of strain gages in Phase II. Location 1 (mid-edge) Location 2 (corner) Intended Position 6 8 10 6 8 10 Top Gage 1 (from top) 0.70.81.00.871.2 Bottom Gage 1 (from bottom) 5.06.65-7.09.0 Table 5-11 Strain Ratios to evaluate the degree of bond in the interface Thickness Mid-edge Corner Phase In R1 = et/ ec R2 = eAC/ et R1 = et/ ec R2 = eAC/ et 4 0.75-5 --I-a 6 0.89-4 --5 0.660.66I-b 6 0.960.836 0.810.830.83 0.65 8 0.880.691.00 0.91 II 10 0.990.641.13 0.60

PAGE 124

124 Figure 5-1 Corner cracks on 4-inch slabs in Phase I-a after 21-kip wheel loads. Figure 5-2 Shrinkage cracks on a 4inch test slab in Phase I-a.

PAGE 125

125 Figure 5-3 Shrinkage cracks on a 5inch test slab in Phase I-a. Figure 5-4 Shrinkage cracks on a 6-in ch concrete slab in Phase I-a.

PAGE 126

126 Figure 5-5 Temperature differential variation in the 4-inch slab in Phase I-a. Figure 5-6 Temperature diffe rential variation in the 5-inch slab in Phase I-a.

PAGE 127

127 Figure 5-7 Temperature differential variation in the 6-inch slab in Phase I-a. Figure 5-8 Temperature differential variation in the 6-inch slabs in Phase I-b.

PAGE 128

128 Figure 5-9 Temperature differential variation in the 5-inch slabs in Phase I-b. Figure 5-10 Temperature differential variat ion in the 4-inch slabs in Phase I-b.

PAGE 129

129 Figure 5-11 Temperature differential variat ion in the 10-inch slab in Phase II. Figure 5-12 Temperature differential vari ation in the 8-inch slab in Phase II

PAGE 130

130 Figure 5-13 Temperature differential variat ion in the 6-inch slab in Phase II. Figure 5-14 Temperature variation on the surface of the asphalt layer for the 4 slab in Phase I-a.

PAGE 131

131 Figure 5-15 Temperature variation on the surface of the asphalt layer for the 5 slab in Phase I-a. Figure 5-16 Temperature variation on the surface of the asphalt layer for the 6 slab in Phase I-a.

PAGE 132

132 Figure 5-17 Temperature on the surface of the AC layer for the 6-inch slab in Phase I-b. Figure 5-18 Temperature on the surface of the AC layer for the 5 slab in Phase I-b.

PAGE 133

133 Figure 5-19 Temperature on the surface of the AC layer for the 4 slab in Phase I-b. Figure 5-20 Temperature on the surface of the AC layer in the 10-inch slab in Phase II.

PAGE 134

134 Figure 5-21 Temperature on the surface of the AC layer in the 8-inch slab in Phase II. Figure 5-22 Temperature on the surface of the AC layer in the 6-inch slab in Phase II.

PAGE 135

135 Temperature Distribution along the Depth of the Concrete Slab4" Slab Phase I-a 0 1 2 3 4 5 6 25272931333537394143 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-23 Temperature distributi on along the depth of the 4 in c oncrete slab in Phase I-a at maximum positive temperature differential Temperature Distribution along the Depth of the Concrete Slab5" Slab Phasa I-a 0 1 2 3 4 5 6 7 15171921232527293133 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-24 Temperature distribut ion along the depth of the 5in c oncrete slab in Phase I-a at maximum positive temperature differential

PAGE 136

136 Temperature Distribution along the Depth of the Concrete Slab6" Slab Phase I-a 0 1 2 3 4 5 6 7 202530354045 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-25 Temperature distribut ion along the depth of the 6in c oncrete slab in Phase I-a at maximum positive temperature differential Temperature Distribution along the Depth of the Concrete Slab4" Slab Phase 1-b 0 1 2 3 4 5 252729313335373941 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-26 Temperature distribut ion along the depth of the 4in c oncrete slab in Phase I-b at maximum temperature differentials

PAGE 137

137 Temperature Distribution along the depth of the Concrete Slab5" slab Phase I-b 0 1 2 3 4 5 6 25 30 35 40 45 50 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-27 Temperature distribut ion along the depth of the 5in c oncrete slab in Phase I-b at maximum temperature differentials Temperature Distribution along the Depth of the Concrete Slab6" slab Phase I-b 0 1 2 3 4 5 6 7 2025303540455055 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-28 Temperature distribut ion along the depth of the 6in c oncrete slab in Phase I-b at maximum temperature differentials

PAGE 138

138 Temperature Distribution along the Depth of the Concrete Slab6" Slab Phase II 0 1 2 3 4 5 6 7 57911131517192123 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-29 Temperature distribut ion along the depth of the 6in c oncrete slab in Phase II at maximum temperature differentials Temperature Distribution along the Depth of the Concrete Slab8" slab Phase II 0 1 2 3 4 5 6 7 8 9 51015202530 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-30 Temperature distribut ion along the depth of the 8in c oncrete slab in Phase II at maximum temperature differentials

PAGE 139

139 Temperature Distribution along the Depth of the Concrete Slab10" Slab Phase II 0 2 4 6 8 10 12 510152025 Temperature (C)Depth (in) Center (+DT) Corner (+DT) Center (-DT) Corner (-DT) Figure 5-31 Temperature distribut ion along the depth of the 10in concrete slab in Phase II at maximum temperature differentials Figure 5-32 Comparison of dynamic and static strain for gage 1 in the 4-inch slab in Phase I-a.

PAGE 140

140 Figure 5-33 Comparison between static and dyn amic strain for Gages 2 and 5 in the 4-inch slab in Phase I-a. Figure 5-34 Measured dynamic and static strains at gage 2 in the 5-in ch slab in Phase I-a

PAGE 141

141 Figure 5-35 Measured dynamic and static strains at gage 4 in the 5-in ch slab in Phase I-a Strain Data collected during HVS loadingOriginal data set Gage at the top 1300 1310 1320 1330 1340 1350 051015202530 Time (sec)Strain (me) Strain Data collected during HVS loadingOriginal data set Gage at the bottom 200 210 220 230 240 250 0 5 10 15 20 25 30 Time (sec)Strain (me) Figure 5-36 Measured strains at two different depths at the mi d-edge of the 6-inch slab in Phase I-a.

PAGE 142

142 Strain Data collected during HVS loadingZeroed data set Gage at the top -30 -20 -10 0 10 20 0 5 10 15 20 25 30 Time (sec)Strain (me) Strain Data collected during HVS loadingZeroed data set Gage at the bottom -10 -5 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Time (sec)Strain (me) Figure 5-37 Zeroed strains profile at two different dept hs at the mid-edge of the 6-inch slab in Phase I-a. < 60 sec strain Time T L Figure 5-38 Strain in the composite pavement as a function of time.

PAGE 143

143 Effect of Temperature Differential on the Peak StrainPhase 1-a 6" slab 12 kips -30 -20 -10 0 10 20 30 -10-5051015 Temperature Differential (F)Strain (me) Top (Mid-edge) Bottom (Mid-edge) Bottom (Corner) Figure 5-39 Effect of temperature differential on the peak strain for the 6 slab in Phase I-a. Effect of the Temperature Differential on the Peak StrainPhase I-b 4in slab 12 kips0 10 20 30 40 50 60 -5 0 5 10 15 20Temperature Differential (F)Strain (me) Mid-Edge (Bottom) Mid-Edge (AC Layer) Corner (Bottom) Figure 5-40 Effect of temperature differential on the peak strain for the 4 slab in Phase I-b.

PAGE 144

144 Effect of the Temperature Differential on the Peak StrainPhase I-b 5in slab 12 kips -50 -40 -30 -20 -10 0 10 20 30 40 -10-50510152025 Temperature Differential (F)Strain (me) Mid-Edge (Top) Mid-Edge (Bottom) Mid-Edge (AC Layer) Corner (Bottom) Corner (AC Layer) Figure 5-41 Effect of temperature differential on the peak strain for the 5 slab in Phase I-b. Effect of Temperature Differential on the peak StrainPhase I-b 6in Slab 12kips -40 -30 -20 -10 0 10 20 30 40 -10-50510152025Temperatrure Differential (F)Strain (me) Mid-Edge (Top) Mid-Edge (Bottom) Mid-Edge (AC Layer) Figure 5-42 Effect of the temperat ure differential on the peak strain for the 6 slab in Phase I-b

PAGE 145

145 Effect of the Temperature Differential on the Peak StrainPhase II 6in slab12 kips-40 -30 -20 -10 0 10 20 30 40 64202468Temperature Differential (F)Strain (me) Mid-edge-Top Mid-edge-Bottom Mid-edgeAsphalt Surface Corner-Top Corner-Bottom Corner-Asphalt Surface Figure 5-43 Effect of the temperature differential on the peak strain for the 6 slab in Phase II Effect of the Temperature Differential in the Peak StrainPhase II 10in slab 12 kips-20 -15 -10 -5 0 5 10 15 20 1 0864202468Temperature Differential (F)Strain (me) Mid-edge-Top Mid-edge-Bottom Mid-edge-Asphalt Surface Figure 5-44 Effect of the temperature differential on the peak strain for the 8 slab in Phase II

PAGE 146

146 Effect of Temperature Differential on the Peak StrainPhase II 8in Slab 12 kips-30 -20 -10 0 10 20 30 -8-6-4-20246810Temperature Differential (F)Strain (me) Mid-edge-Top Mid-edge-Bottom Mid-edge-Asphalt Surface Corner-Top Corner-Bottom Corner-Asphalt Surface Figure 5-45 Effect of the temperature differential on the peak strain for the 8 slab in Phase II Effect of AC Temperature on the Peak StrainPhase 1-a 6" slab 12 kips -30 -20 -10 0 10 20 30 27282930313233 Temperature (C)Strain (me) Top (Mid-edge) Bottom (Mid-edge) Bottom (Corner) Figure 5-46. Effect of AC temperature on th e peak strain for the 6 slab in Phase I-a.

PAGE 147

147 Effect of the AC Temperature on the Peak StrainPhase I-b 4in slab 12 kips20 25 30 35 40 45 50 55 60 2828.52929.53030.53131.5AC Temperature (C)Strain (me) Mid-Edge (Bottom) Mid-Edge (AC Layer) Corner (Bottom) Figure 5-47 Effect of AC temperature on the peak strain for the 4 slab in Phase I-b. Effect of the AC Temperature on the Peak StrainPhase I-b 5in slab 12 kips -50 -40 -30 -20 -10 0 10 20 30 40 2527293133353739 AC Temperature (C)Strain (me) Mid-Edge (Top) Mid-Edge (Bottom) Mid-Edge (AC Layer) Figure 5-48 Effect of AC temperature on the peak strain for the 5 slab in Phase I-b.

PAGE 148

148 Effect of AC Temperature on the peak StrainPhase I-b 6in Slab 12kips -40 -30 -20 -10 0 10 20 30 40 252729313335373941AC Temperatrure (C)Strain (me) Mid-edge (Top) Mid-edge (Bottom) Mid-edge (AC Layer) Corner (Bottom) Figure 5-49 Effect of AC temperature on the peak strain for the 6 slab in Phase I-b. Effect of the AC Temperature on the Peak StrainPhase II 6in slab12 kips-40 -30 -20 -10 0 10 20 30 40 10121416182022AC Temperature (C)Strain (me) Mid-edge-Top Mid-edge-Bottom Mid-edgeAsphalt Surface Corner-Top Corner-Bottom Corner-Asphal Surface Figure 5-50 Effect of AC temperature on the peak strain for the 6 slab in Phase II.

PAGE 149

149 Effect of AC Temperature on the Peak StrainPhase II 8in Slab 12 kips-30 -20 -10 0 10 20 30 10121416182022AC TemperatureStrain (me) Mid-edge-Top Mid-edge-Bottom Mid-edge-Asphalt Surface Corner-Top Corner-Bottom Corner-Asphalt Surface Figure 5-51 Effect of AC temperature on the peak strain for the 8 slab in Phase II. Effect of the AC Temperature on the Peak StrainPhase II 10in slab 12 kips-20 -15 -10 -5 0 5 10 15 20 10121416182022242628AC Temperature (C)Strain (me) Mid-edge-Top Mid-edge-Bottom Mid-edge-Asphalt Surface Figure 5-52 Effect of AC temperature on the peak strain for the 10 slab in Phase II.

PAGE 150

150 Strain v/s Wheel Load Phase IIGages at mid-edge -50 -40 -30 -20 -10 0 10 20 30 101214161820 Load (kips)Strain (me) 6in-top 8in-top 10in-top 6in-bottom 8in-bottom 10in-bottom Figure 5-53 Relationship between strain a nd load in Phase II. Gages at the Mid-Edge. Strain v/s Wheel Load Phase IIGages at the corner -60 -40 -20 0 20 40 60 10 12 14 16 18 20 Load (kips)Strain (me) 6in-Top 8in-Top 10in-Top 6in-Bottom 8in-Bottom 10in-Bottom Figure 5-54 Relationship between strain a nd load in Phase II. Gages at the corner.

PAGE 151

151 Strain v/s Wheel Load Phase IIGages in AC Layer0 5 10 15 20 25 30 35 40 45 10111213141516171819Load (kips)Strain (me) 6in-Edge 8in-Edge 10in-Edge 6in-Corner 8in-Corner 10in-Corner Figure 5-55 Relationship between strain and load in Phase II. Gages in the AC layer. Figure 5-56 Variation of peak strains during the HVS test in the 6-inch slab in Phase I-a.

PAGE 152

152 Figure 5-57 Variation of maximum strains during the HVS test in the 4-inch slab in Phase I-a. Figure 5-58 Variation of maximum strains in th e 6-inch concrete slab and on the surface of the asphalt layer at Location 1 (mid edge of the slab) during HVS test in Phase I-b.

PAGE 153

153 Strain v/s # of passes 6"4'x4' slab Phase I-bLocation 2 0 10 20 30 40 0500010000150002000025000 # of passesStrain (me) 5.5" Depth Asphalt Surface Figure 5-59 Variation of peak strains in the 6-inch concrete slab and on the surface of the asphalt layer at Location 2 (corner of the slab) during HVS test in Phase I-b. Figure 5-60 Variation of maximum strains in th e 5-inch concrete slab and on the surface of the asphalt layer at Location 1 (mid-edge of the slab) during HVS test in Phase I-b.

PAGE 154

154 Figure 5-61 Variation of peak strains in the 4-inch concrete slab and on the surface of the asphalt layer at Location 2 (corner of the slab) during HVS test in Phase I-b. Strain v/s # of passes 12 kips 6" slab Phase IILocation 1 (mid-edge, longitudinal) -50 -40 -30 -20 -10 0 10 20 30 40 50 010,00020,00030,00040,00050,00060,000 # of passesStrain (me) Gage 1 (0.7" Depth) Gage 2 (5" Depth) AC surface Bottom Top Figure 5-62 Variation of peak strains in the 6-inch slab at Location 1 during HVS test in Phase II.

PAGE 155

155 Strain v/s # of passes 12 kips 6" slab Phase IILocation 5 (corner, transversal) -50 -40 -30 -20 -10 0 10 20 30 40 50 010,00020,00030,00040,00050,00060,000 # of passesStrain (me) Gage 13 (1" Depth) Gage 14 (5" Depth) AC surface Bottom Top Figure 5-63 Variation of peak strains in the 6 slab at Location 5 during HVS test in Phase II. Strain v/s # of pass es 12 kips 8" slab Phase IILocation 1 (mid-edge, longitudinal)-40 -30 -20 -10 0 10 20 30 40 010,00020,00030,00040,00050,00060,00070,00080,000 # of passesStrain (me) Gage 1 (0.8" Depth) Gage 2 (6.65" Depth) AC Surface Bottom Top Figure 5-64 Variation of peak strains in the 8inch slab at Location 1 during HVS test in Phase II.

PAGE 156

156 Strain v/s # of passes 12 kips 8" slab Phase IILocation 5 (corner, transversal)-30 -20 -10 0 10 20 30 010,00020,00030,00040,00050,00060,00070,00080,000 # of passesStrain (me) Gage 13 (0.87" Depth) Gage 14 (7.0" Depth) AC surface Bottom Top Figure 5-65 Variation in peak strains in the 8-inch slab at Lo cation 5 during HVS test in Phase II. Strain v/s # of passes 12 Kips 10" slab Phase IILocation 1 (mid-edge, longitudinal) -25 -20 -15 -10 -5 0 5 10 15 20 25 0 20,000 40,000 60,000 80,000 # of passesStrain (me) Gage 1 (1" Depth) Gage 2 (9" Depth) AC Surface Bottom Top Figure 5-66 Variation of peak strains in th e 10-inch slab at Location 1 during HVS test in Phase II.

PAGE 157

157 Strain v/s # of passes 12 Kips 10" slab Phase IILocation 5 (corner, transversal) -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 0 20,000 40,000 60,000 80,000 # of passesStrain (me) Gage 13 (1.2" Depth) Gage 14 (9" Depth) AC Surface Bottom Top Figure 5-67 Variation of peak strains in th e 10-inch Slab at Locati on 2 during HVS test in Phase II.

PAGE 158

158 CHAPTER 6 DEVELOPMENT OF A 3-D FINITE ELEMENT MODEL 6.1 Finite Element Program The m ulti-purpose finite element program ADINA version 8.2 was used to build the model for the analysis of whitetopping pavements in th is study. The capability of the ADINA program for 3-D finite element analysis its versatility in modeling ma terials behaviors under load and temperature effects, and its capability in mode ling the interface condition between two layers make this program very appropriate to model composite pavements. The ADINA program has a very friendly user interface to build the needed models for specific applications. It has routines to automa tically create finite element meshes based on the boundary definitions and density specifications. The program also has a complete post-process routine to generate the results both numerically and graphically. 6.2 Six-Slab and Twelve-Slab 3-D Finite Element Models Figures 6-1 shows a 6-slab 3-D FE model de veloped for the analysis of the com posite pavement test sections with a joint spacing of 6 feet. The 6-slab model was used to analyze the test sections in Phase I-a and Phase II. The num ber of slabs used in this model corresponds to the actual number of slabs in each test se ction in these two phases of the study. Figure 6-2 shows a 12-slab 3-D FE model de veloped for analysis of the composite pavement test sections with a joint spacing of 4 feet which were used in Phase I-b. The number of slabs used in this model also corresponds to the actual number of slabs in each test section in Phase I-b of the study. As shown in Figures 6-1 and 6-2, the sub-grade is modeled by a 100-inch thick layer at the bottom of the model. This layer is modeled as an assemblage of 3-D solid elements whose

PAGE 159

159 vertical dimension decreases towa rds the top. The bottom of this layer is modeled as fixed with no rotation or translation allowed. On top of the sub-grade layer is a 12-inch thick layer modeling the lime rock base, which is modeled as bonded to the sub-grade layer. The vertical dimension of the 3-D elements in this layer also decreases towards the top. On top of the base layer is a 4.5-inch layer modeling the AC layer, which is modeled as bonded to the base layer. The AC layer was modeled by three layers of 3-D elements, 1.5 inches thick each. The layer at the top models the concrete slabs. The thickness of this layer is variable to represent the 4, 5, 6, 8 and 10 inches thickness of the concrete slabs used in the test sections in this study. This layer is modeled by four equal la yers of 3-D element. The thicknesses of the finite elements in this layer are 1, 1.25, 1.5, 2 and 2.5 inches for the 4-, 5-, 6-, 8and 10inch concrete slabs, respectively. The mesh pattern in the XY plane is the same fo r all four layers. Fi gures 6-3 and 6-4 show the mesh patterns in the XY plane used for the 6slab model and the 12-slab model, respectively. Finer meshes are used in the areas of maximum an ticipated stresses in the test sections where the strain gages were placed, in or der to obtain more accurate comput ed strains in these areas for verification with the measured strains. 6.3 Solid 20-Node Finite Element All 3-D solid elem ents were modeled as 20-nod e elements. Figure 6-5 shows the 3-D solid element, along with the node c onfiguration. The 20-node element has a hexahedral shape, with one node at each of its 8 vertices and one node at the middle of each of its 12 edges. Each node has three degrees of freedom (tra nslations along three pe rpendicular directions ). This type of node configuration has been shown to give a high level of accuracy in combination with an

PAGE 160

160 acceptable computing time demand. Considering the mesh patterns as shown in Figures 6-3 and 6-4 and the node configuration in the 3-D solid elements as shown in Figure 6-5, the model response in terms of strains and stresses can be ob tained at locations as close as 2 horizontally and 1.25 vertically from one another. 6.4 Modeling of Concrete Slab Joints Load transf er across the joints between two adjoining concrete slabs is modeled by translational springs connecting the slab s at the nodes of the finite elements along the joint. Three values of spring constants are used to represent the sti ffnesses along three different directions. Figure 6-6 shows the load transfer elements used in the model. Kt, Ks and Kn are stiffness components in the Z, Y and X direction respectively. 6.5 Modeling of Materials The concrete, AC, base and subgrade m ateri als are modeled as isotropic and linearly elastic, and are characterized by th eir elastic modulus and Poisson s ratio. The AC material is also modeled as a temperature dependent material with an elastic modulus which varies with temperature. The contraction and expansion of the concrete due to temperature effects is also considered in the analysis and characterized by the coefficient of thermal expansion of the concrete. 6.6 Modeling of Concrete-Asphalt Interface The concrete-asphalt interface in the com posite pavement can be modeled as fully bonded, partially bonded or un-bonded. By default, the ADINA program treats adjacent nodes as rigidly connected to one another, so that the fully bonded condition would be implicitly considered in th e model with no special treatment of the interface between the concrete slab and the AC layer. The composite pavement

PAGE 161

161 test sections in Phase I-a and Phase I-b, which were constructed to have bonded interface between the concrete and the AC, were modeled to have a fully bonded interface in this fashion. To model the partial bond condi tion in the interface, transl ational spring elements were used to connect the bottom of the concrete layer with the top of th e AC layer at the nodes, with zero distance between the two layers. Each spri ng was modeled with three spring constants to represent stiffness along the three different direct ions. Figure 6-7 shows the configuration of the springs used at the interface. Kt and Ks represent the stiffness in the interface plane (X and Y directions), while Kn represents stiffness perpendicula r to that plane (Z direction). When the concrete-asphalt interface is fully un-bonded, there can not be any tensile load transfer across the interface, but there can still be transfer of compressive load across the interface. This is modeled by a special non-lin ear spring connecting the concrete and the asphalt layer at the interface. This special spring has an infinite stiffness value when the spring is in compression; however when the spring is in tension, the stiffness value is zero. Cores taken from the test sections in Phas e II indicated that, though the concrete-asphalt interface was intended to be un-bonded, there was partial bonding between the concrete and asphalt layers in these test sec tions. Analysis of the results of FWD tests on these test sections also indicated that the computed FWD deflections matched better with the measured deflections when the concrete-asphalt interface was modele d as partially un-bonded rather than fully unbonded. Similarly, the strains calculated using the fully un-bonded model resulted in strain values 25% higher than the measured strains. Thus, the model for the fully un-bonded condition was not used in the analysis of the strain da ta in this study. However the non-linear model was used to evaluate the performance of the 4 slab in Phase I-a, which presented a corner crack after the application of 9, 12, 15, 18 and 21 kips wheel load. Figure 6-8 shows the stiffness parameter

PAGE 162

162 for the non-linear springs. As mentioned before this spring has the characteristic of being infinitely rigid when it is s ubjected to compression, but it stiffness drop to zero when it is tension. An intermediate case was also consider ed in the model with the springs presenting certain stiffness before they reach certain level of tension force. This threshold can be estimated based on laboratory test (pull-up and shear test ). Pull-up test can help to determine normal tension at which the interface bond breaks. Shear tests allow to de termine the shear strength that break the bond in the interface. Because these tests usually do not measure deformation, a forcedeflection relationship cannot be accurately determined. The use of non-linear springs c onverts the static analysis into a time-dependent analysis. Time steps have to be specified for the model to evaluate the stress state on the springs at each time interval. The iteration method and the tolerance crite ria have to be specified for the model to converge. The model strain-ene rgy was used as the tolerance cr iteria in all an alysis involving time dependant runs. The wheel load pressure wa s divided in five time increments while the temperature differential was consid ered a constant load. Chapter 7 shows more details about the use of the non-linear model. 6.7 Modeling of Loads and Temperature Effects To m odel a moving HVS wheel load on a test section in this study, computations were done for the responses of the pavement subjected to static loads placed at different consecutive locations along the wheel path. The calculated response, such as strain at a particular location, could then be plotted versus time, as the wheel passed over the various lo cations at the various times. This computed strain versus time plot could then be compared to the measured strain versus time plot as obtained from the stra in gage measurements during HVS loading. The applied HVS wheel load was modeled as a uniformly distribute d load over a square area. The contact area was taken to be the whee l load divided by the contact tire pressure. For

PAGE 163

163 example, for a 12-kip wheel load with a tire pr essure of 120 psi, the contact area would be 100 in2. Because the model is discrete and the pressure load in the ADINA program has to be applied on element faces, the final contact area wa s adjusted considering the geometry of the finite elements. The concrete is modeled as a material which contracts or expands according to its temperature change and coefficient of th ermal expansion. An initial temperature is first specified for the entire concrete layer. The effects of the change in temperature from the initial temperature were then duly considered in the analysis when the information on the temperature distribution in the concrete layer was provided in th e input. The temperature at the bottom of the concrete slab is set to be unchanged from the initi al temperature. The temperat ure in the concrete layer was assumed to vary linearly from the top to the bott om of the slab. From the temperature at the top and the bottom of the concrete slabs, the temperatur e of the concrete at all the nodes of the finite element mesh for the concrete slabs were computed and entered as inputs to the model.

PAGE 164

164 Figure 6-1 Six-slab 3-D finite element model. Figure 6-2 Twelve-slab 3D finite element model.

PAGE 165

165 Figure 6-3 Mesh pattern in the XY Plane for the 6-slab model.

PAGE 166

166 Figure 6-4 Mesh pattern in the XY plane for the 12-slab model. Figure 6-5 Twenty-node 3D solid elem ent used in the analytical model.

PAGE 167

167 Figure 6-6 Springs to model load tr ansfer at concrete slab joints. Figure 6-7 Springs in the concrete-aspha lt interface to model the partial bond condition.

PAGE 168

168 K Deformation (compression) Deformation + (tension) K Force Figure 6-8 Non-linear springs to model the fully un-bonded condition in the interface.

PAGE 169

169 CHAPTER 7 MODEL CALIBRATION AND VERIFICATI ON 7.1 Overview of Model Calibration In order for the 3-D analytical m odel to accura tely analyze the behavior of WT pavements, it needs to have the correct properties of the pavement materials and the correct values of spring stiffness for modeling the behavior of joints a nd concrete-asphalt interf ace. The elastic moduli of the concrete and asphalt materials were initiall y estimated from the results of laboratory tests on these materials, as described in Chapter 4. The results of the FWD tests on the composite pavement test sections were used to estimate th e elastic moduli of the other pavement materials and the joint and interface sp ring stiffnesses by back-calcu lation method (of matching the analytically computed deflections with the meas ured FWD deflections). FWD test results were also used to verify the values of elastic moduli of th e concrete and asphalt materials. This process is referred to as deflection-based calibration of the model in this research. The analytical model was further refined by matching the analytical ly computed strains with the measured strains in the test sections caused by HVS wheel load s. This process is referred to as strain-based cal ibration in this research. Chapter 4 showed that AC Elas tic Modulus of the test sectio ns can be estimated based on temperature using the following expression: MR = 2321.9 e-0.0473T (7.1) Where: MR= Resilient Modulus of the AC layer (ksi) T = Temperature (C) To effectively calibrate the model and because there are too many variables involved in the calibration process, the AC elas tic modulus first estimated from the Eq.7-1, for the range of temperatures at the time of the HVS testing. Du ring the HVS testing, the temperature in the AC

PAGE 170

170 layer was measured in several points using thermocouples as indicated in Chapter 3. Temperature in the asphalt layer was well know n during the HVS testing period, since it was collected every 5 minutes during th e entire experiment. Unfortunate ly the temperature in the AC layer at the time of the FWD test is unknown. However it can be estimated based on the surface temperature measured during the test and the relationship between surface temperature and AC layer temperature estimated from the full scale experiment. Table 7-1 shows the maximum and minimum values of the AC Resilient Modulus using Eq.7.1, based on the minimum and maximum temper atures during the HVS test, respectively. 7.2 Deflection-Based Calibration of Model Parameters 7.2.1 Phases I-a and I-b As m entioned in Chapter 5, FWD tests were run on the test sections to estimate the values of the elastic moduli for the pavement layers and the stiffness of the springs used to model the load transfer at the joints and concrete-as phalt interface (for the pa rtially-bonded condition). To better estimate the elastic moduli of the pave ment layers independently of the effects of joints, FWD tests were run on large slabs which we re located at both ends of the test track, and had a size of 12 ft x 18 ft. Pavement surface deflection basins caused by a 12-kip FWD load were used to estimate the elastic moduli of th e pavement layers by th e back-calculation method. A special analytical model with a 12 ft x 18 ft slab was developed to estimate the deflections generated by the FWD. Figures 71 through 7-5 show examples of the matched deflection basins from the back-calculation proce ss in which the elastic moduli of the concrete and asphalt layer from material testing were verified and the el astic moduli of the limerock base and subgrade materials were estimated. The estimated elastic modulus had a range of 130,000 160,000 psi and 28,000 30,000 psi for the lime rock ba se and the subgrade, respectively. As mentioned before, the value of the elastic modulus of the AC layer was seen to significantly vary

PAGE 171

171 with temperature. From the results of laboratory tests on the AC samples (as presented in Chapter 4), the resilient modulus of the AC varied from 325 ksi at 40 C (representing a very hot summer condition) to 1,780 ksi at 5 C (representing winter condition). FWD tests for Phase I-a and Phase II were run in the summer time, while FWD test for Phase I-b were run in the Fall (at the end of October of 2005). The elastic modulus of the AC layer, estimated from the backcalculation method using the FWD data from Phase I-a, was very close to that obtained in the laboratory (750 ksi). By using the data from th e FWD tests run in a co lder condition, the backcalculation process gave a value of 1,100 ksi for the elastic modulus of the AC. The elastic modulus of the concrete was estimated as 4,350 ksi, which compares very well with the one obtained by laboratory testing (see Table 4-6, Chapter 4). After the elastic moduli of the pavement layers had been defined, the joint spring stiffnesses, which were used to model the load tr ansfer at the joints, were estimated by matching the analytical with the measured deflection ba sins caused by a FWD load. In this case the FWD test (12 kips load) was run in two locations, at the corner and the mid-edge of the test slab. Deflection basins were recorded along the edge of the slab on both the loaded and the un-loaded slab, as described in Chapter 4. Figures 7-6 through 7-9 show examples of th e matched deflection basins from the backcalculation process for the estimation of the joint spring stiffnesses. From the results of FWD tests run on a 4-inch slab on Lane 6 (Phase I-a) as seen from Fi gures 7-6 and 7-7, an appropriate match between the measured and the calculated deflection basin was achieved with a single vertical stiffness in the order of 100,000 lb/in, and with stiffnesses of zero in the other directions. Back-calculations performed using FWD data from the tests run in Phase I-b (using 4 x 4 slabs) showed that the effect of the load transfer can be properly modeled with a spring constant

PAGE 172

172 in the range of 100,000 to 1,000,000 lb/in for the vertical spring, and that it is not necessary to use the springs in the other two directions, similar to what happened in Phase I-a. 7.2.2 Phase II The elastic moduli of the pavem ent layers for th e test sections in Phase II were assumed to be similar to those in Phase I-a and Phase I-b, du e to the fact that the same pavement materials were used. The deflection basins along the edge of the slabs caused by a 12-kip FWD load applied to the corner and the mid-edge of the slabs were us ed to estimate the joint spring stiffnesses, and the interface spring stiffnesses by back-calculation method. In th is phase, many spring constants had to be calibrated since three springs were used to model the load transfer at the joints and three springs were used to model the interaction between the concrete slab and the AC layer in the interface. Also, the springs used in both the transverse and longitudinal joints were considered separately (as they might have different stiffness values), which gave more flexibility when matching the FWD deflection basins firs t and the measured HVS load-induced strains later. Figures 7-10 through 7-19 show examples of the matched deflection basins from the backcalculation process for the estimation of the joint spring stiffnesse s and interface spring stiffnesses for the test sections in Phase II. The analytical deflections matched well with the measured deflections in all th e cases shown in these figures. During the deflection-based calibration process, it was found that the parameter that had the greatest effect on the FWD deflection of th e composite pavement was the stiffness of the vertical springs modeling the vertical load tran sfers at the concrete-asphalt interface. Using a vertical spring stiffness ranging between 1018 and 1019 lb/in appeared to produce calculated deflections that reasonably matc h the measured FWD deflections.

PAGE 173

173 The deflection basin from the case when the sensors were located on the un-loaded slab was used to calibrate the load transfer at the l ongitudinal joint. In this test, the FWD load was applied at the corner and at the mid-edge of the slab and all the sensors measured deflections on the adjacent slab. Transversal load transfer were calibrated using the FWD test with the drops at the corner, with one of the sensor (d-12) located in an adjacent slab. The horizontal springs (in the X and Y directions) modeling the horizontal load transfers at the interface were found to have no effect on the calculated deflections. It is to be noted that a value of 500,000 lb/in was used for the interface hor izontal springs in the calculation of the FWD deflection basins, which are presented in Figures 7-10 through 7-19. If a different value of horizontal spring stiffness were use d, the calculated FWD deflections w ould essentially be the same. However, it is to be pointed out that in the strain-based calibration process, which is presented in the next section, the horizontal springs at the interface were found to have significant effects on the load-induced strains, and thus need to be properly calibrated. Similar to the cases in Phases I-a and Ib, a vertical spring stiffness of 100,000 lb/in appeared to work well in modeling the joint behavi or of the composite pavement test sections in Phase II. In fact the value of this parameter showed to have little effect in matching the deflection basin. For some test sections, the va lue of this parameter ranged form 0 to 1,000,000 kips/in with no significant effect in the load tr ansfer. It seems that the load transfer in bonded pavements is also affected by bond between the la yers in addition to the lateral interaction between slabs. Because the asphalt layer is continuous at slabs joints, each time a slab is loaded near the edge, the asphalt under th e portion of the slab being loaded transfer the load to the adjacent asphalt, which is under the unloaded slab. Due to the bond in terface the asphalt under the unloaded slab pushes (pulls) the slab generating the same effect as if there was a lateral load

PAGE 174

174 transfer between adjacent slabs. As a re sult, the bond condition in the interface may be responsible of a significant part of the load transfer at joints, collaborating with the interlock mechanism. The large range of th e spring values with little e ffect in matching the deflection basin at joints can be explained by this issue, since the full bond in the interface could have already taken care of the load transfer. 7.3 Strain-Based Calibration of Model Parameters 7.3.1 General Approach The pavem ent parameters and spring stiffness va lues of the 3-D models for analysis of the composite pavement test sections were further calibrated by matching the computed strains caused by the HVS wheel loads to the strains measured by the strain gages. In the HVS experiment, the wheel load traveled at a speed of 8 mph (140 in/sec). In the analysis for the loadinduced strains in the test sec tions using the analytical model, a static load was positioned in different locations along the wheel path to represent a moving load. The distances between the load positions were converted to time using this speed. The computed st rains at a particular strain gage location under diffe rent load positions were then plotted on a time scale, and compared with the measured strain versus time plot. Table 7-2 presents the elastic moduli and Poisso ns ratios of the pavement materials of the test sections as determined from the materi al characterization in Chapter 4 and from the deflection-based calibration procedure as presented in the previous section. A range of values was given for the elastic modulus of the AC since it varied with temperature. The lower and the upper values correspond to the el astic modulus of th e AC at 40 and 5 C, respectively. In Phase I, after estimating and verifying the material pavement properties and calibrating the springs located at the joints using FWD test s, there were no more unknown variables in the model. The elastic moduli of the subgrade, base and concrete seem not to vary during the

PAGE 175

175 experiment since they are basically not temper ature-dependent material. For this reason the strain-based calibration of the model in Phase I was in fact a verification process. When the model could not appropriately replic ate the strain measured in the test sections using the average Elastic Modulus of the AC shown in Table 7-1, a fine-tune calibration pr ocess was undertaken by slightly varying the AC elastic modulus within the range shown in Table 7-1. The elastic moduli of the concre te, base and subgrade material s as given in Table 7-2 were used in the 3-D models in computing the load-ind uced strains. As menti oned before, the elastic modulus of the AC layer was chosen to be in the ranges shown in Table 7-1, which represent the Elastic Modulus values during th e HVS test. The stiffness of th e joint vertical spring modeling the load transfer at the jo int was fixed at a value of 100,000 lb/in in the analysis. 7.3.2 Phase I-a Table 7-3 shows the HVS loading periods for th e 4-, 5and 6-inch slabs when they were loaded with 12 kips in P hase I-a. While the 4and 6-inch slabs were loaded mainly during hot weather time, the 5-inch slabs were loaded in the winter time. Figures 7-20 through 7-26 show the comparison of the computed strains with the measured strains for various test sections and gage locations. The values of the pavement parameters used in the analyses are also given in these figures. It can be seen that the elas tic modulus of the AC were in the range and very close to the av erage shown in Table 7-1. Values of 300,000 and 500,000 psi gave a reasonable match for the 4-inch slab (which was tested in the summer time), and a value of 950,000 psi worked well for the 5-inch slab (which was tested in the winter time), representing the two extreme conditions. It is to be pointed out that in Phase I-a the intended vertical positions of the strain gages were used in the analysis. As indicated in Chap ter 5, some gage locations in Phase I-b and Phase II were verified by mean of cores taken from th e test sections after the HVS loading. In that

PAGE 176

176 process it was found that most of the strain gages had shifted a littl e bit during concrete placement. The differences in the strains can be more substantial for thinner slabs, as a small variation in vertical position in the slab could result in la rge percentage ch ange in strain. Unfortunately, cores were not able to be taken to check the positions of the strain gages in the test slabs in Phase I-a. The possible shifting of the positions of the stra in gages could explain why some of the computed strains did not matc h very well with the measured strains. For example, in the case of the 6 slab, two very di fferent values of the AC elastic modulus were necessary to match the measured strains and one of these values was out of the range shown in Table 7-1. Even though the model could replicate the measured peak strain, the strain reversal phenomenon was not completely rep licated by the model. The peak strain reversal at the midedge occurs when the wheel load is located in the slab corner and the peak strain reversal at the corner happens when the load is at the mid-e dge. It seems that the magnitude of the strain reversal is affected also by th e dynamic behavior of the slab under the passing of the wheel load. The analytical model computes th e strains due to static loads a pplied at different locations, and does not analyze the possible dynamic effects. For the strain ga ges located at the mid-edge, the analytical strain versus time plots were all symmetrical around the poi nt of maximum strain, while the measured strains plots were not enti rely symmetrical. The measured strains were observed to have higher reversal strains when the wheel load moved away from the gage location as compared with the strains when the wheel approached the location. In many researches involving UTW and TWT pavement, it has been hypothe sized that shorter slabs have to be used to minimize the effect of bending in thin slabs. Th e fact that in all cases a strain reversal was observed during the passing of the wheel load at the mid-edge indi cates that the slab bends and

PAGE 177

177 thus the flexural strain is still important. Considering the calculated strain from the figures, it can be noticed that the value of the AC elastic modul us has more effect on the tensile strain at the bottom rather than on the compressive strain at the top. 7.3.3 Phase I-b The strain-based calibration of the model para m eters in Phase I-b was undertaken in the same way as in Phase I-a, except that in this case, some vertical positions of the gages were verified through cores taken at th e locations of the strain gage s after HVS testing. The seven known gages locations indicated in Table 5-9 in Chapter 5 were first used in the calibration process. When the exact vertical position of a strain gage is known, then more attempts to exactly match the strains can be done, since the uncertainties in th is variable are now excluded. In all those cases where the ex act vertical position of a gage was known, the gage position was corrected in the analytical model and the strain in the right place was calculated. In the cases where the positions of the gages were not verified, the original intended positions of the gages were used in the analysis. HVS loading for all the test sections in Phase I-b were run at the be ginning of the summer of 2005 as shown in Table 7-4. Thus, the values of the elastic modulus of the AC layer were lower than those for the winter time. Figures 7-27 to 7-32 show the comparison of the computed strains with the measured strains for various test sections and gage locations, for the case s where the exact gage locations were known. In each graph, the exact vertical pos ition of the gage is shown at the top of the figure and the values of the pavement parameters used in the analyses are shown in the boxes. From the strain-based calibration process consid ering only the cases where the exact position of the gage was known, the estimated AC Elas tic Moduli were between 450,000 psi and 550,000 psi for the 4 slab, 400,000 psi for the 5 sl ab, and between 300,000 psi and 400,000 for the 6

PAGE 178

178 slab. All these values fall within the ranges shown in Table 7-1. It can be seen that the analytical strains match well with the measured strain at their maximum values by only considering the right value of the AC elastic moduli, which were first estimated based on temperature. For the cases where the exact location of the strain ga ges where unknown, a different approach was undertaken to matc h the calculated strain and the measured strain, which also served as a mean to demonstrate the important effect of the position of the gages. Figures 7-33 and 7-34 show the comparison of the computed st rains with the measured strains for two gages in the 5 slab where the exact gage locations were unknown. These two gages were located at the same horizontal location in the same slab, w ith one gage at the top and the other at the bottom. Because the exact positions of the gages were not known, the figures also include the cases if the gages had been disp laced up by 0.2. These figures show examples of the effects of the vertical position of the gage on the load-induced strains at the gage For the case of the gage at the top (as shown in Figure 733), the best match was achieved with an elastic modulus of the AC layer of 400,000 psi, whereas for the case of the gage at the bottom (as shown in Figure 734), the best match was achieved with a value of 700,000 psi. When a diffe rent vertical position of the gage at the top was used, an exact matc h in maximum strain was obtained, as shown in Figure 7-33. However, the consideration of a diffe rent vertical position of the gage at the bottom made the match worse, as shown in Figure 7-34. That means that the gage at the bottom was probably displaced down rather than up. This demonstrates the importance of exactly knowing the position of the gages when considering very thin slabs (UTW pavements). In any case, the range of the AC elastic modulus fo und in this analysis reasonably agrees with the range shown in Table 7-1 for this par ticular test section.

PAGE 179

179 As a summary for this phase, the model could re plicate very well the strain measured in the test sections by only using the appropriate AC elastic modulus depending on the temperature of the AC layer. The strain reversal was not completely captured by the model since the calculated strain profile was symmetric with respect to the peak strain when matching the measured strain at the mid-edge of the slab. The non-symmetrical sh ape of the strain profile for the gages located at the slab corners was reasonabl y replicated by the model but the calculated strain reversal was still lower than the measured one. Similar to Phase I-a, a large variation of the AC elastic modulus showed to have a significa nt effect on the analytical tensil e strain compared to its effect on the analytical compressive strain. 7.3.4 Phase II The strain-b ased calibration of the model parameters in Phase II was performed in a similar manner as in Phases I-a and I-b, except that in this case the springs modeling the partial bonding condition of the concrete-asphalt interface had to be cal ibrated. Similar to the case in Phase I-b, the vertical positions of a few strain gages were verified through cores taken at the strain gages location after HVS testing. Those locations were previously shown in Table 5-10 in Chapter 5. In the case where the exact vertical position of a gage was known, the verified gage position was used in the analytical model to calculate the stra in. In the cases where the positions of the gages were not verified, the original intended positions of the gages were used in the analysis. HVS loading for all the slabs were run duri ng the winter of 2005-2006, as shown in Table 7-5. Thus, the elastic moduli of the AC layer considered in the analysis were higher than those estimated for the summer time according with Eq.7.1 and Table 7-1 Figures 7-35 through 7-52 show the comparison of the computed strains with the measured strains for various test sections and gage locations. The value of the AC elastic modulus was estimated based on Eq-7.1 and using the range of temperatures during the HVS testing.

PAGE 180

180 According with Table 7-1 this range was between 800 and 1,500 ksi, representing a cold condition. The verified vertical positions of the ga ges are shown at the top of the figures and the model parameters are shown in the boxes. It can be seen that in general, the analytical strains matched well with the measured strain at their maximum values. As it happened in the previous phases, the mode l underestimated the stra in reversal in all analyzed cases and presented a sy mmetrical shape around the peak st rain. It can also be observed that the model could reasonably replicate the strain in the asphalt layer. In Phase II there is no strain reversal at the corner since the strain ga ges at that location were measuring strain in the direction perpendicular to the traveling wheel lo ad. In that direction th e slab deflection has a single curvature at any moment during the passing of the load. From the results of the calibration process, it can be observed that the values of the interface horizontal springs modeling the concrete-asphalt interface were consistently higher at the mid-edge than at the slab corners. This indi cates that there was less interface bond at the slab corner than at the slab edge. The values of these springs also varied from one test slab to another, indicating the non-uniform ity of the partial bond condition. The calibrated model parameters for the three te st sections in Phase II are presented in the next section, along with those for the test sections in Phases I-a and I-b. 7.4 Summary of Calibration Results The best estim ated model parame ters of the 3-D model for all the test sections in this study, based on the results of the deflection-base d and strain-based calibration, are summarized in Table 7-6. It is to be poi nted out that these model parame ters are only applicable to the conditions at the time of the HVS loading when the strain data were taken. The variation in the elastic modulus of the AC layer from one test sec tion to another, or within the same test section was due to the different temperatures at the time of the tests. From the an alytical results in all

PAGE 181

181 test sections, the value of the AC elastic modulus was found to have more effect on the tensile strain at the bottom of the slab compared to its effect on the comp ressive strain at the top. For the test sections in Phase II, the stiffness values of the horizontal springs modeling the interface can be observed to decrease as the slab thickness increas es from 6 inches to 10 inches. For each test section, the range of stiffness values for the join t and interface hor izontal springs are given. For the joint horizontal springs, the higher values generally represent the condition at the slab corners, while the lower values represent the cond ition at the edge of the slabs. Conversely, for the interface horizontal springs, th e higher values generally represent the condition at the edge of the joints, while the lower values represent the condition at the slab corners. In Phase II the variables that control the strain in the analytical model are, in order of importance, the horizontal springs at the interface, the horizontal springs at the joints, and the AC Elastic Modulus. Since the model could satisfactorily replicate th e strain at many differe nt locations in the test sections and considering many different sl ab conditions (thickness, joint spacing, interface bond), it is possible to use the model to estimate st resses in both the concrete slab and the AC layer. The model and the parameters indicated in Table 7-6 will be used in Chapter 8 to estimate the level of stresses in the test section and th e potential performance of the composite pavement.

PAGE 182

182 Table 7-1 Extreme values of the AC resilient modulus based on extreme temperature during the HVS test, using Eq. 7.1 Temperature (C) MR (ksi) Phase Thickness (in) Min Max MR(max) MR(min) Average 4 25.6036.50692413 552 5 10.2730.101428559 993 I-a 6 23.6031.80760516 638 4 27.3033.90638467 553 5 26.0037.70679390 534 I-b 6 25.7040.90688335 512 6 8.9022.001524820 1172 8 12.0022.701316793 1054 II 10 13.7020.701214872 1043 Table 7-2 Elastic modulus and Poissons ratio of the pavement materials used in the 3-D finite element model. Material Modulus of Elasticity (psi) Poissons Ratio Sub-grade 30,000 0.35 Base 160,000 0.35 Asphalt 325,000 1,780,000 0.35 Concrete 4,350,000 0.20 Table 7-3 HVS loading periods for Phase I-a. Thickness Joint Spacing From To 4 6 6 07/11/04 10/03/04 5 6 6 11/01/04 11/24/04 6 6 6 05/23/05 05/26/05 Table 7-4 HVS loading periods for Phase I-b. Thickness Joint Spacing From To 4 4 4 06/09/05 06/12/05 5 4 4 06/14/05 06/17/05 6 4 4 06/20/05 06/22/05

PAGE 183

183 Table 7-5 HVS loading periods for Phase II. Thickness Joint Spacing From To 6 6 6 01/30/06 02/15/06 8 6 6 01/09/06 01/28/06 10 6 6 11/14/05 12/16/05

PAGE 184

184 Table 7-6 Summary of the best estimated parameters of the 3-D model for all test sections. Phase I-a Phase I-b Phase II 4 5 6 4 5 6 6 8 10 Concrete 4,350 4,350 4,350 4,350 4,350 4,350 4,200-4,3504,350 4,350 AC Layer 300-500 950-1,000750-1,400450-550 400-700 300-400 800-1,000 1,000 800-1,000 Base 160 160 160 130 130 130 160 160 160 Material Elastic Moduli (ksi) Subgrade 30 30 30 28 28 28 30 30 30 Interface X 3-3.5 2-3 1-3 Interface Y 3-3.5 2-3 1-3 Interface Z 5x1012 1013 1013 Trans. Joint X Trans. Joint Y 0-0.1 0 1-0 Trans. Joint Z 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1-1 0.1 Long. Joint X 0.01-1 0.01-100.01-1 Long. Joint Y 0 0 0-0.1 Spring Constants (lb/in 106 ) Long. Joint Z 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1-5 3

PAGE 185

185 Figure 7-1 Matching of deflecti on basin in the longitudinal di rection caused by a 12-kip FWD load applied to the center of a 4 slab in Phase I-a. Figure 7-2 Matching of deflecti on basin in the longitudinal di rection caused by a 12-kip FWD load applied to the center of a 6 slab in Phase I-b. Determination of Elastic Modulus Phase I-aFWD test run on the center of the 4" 12' x 18' slab, sensors in the longitudinal direction0 1 2 3 4 5 6 7 06121824303642485460 Distance (in)Deflection (mils ) FWD 3D Model Concrete: 4,350,000 psi A C: 700,000 psi Base: 160,000 psi Subgrade: 30,000 psi Determination of Elastic Modulus Phase I-bFWD test run on the center of the 6"12' x 18' slab, sensors in the longitudinal direction0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 01 02 03 04 05 06 07 08 0Distance (in)Deflection (mils ) FWD Case 1 Case 2 Case 1 : Concrete: 4,350,000 psi A C: 950,000 psi Base: 160,000 psi Subgrade: 28,000 psi Case 2: Concrete: 4,350,000 psi A C: 1,100,000 psi Base: 180,000 psi Subgrade: 27,000 psi

PAGE 186

186 Figure 7-3 Matching of deflection basin in the transverse direc tion caused by a 12-kip FWD load applied to the center of a 6 slab in Phase I-b. Figure 7-4 Matching of deflecti on basin in the longitudinal di rection caused by a 12-kip FWD load applied to the center of a 4 slab in Phase I-b. Determination of Elastic modulus Phase I-bFWD test run on the center of the 6"12' x 18' slab, sensors in the transversal direction0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 01020304050607080Distance (in)Deflection (mils) FWD Case 1 Case 2 Case 3 Case 2: Concrete: 4,350,000 psi AC: 1,100,000 psi Base: 180,000 psi Subgrade: 27,000 psi Case 3: Concrete: 4,350,000 psi A C: 850,000 psi Base: 180,000 psi Subgrade: 27,000 psi Case 1: Concrete: 4,350,000 psi A C: 1,100,000 psi Base: 160,000 psi Subgrade: 27, 000 psi Determination of Elastic Modulus Phase I-bFWD Test run on the center of the 4"12' x 18' slab, sensors in the longitudinal direction0 1 2 3 4 5 6 01 02 03 04 05 06 07 08 0Distance (in)Deflection (mils ) FWD Case 1 Case 2 Case1: Concrete: 4,350,000 psi A C: 1,100,000 psi Base: 130,000 psi Subgrade: 28,000 psi Case2: Concrete: 4,350,000 psi A C: 600,000 psi Base: 115,000 psi Subgrade: 28,000 psi

PAGE 187

187 Figure 7-5 Matching of deflection basin in the transverse direc tion caused by a 12-kip FWD load applied to the center of a 4 slab in Phase I-b. Figure 7-6 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD load applied to the corner of a 4 slab in Phase I-a. Determination of Elastic Modulus Phase I-bFWD test run on the center of the 4" 12' x 18' slab, sensors in the transversal direction0 1 2 3 4 5 6 7 01 02 03 04 05 06 07 08 0Distance (in)Deflection (mils ) FWD Case 1 Case 2 Case 1: Concrete: 4,350,000 psi A C: 600,000 psi Base: 115,000 psi Subgrade: 28,000 psi Case 2: Concrete: 3,900,000 psi A C: 600,000 psi Base:125,000 psi Subgrade: 28,000 psi Calibration of springs at joints Phase I-aFWD test run on the 4" 6' x 6' slab, drops at the corner, sensors along the edge of the loaded slab -7 -6 -5 -4 -3 -2 -1 0 -16-10-4281420263238445056 Distance (in)Deflection (mils ) FWD 3D Model Concrete: 4,350,000 psi A C: 700,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-spring: 100,000 lb/in

PAGE 188

188 Figure 7-7 Matching of deflecti on basin along the edge of unl oaded slab caused by a 12-kip FWD load applied to the corner of a 4 slab in Phase I-a. Figure 7-8 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD load applied to the corner of a 4 slab in Phase I-b. Calibration of springs at joints Phase I-aFWD test run on the 4" 6' x 6' slab, drops at the corner, sensors along the edge of the unloaded slab-9 -8 -7 -6 -5 -4 -3 -2 -1 0-18-12-606121824303642485460Distance (in)Deflection (mils) FWD 3D Model Concrete: 4,350,000 psi A C: 700,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-spring: 100,000 lb/in Calibration of springs at joints Phase I-bFWD test run on 4" 4' x 4' slab, drops at the corner, sensors along the edge of the loaded slab0 1 2 3 4 5 6 -60612182430364248546066Distance (in)Deflection (mils ) Case 1 Case 2 Case 3 FWD Case 2: Concrete: 4,300,000 psi A C Layer: 700,000 psi Base: 130,000 psi Subgrade: 29,000 psi K-joint = 1,000,000 lb/in Case 1: Concrete: 4,300,000 psi A C Layer: 850,000 psi Base: 160,000 psi Subgrade: 30, 000 psi K-joint = 10,000 lb/in Case 3: Concrete: 3,900,000 psi A C Layer: 600,000 psi Base: 125,000 psi Subgrade: 28, 000 psi K-joint = 1,000,000 lb/in

PAGE 189

189 Figure 7-9 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD load applied to the corner of a 5 slab in Phase I-b. Figure 7-10 Matching of deflecti on basin along the edge of loaded slab caused by a 12-kip FWD load applied to the corner of a 6 slab in Phase II. Calibration of Springs at joints Phase I-bFWD test run on 5" 4' x 4' slab, drops at the corner, sensors along the edge of the loaded slab0 1 2 3 4 5 6 7 -60612182430364248546066Distance (in)Deflection (mils ) FWD-1 Case 1 Case 2 Case 1: Concrete: 4,350,000 psi A C: 850,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-spring: 100,000 lb/in Case 2: Concrete: 4,350,000 psi AC: 1,100,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-spring: 1,000 lb/in Calibration of vertical springs Phase IIFWD test run on 6" slab, drops at the corner, sensors along the edge of the loaded slab0 1 2 3 4 5 6 -60612182430364248546066 Distance (in)Deflection (mils) FWD Case 1 Case 2 Case 1: Interface Z: 2.5x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Case 2: Interface Z: 5.0x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in

PAGE 190

190 Figure 7-11 Matching of deflecti on basin along the edge of load ed slab caused by a 12-kip FWD load applied to the mid-edge of a 6 slab in Phase II. Figure 7-12 Matching of deflec tion basin along the edge of unloaded slab caused by a 12-kip FWD load applied to the mid-edge of a 6 slab in Phase II. Calibration of vertical springs Phase IIFWD test run on the 6" slab, drops at the mid-edge, sensors along the edge of the loaded slab 0 1 2 3 4 5 -12-606121824303642485460 Distance (in)Deflection (mils ) FWD Case 1 Case 2 Case 1: Interface Z: 2.5x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Case 2: Interface Z: 5.0x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Calibration of vertical springs Phase IIFWD test run on the 6" slab, drops at the mid-edge, sensors along the edge of the unloaded slab0 0.5 1 1.5 2 2.5 3 3.5 4 -12-606121824303642485460Distance (in)Deflection (mils ) FWD Case 1 Case 2 Case 1: Interface Z: 5.0x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Case 2: Interface Z: 2.5x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in

PAGE 191

191 Figure 7-13 Matching of deflecti on basin along the edge of load ed slab caused by a 12-kip FWD load applied to the corner of an 8 slab in Phase II. Figure 7-14 Matching of deflec tion basin along the edge of unloaded slab caused by a 12-kip FWD load applied to the corner of an 8 slab in Phase II. Calibration of vertical springs Phase IIFWD test run on 8" slab, drops at the corner, sensors along the edge of the loaded slab0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 -60612182430364248546066Distance (in)Deflection (mils ) FWD Case 1 Case 2 Case 1: Interface Z: 5.0x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Case 2: Interface Z: 1.0x 1019 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Calibration of vertical springs Phase IIFWD test run on the 8" slab, drops at the corner, sensors along the edge of the unloaded slab 0 0.5 1 1.5 2 2.5 3 3.5 4 -60612182430364248546066 Distance (in)Deflection (mils ) FWD Case 1 Case 2 Case 1: Interface Z: 5.0x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Case 2: Interface Z: 1.0x 1019 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in

PAGE 192

192 Figure 7-15 Matching of deflecti on basin along the edge of load ed slab caused by a 12-kip FWD load applied to the mid-edge of an 8 slab in Phase II. Figure 7-16 Matching of deflec tion basin along the edge of unloaded slab caused by a 12-kip FWD load applied to the mid-edge of an 8 slab in Phase II. Calibration of vertical springs Phase IIFWD test run on the 8" slab, drops at the mid-edge, sensors along the edge of the loaded slab 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 -12-606121824303642485460 Distance (in)Deflection (mils ) FWD Case 1 Case 2 Case 1: Interface Z: 2.5x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Case 2: Interface Z: 1.0x 1019 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Calibration of vertical springs Phase IIFWD test run on the 8" slab, drops at the mid-edge, sensors along the edge of the unloaded slab 0 0.5 1 1.5 2 2.5 3 -12-606121824303642485460 Distance (in)Deflection (mils ) FWD 3D Model 3D Model: Interface Z: 1.0x 1019 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in

PAGE 193

193 Figure 7-17 Matching of deflecti on basin along the edge of load ed slab caused by a 12-kip FWD load applied to the corner of a 10 slab in Phase II. Figure 7-18 Matching of deflection basin along the edge of loaded slab caused by a 12-kip FWD load applied to the mid-edge of a 10 slab in Phase II. Calibration of vertical springs Phase IIFWD Test run on the 10" slab, drops at the corner, sensors along the edge of the loaded slab 0 0.5 1 1.5 2 2.5 3 3.5 -60612182430364248546066 Distance (in)Deflection (mils ) FWD 3D Model 3D Model: Interface Z: 1.0x 1019 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Calibration of vertical springsPhase IIFWD test run on the 10" slab, drops at the mid-edge, sensors along the edge of the loaded slab 0 0.5 1 1.5 2 2.5 3 3.5 -12-606121824303642485460 Distance (in)Deflection (mils ) FWD 3D Model 3D Model: Interface Z: 5.0x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in

PAGE 194

194 Figure 7-19 Matching of deflection basin along the edge of unl oaded slab caused by a 12-kip FWD load applied to the mid-edge of a 10 slab in Phase II. Figure 7-20 Strain comparison at Gage 1 in the 6 slab in Phase I-a. Calibration of vertical springs Phase IIFWD test run on the 10" slab, drops at the mid-edge, sensors along the edge of the unloaded slab 0 0.5 1 1.5 2 2.5 3 -12-606121824303642485460 Distance (in)Deflection (mils ) FWD test 3D Model 3D Model: Interface Z: 5.0x 1018 lb/in Interface X: 500,000 lb/in Interface Y: 500,000 lb/in Longitudinal Joint Z: 100,000 lb/in Transversal Joint Z: 100,000 lb/in Strain comparison 6" 6' x 6 slab Phase I-aLocation 1 Gage 1 (top), 12 kips load -40 -30 -20 -10 0 10 20 16.5 17 17.5 18 18.5 19 19.5 20 Time (sec)Strain (me) Gage1 3D Model (Case 1) 3D Model (Case 2) Case 1: Concrete: 4,350,000 psi A C: 750,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in Case 2 (Best Match): Concrete: 4,350,000 psi A C: 1,400,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in

PAGE 195

195 Figure 7-21 Strain comparison at Gage 2 in the 6 slab in Phase I-a. Figure 7-22 Strain comparison at Gage 3 in the 6 slab in Phase I-a. Strain comparison 6" 6' x 6 slab Phase I-aLocation 1 Gage 2 (bottom), 12 kips load -15 -10 -5 0 5 10 15 20 25 30 16.51717.51818.51919.520Time (sec)Strain (me) Gage2 3D Model (Case 1) 3D Model (Case 2) Case 1 (best Match): Concrete: 4,350,000 psi A C: 750,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in Case 2: Concrete: 4,350,000 psi A C: 1,400,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in Strain comparison 6" 6' x 6 slab Phase I-aLocation 2 Gage 3 (top), 12 kips load -30 -25 -20 -15 -10 -5 0 5 10 15 16.51717.51818.51919.520 Time (sec)Strain (me) Gage3 3D Model (Case 1) 3D Model (Case 2) Case 1: Concrete: 4,350,000 psi A C: 750,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in Case 2 (Best Match): Concrete: 4,350,000 psi A C: 1,400,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in

PAGE 196

196 Strain comparison 5" 6' x 6' slab Phase I-aLocation 1 Gage 2 (bottom), 12 kips load -10 -5 0 5 10 15 20 25 10.51111.51212.51313.5 Time (sec)Strain (me) Gage 2 3D Model (Case 1) 3D Model (Case 2) Case 1: Best Match Concrete: 4,350,000 psi A C: 950,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 psi Case 2: Concrete: 4,350,000 psi A C: 1,100,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 psi Figure 7-23 Strain comparison at Gage 2 in the 5 slab in Phase I-a. Figure 7-24 Strain comparison at Gage 1 in the 4 slab in Phase I-a. Strain comparison 4"6'x6' Slab Phase I-aLocation 1, Gage 1 (top), 12 kips load-40 -30 -20 -10 0 10 20 11 522 533 54Time (sec)Strain (me) 3D Model (Case 1) Gage 1 3D Model (Case 2) Case 1 (best Match): Concrete: 4,350,000 psi A C: 300,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in Case 2: Concrete: 4,350,000 psi A C: 500,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in

PAGE 197

197 Figure 7-25 Strain comparison at Gage 2 in the 4 slab in Phase I-a. Figure 7-26 Strain comparison at Gage 3 in the 4 slab in Phase I-a. Strain comparison 4" 6'x6' slab Phase I-aLocation 1, Gage 2 (bottom), 12 kips load-20 -10 0 10 20 30 40 5011.21.41.61.822.22.42.62.83 Time (sec)Strain (me) 3D Model (Case 1) Gage 2 3D Model (Case 2) Case 1 (Best Match): Concrete: 4,350,000 psi A C: 300,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in Case 2: Concrete: 4,350,000 psi A C: 500,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in Strain comparison 4" 6'x6' slab Phase I-aLocation 2, Gage 3 (top), 12 kips load -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 11.21.41.61.822.22.42.62.83 Time (sec)Strain (me) Gage 3 3D Model (Case 1) 3D Model (Case 2) Case 1: Concrete: 4,350,000 psi A C: 300,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in Case 2 (Best Match): Concrete: 4,350,000 psi A C: 500,000 psi Base: 160,000 psi Subgrade: 30,000 psi K-joint: 100,000 lb/in

PAGE 198

198 Strain comparison 6" 4'x4' slab Phase I-bLocation1 Gage 1 (top, depth 1.25"), 12 kips load -30 -25 -20 -15 -10 -5 0 5 10 15 1717.217.417.617.81818.218.418.618.819 Time (sec)Strain (me) Gage 1 3D Model (Case1) 3D Model (Case2) Case 1: Concrete: 4,350,000 psi A C: 700,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Case 2 (Best Match): Concrete: 4,350,000 psi A C: 400,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Figure 7-27 Strain comparison at Gage 1 in the 6-inch slab in Phase I-b. Figure 7-28 Strain comparison at Gage 2 in the 6-inch slab in Phase I-b. Strain comparison 6" 4'x4' slab Phase I-bLocation1 Gage 2 (bottom, depth 5.5"), 12 kips load -20 -10 0 10 20 30 40 1717.217.417.617.81818.218.418.618.819Time (sec)Strain (me) Gage 2 3D Model (Case1) 3D Model (Case2) Case 1: Concrete: 4,350,000 psi A C: 700,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Case 2 (Best Match): Concrete: 4,350,000 psi A C: 300,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in

PAGE 199

199 Figure 7-29 Strain comparison at Gage 3 in the 6-inch slab in Phase I-b. Figure 7-30 Strain comparison at Gage 1 in the 5-inch slab in Phase I-b. Strain comparison 6" 4'x4' slab Phase I-bLocation2 Gage 3 (bottom, depth 5.3"), 12 kips load -15 -10 -5 0 5 10 15 20 25 16.51717.51818.519 Time (sec)Strain (me) Gage 3 3D Model (Case1) 3D Model (Case2) Case 1: Concrete: 4,350,000 psi A C: 700,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Case 2 (Best Match): Concrete: 4,350,000 psi A C: 300,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Strain comparison 5" 4'x4' slab Phase I-bLocation1 Gage 1 (top), 12 kips load -50 -40 -30 -20 -10 0 10 20 1212.212.412.612.81313.213.413.613.814 Time (sec)Strain (me) Gage 1 3D Model (Case1, 1") 3D Model (Case2, 1") 3D Model (Case 2, 0.8") Case 1 Concrete: 4,350,000 psi A C: 700,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Case 2 (Best Match) Concrete: 4,350,000 psi A C: 400,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in

PAGE 200

200 Strain comparison 5" 4'x4' slab Phase I-bLocation2 Gage 3 (bottom, depth 4.3"), 12 kips load -15 -10 -5 0 5 10 15 20 25 1212.212.412.612.81313.213.4 Time (sec)Strain (me) Gauge 3 3D Model (Case1) 3D Model (Case2) Case 1 Concrete: 4,350,000 psi A C: 700,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Case 2: Best Match Concrete: 4,350,000 psi A C: 400,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Figure 7-31 Strain comparison at Gage 3 in the 5-inch slab in Phase I-b Figure 7-32 Strain comparison at Gage 2 in the 5-inch slab in Phase I-b. Strain comparison 5" 4'x4' slab Phase I-bLocation1 Gage 2 (bottom), 12 kips load -20 -10 0 10 20 30 40 1212.212.412.612.81313.213.413.613.814 Time (sec)Strain (me) Gage 2 3D Model (Case1, 4.5") 3D Model (Case2, 4.5") 3D Model (Case2, 4.3") Case 1 (Best Match) Concrete: 4,350,000 psi A C: 700,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Case 2 Concrete: 4,350,000 psi A C: 400,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in

PAGE 201

201 Figure 7-33 Strain comparison at Gage 2 in the 4-inch slab in Phase I-b. Figure 7-34 Strain comparison at Gage 3 in the 4-inch slab in Phase I-b. Strain comparison 4" 4'x4' slab Phase I-b Location 1 Gage 2 (bottom, depth 3.2), 12 kips load -15 -10 -5 0 5 10 15 20 25 30 1414.214.414.614.81515.215.4 Time (sec)Strain (me) Gage 2 3D Model (Case1) 3D Model (Case2) Case 2 (Best Match): Concrete: 4,350,000 psi A C: 550,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Case 1: Concrete: 4,350,000 psi A C: 700,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Strain comparison 4" 4'x4' slab Phase I-b Location 2 Gage 3 (bottom, 3.3" depth), 12 kips load -15 -10 -5 0 5 10 15 20 25 1414.114.214.314.414.514.614.714.814.915 Time (sec)Strain (me) Gage 3 3D Model (Case1) 3D Model (Case2) Case 1: Concrete: 4,350,000 psi A C: 700,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in Case 2 (Best Match): Concrete: 4,350,000 psi A C: 450,000 psi Base: 130,000 psi Subgrade: 28,000 psi K-joint: 100,000 lb/in

PAGE 202

202 Figure 7-35 Strain comparison at to p of Location 1 (mid edge) of the 10-inch slab in Phase II. Figure 7-36 Strain comparison at bottom of Locati on 1 (mid edge) of the 10-inch slab in Phase II. Strain comparison 10" 6'x6' slab Phase IILocation 1 Top (depth 1"), 12 kips load-25 -20 -15 -10 -5 0 5 101717.217.417.617.81818.218.418.618.819Time (sec)Strain (me) Gage 1 3D Model (Case 1) 3D Model (Case 2) CASE 1 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 1,000,000 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 100,000 lb/in Long Joint Z: 3,000,000 lb/in CASE 2 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in Strain comparison 10" 6'x6' slab Phase IILocation 1 Bottom, 12 kips load -5 0 5 10 15 201717.217.417.617.81818.218.418.618.819Time (sec)Strain (me) Gage 2 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Lon g Joint Z: 100,000 lb/in CASE 2 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 1,000,000 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 100,000 lb/in Long Joint Z: 3,000,000 lb/in

PAGE 203

203 Figure 7-37 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 10inch slab in Phase II. Figure 7-38 Strain comparison at to p of Location 5 (slab corner) of the 10-inch slab in Phase II. Strain comparison 10" 6'x6' slab Phase IILocation 1 AC surface, 12 kips load -5 0 5 10 15 201717.217.417.617.81818.218.418.618.819Time (sec)Strain (me) Gage 3 3D Model (Case 1) 3D Model (Case 2) CASE 1 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in CASE 2 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 1,000,000 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 100,000 lb/in Long Joint Z: 3,000,000 lb/in Strain comparison 10" 6'x6' slab Phase IILocation 5 Bottom (depth 1.25"), 12 kips load -20 -15 -10 -5 0 51717.217.417.617.81818.218.418.618.819Time (sec)Strain (me) Gage 13 3D Model (Case 1) 3D Model (Case 2) CASE 1 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 (Best Match) Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10.000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 3,000,000 lb/in CASE 2 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in

PAGE 204

204 Figure 7-39 Strain comparison at bottom of Location 5 (slab corner) of the 10-inch slab in Phase II. Figure 7-40 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the 10-inch slab in Phase II. Strain comparison 10" 6'x6' slab Phase IILocation 5 Bottom (depth 9"), 12 kips load -5 0 5 10 15 20 25 1717.217.417.617.81818.218.418.618.819 Time (sec)Strain ( me ) Gage 14 3D Model Series3 CASE 1 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10.000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 3,000,000 lb/in CASE 2 Material properties: Concrete: 4,350,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in Strain comparison 10" 6'x6' slab Phase IILocation 5 AC surface, 12 kips load -5 0 5 10 15 201717.217.417.617.81818.218.418.618.819Time (sec)Strain (me) Gage 15 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10.000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 3,000,000 lb/in CASE 2 Material properties: Concrete: 4,350,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in

PAGE 205

205 Figure 7-41 Strain comparison at to p of Location 1 (mid edge) of the 8-inch slab in Phase II. Figure 7-42 Strain comparison at bottom of Location 1 (mid edge) of the 8-inch slab in Phase II. Strain comparison 8" 6'x6' slab Phase IILocation 1 Top (depth 0.8"), 12 kips load-30 -20 -10 0 10 2011 11.5 12 12.5 13 13.5 14Time (sec)Strain (me) Gage 1 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 1,000,000 lb/in CASE 2 Material properties: Concrete: 4,200,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in Strain comparison 8" 6'x6' slab Phase IILocation 1 Bottom (depth 6.65"), 12 kips load -10 -5 0 5 10 15 20 1111.51212.51313.514 Time (sec)Strain (me) Gage 2 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 1,000,000 lb/in CASE 2 Material properties: Concrete: 4,200,000 psi AC: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in

PAGE 206

206 Figure 7-43 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 8inch slab in Phase II. Figure 7-44 Strain comparison at to p of Location 5 (slab corner) of the 8-inch slab in Phase II. Strain comparison 8" 6'x6' slab Phase IILocation 1 AC surface, 12 kips load -10 -5 0 5 10 15 20 25 11 11.5 12 12.5 13 13.5 14 Time (sec)Strain (me) Gage 3 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 1,000,000 lb/in Interface Y: 1,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 1,000,000 lb/in CASE 2 Material properties: Concrete: 4,200,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in Strain comparison 8" 6'x6' slab Phase IILocation 5 Top (depth 0.87"), 12 kips load -25 -20 -15 -10 -5 0 5 11.51212.51313.514 Time (sec)Strain (me) Gage 13 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 2,000,000 lb/in Interface Y: 2,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in CASE 2 Material properties: Concrete: 4,200,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in

PAGE 207

207 Figure 7-45 Strain comparison at bottom of Location 5 (slab corner) of the 8-inch slab in Phase II. Figure 7-46 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the 8-inch slab in Phase II. Strain comparison 8" 6'x6' slab Phase IILocation 5 Bottom (depth 7"), 12 kips load-5 0 5 10 15 20 11.51212.51313.514Time (sec)Strain (me) Gage 14 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 2,000,000 lb/in Interface Y: 2,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in CASE 2 Material properties: Concrete: 4,200,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in Strain comparison 8" 6'x6' slab Phase IILocation 5 AC surface, 12 kips load -5 0 5 10 15 20 25 11.5 12 12.5 13 13.5 14 Time (sec)Strain (me) Gage 15 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 2,000,000 lb/in Interface Y: 2,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in CASE 2 Material properties: Concrete: 4,200,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 1019 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in

PAGE 208

208 Figure 7-47 Strain comparison at to p of Location 1 (mid edge) of the 6-inch slab in Phase II. Figure 7-48 Strain comparison at bottom of Location 1 (mid edge) of the 6-inch slab in Phase II. Strain comparison 6" 6'x6' slab Phase IILocation 1 Top (depth 0.75") 12 kips load-40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 1010.210.410.610.81111.211.411.611.812Time (sec)Strain (me) Gage 1 3D Model (Case 1) 3D Model (Case 2) CASE 1 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 (Best Match) Spring constants: Interface X: 3,500,000 lb/in Interface Y: 3,500,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in CASE 2 (Best Match) Material properties: Concrete: 4,200,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in Strain comparison 6" 6'x6' slab Phase IILocation 1 Bottom (depth 5.5"), 12 kips load-10 -5 0 5 10 15 20 25 1010.210.410.610.81111.211.411.611.812Time (sec)Strain (me) Gage 2 3D Model (Case 1) 3D Model (Case 2) CASE 1 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30, 000 psi CASE 2 (Best Match) Material properties: Concrete: 4,200,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30, 000 psi CASE 1 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in CASE 2 (Best Match) Spring constants: Interface X: 3,500,000 lb/in Interface Y: 3,500,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in

PAGE 209

209 Figure 7-49 Strain comparison on the surface of the AC layer at Location 1 (mid edge) of the 6inch slab in Phase II. Figure 7-50 Strain comparison at to p of Location 5 (slab corner) of the 6-inch slab in Phase II. Strain comparison 6" 6'x6' slab Phase IILocation 1 AC surface, 12 kips load-15 -10 -5 0 5 10 15 20 25 30 35 1010.210.410.610.81111.211.411.611.812Time (sec)Strain (me) Gage 3 3D Model Series3 CASE 1 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Material properties: Concrete: 4,200,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in CASE 2 Spring constants: Interface X: 3,500,000 lb/in Interface Y: 3,500,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 100,000 lb/in Strain comparison 6" 6'x6' slab Phase IILocation 5 Top, 12 kips load-30 -25 -20 -15 -10 -5 0 5 10 10.5 11 11.5 12 12.5Time (sec)Strain (me) Gage 13 3D Model (Case 1) 3D Model (Case 2) CASE 1 Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30, 000 psi CASE 2 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in CASE 2 (Best Match) Material properties: Concrete: 4,200,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30, 000 psi CASE 1 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in

PAGE 210

210 Figure 7-51 Strain comparison at bottom of Location 5 (slab corner) of the 6-inch slab in Phase II. Figure 7-52 Strain comparison on the surface of the AC layer at Location 5 (slab corner) of the 6-inch slab in Phase II. Strain comparison 6" 6'x6' slab Phase IILocation 5 Bottom, 12 kips load-5 0 5 10 15 20 25 1010.51111.51212.5Time (sec)Strain (me) Gage 14 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in CASE 2 Material properties: Concrete: 4,200,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in Strain comparison 6" 6'x6' slab Phase IILocation 5 AC surface, 12 kips load-5 0 5 10 15 20 25 10 10.5 11 11.5 12 12.5Time (sec)Strain (me) Gage 15 3D Model (Case 1) 3D Model (Case 2) CASE 1 (Best Match) Material properties: Concrete: 4,350,000 psi A C: 1,000,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 2 Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in CASE 2 Material properties: Concrete: 4,200,000 psi A C: 800,000 psi Base: 160,000 psi Subgrade: 30,000 psi CASE 1 (Best Match) Spring constants: Interface X: 3,000,000 lb/in Interface Y: 3,000,000 lb/in Interface Z: 5x1018 lb/in Transv. Joint X: 0 lb/in Transv. Joint Y: 0 lb/in Transv. Joint Z: 100,000 lb/in Long. Joint X: 10,000 lb/in Long. Joint Y: 0 lb/in Long Joint Z: 5,000,000 lb/in

PAGE 211

211 CHAPTER 8 EVALUATION OF POTENTIAL PERFORMANCE OF THE WT DESIGNS 8.1 Overview This chapter presents the evaluation of the potential perform ance of WT pavements with the same designs as those used in the test sections in this study. The 3-D finite element model with the model parameters for each test section, as determined from the deflection-based and strain-based calibration (as presented in Chapter 7), was used to perform a stress analysis to determine the maximum stresses in each WT pavement under typical critical temperature-load conditions in Florida. The potential performance of each WT pavement was assessed based on (1) the maximum tensile stress in the concrete, (2) the maximum shear stress at the concrete-asphalt interface, and (3) the maximu m tensile stress in the AC. The non-linear version of the 3D analytical model, which was shown at the end of Chapter 6, was used in combination with the linear model to estimate the level of stresses in the 4 slab in Phase I-a. This was the only test sections that had a corner crack during the HVS loading period. The purpose of this analysis was to determine if the analytical model can also predict failure in the composite pavement. 8.2 Assumptions for the evaluation of the pote ntial performance of the test sections The potential performance of the test sect ions was evaluated using the 3D analytical model calibrated in the previous chapter. After the model could satisfactorily predict deflections and strains for different conditions and slab characteristics, it is possible to use the model to estimate the level of stresses in the composite pavement. 8.2.1 Critical loading conditions A 24-kip single axle load, which is slightly hi gher than the m aximum legal single axle load of 22 kips in Florida, was used as the applied load in the analysis. The two critical loading

PAGE 212

212 positions used in the analysis were (1) the mid-edge and (2) the corner of the slab. Figures 8.1 and 8.2 show the positions of the axle load used fo r the slabs with joint spacings of 4 ft and 6 ft, respectively. These figures also show the co mparison between a typical roadway load position and the critical load position considered in the analysis. The minimum and maximum temperature differe ntial in the concrete slab as observed during the HVS loading were around -10 F a nd +20 F, respectively. These two extreme temperature differentials were used in the critical stress analysis. 8.2.2 Model parameters The m odel parameters of the 3-D model for each test section used in the critical stress analysis are displayed in Table 8.1. It is to be noted that all the model parameters, except the joint spring stiffnesses, are from the results of deflection-based and strain-based calibration as presented in Chapter 7. All the jo int spring stiffnesses were set to zero in this analysis, based on the expectation that all the joints will even tually crack all the wa y through, and there will eventually be less load transfer across the joints as compared with their initial conditions. To represent different temperat ure conditions, which affect th e elastic modulus of the AC layer, all the test sections were analyzed using three different values of the AC elastic modulus, namely 300,000 psi, 700,000 psi and 1,100,000 psi. For the interface horizontal springs, two spring stiffness values were used. The higher values were used for the condition at the edge of the joints, while the lower values were used for the condition at the slab corners. 8.3 Results of Critic al Stress Analysis 8.3.1 Maximum stresses in the concrete slabs The m aximum computed tensile stresses in th e various bonded concrete slabs (in Phases Ia and I-b) and partially bonded concrete slabs (in Phase II) caused by a 24-kip single axle load

PAGE 213

213 placed at two different critical positions (mid-edge or corner), and for three different temperature differentials in the concrete slab (-10, 0 or + 20 F) are shown in Table 8.2. Three different AC moduli, namely 300,000 psi, 700,000 psi or 1,100,000 ps i, which represent the condition of the AC at different temperatures, were used in th e analysis. The values shown in Table 8.2 are principal stresses and they can be located at the top or bottom of the concrete slab. From table 8.2, the relationship between many pavement parameters and pavement behavior can be observed. Following paragraphs describe the eff ect of the AC Elastic Modulus, Temperature differential, Slab size and bond in the interface. 8.3.1.1 Effects of elastic modulus of AC layer Figure 8.3 shows the effects of the elastic m odulus of the AC layer on the maximum stresses in the concrete caused by a 24-kip ax le load applied at mid edge for 4-inch bonded concrete slabs with 6 ft joint spacing. It can be seen that at the conditi on of temperature differential of +20 F, an increase of 55% in tensile stress (from 367 to 568 psi) in the concrete was obtained when the elastic modulus of the AC layer dropped from 1,100,000 psi to 300,000 psi. However, at the condition of temperature differen tial of -10 F, a decrease of 35% in tensile stress (from 380 to 246 psi) in the concrete was obtained when the elastic modulus of the AC layer changed from 1,100,000 psi to 300,000 psi. Figure 8.4 shows similar plots for 5-inch bonded conc rete slabs with 4 ft joint spacing. In this case, a decrease in the elas tic modulus of the AC caused an in crease in the tensile stress in the concrete for all temperature conditions. 8.3.1.2 Effects of temperature differential Figures 8.5 and 8.6 show the effects of tem perature differentials on the maximum stresses in the bonded slabs with 6 ft joint spacing (test slabs in Phase I-a) caused by a 24-kip single axle load placed at mid-edge, and corner of the sl ab, respectively. An AC elastic modulus of 300,000

PAGE 214

214 psi was used in these analyses. Figures 8.7 and 8.8 show similar plots for the bonded slabs with 4 ft joint spacing (test slabs in Phase I-b). It can be seen from these figures that higher stresses in the concrete were obtained at a temperature di fferential of +20 F th an at a temperature differential of -10 F. For the c ondition of temperature differentia l of +20 F, loads at slab midedge produced higher stresses than those produced by loads at slab corner. Similar observation about the effects of temp erature differential and loading positions can be made for the partially bonded slabs in Phase II. Figure 8.9 shows the effects of temperature differentials on the maximum stresses in the par tially bonded slabs with 6 ft joint spacing (test slabs in Phase II), caused by a 24-ki p single axle load placed at midedge of the slab. It can be seen that the condition of temperature differentia l of +20 F produced much higher stresses than a temperature differential of -10 F. 8.3.1.3 Effects of panel size Figures 8.10 and 8.11 show the effects of pa nel size on the m aximum stresses in the bonded concrete slabs caused by a 24 -kip single axle load at midedge and corner of the slabs, respectively. An AC elastic modulus of 300,000 psi was used in these analyses. It can be seen that at the most critical temperature and load condition (when the temperature differential was +20 F and the load was applied at mid-edge of sl ab), the 4 ft X 4 ft panels had slightly lower stresses than the 6 ft X 6 ft panels. The reduction in stress ranges from 2% for the 4-inch slabs to 15% for the 6-inch slabs. When the load was applied to the slab co rner, the 4 ft X 4 ft panels had stresses significantly higher than the 6 ft X 6 ft panels. For example, when the temperature differential was +20 F and the load was applied at the slab corner, the 4 ft X 4 ft pa nels had higher stresses than the 6 ft X 6 ft panels by 26% for the 5-inch slabs.

PAGE 215

215 8.3.1.4 Effects of bonded versus partially bonded interface A direct co mparison of the effects of a bonded concrete-asphalt interf ace versus a partially bonded interface can be made by comparing the co mputed maximum tensile stresses in the 6inch bonded slabs in Phase I-a with those in th e 6-inch partially bonded slabs in Phase II. Figures 8.12 and 8.13 show the comparison of maximum computed tensile stresses in concrete for these two test slabs under a 24-kip single axle lo ad applied at the mid-edge and corner of the slabs, respectively. An AC elastic modulus of 300,000 psi was used in these analyses. From these two figures, it can be seen that, for the most critical condition of a temperature differential of +20 F, the bonded slabs have about the same maximu m tensile stresses as those in the partially bo nded slabs. However, fo r the condition of a temperature differential of -10 F, the bonded sl abs have slightly lower maximum te nsile stresses than the partially bonded slabs. Analyses were also performed to determine the maximum stresses in concrete under critical loading conditions for the hypothetical cases if the test sl abs in Phase II were constructed as fully bonded to the asphalt la yer. These computed stresse s are also shown in Table 8.2. Figure 8.14 shows the comparison of the maximum stresses in concrete caused by a 24-kip single axle load at mid-edge of slab for the test sections in Phase II with those for the hypothetical cases if the same slabs were constructe d bonded to the asphalt layer. Similar trends can be observed here. For the condition of a temperature differe ntial of +20 F, the bonded slabs have about the same maximum tensile stresses as those in th e partially bonded slabs. However, for the condition of a temperature differential of -10 F the bonded slabs have slightly lower maximum tensile stresses than the partially bonded slabs.

PAGE 216

216 8.3.2 Maximum shear stresses at the interface The 3-D finite elem ent model was also used to calculate the shear stre sses at the concreteasphalt interface under critical loading conditi ons. Table 8.3 displays the maximum shear stresses at the interface caused by a 24-kip single axle load at a temperature differential of +20 F for the bonded slabs in Phases I-a and I-b. Figures 8.15 and 8.16 show the plots of these maximum shear stresses at the in terface for the bonded slabs with 6 ft joint spacing and 4 ft joint spacing, respectively. Three AC elastic moduli, namely 300,000, 700,000 and 1,100,000 psi, were used in the analyses. From these two figures, it can be observed that for both load locations (mid-edge and corner), the shear stress is higher when the AC layer is stiffer, which is a consequence of the degree of restriction that the AC layer represent for the concrete slab. It can also be observed that the smaller 4 ft X 4 ft slabs (Phase I-b) had lower maximum shear stresses. In all cases, the maximum computed shear stress at the interface are very low compared with the shear strength measured from the core samples from the test sections using the Iowa Shear test. From Table 4.4 in Chapter 4, it can be seen that the average shear strength for the 4 ft 4 ft slabs was 194.5 psi, and the average shear strength for the 6 ft X 6 ft slabs was 220 psi. The maximum shear stress among all cases was only 84.8 psi. The fact that the shear stress developed at the interface is much lower than the shear strength indicates that the fully bonded condition at the interface may remain in pl ace for a long time during the life time of the composite pavement. It has to be pointed out that the Iowa Shear test is performed in such a way that no tensile stress is developed in the interface In real composite pavements, some critical combination of shear stress and vertical tensile stress can negatively affect the shear strength in the interface. However, the appli cation of the analytical model showed that that particular

PAGE 217

217 combination does not happen in the composite pavements investigated in this research. Effectively, vertical stresses calcu lated using the analytical model resulted in maximum values of 50 psi in places where shear stress was less than 30 psi. In places where the shear stress was maximum, the vertical stress in the interface was generally compression stress or a very low tensile stress. 8.3.3 Maximum stresses in the AC layer The 3-D analytical m odel was also used to de termine the tensile stre sses in the AC layer under critical loading conditions. Table 8.4 displays the maximum tensile stresses in the asphalt concrete layer caused by a 24-kip single axle load at various critical lo ading conditions. Figures 8.17 and 8.18 show the maximum tensile stresses in the AC layer caused by a 24-kip single axle load at a temperature differential of +20 F for the bonded slabs with 6 ft joint spacing and 4 ft joint spacing, respectively. Two AC elastic moduli, namely 300,000 psi and 1,100,000 psi, were used in the analyses. From these two figures, it can be observed that the tensile stress in the AC layer increases as the elastic modulus of the AC increases and that the size of the slab concrete has little effect on the tensile stresses in the AC layer. The tens ile stress in the AC layer increases since the relative stiffness between concrete and asphalt change as the aspha lt become more rigid. In other words the asphalt layer is helpi ng the concrete slab to carry the effect of the load. The maximum calculated tensile stresses in the AC layer are lower than the tensile strength of the AC measured in the IDT test (Table 4.9, Chapter 4). The stress calculated using an AC elastic modulus of 1,100 ksi is lowe r than strength measured at 15 C, which can be considered a low temperature for the AC layer. Similarly, the stress calculated using an AC elastic modulus of 300 ksi is lower than the strength measured at 40 C, which can be considered as a very high temperature for the AC layer. The highest tensil e stresses in the AC la yer were obtained when a

PAGE 218

218 high value of the AC elastic modulus was used. Since high values of the AC elastic modulus occur at low temperatures when the tensile stre ngth of the AC layer is high, this creates a favorable condition where the AC is less likely to crack. Thus, st resses in the AC layer should not be a controlling factor in the performance of these WT pavements. 8.4 Potential performance of the test sections The m aximum computed stresses in the conc rete slab caused by th e critical loading condition (of a 24-kip single axle load, placed at th e mid-edge of the slab, and at a temperature differential of +20 F in the conc rete slab) were used to assess th e potential performance of the WT pavement test sections evaluated in this study. The fatigue curve given by the PCA, which relates the stress/strength ratios with the numbe r of repetitions to produce fatigue failure in concrete, was used to estimate th e number of load repetitions to failure. The following equations were used to calculate the maximum number of load repetitions as a function of the stress/strength ratio: NLR = 10(96.5 100 r)/8.1 If r > 0.5 (8-1) NLR = infinite If r < 0.5 (8-2) Where NLR = number of load re petitions to failure, and r = stress/strength The average flexural strength at 56 days of the concrete used in the test sections was 842 psi, as presented in Tables 4.6 and 4.7 in Chapter 4. This flexural strength value was used in the computation of the stress to flexural strength ratios for the WT pavement test sections. Table 8.5 displays the computed maximum stresses and stress to flexural strength ratios for all the WT pavement test sections evaluated in this study. The allowable number of 24-kip single axle loads under critical load ing conditions were also computed and shown in this table. It

PAGE 219

219 is to be stressed that the results of these analys es are only applicable to the condition of the test sections, which had 4.5 inches of AC layer over 12 inches of limerock base. It can be seen that the maximum computed stre sses were all below the flexural strength of the concrete for all the test sections. This mean s that all the WT pavement test sections with a concrete slab thickness of 4 inch es or higher can withstand certain number of repetitions of the 24-kip single axle load under the critical loading condition w ithout cracking. The allowable number of repetitions of this critical load increases with slab th ickness. The allowable number of load repetitions also increases with smaller joint spacing. In order to be able to withstand the critical load without fear of fatigue failure (for an infinite number of critical load repetitions), a minimum slab thickness of 6 inches would be needed for a joint spacing of 4 ft, and a minimum slab thickness of 8 inches would be needed for a joint spacing of 6 ft. 8.5 Evaluation of the actual performance of th e test sections During the HVS testing, only one slab failed. Th e 4 slab in Phase I-a was loaded with 36,407 passes of a 9-kips wheel load, followed by 146,748 passes of a 12-kips wheel load. Then the load was increased to 15 kips with a to tal of 35,918 passes and then to 18 kips with 21,727 passes. Finally, the load was increased to 21 kips, and corner cracks developed after 12,187 passes. The crack was generated at the slab corner which represents a typical type of failure for the WT pavements. Using both the linear and th e non-linear model, an investigation of the failure mechanism was undertaken. According with the results of WT experime nts performed nationwid e and in the current research, if this type of composite pavement is constructed accordingly, w ith the concrete placed bonded to the AC layer, which is at least 3 thick, then the pavement is able to carry a significant number of load repetitions wit hout failure. It has been hypothesi zed that prior to failure the bond

PAGE 220

220 interface has to be broken, so that the co mposite pavement becomes unbonded. Temperature differential plays an importan t role in making the pavement unbonded. Negative temperature differential tend to curl the slab up at corners and edges. Because the pavement is bonded, a high vertical tensile strain is gene rated at that location that even tually will break the bond in the interface. Since the shear strengt h measured at the interface was in the order of 200 psi, it is expectable that the tensile strength at the interface had a similarly high value. When the temperature differential is positive, another critical stress condition appears in the concrete. At positive temperatur e differential, the slab tends to curl down at the corners and edges, and because the slab is restricted to curl due to the bond in the interface, tensile stress is generated at the bottom of the slab. The wheel lo ad positioned at the corner or edge will also generate tensile stress at the bottom, making the combination of load and positive temperature differential the worst scenario for tensile stress. According to this, a high value of the tensile stress can be reached at the bottom of the slab, even though the composite pavement is bonded and the slab does not curl. The linear model for a slab thickness of 4 wa s analyzed for a load at the corner with wheel loads of 12, 15, 18 and 21 kips for thr ee temperature differentials (-10, 0, and +20 F). Table 8.6 shows the maximum stresses due to thes e loads. The model was also analyzed for the condition of only temperature differential to estimate the level of tensile vertical stress in the interface. The vertical tensile stress in the corn er interface was in the or der of 200 psi for a -10 F temperature differential. This means that critical points at the corner of the slab might become unbonded as soon as the slab reaches a critical temperature differe ntial. Using Eq-8.2 it can be seen that the slab can withstand the 12-kip, 15-kip, and 18-kip wheel load without experiencing fatigue cracking (the stress/stre ngth ratio is less than 0.5), wh ile it would fail after a limited

PAGE 221

221 number of repetition of the 21-kip load. Table 8.6 also shows the maximum tensile stress in the concrete due to the application of a 21-kip wheel load along with a temper ature differential using the non-linear model. As indicated in Chapter 6, the non-linear model has the characteristic that as soon as the interface is in a tensile state of stresses, the co nnection between the layers is broken and the layers are no longer bonded. Figure 8-19 shows the case when the slab is affected by only negative temperature differential using the non-linear m odel. The entire temperature differential was applied at the beginning of the computational analysis while the wheel load was applied at 20% load intervals. The total energy was used as a convergence criterion in the computational analysis. At the beginning of the an alysis, the corners of the slab curl up and the slab is no longer in contact with the asphalt. The incremental app lication of the 21 kips at the corner put the slab immediately back in cont act with the asphalt (Figure 8-20) not allowing complete development of tensile stress at the top of the slab as normally happens in a cantilever situation. Maximum tensile stress appears at the bo ttom of the slab (corners) similar to the case of the bonded linear model. Using Eq-8.1, the maximum number of load repetitions due to the application of the 21kips wheel load in both the linear an d the non-linear model can be estimated. NLR (21 kips, linear) = 112,201 NLR (21 kips, non-linear) = 14,791 It is to be pointed out that the 4 slab failed after 12,187 app lications of a load of 21 kips. This result shows that the com posite pavement must have beco me unbonded before it fails since the non-linear model representing the fully unbonde d condition could satisf actorily predict the number of load repetitions to failure in the HVS testing.

PAGE 222

222 Table 8-1 Model parameters of the 3-D model for each test section used in the analysis. Phase I-a Phase I-b Phase II 4 5 6 4 5 6 6 8 10 Concrete 4,350 4,3504,3504,3504,3504,3504,200 4,3504,350 AC Layer 3-11 102 3-11 1023-11 1023-11 1023-11 1023-11 1023-11 102 3-11 1023-11 102Base 160 160160130130130160 160160 Material Elastic Moduli Ksi Subgrade 30 303028282830 3030 Interface X 3-3.5 2-31-3 Interface Y 3-3.5 2-31-3 Interface Z 5 1012 10131013Trans. Joint X 0 000000 00 Trans. Joint Y 0 000000 00 Trans. Joint Z 0 000000 00 Long. Joint X 0 000000 00 Long. Joint Y 0 000000 00 Spring Constants lb/in 106 Long. Joint Z 0 000000 00

PAGE 223

223 Table 8-2 Maximum tensile stresses in the concrete slabs caused by a 24-kip single axle Load at various critical loading conditions. AC Elastic Modulus (psi) Tensile Stress (psi) Mid-Edge Corner Phase Slab Temp 300,000 700,000 1,100,000300,000 700,000 1,100,000 -10 246.5T307.7379.6259.6347.5 418.5 0 333.8221.2158.7204.3134.2 95.9 4 20 568.3441.7366.7325.8291.0 268.0 -10 223.7T285.4360.9251.5333.1 409.2 0 293.5208.2157.9171.0119.6 89.8 5 20 531.4433.3371.6290.9267.8 249.9 -10 202.0T260.5336.6248.3315.6 391.5 0 256.6190.7149.9146.0104.0 80.5 I-a (bonded) (6 6) 6 20 488.6412.1361.4279.2260.9 245.9 -10 323.6298.9304.3331.0303.6 341.3 0 286.3186.7131.1206.5152.7 125.2 4 20 555.1435.9364.8411.7340.0 295.9 -10 315.2295.8289.8340.9315.5 328.7 0 243.3170.9127.7200.8160.8 135.5 5 20 486.9405.4352.0366.5316.7 283.6 -10 296.8281.6277.1340.4315.9 308.9 0 209.7154.2119.7196.3161.3 138.8 I-b (bonded) (4 4) 6 20 416.2362.9324.2318.7284.5 260.2 -10 227.8 217.6 0 253.6 182.9 6 20 476.0398.2361.4272.4 -10 186.0 0 200.1 8 20 400.4353.1323.9 -10 165.3 0 158.6 II (partially bonded) 6 6 10 20 318.4290.3273.9 -10 172.7 0 197.5 8 20 398.8 -10 148.6 0 156.1 Hypothetical case (bonded) 6 6 10 20 316.8 T = stress located at the top of the slab Where no indicated, the stress is lo cated at the bottom of the slab

PAGE 224

224 Table 8-3 Maximum shear stress at the concre te-asphalt interface caused by a 24-kip single axle load at a temperature differential of +20 F for the bonded slabs. AC Elastic Modulus (psi 103 ) Shear Stress in the Interface (psi) Load at the Mid-Edge Load at the Corner Phase Slab 300 700 1100 300 700 1100 4 36.047.053.059.3 71.7 78.8 5 34.744.249.663.0 75.5 82.7 I-a 6 33.641.846.665.5 77.7 84.8 4 35.846.752.345.1 56.9 62.8 5 34.343.548.546.0 56.8 62.4 I-b 6 33.140.945.346.8 56.5 61.7

PAGE 225

225 Table 8-4 Maximum tensile stresses in the asph alt concrete layer caused by a 24-kip single Axle load at various critical loading conditions. AC Elastic Modulus (psi) Tensile Stress (psi) Mid-Edge Corner Phase Slab Temp Diff. F 300,000 700,000 1,100,000300,000 700,000 1,100,000 -10 97.8172.3230.9103.6182.1 243.4 0 21.865.697.825.379.0 122.3 4 20 40.8127.8202.745.5150.8 246.7 -10 100.7179.5242.6107.5191.4 258.0 0 18.355.9584.322.674.3 116.4 5 20 40.5129.8208.344.8156.8 260.7 -10 101.9182.8248.0109.8196.8 266.8 0 15.748.273.219.366.2 104.6 I-a (bonded) (6 6) 6 20 39.7129.1208.643.9160.0 268.8 -10 91.07161.5216.490.0159.8 214.2 0 21.358.785.624.068.2 101.4 4 20 48.4135.3206.857.8169.4 266.7 -10 92.6165.4222.991.6163.5 220.3 0 17.348.872.121.563.3 95.6 5 20 46.9133.5206.155.9173.2 276.4 -10 91.9165.9222.990.9163.0 220.0 0 14.441.261.519.558.8 90.0 I-b (bonded) (4 4) 6 20 44.6128.7206.863.4173.0 278.5 -10 107.6 86.7 0 15.7 18.6 6 20 37.0121.1197.942.9 -10 110.3 0 14.1 8 20 33.7114193.8 -10 104.3 0 16.2 II (partially bonded) 6 6 10 20 29.8105.9188.9 -10 100.4 0 11.8 8 20 37.0 -10 95.0 0 9.2 Hypothetical Bonded 6 6 10 20 33.7

PAGE 226

226 Table 8-5 Computed stress ratio in the concrete and allowable number of 24-kip single axle loads under critical loading c onditions for the test secti ons evaluated in this study. Phase Slab Thickness Stress (psi) Stress-strength Ratio # of Repetitions of 24-kip Axle Loads to Failure 4 568.30.675 3,810 5 531.40.631 13,231 I-a 6 488.60.580 56,178 4 555.10.659 5,958 5 486.90.578 59,416 I-b 6 416.20.494 no limit 6 476.00.565 85,963 8 400.40.476 no limit II 10 318.40.378 no limit Table 8-6 Level of tensile stresse s in the 4 slab that failed in Phase I-a. Load applied at the corner Temp. Diff 12 kips, linear analysis 15 kips, linear analysis 18 kips, linear analysis 21 kips, linear analysis 21 kips, non-linear analysis -10 285 289 293 316 355 0 198 248 297 347 +20 320 370 419 468 528

PAGE 227

227 48" 72" 48" Real Situation in the Road 48" 72" 48" Critical condition in the 3D Model 48" 48" Figure 8-1 Axle load positioned on slabs with 4-ft joint spacing. 72" 72" Real Situation in the Road 72" 72" 72" Critical condition in the 3D Model 72" Figure 8-2 Axle load positioned on slabs with 6-ft joint spacing.

PAGE 228

228 Stress Comparison Effect of the AC Elastic Modulus4"6'x6' slab Bonded condition Load at the mid-edge0 100 200 300 400 500 600 300,000700,0001,100,000MR (psi)Stress (psi) dT=-10 dT=0 dT=+20 T Stress is at the Bottom unless Top (T) is indicated Figure 8-3 Effect of AC elastic modulus on ma ximum tensile stress in concrete caused by a 24kip axle load at mid-edge of 4-inch bonde d concrete slabs with 6 ft joint spacing. Stress Comparison Effect of the AC Elastic Modulus5"4'x4' slab Bonded condition Load at the mid-edge 0 100 200 300 400 500 600 300,000700,0001,100,000 MR (psi)Stress (psi) dT=-10 dT=0 dT=+20 Stress is at the Bottom unless Top (T) is indicated Figure 8-4 Effect of AC elastic modulus on ma ximum tensile stress in concrete caused by a 24kip axle load at mid-edge of 5-inch bonde d concrete slabs with 4 ft joint spacing.

PAGE 229

229 Figure 8-5 Effect of temperature differential on maximum tensile stresses in concrete caused by a 24-kip single axle load at mid-edge of bonded slabs with 6 ft joint spacing. Figure 8-6 Effect of temperature differential on maximum tensile stresses in concrete caused by a 24-kip single axle load at corner of bonded slabs with 6 ft joint spacing. Stress Comparison Effect of Temperature DifferentialBonded condition 6' x 6' slab Load at the mid-edge 0 100 200 300 400 500 600 4" 5" 6" Slab ThicknessStress (psi) -10 F 0 F +20 F Stress Comparison Effect of Temperature DifferentialBonded condition 6' x 6' slab Load at the corner 0 50 100 150 200 250 300 350 4"5"6" Slab ThicknessStress (psi) -10 F 0 F +20 F

PAGE 230

230 Figure 8-7 Effect of temperature differential on maximum tensile stresses in concrete caused by a 24-kip single axle load at mid-edge of bonded slabs with 4 ft joint spacing. Figure 8-8 Effect of temperature differential on maximum tensile stresses in concrete caused by a 24-kip single axle load at corner of bonded slabs with 4 ft joint spacing. Stress Comparison Effect of Temperature DifferentialBonded condition 4' x 4' slab Load at the corner 0 100 200 300 400 500 4" 5" 6" Slab ThicknessStress (psi) -10 F 0 F +20 F Stress Comparison Effect of Temperature DifferentialBonded condition 4' x 4' slab Load at the mid-edge 0 100 200 300 400 500 600 4" 5" 6" Slab ThicknessStress (psi) -10 F 0 F +20 F

PAGE 231

231 Figure 8-9 Effect of temperature differential on maximum tensile stresses in concrete caused by a 24-kip single axle load at mid-edge of partially bonded slabs with 6 ft joint spacing. Stress comparison Effect of Slab SizeBonded Condition Load at the mid-edge 100 200 300 400 500 600 4" 5" 6" Slab Thickness (in)Stress (psi) 6'x6', T=-10 6'x6', T=0 6'x6', T=20 4'x4', T=-10 4'x4', T=0 4'x4', T=20 T T T Stress is at the Bottom unless Top (T) is indicated Figure 8-10 Effect of slab si ze on the maximum tensile stresses in concrete caused by a 24-kip single axle load at mid-edge of bonded slabs. Stress Comparison Effect of Temperature DifferentialUnbonded condition 6' x 6' slab Load at the mid-edge 0 100 200 300 400 500 6" 8" 10" Slab ThicknessStress (psi) -10 F 0 F +20 F

PAGE 232

232 Stress comparison Effect of Slab SizeBonded Condition Load at the corner 0 50 100 150 200 250 300 350 400 450 4" 5" 6" Slab Thickness (in)Stress (psi) 6'x6', T=-10 6'x6', T=0 6'x6', T=20 4'x4', T=-10 4'x4', T=0 4'x4', T=20 Stress is at the Bottom unless Top (T) is indicated Figure 8-11 Effects of slab size on the maximum tensile stresses in concrete caused by a 24-kip single axle load at corner of bonded slabs. Stress comparison for the 6" slabLoad applied at the mid-edge 0 100 200 300 400 500 600 -10 0 20 Temperature Differential (F)Stress (psi) 6" 6' x 6' Bonded 6" 6' x 6' Unbonded Stress is at the Bottom unless Top (T) is indicated T Figure 8-12 Effects of interface condition on maxi mum tensile stresses in concrete caused by a 24-kip single axle load at midedge of 6-inch slabs with 6 ft joint spacing.

PAGE 233

233 Stress comparison for the 6" slabLoad applied at the corner 0 50 100 150 200 250 300 -10 0 20 Temperature Differential (F)Stress (psi) 6" 6' x 6' Bonded 6" 6' x 6' Unbonded Stress is at the Bottom unless Top (T) is indicated Figure 8-13 Effects of interface condition on maximu m tensile stresses in concrete caused by a 24-kip single axle load at corner of 6-inch slabs with 6 ft joint spacing. Stress comparison Effect of bond in the interfaceLoad applied at the mid-edge 0 100 200 300 400 500 600 6" 8" 10" Slab ThicknessMaximum Stress (psi ) -10 F Phase II 0 F Phase II +20 F Phase II -10 F Fully bonded 0 F Fully bonded +20 F Fully bonded Stress is at the Bottom unless Top (T) is indicated Figure 8-14 Effects of interface condition on maxi mum stresses in concrete caused by a 24-kip single axle load at mid-edge of slab for the test sections in Phase II.

PAGE 234

234 Figure 8-15 Maximum shear stresses at the in terface caused by a 24-kip single load at a temperature differential of +20 F for the bonded slabs with 6 ft joint spacing. Figure 8-16 Maximum shear stresses at the interface caused by a 24-kip single load at a temperature differential of +20 F for the bonded slabs with 4 ft joint spacing. Shear Stress in the interfaceBonded Condition 4' x 4' slabs 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 4" 5" 6" Slab Thickness (in)Shear stress (psi) E=300,000 psi, Load at the edge E=1,100,000 psi, Load at the edge E=300,000 psi, Load at the corner E=1,100,000 psi, Load at the corne r Shear Stress in the interfaceBonded Condition 6' x 6' slabs 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 4" 5" 6" Slab Thickness (in)Shear stress (psi) E=300,000 psi, Load at the edge E=1,100,000 psi, Load at the edge E=300,000 psi, Load at the corner E=1,100,000 psi, Load at the corner

PAGE 235

235 Figure 8-17 Maximum tensile stre sses in the AC layer caused by a 24-kip single axle load at a temperature differential of +20 F for the bonded slabs with 6 ft joint spacing. Figure 8-18 Maximum tensile stre sses in the AC layer caused by a 24-kip single axle load at a temperature differential of +20 F for the bonded slabs with 4 ft joint spacing. Tensile Stress in the AC layerBonded Condition 6' x 6' slabs +20 F Temp diff. 0 50 100 150 200 250 300 4"5"6" Slab Thickness (in)Stress (psi) E=300,000 psi, Load at the edge E=1,100,000 psi, Load at the edge E=300,000 psi, Load at the corner E=1,100,000 psi, Load at the corner Tensile Stress in the AC layerBonded Condition 4' x 4' slabs +20 F Temp diff. 0 50 100 150 200 250 300 4" 5" 6" Slab Thickness (in)Stress (psi) E=300,000 psi, Load at the edge E=1,100,000 psi, Load at the edge E=300,000 psi, Load at the corner E=1,100,000 psi, Load at the corner

PAGE 236

236 Figure 8-19 4 slab of the non-linear model curled up at corn ers due to temperature negative temperature differential Figure 8-20 4 slab of the non-linear model loaded with 21 ki ps in the corner and negative temperature differential

PAGE 237

237 CHAPTER 9 CONCLUSION 9.1 Summary of Findings A f ull scale experiment was performed at the APT facility located at the FDOT Material Research Park to evaluate the feasibility of us ing whitetopping (WT) pavements in Florida. A total of nine instrumented WT test sections were constructed and tested using a Heavy Vehicle Simulator (HVS). Analysis for both strain and temperature from the full scale experiment was performed to evaluate pavement behavior. Labora tory testing was also performed to characterize both material properties and pavement response. A 3-D finite el ement model was developed to analyze the behavior of the WT pavement test sections. The model was verified and calibrated using the measured FWD deflections and HVS load -induced strains from the test sections. The model was then used to evaluate the potential pe rformance of these test se ctions under a typical critical temperature-load conditi on in Florida. A summary of th e findings from this study is presented in the following section. 9.1.1 Bond Strength at the Concrete-Asphalt Interface In the construction of the test sections with a bond ed concre te-asphalt interface (in Phases I-a and I-b), the asphalt surface was milled, cleane d and sprayed with water before the placement of concrete. This method was found to produce ex cellent bonding at the in terface. The average shear strength from the Iowa shear test on the cores from these test sections was 220 psi for Phase I-a, and 195 psi for Phase I-b. The maximum computed shear stress at the interface under the critical temperature-load condition for all ca ses is only 85 psi. No critical combination of tensile stress and shear stress was found at the interface. In the construction of the test sections with an unbonded c oncrete-asphalt interface (in Phase II), a white-pigmented curing compound was sprayed on the surface of the asphalt to act

PAGE 238

238 as a debonding agent before the placement of concre te. Results of Iowa shear test on the cores from these test sections indicated an average shear strength of 119 psi before the HVS loading and 135 psi after the HVS loading. This indicates that partial bonding exis ted at the interface though an unbonded condition was intended, and th at the bonding improved with additional loading on the pavement. Analysis of the strain and temperature data fr om the test track demonstrated that no curl occurred in the slabs in either of the phases. The bond strength in the interface was of such magnitude that the concrete slab was always in contact with asphalt even where there was a severe temperature differentia l in the concrete slab. 9.1.2 Development of the 3-D Finite Element Model A 3-D finite elem ent model, as described in Ch apter 6 of this report, was developed for the analysis of WT pavements. The model was verified and calibrated with the measured FWD deflections and HVS load-induced strains. It was found that the bonded interface (as existed in the test sections in Phases I-a and I-b) could be modeled we ll by modeling the concrete as perfectly bonded to the asphalt. The partially-bonde d interface (as existed in the test sections in Phase II) could be modeled well by vertical and horizontal springs connec ting the concrete layer with the asphalt layer. 9.1.2.1 Deflection-based calibration For the fully bonded condition in the interface, the model could satisfactorily replicate the deflection basin from FWD test by mainly considering material properties obtained from standard laboratory test. For the partially bonded condition, the model could replicate the deflection basin by adjusting mainly the value of the vertical springs in the interface.

PAGE 239

239 9.1.2.2 Strain-based calibration The m odel could satisfactorily re plicate not only the peak stra ins but also the shape of the strain profile in all analyzed cas es. The model was able to replic ate the strains in the fully bonded condition when the effect of temperature in the AC layer was considered in the model. In the case of the partially bonded condition, both the AC elastic modulus and the values of the springs modeling the interface need to be fully considered in the model. 9.1.2.3 Load transfer For the conditions at the tim e of the HVS loading on the test sections, it was found that joints could be modeled well by vert ical springs connecting the slabs at the joint. It is believed that the bond in the interface plays a major role in transferring th e load between adjacent slabs. It is postulated that load transfer at the joints will eventually decr ease with time. In the analysis for long-term performance of the test sections under critical conditions, the worst joint condition was assumed, and thus springs of zero stiffness were used to model the joint behavior in this analysis. 9.1.2.4 Interface bond The concrete slabs in Phase I were effectiv ely well bonded to the as phalt layer, and those from the Phase II were partially bonded. In bot h cases, the bond was strong enough to keep the slab in contact with the asphalt la yer, which justifies the use of linear springs in the interface for the model in Phase II. 9.1.3 Stress analysis of the test sections 9.1.3.1 Effects of elastic modulus of AC In both the analysis of the m easured strain da ta and the results from the analytical model, the elastic modulus of the AC layer was found to have great influence on the maximum tensile

PAGE 240

240 stresses in the concrete slab. Thus, for the an alysis for the most crit ical loading condition, the lowest possible elastic modulus of the AC (at the highest temperature) was used. 9.1.3.2 Effects of concrete panel size Maxim um stresses in the concrete were found to decrease as the join t spacing decreases. At the most critical loading condition, the conc rete slabs with 4 ft joint spacing had lower maximum stresses than thos e with 6 ft joint spacing. 9.1.3.3 Effects of bonded versus partially bonded interface At the condition of negative tem perature differe ntials in the concrete slab, the concrete slabs with a partially bond inte rface were found to have higher maximum stresses than those with a fully bonded interface. However, at th e condition of zero or positive temperature differential in the slab, the maximum stresses in the partially bonded slabs are about the same as those in the fully bonded slabs. 9.1.4 Performance of the Test Sections 9.1.4.1 Potential performance The verified and calibrated 3-D finite elemen t model was used to evaluate the potential performance of the nine test sections under a critical temperature-load condition. Maximum tensile stresses in the pavement were computed for the critical condi tion when a 24-kip single axle load (which is higher than the legal limit of 22 kips in Florida) was placed at the mid-edge of the slab (which is the most critical loadi ng position) and when the temperature differential in the concrete slab was +20 F (which is a typical severe temperatur e condition in the summer time in Florida.) The maximum computed stresses in the concrete slabs were all below the flexural strength of the concrete for all the 9 test sections. Based on the computed maximum stresses in the concrete, the expected numbers of repetitions of the 24-kip single axle loads at the critical

PAGE 241

241 thermal condition were computed for the nine test sections. The results show that the 4-inch slabs can be used for heavy (24-kip single axle ) load but only for low-volume traffic condition. The allowable traffic volume increases as the concre te slab thickness increases. In order to be able to withstand the critical load without fear of fatigue failure (f or an infinite number of critical load repetitions), a minimum slab thickness of 6 inches would be needed for a joint spacing of 4 ft, and a minimum slab thickness of 8 inches wo uld be needed for a joint spacing of 6 ft. 9.1.4.2 Mechanism of failure The actual perform ance of the 4 slab that failed during the HVS testing was evaluated using the 3D analytical model. A variation of the model with nonlinear springs in the interface was used to model the situation when the slab becomes unbonded due to temperature effect. The results of the analysis indi cated that the slab had become unbonded when it failed. 9.2 Limitations of the Research Som e of the limitations of the research are as follows: 1. During the loading process, a significant po rtion of the test se ctions were under the shade of the HVS. For this reason, the temperatur e differential measured in the experiment may not represent the maximum achie vable in a real pavement. 2. Only one of the test sections was loaded unt il failure, and this required a high number of load repetitions using a very heavy wheel load. Th e analysis of the perfor mance of the actual WT pavement was limited to this test section and it may not represent the general case. 3. Only two cases of bond conditions were considered in the full scale experiment, and even though they were intended to represent the extreme conditions of bond strength (fully bonded and fully un-bonded), both of them were in the range of partially bonded condition.

PAGE 242

242 4. The structural condition of th e asphalt layer under the concrete slabs in the test track can be considered between fair and good at the time of the HVS loading. This condition might not be representative of deteriorated asphalt layers considered as candidates for WT resurfacing. 5. The stress estimation using the analytical model should be considered only as representative of the test sections loaded in th e full scale experiment. More field experimentation needs to be done to extrapolate thes e results to other field conditions. 6. Due to the high demand on the use of the HV S, it was not possible to load the test sections for an extended period of time to ev aluate the long-term performance of the WT pavements and the modes of failures under actual traffic and weather conditions. 7. The analysis of the composite pavement did not consider the dynamic effect of the wheel load traveling at sp eeds faster than the one use in the HVS testing. 9.3 Recommendations and future work The developed 3-D finite elem ent model is recommended for use for analysis of WT pavements subjected to load and temperature effects. The model parameters needed in the analysis include the elastic moduli of the concrete, AC, ba se and effective subgrade layers. For analysis of long-term behavior, the join t stiffness can be assumed to be zero. For partially bonded interface condition, the stiffness va lues of the springs for modeling the interface are also needed as model parameters. The elastic moduli of concrete and AC can be determined by testing in the laboratory, while the other parameters can be determined through the back -calculation method of matching analytical deflections a nd strains due to an applied lo ad with measured values. It is recommended that experime ntal WT pavement test sec tions of various designs be constructed on actual roadways in Florida to evaluate their behavior and performance under actual environmental and traffic conditions. The e xperimental pavement sections will be instrumented for monitoring of temperature and strains on a long-term basis. This will enable the

PAGE 243

243 monitoring of the behavior of the WT pavement s under critical load a nd temperature conditions, and the verification of the predicte d response from the analytical mode l. It will also enable the evaluation of the long-term behavior of the WT pavements under actual traffic and weather conditions. In addition, the following recommendations for further research in whitetopping pavements are offered: 3D Finite element micromodel to study in more detail the interaction between the concrete slab and the asphalt layer in the interface Development of laboratory test s to estimate both the shear and tensile strength in the interface Full scale experiments that consider many types of interface bond from fully bonded to fully unbonded condition and also differe nt conditions in the asphalt layer HVS test long enough to produce cracking in the composite pavement so that design guidelines can be developed based on observed failure. A computer model to analyze the dynamic effect of the wheel traveling at typical speeds in highways.

PAGE 244

244 LIST OF REFERENCES Am erican Association of State Highway and Tran sportation Officials, (1 993). Guide for Design of Pavement Structures, Washington, DC. American Concrete Pavement Association, ( 1998). Whitetopping State of the Practice, ACPA Engineering Bulletin EB210P, Skokie, IL. Armaghani, J.M. and Tu, Diep. (1999). Rehabili tation of Ellaville Weigh Station with UltraThin Whitetopping, Transportation Resear ch Record 1654, Tran sportation Research Board, National Research Counc il, Washington, DC, pp. 3-11. Brown, D. (1995). Ultra-Thin Whitetopping Emerges as Reha b Technique, Transportation Builder, V7, No. 1, Jan. 1995, pp 37-41. Cable, J.K., Grove, J.D., and Heyer, M. (1997 ). Ultra-thin Pavements Making the Grade, Proceedings, Sixth International Purdue C onference on Concrete Pavement Design and Materials for High Performance, Volume II, Indianapolis, IN, pp. 245-266. Cable, J.K, (1998). Iowa Ultra-thin Whiteto pping Research, A Performance Update, Paper Presented at 1998 Transportation Research Board Annual Meeting, Washington DC. Cable, J.K. and Ciha, T. (2001). The Ultr a-thin Whitetopping Option, Proceedings of the Seventh International Conference on Concrete Pavements, Vol. 2, Orlando, Florida, September, pp. 969-975. Cole, L.W., and Mohsen, J.P. (1993). Ultra -Thin Concrete Overlays on Asphalt, Paper prepared for presentation at the 1993 TAC Annual Conference, Ottawa, Ontario, Canada. Cole, L.W., Sherwood, J., Qi, X. (1999). Accel erated Pavement Testing of Ultra-Thin Whitetopping, Accelerated Pavement Te sting Internationa l Conference, Reno. Cown, R.M. (1993). Experimental Concrete Inlay on Existing Asphalt Pavement, Georgia, Department of Transportation, Office of Materi als and Research, Conc rete Branch, Forest Park, GA. Dumitru, N. I., Hossain, M., and Wojakowski, J. (2002). Construction and Performance of Ultra-Thin Whitetopping in Kansas, Paper Presented at the 2002 Tr ansportation Research Board Annual Meeting, Washington, DC. Edwards, W. F. and Sargand, S. M., (1999). Response of an Ultra-Thin Whitetopping Pavement to Moving Wheel Loads, Accelerated Pa vement Testing International Conference, Athens, Ohio. Galal, K. A., Newbolds, S.A., Olek, J., Weiss, W. J., and Nantung, T. (20 04). Stress and Strain Analysis of Ultra-Thin Whitetopping over Co mposite Pavement Section using Accelerated Pavement Testing, Paper Presented at 2004 Transportation Re search Board Annual Meeting, Washington, DC.

PAGE 245

245 Hutchinson, R.L., (1982). Resurfacing with Port land Cement Concrete, Synthesis of Highway, Practice, NCHRP 99, Transportation Research Board, Washington, DC. Khazanovich, Lev, Gotlif, Alex. (2002). ISL AB2000 Simplified Friction Model. Paper Presented at 2002 Transportation Research Board Annual Meeting, Washington, DC. Mack, J.W., Wu, C.L., Tarr, S.M., and Refai, T. (1997). Model Development and Interim Design Procedure Guidelines for Ultra-thin Whitetopping Pavements, Proceedings, Sixth International Purdue Confer ence on Concrete Pavement Design and Materials for High Performance, Volume I, Indianapolis, IN, pp. 231-256. McGhee, Kenneth, H. (1994). Portland Cement Concrete Resurfacing, Synthesis of Highway Practice, NCHRP 204, Transportation Research Board, Washington, DC. Middleton, Brent, Fall, Lynne, Day, Robert. (2005). Durability and mechanical Properties of High Performance Concrete for Ultra-Thin Whitetopping Pavements. Paper Presented at 2005 Transportation Research Board Annual Meeting, Washington, DC. Nelson, Patricia K., Rasmussen, R obert O., (2002). Delamination Stresses at the Interface of Bonded Concrete Overlays. Paper Presente d at 2002 Transportation Research Board Annual Meeting, Washington, DC. Nishiyama, T., Lee, H., Asghar, B.M., (2005). Investigation of Bonding Condition in Concrete Overlay by Laboratory Testing, Finite Element Modeling and Field Evaluation, Transportation Research Record No. 1933, Washington, DC, pp. 15-23. Nishizawa, Tasuo, Murata, Yoshiki, Kokubu, Katsuro. (2003). Mechanical behavior of UltraThin whitetopping structure under stationary and moving loads, Tran sportation Research Record No. 1823, Washington, DC, pp. 102-110. Portland Cement Association, (1984). "Thickne ss Design for Concrete Highway and Street Pavements," Publication No. EB109.01P, Skokie, IL Rajan, S., Olex, J., Robertson, T.L., Galal, K ., Nantung, T., and Weis, J. (2001). Analysis of Performance of Ultra-Thin Whitetopping S ubjected to Slow Moving Loads in an Accelerated Pavement Testing Facility, 7th International Conf erence on Concrete Pavements, Orlando, Florida. Rasmussen, Robert O., Rozycki, Dan K. (2001). Characterization and Modeling of Axial SlabSupport Restraint, Paper Presented at 2001 Transportation Research Board Annual Meeting, Washington, DC. Risser, R.J., LaHue, S.P., Voigt, G.F., and Mac k, J. (1993). Ultra-Thin Concrete Overlays on Existing Asphalt Pavement, 5th Internationa l Conference on Concrete Pavement Design and Rehabilitation, Vol. 2, April, Pur due University, Lafayette, IN., pp. 247-254. Saeed A., Hammons, M.I., and Hall, Jr., J.W. (2001). Design, Construction, and Performance Monitoring of Ultra-thin Whitetopping at a General Aviation Airport, proceedings, 2001 ASCE Airfield Pavement Specialty Conferen ce, American Society of Civil Engineers, Chicago, Illinois.

PAGE 246

246 Saeed, A. and Hall, Jr., J.W. (2001). Non-destruc tive Pavement Evaluation and Design of Ultrathin Whitetopping at a General Aviation Airport in Tennessee, Proceedings Second International Conference on Maintenance and Rehabilitation of Pavements and Technical Control, Auburn University, Alabama. Speakman, J., and Scott, III, H. (1996). Ultra-Thin, Fiber-Reinforced Concrete Overlays for Urban Intersections, Transportation Research Record 1532, Advancements in Concrete Materials Technology, TRB, National Research Council, Washington, DC. Sprinkel, M.M., and Ozyildirim C., (1999). Evalu ation of the Installation and Initial Condition of Hydraulic Cement Concrete Overlays Plac ed on Three Pavements in Virginia, Interim Report. Report No. VTRC 99-IR3, April. Tarr, Scott M., Sheehan, Mattew J., Ardani, Ahmad. (2000). Mechanistic Design of Thin Whitetopping Pavements in Colorado, Tr ansportation Research Record No. 1730, Washington, DC., pp. 64-72. Tia, M., and Kumara, W. (2003) Evaluation of performan ce of Ultra-thin whitetopping by means of Heavy vehicle simulator (Analysi s, Planning and Design phase). Final Report UF Project No: 49104504863-12. Gainesville, Florida. Tia, M., Wu, C.L., Kumara, W. (2002). Forensi c Investigation of the Ellaville Weigh Station UTW Pavements, UF Project No : 49104504831-12. Gainesville, Florida. Tritsch, S. (1995). Whitetopping, Technique Revives Burge oning Kansas Thoroughfare, Roads and Bridges, September, pp. 52-55. Vandenbossche, J.M. and Fagerness, A.J., (2002) Performance and Repair of Ultra-Thin Whitetopping: The Minnesota Experience, Transportation Research Record No. 1809, Washington, DC., pp.191-198. Winkelman, Thomas J., (2005). The Illinois Wh itetopping Experience: A Practical Approach, Proceeding International Conference on Be st Practices for Ultra-thin and Thin Whitetopping, Denver, Colorado. Wu, C.L., Tarr, S.M., Refai, T.M., Nagi, M. N., and Sheehan, M.J. (1997). Development of Ultra-Thin Whitetopping Design Procedure, Report prepared for Portland Cement Association (PCA), PCA Serial No. 2124, Skokie, Illinois, January. Wu, C.L., Tayabji, S.D., Sheehan, M.J., and Sh erwood, J. (2001). Performance and Repair of UTW Pavements, Proceedings of the Seventh International Conference on Concrete Pavements, Vol. 2, Orlando, Florida, September, pp. 839-856. Wu, C.L., and Sheehan, M.J. (2002). Testing and Performance Evalua tion of UTW Pavements at the Spirit of St. Louis Airport, Transportation Resear ch Record No. 1809, Washington, DC., pp. 218-227.

PAGE 247

247 BIOGRAPHICAL SKETCH Patricio Tapia was born in 1965 in C alama, Chile. He is the son of Nelson Tapia and Carmen Gutierrez. He graduated with a bach elor degree from the Department of Civil Engineering of the Universidad Catolica del Norte, Chile in 1992. At the same time he received his PE license. One year after graduation he wa s recruited as an assistant professor in the Department of Civil Engineering at the Univer sidad Catolica del Norte, Chile. In 1992 he married Juana Mora. In 1993 his son Pablo was born. In 2002, after teaching for 10 years and attending non-degree program in Japan and USA, he was granted the Fulbright-Laspau Scholarship to pursuit doctoral studies in the USA. The same year, he enrolled in the Ph.D. program of the Department of Civil and Coastal E ngineering at the University of Florida. In May 2004 he received his masters degree in Engineering from UF. In August 2004 he earned a teaching assistantship in the material division in the Department of Civil and Coastal Engineering at UF. In 2006 he received the Outstanding Internati onal Student award for academic excellence at UF. The same year he wa s proposed as a department candidate for the Teaching Assistant award.