<%BANNER%>

Composite Specimen Testing to Evaluate the Effects of Pavement Layer Interface Characteristics on Cracking Performance

Permanent Link: http://ufdc.ufl.edu/UFE0042810/00001

Material Information

Title: Composite Specimen Testing to Evaluate the Effects of Pavement Layer Interface Characteristics on Cracking Performance
Physical Description: 1 online resource (160 p.)
Language: english
Creator: CHEN,YU
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ASPHALT -- BOND -- CRACKING -- FRACTURE -- INTERFACE
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: Open graded mixture which is widely used for Open Graded Friction Course (OGFC) has considerably lower values of fracture energy density and dissipated creep strain energy density to failure than dense-graded mixture. Since those OGFC may be ?first front? in resisting top-down cracking, previous research has suggest that the quality of the OGFC mixture and the bond between OGFC and the structural layer affects top-down cracking performance. The primary objective of this study is to identify a method to evaluate the effect of interface bonding condition on top-down and reflective cracking within the composite pavement system. A composite specimen interface cracking (CSIC) test was developed to evaluate the effects of pavement layer interface characteristics on cracking performance. The following factors were considered and evaluated when developing the testing system: loading mode, specimen symmetry, stress concentration method and specimen curved end surface reinforcement. Proper data collection and interpretation methods were developed to properly account for the complex stress distribution and progressive damage during the test. Three types of interface conditions were evaluated for top-down cracking, while two types of interface conditions were evaluated for reflective cracking. In addition, the effect of OGFC on top-down cracking was evaluated using this newly developed test method. Novabond? when used as a bonding agent between the OGFC and underlying structural layer, polymer modified asphalt emulsion (PMAE), increased the top-down cracking performance of composite pavement. More generally, it was determined that the effectiveness of bonding agent is closely related to its brittleness. Trackless tack coat, which is also polymer modified, but much more brittle than Novabond?, was determined to have a negative effect on top-down cracking performance. As compared with conventional tack coat, Novabond? was determined to improve the reflective cracking resistance of composite pavement. Novabond? application rate was determined to have a significant effect on the top-down and reflective cracking resistance. The introduction of an OGFC with conventional tack coat on dense-graded mixture was determined to reduce the cracking resistance of composite pavement as compared with pavements without OGFC. The research clearly illustrated the importance of interface bond and flexibility on top-down and reflective cracking performance.
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 YU CHEN.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Roque, Reynaldo.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042810:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042810/00001

Material Information

Title: Composite Specimen Testing to Evaluate the Effects of Pavement Layer Interface Characteristics on Cracking Performance
Physical Description: 1 online resource (160 p.)
Language: english
Creator: CHEN,YU
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ASPHALT -- BOND -- CRACKING -- FRACTURE -- INTERFACE
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: Open graded mixture which is widely used for Open Graded Friction Course (OGFC) has considerably lower values of fracture energy density and dissipated creep strain energy density to failure than dense-graded mixture. Since those OGFC may be ?first front? in resisting top-down cracking, previous research has suggest that the quality of the OGFC mixture and the bond between OGFC and the structural layer affects top-down cracking performance. The primary objective of this study is to identify a method to evaluate the effect of interface bonding condition on top-down and reflective cracking within the composite pavement system. A composite specimen interface cracking (CSIC) test was developed to evaluate the effects of pavement layer interface characteristics on cracking performance. The following factors were considered and evaluated when developing the testing system: loading mode, specimen symmetry, stress concentration method and specimen curved end surface reinforcement. Proper data collection and interpretation methods were developed to properly account for the complex stress distribution and progressive damage during the test. Three types of interface conditions were evaluated for top-down cracking, while two types of interface conditions were evaluated for reflective cracking. In addition, the effect of OGFC on top-down cracking was evaluated using this newly developed test method. Novabond? when used as a bonding agent between the OGFC and underlying structural layer, polymer modified asphalt emulsion (PMAE), increased the top-down cracking performance of composite pavement. More generally, it was determined that the effectiveness of bonding agent is closely related to its brittleness. Trackless tack coat, which is also polymer modified, but much more brittle than Novabond?, was determined to have a negative effect on top-down cracking performance. As compared with conventional tack coat, Novabond? was determined to improve the reflective cracking resistance of composite pavement. Novabond? application rate was determined to have a significant effect on the top-down and reflective cracking resistance. The introduction of an OGFC with conventional tack coat on dense-graded mixture was determined to reduce the cracking resistance of composite pavement as compared with pavements without OGFC. The research clearly illustrated the importance of interface bond and flexibility on top-down and reflective cracking performance.
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 YU CHEN.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Roque, Reynaldo.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042810:00001


This item has the following downloads:


Full Text

PAGE 1

COMPOSITE SPECIMEN TEST ING TO EVALUATE THE EFFECTS OF PAVEMENT LAYER INTERFACE CHARACTERISTI CS ON CRACKING PERFORMANCE By YU CHEN 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 2011 1

PAGE 2

2011 Yu Chen 2

PAGE 3

To my pare nts 3

PAGE 4

ACK NOWLEDGMENTS It is an honor and pleasure for me to thank and ack nowledge many individuals who have assisted and supported me during the course of my doctoral program. First of all, I would like to express my sincere appreciation to my committee chairman, Dr. Reynaldo Roque for his invaluable guidance and support throughout my studies at the University of Florida. I would not have been abl e to reach this milestone if it was not for his advice and understanding. I would also li ke to express my gratitude to other committee members, Dr. Denni s R. Hiltunen, Dr. Mang Tia, and Dr. Bhavani V. Sankar, for their support in acco mplishing my work. Many thanks should also be given to a ll former and current colleagues in the material and geotechnical group for their hel p and friendship during my stay in the Department of Civil and Coastal Engineerin g at the University of Florida. And lastly, I would like to thank my par ents in China and here in America for supporting my long pursuit of the Ph.D. degree. 4

PAGE 5

TABL E OF CONTENTS page ACKNOWLEDG MENTS..................................................................................................4 LIST OF TABLES............................................................................................................9 LIST OF FI GURES ........................................................................................................10 ABSTRACT ...................................................................................................................15 CHAPTER 1 INTRODUC TION....................................................................................................17 1.1 Background .......................................................................................................17 1.2 Hypot hesis ........................................................................................................18 1.3 Object ives.........................................................................................................19 1.4 Sc ope................................................................................................................19 1.5 Research Approach..........................................................................................20 2 LITERATURE REVIEW..........................................................................................21 2.1 Evaluation of Pavem ent Layer In terface...........................................................21 2.1.1 Background...........................................................................................21 2.1.2 Influencing Factors on Pavement Layer Interface Performance............23 2.1.2.1 Tack c oat ty pe.........................................................................23 2.1.2.2 Applic ation ra te........................................................................24 2.1.2.3 Curing time..............................................................................26 2.1.2.4 Tem perature............................................................................26 2.1.2.5 Surface conditi ons...................................................................28 2.1.2.6 Pavement materials and stru cture...........................................30 2.1.2.7 Testing methods and loading c onditions..................................32 2.1.3 Effects of Bonding Conditi ons on Pavement Performance....................33 2.1.4 Su mmary...............................................................................................35 2.2 Top-down and Reflective Cracking Me chanism ................................................36 2.2.1 Top-down Cra cking Mechan ism............................................................36 2.2.2 Reflective Cr acking Mech anism............................................................38 2.2.3 Su mmary...............................................................................................39 2.3 Testing Methods for Pavement Layer Interfac e Evaluat ion...............................40 2.3.1 Interface Shear Resistance Te sting ......................................................40 2.3.1.1 ASTRA te st set-up...................................................................40 2.3.1.2 Double shear te st....................................................................41 2.3.1.3 FDOT s hear tester...................................................................43 2.3.1.4 Layer-Parallel Direc t Shear (LPD S) te st..................................44 2.3.1.5 Leutner test..............................................................................45 2.3.1.6 Shear fatigue te st....................................................................46 5

PAGE 6

2.3.1.7 Superpave Shear Te ster ..........................................................46 2.3.1.8 Tor que bond te st.....................................................................48 2.3.1.9 UTEP Pull-Off Device (UPOD)................................................51 2.3.1.10 Wedge splitting te st...............................................................52 2.3.2 Interlayer Cracking Resistance Testing.................................................54 2.3.2.1 TTI over lay test er.....................................................................54 2.3.2.2 Interlayer Stress Absorbing Composite (ISAC) system testing equi pment ...................................................................56 2.3.2.3 Beam tests...............................................................................57 2.3.3 Su mmary...............................................................................................60 3 COMPOSITE SPECIMEN PREPARATION AND E VALUATION............................61 3.1 Composite Spec imen Preparat ion....................................................................61 3.2 Evaluation of the Effect of Overlay Compaction on the Integrity of Lower Layer ..............................................................................................................65 3.2.1 Materials and Testing Me thods.............................................................65 3.2.2 Analysis of Test Result ..........................................................................66 3.3 Summa ry..........................................................................................................72 4 DEVELOPMENT OF COMPOSIT E SPECIMEN TENS ION TEST..........................73 4.1 Background .......................................................................................................73 4.2 Prototype Compos ite Specimen Tensi on Test System.....................................73 4.2.1 Slicing, Cutting and Groovi ng of Composit e Specim en.........................75 4.2.2 Sanding, Gluing, and Gage Poin ts Attachment of Composite Specim en............................................................................................78 4.3 Monotonic Strength Te sts on Asymmetrical Co mposite Sp ecimens.................80 4.3.1 Mate rials................................................................................................80 4.3.2 Monotonic Strength Test at 25 mm/min.................................................81 4.3.3 Monotonic Strength Test at 2.5 mm/min and 0.25 mm/min...................85 4.3.4 Conclusion for Asymmetric Monotonic Str ength Test s ..........................88 4.4 Repeated Loading Test on Asymmetrical Composit e Specim ens....................89 4.4.1 Repeated Loading Test on Asymmetrical Composite Specimen with 1 inch Dense-graded Mixture La yer.....................................................89 4.4.1.1 Mate rials..................................................................................89 4.4.1.2 Analysis of resu lts....................................................................90 4.4.2 Repeated Loading Test on Asymmetrical Composite Specimen with 3 inch Dense-graded Mixture La yer.....................................................93 4.4.2.1 Mate rials..................................................................................94 4.4.2.2 Analysis of resu lts....................................................................95 4.4.3 Repeated Loading Test on Asymmetrical Composite Specimen with Rectangular Groove ............................................................................97 4.4.3.1 Mate rials..................................................................................97 4.4.3.2 Analysis of resu lts....................................................................98 4.5 Repeated Loading Test on Symmetr ical Composite Specimen ......................101 4.5.1 Materials and Te sting Met hod.............................................................101 6

PAGE 7

4.5.2 Alignment Syste m for Com posite Specimen Pr eparation....................103 4.5.3 Results of Symmetrica l Composite Sp ecimen.....................................103 4.6 Monotonic Partial Loadi ng on Symmetric Com posite Spec imen.....................105 4.6.1 Optimum Geometry of Symmetr ic Composite Specimen for Partial Loading.............................................................................................. 105 4.6.2 Analysis of Result s..............................................................................108 4.7 Monotonic Internal Loading on Sy mmetric Composit e Specimen...................111 4.7.1 Materials and Lo ading Assembly.........................................................111 4.7.2 Analysis of Result s..............................................................................111 4.8 Monotonic Internal Loading on Sy mmetric Composite Specimen with Carbon Fiber Sheet Reinforcem ent..............................................................115 4.8.1 Materials and Spec imen Preparat ion..................................................115 4.8.2 Analysis of Result s..............................................................................115 4.9 Repeated Internal Loading on Symme tric Composite Specimen....................118 4.9.1 Materials and Loadi ng Configurat ion...................................................118 4.9.2 Analysis of Result s..............................................................................119 5 DATA COLLECTION AND IN TERPRETATION METHOD...................................121 5.1 Data Collect ion Met hod...................................................................................121 5.2 Data Interpre tation Met hod.............................................................................122 5.2.1 Data Interpre tation Method..................................................................122 5.2.2 Damage Rate as a Diffe rentiation Para meter......................................125 6 INTERFACE CRACKING PERFORMANCE EVALUA TION.................................127 6.1 Effects of Interface on Top-down Cr acking.....................................................127 6.1.1 Effects of Novabond on Top-down Cr acking .....................................127 6.1.2 Effects of Trackless Tack on Top-down Cracking ...............................128 6.2 Effects of Interface on Reflective Cr acking.....................................................132 6.2.1 Effects of Novabond on Reflective Cracking......................................132 6.2.2 Effects of Novabond on Reflective Cracking on Specimens with Teflon Spac er....................................................................................135 6.2.3 Effects of Novabond on Double Novachip Reflective Cracking.........138 6.3 Effects of OGFC on Top-down Cra cking.........................................................142 7 SUMMARY, CONCLUSIONS AN D RECOMMENDATI ONS.................................145 7.1 Summary and C onclusion s.............................................................................145 7.2 Recommendat ions..........................................................................................147 APPENDIX A THE AMOUNT OF INTERFACE MATERIAL CALC ULATION..............................148 B THE AMOUNT OF OVERLAY MATERIAL CALC ULATION.................................149 LIST OF REFE RENCES.............................................................................................150 7

PAGE 8

BIOGRAPHICAL SKETCH .......................................................................................... 160 8

PAGE 9

LIST OF TABLES Table page 3-1 Dense-graded mixtur e aggregate gr adation .......................................................67 3-2 Bulk specif ic gravity............................................................................................67 3-3 Superpave IDT test results.................................................................................67 3-4 Damage rate re calculat ion..................................................................................71 4-1 Oolitic limestone FC-5 mixture aggregate gradatio n...........................................81 4-2 Nova scotia-granite FC-5 mixture aggregat e gradatio n......................................98 4-3 Material properties of composite s pecimen ......................................................105 6-1 Novabond, conventional tack and trackle ss tack applicati on rate...................130 6-2 Novabond application rate of double Novachip.............................................139 6-3 Double Novachip lift thick ness........................................................................140 9

PAGE 10

LIST OF FIGURES Figure page 2-1 Stress distribution for fully bonded (left) and fully slipped (right) pavement system (after Gom ba et al 2005) .......................................................................34 2-2 Reflective cracking in HMA overlay of PCC base (after Von Quintus et al. 2010).................................................................................................................. 39 2-3 Configuration of the ASTRA test device (after Canestrar i et al. 2005)................41 2-4 Schematic of the double shear te st (after Diakhat e et al 2011).........................42 2-5 Simple direct shear device developed by FDOT (after Sholar et al. 2002).........43 2-6 Schematic view of the LPDS test device with pneumatic clamping (after Raab and Partl 2008)...................................................................................................44 2-7 Photograph and schematic diagram of Leutner load frame (after Collop et al. 2009).................................................................................................................. 45 2-8 Schematic of shear fatigue test (after Romanoschi and Metcalf 2001)...............47 2-9 Design shear mold (left) and mold with a sample inside (right) (after Mohammad et al. 2005) ......................................................................................48 2-10 Test arrangement for the shearing test (after Mohamma d et al. 2005)...............48 2-11 Photograph and schematic diagram of the automatic torque equipment (after Collop et al. 2010) ..............................................................................................50 2-12 UTEP pull-off device test set-up (after Deysarkar and Tandon 2004)................52 2-13 Schematic view of the wedge splitting test (after Tschegg et al. 2007)..............53 2-14 Schematic diagram of TTI overlay tester (after Zhou and Scul lion 2004)...........54 2-15 Typical displacement used in over lay tester (after Zhou and Scullion 2004)......55 2-16 Schematic diagram of ISAC system testing e quipment (after Mukhtar and Dempsey 1996)..................................................................................................57 2-17 Test arrangement (aft er Kim et al. 1999) ............................................................58 2-18 Schematic diagram of the testing set-up (aft er Brown et al. 2001).....................59 2-19 Schematic diagram of the test setup (afte r Khodaii et al. 2009) .........................60 10

PAGE 11

3-1 Half sliced specim en...........................................................................................62 3-2 Silicone rubber mold on level shelf.....................................................................62 3-3 Dense-graded mixture plac ed on bonding agent residue...................................63 3-4 Dense-graded specimens with applied conventional tack coat (left) and Novabond (right)...............................................................................................63 3-5 Compacted base material being pushed into SGC co mpaction mold.................64 3-6 Newly compacted composite s pecimen ..............................................................65 3-7 Creep compliance versus time fo r no re-compaction replicate 1........................68 3-8 Creep compliance versus time fo r no re-compaction replicate 2........................69 3-9 Creep compliance versus time for 50 additional gyrations.................................69 3-10 Creep compliance versus time for 100 additional gyrations...............................70 3-11 Creep compliance versus time for 150 additional gyrations...............................70 3-12 Re-calculated creep rates fo r no re-compaction and re-compacted specim ens.......................................................................................................... 71 4-1 Prototype of composit e specimen tens ion te st...................................................74 4-2 Diamond-tip saw used for specimen slicin g.......................................................75 4-3 Sliced compos ite spec imen................................................................................76 4-4 Diamond-tip saw used for composite spec imen cutti ng......................................76 4-5 Composite specim en after cu tting......................................................................77 4-6 Composite specimen stress concentrator dr illing setu p......................................77 4-7 Grooved compos ite spec imen............................................................................78 4-8 Spindle sander and sanded specimen................................................................79 4-9 Strain gage distributions on the composit e specim en.........................................79 4-10 Loading head and epoxy used............................................................................80 4-11 Geometry of the five cu ts to form the groov e......................................................82 4-12 Prepared composite specimen wi th diamond saw cut groov e............................82 11

PAGE 12

4-13 Cracks in the specimen at loading rate of 25 mm/m in........................................83 4-14 Strain gage displacement at loading rate of 25 mm/min.....................................84 4-15 Strain gage displacement at loading rate of 25 mm/min.....................................84 4-16 Cracks in the specimen at loading rate of 2.5 mm/ min.......................................85 4-17 Strain gage displacement at loadi ng rate 2.5 mm/min-broke near the end.........86 4-18 Strain gage displacement at loading rate 2.5 mm/min-broke at the center.........86 4-19 Cracks in the specimen at loading rate of 0.25mm/min......................................87 4-20 Strain gages displacement at loading rate of 0.25mm/ min.................................88 4-21 Repeated loading sc hematic di agram................................................................89 4-22 Composite specimen broke near the end ...........................................................91 4-23 Compacted open-graded mixture surface with r ough surfac e............................91 4-24 Compacted open-graded mixture surf ace after rough su rface s liced.................92 4-25 Composite specimen wit h top 3/8 inch rough surface slic ed off.........................92 4-26 Total recoverable deformation for OGFC and dense-gr aded mixture.................93 4-27 Strain gage distributions on specimen with 3 inch dens e-graded mixture..........94 4-28 Total recoverable deformation fo r OGFC and dense-graded mixture with tack coat interface (3 inch dens e-graded mixtur e layer).............................................96 4-29 Total recoverable deformation for OGFC and dense-gr aded mixture with Novabond interface (3 inch dens e-graded mixtur e layer).................................96 4-30 Composite specimen wit h rectangular groove....................................................97 4-31 Total recoverable deformation for OGF C with tack coat interface (rectangular groove) ...............................................................................................................99 4-32 Total recoverable deformation for dense-graded with tack coat interface (rectangular groove) ...........................................................................................99 4-33 Total recoverable defo rmation for OGFC with Novabond interface (rectangular gr oove).........................................................................................100 4-34 Total recoverable deformati on for dense-grad ed with Novabond interface (rectangular gr oove).........................................................................................100 12

PAGE 13

4-35 Symmetric al com posite spec imen....................................................................102 4-36 Cracking surfaces for fiber glass (l eft) and cardboard (r ight) spacer................102 4-37 Loading heads with tw o steel angl es................................................................103 4-38 Total recoverable deformation for symmetrical composite specim en...............104 4-39 Half composite spec imen modeling sketch.......................................................106 4-40 Effect of composite specimen di ameter on stress di stributio n..........................107 4-41 Effect of constraint condi tion on stress dist ribution...........................................107 4-42 Schematic diagram of the 3 inch diameter specimen co ring............................109 4-43 Prepared 3 inch diameter symme trical composit e specim en............................109 4-44 Loading heads, alignment bar and shim blocks...............................................110 4-45 Failure mode for composite specimen under monotonic parti al loading...........110 4-46 Loading assembly for in ternal load ing..............................................................112 4-47 Test setup for in ternal load ing..........................................................................113 4-48 Failure mode for composite specimen with tack coat interface........................114 4-49 Failure mode for composite specimen with Novabond interface.....................114 4-50 Composite specimen curved end surfac e, epoxy, and carbon fiber sheet........116 4-51 Composite specimen curved end surfac e with glued carbon fiber sheet..........117 4-52 Failure mode of composite specimen with curved end surface reinforced.......117 4-53 Load-displacement curve for specimens with curved end surface reinforced...118 4-54 Failure mode for composite specimen under repeated inter nal loading...........120 4-55 Number of cycles to failure of Novabond and conventio nal tack.....................120 5-1 Typical repeated load ing versus time...............................................................123 5-2 Data points recorded for recove rable deformation calculation..........................123 5-3 Total recoverable def ormation defin ition...........................................................124 5-4 Typical total recove rable deforma tion...............................................................125 13

PAGE 14

5-5 Normalized total recoverable defo rmation........................................................126 6-1 Stain gage distribution on composite s pecimen ................................................128 6-2 Damage rate of Novabond and conventional tack c oat interface....................129 6-3 Cracking surfaces of specimens with Novabond and tack coat interface.......129 6-4 Number of cycles to failure of trackless tack and conv entional tack.................131 6-5 Damage rate of trackless tack and conventio nal tack.......................................131 6-6 Cracking surfaces of specimens with trackless tack interface..........................132 6-7 Composite specimen geometry and strain gage di stribution............................133 6-8 Number of cycles to failure of Novabond and diluted conventional tack.........133 6-9 Damage rate of Novabond and diluted convent ional tack...............................134 6-10 Cracking surfaces of specimens with Novabond and diluted conventional tack interface....................................................................................................135 6-11 Composite specimen with teflon s pacer...........................................................136 6-12 Number of cycles to failure of Novabond and diluted conventional tack for specimens with te flon spac er............................................................................137 6-13 Damage rate of Novabond and diluted conventional tack for specimens with teflon spac er.....................................................................................................137 6-14 Double Novachip composite s pecimen ...........................................................139 6-15 Cracking surfaces of double Novachip composite s peciemn..........................140 6-16 Number of cycles to failure for double Novachip specimen............................141 6-17 Damage rate of double Novachip specim en...................................................141 6-18 Uneven concrete thicknes s..............................................................................142 6-19 All dense-graded mixt ure specim en.................................................................143 6-20 Number of cycles to failure of OGFC on dense with conventional tack interface and all dense-graded .........................................................................143 6-21 Damage rate of OGFC on dense with c onventional tack interface and all dense-grad ed...................................................................................................144 14

PAGE 15

Abstract of Dissertation Pr esented to the Graduate School of the University of Fl orida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy COMPOSITE SPECIMEN TESTING TO EVALUATE THE EFFECTS OF PAVEMENT LAYER INTERFACE CHARACTERISTI CS ON CRACKING PERFORMANCE By Yu Chen May 2011 Chair: Reynaldo Roque Major: Civil Engineering Open graded mixture which is widely used for Open Graded Friction Course (OGFC) has considerably lower values of fracture energy density and dissipated creep strain energy density to failure than dens e-graded mixture. Since those OGFC may be first front in resisting topdown cracking, previous research has suggest that the quality of the OGFC mixture and the bond between OGFC and the structural layer affects topdown cracking performance. The primary objective of this study is to identify a method to evaluate the effect of interface bonding c ondition on top-down and reflective cracking within the composite pavement system. A composite specimen interface cracking (CSIC) test was developed to evaluate the effects of pavement layer interface characteristics on cracking performance. The following factors were cons idered and evaluated when dev eloping the testing system: loading mode, specimen symmetry, stress concentration method and specimen curved end surface reinforcement. Proper data collection and interpretation methods were developed to properly account for the comp lex stress distribution and progressive damage during the test. Three types of interf ace conditions were evaluated for top-down cracking, while two types of interface conditi ons were evaluated for reflective cracking. 15

PAGE 16

16 In addition, the effect of OGFC on top-down cracking was evaluated using this newly developed test method. Novabond when used as a bonding agent between the OGFC and underlying structural layer, polymer mo dified asphalt emulsion (PMAE), increased the top-down cracking performance of co mposite pavement. More generally, it was determined that the ef fectiveness of bonding agent is closely related to its brittleness. Trackless tack coat, which is also polymer modified, but much more brittle than Novabond, was determined to have a negative effect on top-down cracking performance. As compared with c onventional tack coat, Novabond was determined to improve the reflective cracking resi stance of composite pavement. Novabond application rate was determined to have a si gnificant effect on the top-down and reflective cracking resistance. The introduc tion of an OGFC with conventional tack coat on dense-graded mixture was determined to reduce the cracking resistance of composite pavement as compared with pavem ents without OGFC. The research clearly illustrated the importance of interface bond and flexibility on top-down and reflective cracking performance.

PAGE 17

CHA PTER 1 INTRODUCTION 1.1 Background Since first constructed by California Department of Highways in 1944, OGFC has been widely used in the United States (NCHRP 2000). With primarily single sized and gap-graded gradation, OGFC has a much higher percentage of air voids, in the range of 15% to 25%, as compared to dense-graded mixture (Kandhal and Mallick 1998). This high air void content gives OGFC a great deal of potential to improve road safety under wet conditions by reduction in hydroplaning and tire splash and spray, especially in southern parts of the United States with frequent high intensity short-duration rainfall events. However, there are several problems associated with OGFC, including low durability and rutting resistance, and decreasi ng porosity through voids clogging. Most of the OGFC research work done so far has been related to function issues, such as skid resistance, permeability and noise redu ction. Most existing pavement design guidelines consider the OGFC as a wearing surface layer with no structural value. However, it is recognized that OGFC mixtures may be the first front in resisting top-down cracking. Thus it is necessary to evaluate the fracture resistance of the OGFC. Open graded mixture has considerably lower values of fracture energy density and dissipated creep strain energy density to failure than dense-graded asphalt mixture based on specimens produced in the labor atory (Koh 2009). However, fracture resistance of open graded mixture from field cores has not been measured because of difficulties in recovering representativ e specimens from thin OGFC layers. Based on the analysis of findings from pavem ent field sections in Florida (e.g. SR16 in Bradford County, FL and US-27 in Highl ands County, Florida), there is suggestive 17

PAGE 18

evidenc e that the quality of OGFC and the effectiveness of the bond between OGFC and the structural mixture affects top-down cracking per formance. A bonded OGFC is an open graded HMA mixture laid on a dense-gr aded mixture using a relatively thick polymer modified asphalt emulsion, such as Novabond. Polymer modified asphalt emulsion seals existing pavement and bonds t he OGFC to the underlying surface. The relatively thick nature of the polymer modi fied asphalt emulsion allows it to migrate upwards into the OGFC, filling voids in the aggregate and creating an interface of high cohesion. Some of the potential adv antages of bonded OGFC are dissipation of stresses and increased fracture resistance of the OGFC near the interface. In addition, recent laboratory work at the University of Florida (Birgisson et al. 2006) suggests that cracks that develop either in the OGFC or the HMA structur al layer can be effectively arrested and/or deterred with appropriate interface conditions and bonded interface formed at the bottom of the OGFC, which reduces the rate of creep and damage of the composite system. However, limited work has been done to develop a suitable fracture performance test (and test conditions) that can characterize the effect of bonded interface on top-down cracking. Therefore, it is necessary to identify and develop a method to evaluate the effect of bonded interface on top-down or reflective cracking within the composite pavement system. This test may also serve as an appr opriate method to evaluate the effect of OGFC or other thin layers on top-down cracking. 1.2 Hypothesis The use of highly cohesive interface formed by introducing a relatively thick polymer modified asphalt emulsion between asphal t mixture lifts (surfa ce to structural, structural to existing) can increase the re sistance to top-down or reflective crack 18

PAGE 19

propagation by dissipating stresses or by reducing stresses transmitted through the interface. 1.3 Objectives The objectives of this research may be summarized as follows: Develop a composite specimen testi ng method for the evaluation of bonded interface on top-down and reflective cracking resistance; Evaluate the effect of bonded interf ace on top-down cracking resistance; Evaluate the effect of bonded interfac e on reflective cracking resistance; Evaluate the effect of OGFC on top-down cracking; 1.4 Scope This study will primarily focus on the i dentification of an appr opriate evaluation method that allows for the characterization of the effect of bonded interface on top-down cracking and reflective cracking resistance and the effect of OGFC or other thin layers on top-down cracking. In order to develop testing protoc ols for composite specimen tests, two Superpave dense-gr aded mixtures were used to evaluate the possible damage to the dense-graded HMA induced during the composite specimen compaction process. One Superpave dense-graded mi xture and two open-graded mixtures were used to produce composite specimens for te st method development. Three types of interface bonding conditions were exami ned: the conventional tack coat, Novabond polymer modified asphalt emulsion and trackle ss tack. For each interface type, only one application rate was examined, as recommen ded by manufacturers. For the effect of bonded interface on reflective cracking resistance, composite specimens consisting of dense-graded mixture placed on dense-graded mixture with two types of interfaces, and double Novachip on concrete field cored composite specimens with 3 different 19

PAGE 20

Novabond application rates will be examined. All tests were conducted at one temperature (10C), which has been determined to work well in prior fracture research at UF. 1.5 Research Approach The approach taken for the development of an appropriate method for the evaluation of the effects of bonded interfac e on top-down and reflective cracking and OGFC on top-down cracking involved the following: Literature review: (1) Evaluation of pavem ent interface; (2)Top-down and reflective cracking mechanism; (3) Existing testing methods for composite specimen cracking resistance evaluation, with or without interface. Test method development: Develop a composite specimen testing method and associated data interpretation method for t he evaluation of the effect of bonded interface on top-down and reflective cracking. Conduct composite specimen testing to (1) ev aluate the effect of different interface conditions on top-down cracking; (2) evaluat e the effect of different interface conditions on reflective cracking; (3) ev aluate the effect of OGFC on top-down cracking. During the development of the appropriate testing method, finite element method analysis will be conducted to optimize the composite specimen geometry. 20

PAGE 21

CHA PTER 2 LITERATURE REVIEW 2.1 Evaluation of Pavement Layer Interface 2.1.1 Background Bonding agent is usually applied on existi ng clean asphalt or concrete surfaces to provide adhesive bond between existing pa vement surface and newly constructed asphalt overlay. Three types of bonding agent will be evaluated in this work, conventional tack coat, polymer modified asphalt emulsion, and trackless tack. Tack coat is the most commonly used bonding agent between pavement layers. According to ASTM D 8-02 Standard Terminology Relating to Materials for Road and Pavements, Tack coat (bond coat) is an application of bituminous material to an existing relatively non absorptive surface to provide a thorough bond between old and new surfacing (ASTM 2003). The tack mate rial can be asphalt emulsion (slow, medium, and fast setting), cutback asphalt, hi gh float emulsion, and polymer modified asphalt emulsion, and paving grade asphalt ce ment (Cross and Shrestha 2005; West et al. 2005; Wheat 2007). Since cutback asphalt cont ain volatile chemicals that evaporate into the atmosphere whereas emulsified asphal ts evaporate water into the atmosphere, its usage has been greatly reduced (Texas Tec hnical Advisory 2001). Several surveys indicated that anionic and cationic slow se tting asphalt emulsions are the most widely used tack material (Chaignon and Roffe 2001; Mohammad et al. 2008; Santucci 2009). Paving grade asphalt cement is occasionally used for night work or work in cool weather or where multiple pavement layers are being placed and closure time is critical because it does not need any time to break, but they need to be heated for spray application (Cross and Shrestha 2005; Tack Coat Guidelines 2006; Santucci 2009). 21

PAGE 22

Polymer modified asphalt emulsion has been widely used in ultrathin friction course (Russell et al. 2008; Kandhal and Lockett 1997; Cooper and Mohammad 2004), ultrathin bonded wearing c ourse (Ruranika 2007), bonded wearing course (MTAG 2007), and bonded fric tion course (Birgisson et al. 2006). In spite of the differences among ultrathin friction course, ultrat hin bonded wearing course, bonded wearing course, and bonded friction course, all are developed from NovaChip with different overlay mixtures and tack coat materials. NovaChip is a paving process utilizing a single piece of equipment (spray-pave r) to place a thin, gap graded HMA onto a relatively thick Novabond membrane (polymer modified asphalt emulsion seal coat) (Kandhal and Lockett 1997; Russell et al. 2008). The polymer modified asphalt emulsion seals the existing pavement surfac e and produces high asphalt binder content at the interface of existing pavement su rface, bonding the gap or open-graded mixture to the original pavement surface (MTAG 2007) The relatively thick nature of the polymer modified asphalt emulsion allows it to migrate upwards into the OGFC, filling voids in the aggregate and creating an interface of high cohesion. Tracking or pickup occurs when the tack c oat is picked up by the rubber tires of construction equipment and removed from the existing pavement surface. The survey conducted by Mohammad et al. (2008) indica ted that 38% of t he respondents required that the tack coat material s hould be completely set before haul trucks are allowed on it to reduce the tracking problem. At present, t here are several trackless tack products on the market (e.g., Blacklidge Emul sion, Inc. trackless tack and Ultrabond, and NTSS1HM emulsion trackless tack). Trackless tack can be driven over by haul trucks and paver without tracking. 22

PAGE 23

2.1.2 Influencing Factors on Pavem ent Layer Interface Performance It is generally believed that pavement laye r interface characteristics between an existing pavement and an overlay are affected by several factors. These factors include the tack coat type, application rate, curi ng time, moistu re, te mperature, surface conditions, pavement structure, and testing method and loading conditions. 2.1.2.1 Tack coat type Various types of materials, includi ng asphalt emulsion, paving grade asphalt binder, cutback asphalt, and trackless tack, hav e been used as interface layer materials between AC layers, and between AC and PCC la yers. Many studies have reported that paving grade asphalt binders, like PG64-22 and PG67-22, have higher shear strength compared to asphalt emulsions when they are used as interface layer materials between HMA layers (Buchanan and Woods 2004; Tayebali et al. 2004; West et al. 2005). The possible reason for this is attri buted to the higher viscosity of paving grade asphalt binder. However, Tayebali et al. ( 2004) pointed out that asphalt emulsion CMS2 performs better than PG 6422 when they are used as interface materials between PCC and AC layers. This phenomenon was explained by s lippage that occurred between PCC and AC layers because of t he impervious ness of the PCC layer. Latex modified asphalt emulsion (CRS-2L) has been reported to have better performance than unmodified asphalt emulsi on (SS-1, CSS-1, SS-1h and SS-1L) and paving grade asphalt binder (P G64-22 and PG76-22M) at 130 F; but statistical analysis indicated the latex modified asphalt emulsion is not significantly different from the same emulsion without latex (Mohammad et al. 2005). Mohammad et al. (2009) indicated that tr ackless tack exhibited higher interface shear strength than asphalt emulsions CRS-1 and SS-1h at three applic ation rates, 0.14 23

PAGE 24

(0.031), 0.28 (0.062) and 0.7 (0.155) l/m2 (gal/sy). These results we re directly related to the viscosity of the residual binder at 25C; this supports the view of Tayebali et al. (2004). The non-tracking tack used by VDOT ex hibited laboratory lower shear strength than asphalt emulsion CSS-1h and CRS-2 (McGhee and Clark 2009). However, the difference was attributed to the different interface layer materials and testing methods. In order to evaluate the effect of inte rface conditions between AC overlay and PCC pavement on the overlay service life, the laboratory work performed by Leng et al. (2008) concluded that asphalt emulsion, SS -1hP, provided better interface shear strength than cut back asphalt, RC-70 at the same application rate; Accelerated Pavement Testing (APT) indicated t hat PG64-22 and SS-1hP have better rutting resistance than RC-70 when they are applied at the same residue application rate (Leng et al. 2009). 2.1.2.2 Application rate Insufficient bond between pavement layers can result in a significant reduction in the shear strength in the pavem ent structure, thus making the whole pavement work in separate layers and leading to many pavement problems such as fa tigue cracking, top down cracking, delamination, and slippage failure (Uzan et al. 1978; Tayebali et al. 2004; Tashman et al. 2006; Ta shman et al. 2008). Howeve r, excessive bonding agent may reduce adhesion and aggregate interlock (Uzan et al. 1978), increase slippage between pavement layers (Tashman et al. 2008) reduce air void content of the overlay mixture because of the migr ation of bonding agent into the HMA mat during compaction (Mohammad et al. 2009). Therefore, it is of great significanc e to estimate the optimum bonding agent application rate for different interface c onditions. Optimum residue application rate has 24

PAGE 25

been determined by Mohammad et al. (2002 and 2005) for modified asphalt emulsion CRS-2P and CRS-2L to be 0.02 gal/sy when it is applied between HMA layers and by Leng et al. (2008) for SS-1hP to be 0.04 gal /sy when it is applied between HMA and PCC. The improvements of bond between pavem ent layers on the interface shear resistance are varied among the studies comple ted to date. Study by Tayebali et al. (2004) indicated that bonded su rfaces have extremely high strengths compared to nonbonded surfaces. Specimens with no tack appli ed failed at the interface during the coring process and all three tack coat mate rials, SS-1h, CRS-1 and trackless tack, exhibited the higher shear strength at 0.16 gal/sy than at 0.03 and 0.06 gal/sy (Mohammad et al. 2009). Specimens with cationi c emulsion interface at the rate of 150g/m2 between HMA layers displayed much higher shear strength than those with no emulsion (Collop et al. 2010). West et al. (2005) report ed that for both CSS-1 and CRS-2 emulsions with three residue application rates, 0.02, 0.05 and 0. 08 gal/sy, lower application rate generally displayed higher bond strength when they ar e applied between HMA layers consisting of 4.75 mm NMAS fine-graded mixt ure; however, application rate has little effect on bond strength when they are app lied between HMA layers c onsisting of 19.0 mm NMAS coarse-graded mixture. Some studies have concluded that the use of tack coat material does not necessarily result in an increase in interlay er shear strength. Sholar et al. (2002) reported that the shear strengt h increase slightly as the tack coat application rate was increased; but the strengths essentially equalized regardless of the application rates 25

PAGE 26

after weeks of traffic fo r pavements consisting of a 12.5 mm fine graded Superpave mixture over a 12.5 mm fine graded Superpave mixture. Non-destructive defection tests on pavements consisting of tack coat at three different applic ation rates, 0.056, 0.046 and 0.04 gal/sy, indicated that the effectivene ss of tack coat on pavement strength is not significant at 95% conf idence level (Mrawira and Yin 2006). It should be noted that even though the use of tack coat at pavement layer inte rface does not increase the shear strength, the benefit of tack coat ma terial against watering of the surface and thermal aging must not be ignored (Raab and Partl 2004). 2.1.2.3 Curing time There was no complete agreement in the lit erature concerning the curing time of tack coat. Some studies and guidelines sug gest that tack coat should be cured before laying the new asphalt overlay (Hot Mix Asphalt Paving H andbook 2000; Texas Technical Advisory 2001; Flexible Pavement of Ohio 2001). It has been reported that interface shear strength increased with long er curing time (Hachiya and Sato 1997; Canestrari et al. 2005; Chen et al. 2008). Howeve r, some studies concluded that curing time has little effect on the interface shear strength (Tashman et al. 2006; Tashman et al. 2008). Study by Buchanan and Woods (2004) indicated that thr ee emulsions (SS-1, CSS-1 and CRS-2) with three application rates (0.05, 0.09, and 0.13 gal/sy) exhibited highest tensile and torque-shear strength at low application rates when emulsions are not fully broken, and highest tensile and torque-s hear strength at the ap plication rate of 0.09 gal/sy after emulsion are fully cured. 2.1.2.4 Temperature It is well known that asphalt mixture is a tem perature susceptible material; thus it is expected that temperature has a signific ant effect on the s hear strength between 26

PAGE 27

pavement layers. Studies with various testing temperatures have all concluded that the interface shear strength increased as the testing temperature decreased (Deysarkar and Tandon 2005; West et al. 2005; Canestr ari and Santagata 2005; Yang et al. 2007; Leng et al. 2008; Collop et al. 2010). However, this relationship may not hold true as the temperature approaches the glass temperatur e of the HMA and / or tack coat (Leng et al. 2008). Average bond strength measur ed by NCAT Bond Strength Device is 2.3 times greater at 50F compared to 77F and it is 6 times greater at 77F compared to 140F (West et al. 2005). Shear reaction modulus defined as nominal shear modulus divided by composite specimen thickness increases as the temperature decreases in the temperature range of 10 C to 30C (Collop et al. 2010). According to Canestrari et al. (2005), the shear strength between pavement layers consists of residual friction, dilatancy, the inner cohesion of layer materials, and adhesion given by interface material. At a higher temperature, different interface treatments will provide similar interface shear strength because the interface material adhesion becomes relatively insignificant and the interface shear resistance is mainly related to the layers characteristics, like surface roughness (Canestrari et al. 2005; West et al. 2005). Thus the effe ct of interface friction is ex pected to be more evident at higher temperature and the effect of normal pressure will be more notable at higher temperature. In the filed, interface shear resistance appear s to be lowest during hot days. 27

PAGE 28

It is interesting to note that increasing the temperature of t he original pavem ent before placing the overlay can increase t he shear strength at the interface (Abd and Easa 2002). 2.1.2.5 Surface conditions Existing pavement surface conditions, like surface roughness, cleanliness, and wetness have a significant effect on the interface shear resistance between pavement layers. It is generally recommended that tack coat materials should be applied on a dry and clean pavement surface (Hot Mix As phalt Paving Handbook 2000; Flexible Pavement of Ohio 2001; Texas Technical Advisory 2001; Cross and Shrestha 2005). This requirement was confirmed by studies of Hachiya and Sato (1997), Collop et al. (2003) and McGhee and Clark (2009) in which interface shear resistance reduction is found for dirty pavement surfaces even when extra tack coat materials are applied. On the other hand, Kruntcheva et al. (2006) found that a dr y and clean interface with no tack coat has similar properties to the sa me interface with a standard quantity of tack coat. However, it is interesting to note that in the study by Mohammad et al. (2009), dusty conditions exhibited higher interface strength than clean conditions especially when tested with a confining pressure. This lik ely resulted from the higher viscosity of the dust and asphalt mastic than the neat residual asphalt. It has been reported that the micro and ma cro texture of the lower pavement surfaces appears to play a significant role in interface bonding when it comes to contact surface roughness (Mrawira and Damude 1999) Roughness characteristics measured by Computer Tomography indicated that hi gher macro-texture (higher roughness) leads to an increase in interface shear resistance (Santagata et al. 2008). 28

PAGE 29

It has been reported that pavement milling pr ovided a b etter bond at the interface between the existing pavement and the new HMA overlay due to the rough longitudinal grooving on the lower pavement surface created during the milling process (West et al. 2005; Tashman et al. 2008). For milled surfaces, the bond st rength at the interface between existing pavement and the overlay is not affected by the application of tack coat materials; in other words, tack coat was not effective at increasing the interface shear resistance for composite pavements with milled lower layer (Sholar et al. 2002; Tashman et al. 2006; Tashman et al. 2008; McGhee and Clark 2009). This indicated that the application of tack coat mate rials on milled pavement surfaces was not necessary. Study by Cooley (1999) point ed out that grooved pavement (from milling operation) in conjunction with the melting of the asphalt within the loose milling materials (left by the lightly sweeping of the milled surface) by heat of the overlay mixture would create a bond between milled pavement surface and overlay. For pavements with HMA overlay placed on PCC slab, the texture of the PCC surface affects not only the interface shear resistance but also the overlay rutting. Among the four types of concrete surface textures evaluated by Leng et al. (2008), including smooth, transverse tinting, longitu dinal tinting and milling, milled PCC surface exhibited the highest interface shear strength. At low tack coat application rate, the tined PCC surface provided higher interface shear strength than smooth surf ace; however, at optimum tack coat application rate, smoot h surface has better bonding than the tined surface at intermediate temperature without no rmal forces applied. The reason is that at low application rate, the interlock between ti ned PCC and HMA over lay is dominant in spite of less contact area between the two surf aces. In the following field work, Leng et 29

PAGE 30

al. (2009) reported that milled surfac es showed lower rutting than smooth and transverse tinting surfaces. Interfaces between AC and AC provide better shear strength than interfaces between AC and P CC due to the absorption of emulsion into the underly ing AC layer (Tayebali et al. 2004). As far as the effect of we tness on the interface shear resistance, Sholar et al. (2002) showed that water applied to the tack coated surface reduced the shear strength of the specimens compared to those wit hout water applied; Raab and Partl (2004) indicated that for specimens without tack coat, the watering of the surface has a negative influence on the adhesion. However, Mohammad et al. (2009) reported that water sprayed on tack coated surfaces pr ior to HMA overlay placement has no significant effect on the bond strength becau se a small amount of water can be flashed away by the hot HMA mat and it has no c onsequential effects on the tack coat quality. 2.1.2.6 Pavement mate rials and structure Since the pavement layer interface is the surface where two pavement layers come into contact, it is well recognized that the materials in contac t play a key role on the interface properties. Kruntcheva et al. (2 006) concluded that the interface properties depend on the type of materials in contact rather than on the amount of tack coat applied and/or the interface conditions. Abd and Easa (2002) also concluded that the interface shear resistance is strongly affect ed by the types of materials in contact. Through the comparison between specim ens from laboratory and in situ, Canestrari et al. (2005) report ed that air voids content is closely related to the inner cohesion of the mixture. Gener ally, it is believed that interface shear resistance will increase with increasing mixtur e density. This was confirmed by the conclusion drew by Kruntcheva et al. (2006); mate rials requiring more compac tion time will create better 30

PAGE 31

bond at the interface. Ponniah et al. (2006) also stated that for composite spec imens with HMA mixture for both top and lower layers, the application of tack coat has no significant effect on the interface shear stre ngth when the top layer is compacted to 125 gyrations, whereas the tack coat will increase the interface shear strength by 10% when the top layer is only compacted to 75 gyrations. Fine-graded, smaller NMAS mixture has been reported to have better interface shear strength than coarse-graded, larger NMAS mixture (West et al. 2005; Leng et al. 2008). The interface cohesion of tack coat between different specimens decreases in the order of between densegraded mixture and dense-graded mixture, between porous mixture and dense-graded mixt ure, between porous mixtur e and SMA mixture (Chen et al. 2008). Sholar et al. (2002) also concluded t hat tack coat at the application rate of 0.02 gal/sy has no effect on the interface shear strength between 12.5mm coarse graded Superpave mixture and 19.0mm co arse graded Superpave mixture. Raab and Partl (2004) report ed the negative influence of thermal ageing of the composite specimen on the interface shear str ength. As far as construction practices, West et al. (2005) pointed out those cores taken from within the wheelpath and those from between wheelpaths showed no significant difference in bond strength; and those cores from sections paved with Novachip spreader showed significantly higher bond strength than those paved with normal paving equi pment due the avoidance of tack coat material tracking during Novachip paving process. Mohammad et al. (2002) reported that monolithic mixture has higher shear strength than specimens jointed at the interface. Analytical analysis indicated that thicker overlay can reduce the interface shear stress (Hachiya and Sato 1997; Tayebali et al. 2004). 31

PAGE 32

2.1.2.7 Testing methods and loading conditions Interface shear strength has been reported to increase with the increasing nor mal pressure applied on the specimen (Uzan et al 1978; Mohammad et al. 2005; West et al. 2005). The effect of normal pressure is expected to be more pronounced at higher temperature because the effect of friction is more evident at higher temperature. However, Romanoschi and Metcalf (2001) conclu ded that shear strength values are not affected by the normal stress levels applied on the interface with tack coat, but they are affected for the interface without tack coa t. Interlayer reaction tangential modulus defined by Canestrari et al. (2005) as initial slope of the stress-displacement curve increases proportionally with normal stress. Interface shear strength was also report ed to increase with higher shear loading rate (Leng et al. 2008; Mohammad et al. 2009), the application of horizontal load (Tayebali et al. 2004), and the application of confinement on specimens during testing (Mohammad et al. 2009). Raab and Partl (2004) indicated that inte rface shear strength measured by layer parallel static direct shear test can not de scribe the effect of tack coats under dynamic loading. Repeated loading tests performed by Collop et al. (2010) showed that a higher fatigue life and greater sensitivity to shear st ress level at the lower temperature; for instance, reduction of the stress level by a fa ctor of approximately 1.9 increases the life by a factor of approximately 15. Collop et al. (2010) also ev aluated two different loading rates (180/min and 600 N m/ min); the nominal shear strength measured at 180/min is approximately 1.9 times higher than t hat measured at 600 N m/ min. Specimens prepared in the l aboratory with no tack coat exhibited substantial shear strength (Uzan et al. 1978; Mohammad et al. 2005; Kruntcheva et al. 2006); however 32

PAGE 33

field cores without tack coat applied at the interface have been reported to de-bond at the interface (Tayebali et al. 2004; West et al 2005). Muench and Moomaw (2009) pointed out that the non-uniform a pplication rate, torsional/normal forces created by the coring equipment, and compaction by construction equipment might have contributed the de-bonding. Three different tests, UTEP pull-off tes t, FDOT shear tester, and torque bond test, evaluated by Tashman et al. (2008) generally showed different results. 2.1.3 Effects of Bonding Conditions on Pavement Performance A good bond between pavement layers can ensure the pavement system act as a uniform composite layer and more effectivel y transfer the external load into the subgrade and distribute the loadi ng over a larger area, thus reduce the potential of pavement distress. On t he other hand, poor bonding or debonding can cause slippage and reduce the shear strengt h between pavement layers, thus reducing the load transferring capability and leading to pavement distress, like cracking, rutting, shoving, and pothole. Debonding has been reported to be caused by either poor tack coat between layers or water infiltration due to distress or inadequate compaction (Muench and Moomaw 2009) Early fatigue cracking has been reported to be related to debonding through layer elastic analysis (Shahin et al. 1986; Willis an d Timm 2006) and field test section results (Harvey et al. 1997; Willis and Timm 2006). Analytical analysis by Shahin et al. (1986), Willis and Timm (2006), Ziari and Khabiri (2007) and Hu et al. (2010) indicated that compressive strains on subgrade surface increased substantially with increasing slippage, leading to higher subgrade rutting. Th rough field test section investigation, 33

PAGE 34

Leng et al. (2009) reported that test sections with higher interface shear strength exhibited lower MHA overlay surface rutting. Study by Shahin et al. (1986) indicated that maximum tensile strain is located at the bottom surface of the original asphalt lay er if pavement layers are fully bonded at the interface. However, if debonding occurs at the interface, the overlay and the underlying structural layer will respond individua lly, leading to greater interface stress (Buchanan and Woods 2004; Ziari and Khabiri 2007), tensile stress at the bottom of the overlay (Shahin et al. 1986; Hu et al. 2010) and critical stress shifted to the bottom of the overlay and is far more critical compar ed to complete bond (Ameri et al. 1990). The stress distribution in the pavement system is shown in Figure 2-1 for fully bonded and fully slipped pavement layers (Gomba et al 2005). Only a small amount of debonding is able to produce strains in the pavement approaching to thos e with full debonding (Shahin et al. 1986). Kruntcheva et al. ( 2005) reported a 20 to 35% reduction in pavement life due to debonding. Figure 2-1. Stress distribution for fully bonded (left) and fully slipped (right) pavement system (after Gomba et al. 2005) Meanwhile, in absence of interface bond, tensile stress at the bottom of the overlay induces a compressive stress at the top surface of the under lying asphalt layer. This allows the occurrence of relative movement between the overlay and the underlying layer at the interface, leading to weaker bond and more slippage (Shahin et 34

PAGE 35

al. 1986) and larger octahedral shear stress due to less confining s tress (Ameri et al. 1990). Analytical analysis indicated that the presence of horizontal force, in the form of acceleration and braking, caused an increase in the interface shear stress (Shahin et al. 1986; Ameri et al. 1990; Hachiya and Sato 1997) It is interesting to note that when debonding occurs, normal acceleration (less than 1.5 m/s2) has no significant effect on the AC layer mechanistic response, whereas deceleration can cause dramatic changes in the maximum shear strains at the surf ace of the pavement st ructure even at the deceleration rate of 1.5 m/s2 (Hu et al. 2010). Chen (2010) reported a premat ure pavement overlay failure only 1 day after it was opened to traffic due to the loss of interf ace bonding. Dynamic cone penetrometer results confirmed that the slip page cracks were not linked to weak base or subgrade. Heavy Vehicle Simulator (HVS) testi ng performed on full scale accelerated pavements with bonded and unbonded interfaces s howed a 10 to 45 fold increase in the estimated loading (ESALs) for the bonded pavement sections over the unbonded sections (Santucci 2009). 2.1.4 Summary The effect of interface materials on pav ement layer interface shear strength has been the focus of most of the research work to date. However, some analytical and field section research has indicated that the interf ace conditions play a key role on the stress and strain distribution in the pavement syst em, especially in surface layers for pavements with overlays (and thus the pavement service life). Research directly related to the effect of interface conditions on pav ement cracking performance is necessary to evaluate the interface conditions on pavement performance. 35

PAGE 36

2.2 Top-dow n and Reflective Cracking Mechanism 2.2.1 Top-down Cracking Mechanism Top-down cracking is a pavement deteriora tion mechanism where cracks initiate at the pavement surface and propagate downward with time. It has been widely reported in the United States of America (Roque et al. 1990; Uhlmeyer et al. 2000; Svasdisant et al. 2002), as well as in Europe (Gerritsen et al. 1987; Dauzats et al. 1987; Nunn et al. 1998; De Freitas et al. 2003) Japan (Matsuno et al. 1992), and other countries (Wambura et al. 1999; Emery 2006; Raju et al. 2008). However, this failure mode can not be explained by the traditional fatigue mechanisms in which the maximum tensile stresses and strains always occur at the bottom of the asphalt structure layer when the pavement system is subjected to external loading on the pavement surface. Of the many mechanisms identified to date, it is well accepted that the critical near surface stresses and strains induced by nonuniform contact stresses between the tire and the pavement surface may cause top-down cracking. Research work by Myers (2000) indicated that surface cracking appears to be initiated by critical tensile stresses induced by non-uniform contact stress between the ribs of radial truck tires and the pavement surface. However, Jacobs (1995) points out that the maximum tensile stresses occur at the edge of a bias ply truck tire on the pavement surface. The difference as explained by Myers (2000) wa s caused by the different stress states induced by different tire structures; and t he radial and wide base tires are potentially more detrimental to pavement surface distress than bias ply tires. Analytical studied by Molenaar (1984) and Gerritsen et al. (1987) conc luded that high surface tensile strains induced at the edge of the tire are the cause of top-down cracking. Nunn et al. (1998) indicated that surface initiated cracking wa s due to horizontal tensile stresses generated 36

PAGE 37

by truck tires at the pavement surface. Groenendijk (1998) also indicated that the combined influence of the non-uniform tensile contact stress and the AC mixture aging could result in critical tensile stress at the pavement surfac e rather than at the bottom. Since asphalt mixture is a temperature susceptible material; temperature has significant effects on the engineering charac teristics of the asphalt pavement. Dauzats et al. (1987) reported that thermal stresses could initia te top-down cracking after a number of repeated temperat ure cycles and it will furt her propagate under traffic loading. Roque et al. (1990) studied the top-down cra cking potential in asphalt pavements due to thermal and load-induced stre sses and concluded that it is the combination of thermal and load-induced stre sses that may induce top-down cracking. Schorsch et al. (2003) reported that nighttime temperatures produce the highest magnitude of surface tensile stress. However, OGFC mixtures may be the first front in re sisting top-down cracking. For pavements surfaced with OGFC, it is recognized that more sever aging can occur in the OGFC due to its higher air voids in combi nation with their direct exposure to UV radiation and heat. Aging will increase t he binder stiffness and brittleness with a corresponding effect on the mixture. Asphal t binder aging has been attributed to be the major cause of top-down cracking in many studies such as Hugo et al. (1985) in South Africa, Gerritsen et al. (1987) in Netherlands, Matsuno et al. (1992) in Japan, Wambura et al. (1999) in Kenya, and Svasdis ant et al. (2002) in Michigan. Stiffness gradients induced by temperat ure gradients along the depth of the pavement, asphalt mixture age-hardening and pavement cooling rate have also been linked to top-down cracking (Roque et al 1988; Rowe et al. 2001). In addition, 37

PAGE 38

construction quality has been identified as a si gnific ant contributor to the top-down cracking (Gerritsen et al. 1987; Schorsch et al. 2003; De Freitas et al. 2005). 2.2.2 Reflective Cracking Mechanism Reflective cracking is the phenomenon wh en a crack reflects up into and through the new pavement overlay just above the di scontinuities in the old pavement. It has been observed in nearly all types of overlays, but it is most common in AC overlays placed on rigid pavements (Mukhtar and Dempsey 1996). The most recognized driving force of reflective cracking is the horizontal movement concentrated at the cracks and joints in the existing pavement, as represented in Figure 2-2. Because of the bond between overlay and existing pavement, tensile stress is induced in the HMA overlay directly above the crack and joint. This horizontal movement is introduc ed by the daily temper ature change. Cracking occurs when the induced tensile stress ex ceeds the breaking strength of overlay mixture. Reflective cracking caused by this mechanism initiates at the bottom of the overlay just above the cracks and joints. A large daily temperat ure drop with a low temperature at the end of the cooling cycle creates the most critical reflective cracking condition (Bozkurt and Buttlar 2002). The HMA overlay contraction due to low temperature will increase the resistance to the joint opening and create additional tensile stress in the overlay. The temperature drop in the evening creat es a temperature gradient with lower temperature in the upper portion of PCC sl ab and higher temperature in the lower portion of PCC slab. This temperature gradient causes the upper portion to contract more than that of the lower portion, caus ing the upward curling of PCC slabs. This 38

PAGE 39

upward curl initiates cracking from the HMA overlay surface where more severe overlay mixture age hardening occurs (Nesnas and Nunn 2004; Von Quintus et al. 2010). Moving traffic loads can cause differentia l vertical movement of the PCC slab across the cracks and joints in the PCC slab. This differential vertical movement can be caused by the development of voids beneath t he PCC slab at the cracks and joints, or poor PCC slab support, or poor load transfer (Bennert 2010; Von Quintus et al. 2010). These vertical movements create bending and/ or shear stresses in the HMA overlay near the cracks and joints, and eventually reflective cracking. Figure 2-2. Reflective cracking in HMA over lay of PCC base (after Von Quintus et al. 2010) 2.2.3 Summary This review of the literatur e indicates that the interface conditions will affect the pavement material properties near the in terface for both top-down and reflective cracking, the stress distribution in the over lay (like OGFC) for t op-down cracking, and the stress distribution near the crack tip for reflective cracking. A good bond at the interface can help dissipate the stresses bu ilt up near the interf ace and increase the 39

PAGE 40

fracture resistance of the material near the interface. Since OGFC mixtures may be the first front in resisting top-dow n cracking, it is also of si gnificant importance to evaluate the effect of OGFC on pavement top-down cracking performance. It can be conclude d that composite specimen testing is needed for the evaluation of cracking initiation and propagation through the interface. 2.3 Testing Methods for Pavemen t Layer Interface Evaluation 2.3.1 Interface Shear Resistance Testing 2.3.1.1 ASTRA test set-up ASTRA Test device, a direct shear box, si milar to the device usually used in soil mechanics, was developed under Anocona Shear Testing Research and Analysis (ASTRA) program (Canestrari et al. 2005). The basic test conf iguration is schematically shown in Figure 2-3. It can accommodate both prismatic (maximum square crosssection area 100 mm2) and cylindrical specimens (d iameters from 94 to 100 mm). The mixture influences on shear resistance ar e eliminated by holding two half-boxes on the mixture with an unconfined interlayer shear zone in the center. A normal load is applied with a lever and weight system; a horiz ontal load is applied with a driving motor and measured by a load cell. Shear force, hor izontal and vertical displacement are recorded during the test. To evaluate the interlayer shear resistance, the test is performed at the horizontal displacement loading rate of 2.5mm/mi n under two temperatures, 20C and 40C. Composite specimens consisting of two types of dense-graded hot mixes (AC16 and AC 11), compliant with Italian technical standards, were prepared from both field trial section and laboratory. Three types of inte rface treatment methods were evaluated, without tack coat, with conventional cationi c emulsion, and wit h polymer modified 40

PAGE 41

cationic emulsion. For both the latter two treatment methods, 300g/m2 residue application rate was used. Short curing time (tested after 2 to 3 weeks) and medium curing time (tested after 7 to 8 weeks) were evaluated under 20C. Figure 2-3. Configuration of the ASTRA test device (after Canestrari et al. 2005) Application of different normal load on the same testing configuration allows the construction of peak envelope that represents the interlayer failure criterion like Coulomb failure law. Test results indicat ed that tack coat does have an effect on the interlayer shear resistance under low temperatur e (20C), but it is more related to the layer characteristics under higher temperature (40C). 2.3.1.2 Double shear test Double shear test (DST) was originally developed by Diakhate et al. (2006) to study the shear fatigue behavior of tack coats. Since this device can not apply either oligocyclic or monotonic test s beyond fatigue, a more versatile device was considered. Diakhate et al. (2011) optimized both the specimen and device geometries to obtain a relatively pure shear loading at the interface. 41

PAGE 42

DST in the laboratory involves a specim en consisting of three layers bonded twoby-two with the same tack coat, as shown in Figure 2-4. The two side layers (AC #1 and AC #3) are fixed during the te st, and the center layer (AC #2) is subjected to either monotonic or repeated loading. Diakhate et al (2011) evaluated two types of interface conditions, with and without tack coat, at a frequency of 10 Hz under load control. Specimens have to be conditioned in the c limatic chamber for at least 6 h before testing. Two temperatures (10 C and 20 C) we re evaluated. The cyclic test was set to automatically stop whenever the measured load exceeds the setting value by 20%. Figure 2-4. Schematic of the double shear test (after Diakhate et al. 2011) The results showed that at 10 C, the absence of tack coat results in a decrease in bonding fatigue performance. It should also be noted that the relationship between applied stress and the number of loading cycles to failure can be expressed by a power law. 42

PAGE 43

2.3.1.3 FDOT shear tester FDOT shear tester was developed to quantify the bond strength of the tack coat that had been wetted by rain (Sholar et al. 2 002). This simple direct shear device can operate in the Materials Testing System (MTS) as shown in Figure 2-5. The six-inc h nominal di ameter specimen can be either roadway cores or laboratory produced specimens. A 3/16 inch es gap between the shearing platens is chosen to reduce the effects of skewness, and bending due to the cantilever effect of the unsupported edges during shearing. The strain controlled load is applied at a displacement rate of 2 in/min, which c an be easily achieved on Marshall apparatus. Specimens have to be conditioned at 77F for at least 2 hours prior to testing. The field cores are oriented in the shearing plates so that the loading direction is parallel to the traffic direction marked on the specimen. Figure 2-5. Simple direct shear device dev eloped by FDOT (after Sholar et al. 2002) Test results from field projects indicated that rainwater has a negative effect on the shear strength of the specimens. Tack coat application rate within the range of 0.02 to 43

PAGE 44

0.08 gal/sy had a slight effect on the shear strength. The gradat ions of the asphalt mixtures bonded by the tack coat played a cr itical role the shear resistance. 2.3.1.4 Layer-Parallel Dir ect Shear (LPDS) test LPDS test device (See Figure 2-6) is an EMPA modified version the Leutner test developed in Germany in the late 1970s (R aab and Partl 2004; Canestrar et al. 2005; Raab and Partl 2008; Santagata et al. 2008). This test is performed without the application of normal load. One part of the specimen is laid on a circular U-bearing and held by a semicircular pneumatic clamp; the other part remains suspended. An interface shear zone, a 5 mm gap, is introduced betw een the yoke and the pneumatic clamp. The test uses 150 mm diameter or prismati c (150 mm mm) specimens comprising at least two layers; it is usually performed at the displacement rate of 50 mm/min through semicircular yoke under 20C temper ature. The test returns a shear forceshear displacement curve that can be used to determine the maximum shear stress and shear reaction modulus. Figure 2-6. Schematic view of the LPDS test device with pneumatic clamping (after Raab and Partl 2008) 44

PAGE 45

2.3.1.5 Leutner test Leutner test was developed in the late 1970s for evaluation of pavement layer interface using simplified direct shear test (Collop et al. 2009; Sutanto 2010). The test is performed on 150 mm diameter specimens enc ompassing at least two layers from either field cores or laboratory produced specimens as shown in F igure 2-7. To make sure good contact between the specimen and the load device, the layer thickness has to be at least 70 mm and 25 mm for lowe r layer and upper layer, respectively. A constant shear displacement rate acro ss the pavement layer interface was applied without any normal force applied. The test is normally conducted at the displacement rate of 50 mm/min at 20C. The displace ment rate 50 mm/min a llow the test to be performed in Marshall or CBR load devices. The disadvantage of this test is that nonuniform shear stresses are induced on the in terface. A 5 mm gap was introduced into the shear plane on the standard Leutner test to reduce the edge damage caused by misalignment and irregular interface. Figure 2-7. Photograph and schem atic diagram of Leutner load frame (after Collop et al. 2009) 45

PAGE 46

Collop et al. (2009) pointed that the average shear strength from field cores tends to be higher for the surfacing/binder cour se interface than the bind er course/base interface. The average pavem ent layer interface shear strength from field cores increases as the class of the road increases. 2.3.1.6 Shear fatigue test Shear fatigue test was developed to eval uate the pavement layer interface shear fatigue behavior under the repetitive mec hanical action of t he traffic loading (Romanoschi and Metcalf 2001). Normal and s hear stresses are applied simultaneously on the specimen; the longitudinal axis of the specimen was tilt ed 25.5 to the vertical so that the shear stress at the in terface is half the normal stress, as shown in Figure 2-8. A vertical load with a minimum value of 10% of the maximum at the frequency of 5Hz was applied on the specimen. 0.2 s l oading period with 0.05 s pulse was used to simulate the vehicle speed at 50 km/h. Four vert ical load levels at 4, 6, 8, and 10 kN were used, introducing normal stress of 0.5, 0.75, 1.0, and 1.25 MPa at the interface. The elastic and permanent deformations at the interface in normal and tangential directions are recorded for each loading cycle. The number of loading cycles leading to an increase of permanent shear displacement of 1mm was used to evaluate the fatigue properties of the layer interfac e. The results from this study indicated that interface with a tack coat exhibited a longer lif e than that without tack coat. 2.3.1.7 Superpave Shear Tester The Superpave Shear Tester (SST) was developed as part of SHRP research to measure mixture properties that can be used to characterize a HMA mixtures resistance to permanent deformation (Witcza k et al. 2004). Vertical, horizontal, and 46

PAGE 47

confining loads can be simu ltaneously applied on the spec imens. The environmental chamber can maintain temperature in t he range of 0C (14F) to 80C (176F). Figure 2-8. Schematic of shear fatigue test (after Romanoschi and Metcalf 2001) A shearing mold, consisting of two parts, was specially designed for the shearing test using the Superpave Shear Tester as shown in Figure 2-9 (Mohammad et al. 2002; Mohammad et al. 2005). Each part has a 150 mm (5.9 inch.) diam eter and 50.8 mm (2 in.) deep cylindrical groove in it for holding the specimen during te sting. The applied horizontal and vertical loads, specimen deformati on (in the axial, horiz ontal and vertical direction) can be recorded during testing. This shearing test was performed in a load controlled manner at a constant rate of 222.5 N/min (50 lb/min). The test arrangem ent (SST environmental chamber with test mold in it) was shown in Figure 2-10. 47

PAGE 48

Among the evaluated tack coat material s (PG 64-22, P G 76-22M, CRS-2P, CRS2L, CSS-1, SS-1H, SS-1L, SS-1), CRS-2P and CRS-2L prov ided significantly higher interface shear strengths (Mohammad et al., 2005). For each of the tack coat material, optimum residue application ra te was determined. The interface shear resistance increases with increasi ng vertical stress. Figure 2-9. Design shear mold (left) and mold with a sample inside (right) (after Mohammad et al. 2005) Figure 2-10. Test arrangement for the shearing test (after Mohammad et al. 2005) 2.3.1.8 Torque bond test Torque bond test was originally develo ped in Sweden for in-situ pavement layer interface conditions evaluation (Walsh and Williams 2001). It has been adopted in UK to 48

PAGE 49

measure interface properties between thin s urf ace layer and underlying structural layer. A torque is applied manually at th e top of the core, introduci ng a twisting shear failure at the layer interface. A 100 mm diameter coring specimen is used to limit the required magnitude of moment within the range of m anual application. The manual torque bond testing procedure in UK (UK Guidelines 2000) requires the test to be performed at a constant torque rate so that the failure can occur in (60) seconds. To avoid the difficulty in controlling the torque rate, a constant torque rate of 600 N m /min was used by synchroni zing the torque dial gauge in the torque wrench with the second hand of an analogue clock (Choi et al. 2005). Since the moment is applied by twisting the top of the core with to rque wrench, the stiffness and thickness of the adjacent pavement layers pl ays an important role on the interface shear resistance. An automatic laboratory-based to rque bond test was developed by Collop et al. (2010) to investigate the two different loading rates (600 N m/min and 180/min) that were used in the manual torque bond test. This testing setup is shown in Figure 2-11. The testing equipment consists of an environmental c hamber with a range between -5C and 40C, a 100kN servo hydraulic actuator, an axially mounted load cell and a Linear Variable Different ial Transformer (LVDT). The hy draulic actuator testing machine can apply either compressive or tens ile load under static or repeated loading conditions. Specimens with 100mm in diameter and 10mm in thickness cylindrical metal platens glued to the top and bottom have to be conditioned in the environmental chamber for at least 5 hours prior to testing. Collop et al. (2010) prepared double-layered cylindrical specimens for interface condition evaluation. The top layer was an 11mm Asphalt Concrete mixture, 30mm thick 49

PAGE 50

and the bottom layer w as a 16mm Asphalt Concrete mixture, 50mm thick. Three interface conditions, no bitumen emulsion, a modified bitumen emulsion, and a cationic bitumen emulsion. Both modified bitumen emulsion and cationic bitumen emulsion has a 150g/m2 residue bitumen application rate. The s hear strength is ca lculated using the following equation, 32/NTR Where N is the nominal shear strength, T is the peak torque and R is the specimen radius. Figure 2-11. Photograph and schem atic diagram of the aut omatic torque equipment (after Collop et al. 2010) Test results reported by Collop et al. (2010) indicated that shear strength increased with deceasing tem perature. And specimen s with no emulsion at the interface showed the lowest shear strengths The material compliance of both upper and lower mixture has a relatively small effect on the rotation and shear reaction modulus of the interface unless the shear stiffness of the mixture is very low and/ or the thickness is very larger. 50

PAGE 51

2.3.1.9 UTEP Pull-Off Device (UPOD) In order to remove the effect of pavem ent layer surface, like surface roughnes s, UPOD was performed in tension (pull-off) rather than shear mode (Deysarkar 2004). It measures the tensile strengt h of the tack coat before the overlay is placed. The instrument weighs about 23 lbs and can be easily leveled with the pivoting feet as shown in Figure 2-12. It has a weight key (4 0 lbs) on the top, providing the stability during placement of loads. A drive torque wr ench is used to the pull the plate up or down from the tack-coated su rface. 3M double-sided tape is attached between the aluminum contact plate and moisture bearing foam to make sure that the device conforms to the rough surface. UPOD is placed on the tack coated pavement after a specified period of set time. The torque wrench is rotated clockwise until the contact plate is firmly set on the tackcoated pavement. After the 40 lbs load has been applied on the weight key for 10 minutes, the load is then removed and the to rque wrench is rotated in the counterclockwise direction to detach the contact pl ate from the tack-coat ed pavement surface. The torque required to detach the contact plate from the pavement is recorded and converted to the strength using a calibration factor. For the equipment used by Deysarkar (2004) the following calibration factor was obtained: F=0.6571T Where T is the torque in in-l b and F is the load in lbs. In the laboratory work conducted by De ysarkar (2004), the following parameters were evaluated: six types of tack coa t: CSS-1h, CSS-1, SS1h, SS-1, PG64-22, and RC-250; three temperatures: 140, 93 and 50F; three set times: 30, 45, and 60 minutes; 51

PAGE 52

0.04 gal/sy residue application rate. Test results indicated that the em ulsion strength depends on the applic ation rate, set time and test temperature. Figure 2-12. UTEP pull-off device test set-up (after Deysarkar and Tandon 2004) 2.3.1.10 Wedge splitting test Wedge splitting test was developed in 1986 to replace the pull-off test for adhesive tensile strength measurement because the pull-off test only returns the adhesive tensile strength and the results are extensive sca ttered (Tschegg et al. 1995). From the loaddisplacement curve recorded during testing, the slope of the curve that describes the elastic properties, the not ched-bar tensile strength fo llowing the adhesive tensile strength, and the fracture energy and cra ck resistance, respectively, can be obtained (Tschegg et al. 2007). 52

PAGE 53

Specimens were prepared with a rectangular groove introduced at the interface and a starter notch placed in the interface at the bottom of the groove. The schematic view of the test is shown in Figure 2-13. The vertical load FM applied by the machine is transmitted by the slender wedge to a high horizontal force FH and a small vertical force FV during testing. The horizontal displaceme nt in the plane of the splitting force FH measures the crack mouth opening displacement. The splitting force FH can be calculated from the meas ured vertical force FV. Notched-bar tensile strength and fracture work (or specific fractu re work) can be obt ained from the FH CMOD curve. Figure 2-13. Schematic view of the wedge splitting test (after Tschegg et al. 2007) Three different pretreat ments of the milled asphal t concrete surface, no pretreatment, cement grout, and cement grout plus disper sion, were evaluated by Tschegg et al. (2007). The prismatic s pecimen dimensions are 100 mm mm mm (70 mm asphalt + 70 mm concrete). Test s were performed at horizontal splitting speed 0.5 mm (0.02 inch.)/ min under -10, 0, 10, 22C. The results indicate that no pretreatment leads to the highest cracking resistance. T he tensile strength and the specific fracture energy show a completely different behavior as a function of 53

PAGE 54

temperature; the tensile st rength deceases with increasi ng temperature whereas the specific fracture energy reaches the maximum in the medium temperature range. 2.3.2 Interlayer Crack ing Resistance Testing 2.3.2.1 TTI o verlay tester TTI overlay tester was developed in t he 1970s to simulate the opening and closing of joints or cracks and it was used to evaluate reflective crack initiation and propagation of the overlay materials (Cleveland et al. 2002; Zhou and Scullion 2004; Chowdhury et al. 2009). The test instrument, as shown in Figure 2-14, feat ures two steel blocks, one is fixed and the other slides horizontally to simulate the tensile and compressive stresses induced in the old pavement as a result of cyclic changes in temperature (Pickett and Lytton 1993). Figure 2-14. Schematic diagram of TTI over lay tester (after Zhou and Scullion 2004) The original overlay tester has two differ ent specimen dimensions: the small one is 15 inch (375 mm) long by 3 inch (75 mm) wide with variable height; the large one is 20 inch (500 mm) long by 6 inch (150 mm) wide with variable height. It has been successfully used to evaluate the effectiveness of different geosynthetic materials on mitigating reflective cracking (Pickett and Lytton 1993; Cleveland et al. 2002; Chowdhury et al. 2009). One of th e disadvantages reported in previous work is the large 54

PAGE 55

specimen size. The specimen siz e was reduced to 6 inch (150 mm) long by 3 inch (75 mm) wide with the height of 1.5 inch (38 mm) in the upgraded overlay tester in 2003 (Zhou and Scullion 2004). The specimens required for the upgraded overlay tester can be prepared either form Superpave Gyratory Compactor or from fiel d cores. Typically, the test is performed at 25C at the loading rate of one cycle per 10 sec with a maximum displacement of 0.025 inch (0.64 mm), as shown in Figure 2-15. This displacement is approximately equal to the displacement experienced by PCC pavements with a 4. 5 m joint or crack spacing undergoing 14C temperatur e changes (Zhou and Scullion 2004). The validation work by Zhou and Scullion (2004) showed that the upgraded tester is repeatable and can effectively differentiate the re flective cracking resistance of different asphalt mixtures. Figure 2-15. Typical displacement used in overlay tester (after Zhou and Scullion 2004) Four large specimens (18 inch long by 6 inch wide) were taken from each of the test sections with different geosynthetic products interlayer at all three test locations and three small specimens (6 inch long by 3 inch) were taken from each of the test sections with different geosynthetic products interlayer at one location. The bottom part of the 55

PAGE 56

specimens was trimmed to obtain a smooth, flat surface with a uniform thickness of leveling lay er. The top part was not disturbed. For a given test section, small and large specimens were trimmed to the same thickness. The small specimen sizes always exhibited failure due to separation at the geosynthetic interlayer. The large specim en occasionally exhibited similar failure mechanism. On average, the geosynthetic pr oducts have significant improvement in reflective cracking resistance of the composite specimens. The small overlay tester was not appropriate for evaluating the specimens with geosynthetic interlayer whereas the large overlay tester can be used for the geo synthetic interlayer reflective cracking resistance on composite specimens. 2.3.2.2 Interlayer Stress Absorbing Com posite (ISAC) system testing equipment ISAC is a composite material of a low stiffness geotextile, viscoelastic membrane layer, and a high stiffness geotextile (Mukhtar and Dempsey 1996). It was introduced between the old PCC slab and new AC overla y to stop the upward propagation of a crack into the overlay and reinforce the over lay. The equipment as shown in Figure 216 was introduced to measure the effectiv eness of ISAC on reflective cracking resistance. The composite specimen consisted of two 3. 75 feet long by 6 inch wide by 5 inch thick PCC slab, ISAC interlayer, and 2.5 inch thick AC overlay on top of the PCC slab. The joint opening between the slabs was inch. The specimen was conditioned to 30F prior to be cycled back and forth by the hydr aulic ram over a distance of 0.063 in at the rate of 0.0016 inch/min. The induced fo rce from hydraulic ram and the relative movement between the two box sections were recorded. Test result s indicated that the ISAC vastly outperformed the AC overlay without ISAC. 56

PAGE 57

Figure 2-16. Schematic diagram of ISAC system testing equipment (after Mukhtar and Dempsey 1996) 2.3.2.3 Beam tests Kim et al. (1999) developed a beam specim en testing technique to evaluate the performance of selected modified and/or re inforced asphalt mixtures under mode I (bending) fatigue fracture. The test arr angement is shown in Figure 2-17. The composite specimen consisted of a dense-graded asphalt mixture as an overlay on top of a concrete block with reinforcing interlayer at the bottom of t he asphalt mixture. The asphalt mixture specimen was bonded to the conc rete block with tack coat; the concrete block has a 10mm gap cut 2/3 the depth from the top and an arbitrarily broken crack in the lower 1/3 depth. The dimensions of the specimens are 340 mm (13.39 inch) long by 120 mm (4.72 inch) wide by 50 mm (1.97 inch ) thick, and 340 mm (13.39 inch) long by 120 mm (4.72 inch) wide by 120 mm (4.72 inch) thick for asphalt mixture and concrete 57

PAGE 58

block respectively. Repeated square loads ar e applied on the top cent er of the beam at 10 Hz using a hydraulic loading device. A circular loading plate with rubber pad attached to its bottom was used to simulate t he tire contact. The maximum pressure of 100 psi (5.4 kN load) was used; and a si tting load of 0.196-kN was applied. The specimen has to be conditioned to 20C in the environmental chamber prior to testing. Horizontal expansion of the asphalt mixture and visualized vertical crack length versus number of l oading cycles were recorded. Figure 2-17. Test arrangement (after Kim et al. 1999) Brown et al. (2001) developed a semi-continuous su pport system to more closely simulate the field conditions as a combination of bending in the specimen and continuity of support. The schematic diagram of t he test is shown in Figure 2-18. The reinforcement was placed 30 mm above the base layer. A 12 mm thick rubber layer was used to provide the support for the composit e specimen over the steel base. The base layer has a gap cut 1/3 depth from the bottom. The tests were carried out at 20C and 5 58

PAGE 59

Hz under 5.5 kN load. Crack propagation reduction was observed in the test for specimens with reinforcement even though there is a significant scatter in results. Figure 2-18. Schematic diagram of the testing set-up (after Brown et al. 2001) Khodaii. A et al. (2009) introduced an experim ental program to evaluate the effects of geosynthetic reinforcement on the refl ective cracking resistance improvement in asphalt overlays. The schematic diagram of the test setup is shown in Figure 2-19. The three-layered pavement struct ure consisted of asphalt overlay, existing pavement, and resilient subgrade modeled with neoprene rubber Repeated square loads are applied on the top center of the beam at 10 Hz. A maximum pressure of 100 psi (6.79 kN load on 112 mm diameter loading plate) was used; and a sitting load of 0.196-kN was applied. The specimens have to be conditioned in the environmental chamber at least 2 h before tested at either 20C or 60C. The existing pavem ent has a 10 mm, 15 mm or 20 mm gap cut 2/3 the depth from the top. Result s indicate that the geogrid inclusion in the asphalt overlay leads to a significant increase in overlay reflective cracking 59

PAGE 60

resistance. And the geogrid is most effect ive when it is placed at the one-third depth from the overlay bottom. Figure 2-19. Schematic diagram of the test setup (a fter Khodaii et al. 2009) 2.3.3 Summary The tests available for pavement layer inte rface evaluation are mainly focused on the interface shear resistance and most of them are perform ed in monotonic mode, which is not representative of the pavement interface loading conditions under traffic. On the other hand, tests used to evaluate t he cracking resistance of the interlayer materials require large specimens; they are relatively difficult to fabricate in the laboratory and more difficult to get from the field. Thus a test needs to be developed for the pavement interface cracking perform ance evaluation using 150 mm diameter specimens which can be easily produced in the laboratory or obtained from field cores. 60

PAGE 61

CHA PTER 3 COMPOSITE SPECIMEN PREPARATION AND EVALUATION 3.1 Composite Specimen Preparation An appropriate composite specimen with bon ding agent applied at the interface is needed to evaluate the effect of pavement layer interface on cracking performance. In the laboratory, it can be prepared by compacting loose overlay material, like opengraded mixture, on top of the pre-compacted lower la yer material using Superpave Gyratory Compactor (SGC). Bonding agents, like conventional tack coat or Novabond can be applied to the base material surface before placing the overlay mixture. The following procedures were followed: Compact dense-graded mixture using SGC to desired air voids as the base material. Dense-graded mixture was designed according to SuperpaveTM volumetric mix design method. Slice compacted dense-graded mixture s pecimen to the desired dense-graded mixture thickness to be used in the compos ite specimen, as shown in Figure 3-1. Apply conventional tack coat or Novabond to the cut side of the dense-graded mixture. The amount of bi nder or emulsion should corre spond to the anticipated field application rate; the calculation is illustrated in Appendix A. For a hot binder tack material, preheat the silicone rubber mold (Figure 3-2) in a 135C (275F) oven for 7 minutes. Pour the hot binder in to the appropriate size silicone rubber mold based on the specimen size to be prepared; a 150mm diameter silicone rubber mold was used in this study. Place the silicone rubber mold on a level shelf in the 135C (275F) oven for 10 minutes to allow the binder to self level. Remove the mold from the oven, place on a level surface and allow the mold to cool to room temperature. For an emulsion tack material, pour the emulsion into the appropriate size room temperature s ilicone rubber mold. Place the mold on a level shelf in a 60C (140F) oven and allow the emulsion to set to a constant weight. Remove the mold from the oven, place on a level surface and allow the mold to cool to room temperature. With the silicone rubber mold sitting on a level surface, place compacted densegraded mixture in the silicone rubber mold as shown in Figure 3-3. Allow the weight of the specimen to remain on the silicone rubber mold for at least 5 minutes. Transfer the compacted specim en with the silicone rubber mold to a freezer. Remove the compacted specimen wit h the attached mold after at least 15 minutes in the freezer. Invert the specimen to where the silicone rubber mold is on 61

PAGE 62

top the specimen. Slowly remove the mold from the dense-graded mixture surface. Allow the specimen to warm to room te mperature while sitting undisturbed. The dense-graded base materials wit h applied conventional tack coat and Novabond are shown in Figure 3-4. Figure 3-1. Half sliced specimen Figure 3-2. Silicone rubber mold on level shelf 62

PAGE 63

Figure 3-3. Dense-graded mixture placed on bonding agent residue Figure 3-4. Dense-graded spec imens with applied conventi onal tack coat (left) and Novabond (right) Heat SGC compaction mold and top mo ld plate in the oven at the required compaction temperature for at leas t 30 minutes prior to compaction. 63

PAGE 64

Push the SGC compaction mold along t he edge surface of the compacted room temperature specimen sitting on a level surface after removi ng the ring holding rotating base from falling off the compacti on mold (see Figure 3-5). Place a room temperature base plate and a paper di sk underneath the compacted dense-graded mixture in the compaction mold. Place desired weight of open-graded mixture into the co mpaction mold on top of the base material. A paper disk was placed on top of the mixture followed by the top mold plate. Compaction mold was loaded and the compaction was initiated. The amount of open-graded mixture was calculated based on the relationship among mixture maximum specific gravity, air void content, and desired height of open-graded mixture in the co mposite specimen after compaction. This calculation is illustrated in Appendix B. After compaction, remove the mold from the gyratory compactor. Extrude the compacted composite specimen from the mold after an appropriate cooling period (around 10 minutes). Compacted spec imen is shown in Figure 3-6. Figure 3-5. Compacted base material bei ng pushed into SGC compaction mold 64

PAGE 65

Figure 3-6. Newly compac ted composite specimen 3.2 Evaluation of the Effect of Overlay Compaction on th e Integrity of Lower Layer Since additional compaction was intr oduced to the compacted dense-graded mixture in the composite specimen preparation process, Superpave IDT tests were performed to evaluate the possible damage to the dense-graded HMA induced during the overlay compaction process. 3.2.1 Materials and Testing Methods A dense-graded mixture commonly used by the FDOT and identified as DenseGA-Granite was selected as the base material for this evaluation. Its aggregate was made up of four components: coarse aggregate, fine aggregate, screenings, and sand. Its gradation is shown in Table 3-1. This mixture was designed according to the Superpave volumetric mix design method. De sign asphalt contents for the mixture was 65

PAGE 66

determined such that each mi xture had 4% air voids at Ndesign= 75 gyrations. PG 67-22 asphalt was used for the mixture. A tota l of nine dense-graded Gyratory specimens were prepared. Table 3-2 shows the bulk spec ific gravity of the prepared specimen. A Superpave mixture available in the lab was used as overlay mixture (2000 grams per specimen) to evaluate the additional compac tion on the integrity of the underlying base material. The following tests and analyses were performed: Dense-graded mixtures were designed according to the SuperpaveTM volumetric mix design method. 4500 grams of aggregat e were prepared for each mixture. Asphalt content was 4.8%. A total of nine specimens were prepared. Compacted dense-graded mixture specimens were cut in half. Nine of them were used for Superpave IDT control test, while the other nine were used for recompaction evaluation. Tack coat was applied on the cut side of dense-graded specimens. The reason that the cut side was select ed was that the mi xture was more uniformly compacted than that of the uncut side. The tack coat was applied in form of pure asphalt (AC20) to save time on emulsified asphalt setting. The reason that AC-20 was used was that the viscosity of AC-20 corre sponded to the residual asphalt of the emulsion commonly used in Florida. 2000 grams of Superpave mixtur e available in the lab were placed on top of each of the nine half cut dense-graded spec imens. Three were compacted to 50 gyrations, three to 100 gyrations and thr ee to 150 gyrations. The ring holding the rotating base was taken off the gyratory compaction mold so the prepared densegraded specimen could be easily pushed back into the mold from the bottom. Three groups of Superpave IDT tests were performed on no re-compaction (control) and re-compacted specimens respectively. Each group had three specimens. A total of 18 IDT tests were performed. All tests were performed at 10C. 3.2.2 Analysis of Test Result Superpave IDT test results are summarized in table 3-3. The results show that additional compaction appeared to have slig htly improved the specimen strength. However, students t-tests show that the di fferences between control and recompaction 66

PAGE 67

for m-value, D1, St, MR, FE, DCSE, ER, were not statistica lly significant; in other words, no damage was induced by the addi tional compaction process. Table 3-1. Dense-graded mixture aggregate gradation Percent Passing Blend 33% 7% 50% 10% 100% Number 1 2 3 4 JMF 3/4" 100 100 100 100 100 1/2" 97 100 100 100 99 3/8" 59 100 100 100 86 # 4 9 30 100 100 65 # 8 4 4 70 100 47 # 16 2 2 42 100 32 # 30 2 1 25 94 23 # 50 1 1 16 53 14 # 100 1 1 10 11 7 Sieve Size # 200 1.0 1.0 7.0 3.0 4.2 Table 3-2. Bulk specific gravity Number 1 2 3 4 5 6 7 8 9 Bulk Specific Gravity 2.45 2.462.452.452. 442.442.45 2.45 2.45 Table 3-3 Superpave IDT test results Project Name 50 Gyrations 100 Gyrations 150 Gyrations Control 1Control 2 Control 3 m-value 0.554 0.564 0.53 0.474 0.618 0.547 D1 5.62E-07 4.70E-07 5.26E-07 1 .03E-06 4.17E-07 5.74E-07 St (Mpa) 2.77 2.8 3 2.6 2.76 2.67 MR (Gpa) 12.37 14.05 13.4 11.21 12.18 14.12 FE (kJ/m3) 4.1 3.2 5.1 5 3.5 3.1 DCSEHMA (kJ/m3) 3.8 2.9 4.8 4.7 3.2 2.8 Stress (psi) 150 150 150 150 150 150 a 4.46E-08 4.44E-08 4.33E-08 4 .56E-08 4.47E-08 4.52E-08 DCSE (kJ/m3) 2.173 1.918 1.835 2.458 2.227 2.11 ER 1.74 1.52 2.6 1.91 1.43 1.35 Damage Rate 1.43E-08 1.30E-08 1. 08E-08 1.29E-08 1.84E-08 1.37E-08 Damage Rate 1.28E-08 1.51E-08 67

PAGE 68

Damage rate was defined as creep rate, wh ich is one of the most critical parameters affecting mi xture performance; t he damage rates presented in table 3-3 was calculated using equation 3-1. 1000 twhere, )( )(11 mtmD cr td tdD (3-1) The damage rates calculated from equati on (3-1) show that there is some improvement in damage resistance. However, careful examinatio n of the results indicated that some measurements were clearly unreliable and appeared to be affected by problems with the gages. Only measurem ents that were clearly reliable were included in the analysis and presented in Figures 3-7 through 3-11. No Re-compaction: Replicate 10.00E+00 5.00E-06 1.00E-05 1.50E-05 2.00E-05 2.50E-05 3.00E-05 3.50E-05 4.00E-05 020040060080010001200t(sec)Creep Compliance (1/psi) Figure 3-7. Creep compliance versus time for no re-compaction replicate 1 68

PAGE 69

No Re-compaction: Replicate 20.00E+00 5.00E-06 1.00E-05 1.50E-05 2.00E-05 2.50E-05 3.00E-05 3.50E-05 020040060080010001200 t(sec)Creep Compliance (1/psi) Figure 3-8. Creep compliance versus time for no re-compaction replicate 2 50 Gyrations0.00E+00 5.00E-06 1.00E-05 1.50E-05 2.00E-05 2.50E-05 3.00E-05 020040060080010001200 t(sec)Creep Compliance (1/psi) Figure 3-9. Creep compliance versus time for 50 additional gyrations 69

PAGE 70

100 Gyrations0.00E+00 5.00E-06 1.00E-05 1.50E-05 2.00E-05 2.50E-05 3.00E-05 3.50E-05020040060080010001200t(sec)Creep Compliance (1/psi) Figure 3-10. Creep compliance versus time for 100 additional gyrations 150 Gyrations0.00E+00 5.00E-06 1.00E-05 1.50E-05 2.00E-05 2.50E-05 3.00E-05 020040060080010001200t(sec)Creep Compliance (1/psi) Figure 3-11. Creep compliance versus time for 150 additional gyrations The creep rates recalculated using only re liable measurements are presented in Table 3-4 and Figure 3-12. 70

PAGE 71

Table 3-4. Damage rate recalculation Project Name Creep Rate (1/psi-sec) Average No Re-compaction: Replicate 1 1.27E-08 No Re-compaction: Replicate 2 1.45E-08 1.36E-08 50 gyrations 1.25E-08 100 gyrations 1.45E-08 150 gyrations 1.32E-08 1.34E-08 The data in Table 3-4 and Figure 3-12 s how that additional compaction had no effect on the integrity of the dense-graded mixture. Replicate 1 50 gyrations Replicate 2100 gyrations Average 150 gyrations Average 0.0E+00 2.0E-09 4.0E-09 6.0E-09 8.0E-09 1.0E-08 1.2E-08 1.4E-08 1.6E-08 No Re-compactionRe-compactedCreep Rate(1/psi-sec) Figure 3-12. Re-calculated creep rate s for no re-compaction and re-compacted specimens 71

PAGE 72

3.3 Summar y The additional compaction of Superpave mi xture or OGFC has no effect on the integrity the bottom densegraded mixture. Therefore, the method presented here appears suitable to prepare appropriate specim ens for composite specimen tests. 72

PAGE 73

CHA PTER 4 DEVELOPMENT OF COMPOSITE SPECIMEN TENSION TEST 4.1 Background As mentioned in the literature review, pavement layer interface plays a significant role on the stress distribution in the overlay for both top-down and reflective cracking and the pavement material properties near t he interface when polym er modified asphalt emulsion is used as the bonding agent. One possible way to evaluate the interface conditions on cracking performance is to perform direct tension tests on composite specimens. The newly developed Dog-Bone Dir ect Tension test (DBDT, Koh 2009) has been successfully used to measure the tensile properties of t he dense and open graded mixture. However, careful examination indicated that although DBDT is ideal for uniform specimens composed of a single mixture, it is not suitable in its current form for tests on composite specimens. The dog-bone specimen re sults in excessively non-uniform and complex stress states onc e composite specimens ar e introduced. As dog-bone specimen is suitable for evaluation of cr acks propagating inwards to the specimen center but not suitable for evaluation of cracks propagating downwards from overlay to base material. Therefore, it was deci ded that a modified specimen geometry and loading configuration was needed to properly evaluate t he interface conditions on cracking performance. 4.2 Prototype Composite Speci men Tension Test System Based on thorough analysis of the available testing syst ems and consideration of the research goals, a composite specim en interface cracking (CSIC) test was conceived, as shown in Figure 4-1. In this composite specimen test system, the crack is expected to initiate from the open-graded mixture side (with t he help of a stress 73

PAGE 74

concentrator) and propagate through the interf ace and into the dense-graded mixture. This crack initiation and pr opagation process can also be used to simulate reflective cracking if the stress concentrator is located on the base material side. The various devices used for and procedures followed in the composite specim en fabrication are presented in the following sections. Figure 4-1. Prototype of com posite specimen tension test 74

PAGE 75

4.2.1 Slicing, Cutting and Gr ooving of Composite Specimen The following procedures were followed: Slice the top and bottom of the compac ted composite specimen to obtain the desired thickness for both open-graded and dense-graded mixtures. Diamond-tip saw used for slicing is shown in Figure 42. A sliced composite specimen is shown in Figure 4-3. Cut the sliced specimen to desired width (3.5 inch was selected). The diamond saw used for cutting is shown in Figure 4-4. The straight cut composite specimen is shown in Figure 4-5. Two specimens were held together by a clamp to core through both OGFC surfaces and obtain one semicircular groove (stress concentrator) on each specimen. The core drill bit diameter is inch. The coring setup is shown in Figure 4-6. The composite specimen with groove is shown in Figure 4-7. Figure 4-2. Diamond-tip saw used for specimen slicing 75

PAGE 76

Figure 4-3. Sliced composite specimen Figure 4-4. Diamond-tip saw used for composite specimen cutting 76

PAGE 77

Figure 4-5. Composite specimen after cutting Figure 4-6. Composite specimen st ress concentrator drilling setup 77

PAGE 78

Stress Concentrator Figure 4-7. Grooved composite specimen 4.2.2 Sanding, Gluing, and Gage Points Attachment of Composite Specimen The following procedure was followed: In order to make sure the loadi ng head and specimen curved end surface made solid contact, loose materials (i.e. mastic ) were sanded off until aggregates were fully exposed. The spindle sander used for this purpose and sanded specimen are shown in Figure 4-8. Four brass gage points (5/16inch diameter by 1/8-inch thick) were affixed with epoxy to each side of the specimen. The locations of the gage points are shown in Figure 4-9. Both the loading head inner surfaces and specimen curved end surfaces need to be uniformly coated with epoxy and the void s on the curved end surfaces need to be filled with epoxy before joining them together. The epoxy used was LOCTITE Hysol Product E-20HP epoxy. It takes about 6 hours for this epoxy to get about 90% of its full strength on aluminum at 25C. The l oading heads and epoxy used are shown in Figure 4-10. 78

PAGE 79

Figure 4-8. Spindle sander and sanded specimen Gage H1(H2) Gage V1(V2) OGFC Dense-Graded Figure 4-9. Strain gage distribut ions on the composite specimen 79

PAGE 80

Figure 4-10. Loading head and epoxy used 4.3 Monotonic Strength T ests on Asymmetrical Composite Specimens Since monotonic strength test has been successfully used to determine strength and fracture energy for specimens with a si ngle mixture, the loading rate applied in DBDT for single dense-graded mixture, 25 mm/min was selected for preliminary testing. All tests were performed at 10C. 4.3.1 Materials The dense-graded mixture used for composite specimen was the same as in section 3.2.1. Oolitic limestone FC-5 ( open-graded friction courses) was used for top layer of the composite specimen. Its aggregat e gradation is shown in Table 4-1. The asphalt binder used for oolitic limestone FC-5 was ARB-12; namely PG67-22 with 12% ground tire rubber and 0.4% mineral fiber by weight of the mixture. The compacted composite specimen consis ted of 1 inch dense-graded mixture and 1 3/8 inch open-graded mixture with approx imately 15% air voids. The open-graded 80

PAGE 81

mixture has natural compacted surfaces without slicing. The amount of each component of open-graded mixtur e needed was 78.4 gr ams of ARB-12 and 1147.0 grams of aggregate with 6.4% binder content. For initial testing to evaluate the effectiveness of this specimen geometry, neither tack coat nor Novabond was introduced to the interface between open-graded mixt ure and dense-graded mixture. Table 4-1. Oolitic limestone FC-5 mixture aggregate gradation Percent Passing Blend 44.7% 49.4% 3.2% 2.7% 100% Number 1 2 3 4 JMF 3/4" 100 100 100 100 100 1/2" 79 100 100 100 91 3/8" 36 92 100 100 67 # 4 7 26 100 100 22 # 8 3 7 68 100 10 # 16 3 3 67 100 8 # 30 3 3 55 100 7 # 50 3 2 35 100 6 # 100 2 2 14 100 5 Sieve Size # 200 1 1 3 100 4 4.3.2 Monotonic Strengt h Test at 25 mm/min While the inch diameter diamond-tip coring bit was being made, an attempt was made to cut the groove as a stress concent rator using a diamond saw as shown in Figure 4-11. Five cuts were made to form the groove. T he specimen with diamond saw cut groove is shown in Figure 4-12. Test results of 3 composite specimens prepared as described above showed that the newly introduced groove successfully conc entrated the stress in middle of the specimen on the OGFC side. This indicat ed that the diamond-tip coring approach can be used to form an effective groove when a diam ond-tip coring bit is not available. The cracks in the specimen are shown in Figure 4-13. 81

PAGE 82

Figure 4-11. Geometry of the five cuts to form the groove Figure 4-12. Prepared com posite specimen with di amond saw cut groove 82

PAGE 83

Figure 4-13. Cracks in the specim en at loading rate of 25 mm/min 83

PAGE 84

Typical dis placements at a loading ra te of 25 mm/min are shown in two consecutive parts for two sets of gages in Figur es 4-14 and 4-15. Gage Displacement for OGFC and Dense-Graded -0.00001 0 0.00001 0.00002 0.00003 0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 00.20.40.60.811.21.41.6 t(sec)Displacement(in) OGFC Dense-Graded Figure 4-14. Strain gage displacement at loading rate of 25 mm/min Gage Displacement for OGFC and Dense-Graded 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005 1.51.71.92.12.32.52.7 t(sec)Displacement(inch) OGFC Dense-Graded Figure 4-15. Strain gage displacement at loading rate of 25 mm/min 84

PAGE 85

The strain gage displacement Figures 4-14 and 4-15 did show that specimens start to crack from the open-graded mixture side and pr opagate all the way through the dense-graded side. All tests were completed in 4 seconds. Unfortunately, these results indicated that it is almost impossible to i dentify the instant of crack initiation. This monotonic loading rate 25mm/min was determined to be too fast for testing to evaluate interface effects. 4.3.3 Monotonic Strength Test at 2.5 mm/min and 0.25 mm/min Since loading rate at 25 mm/min was determined to be too fast, loading rates were reduced to 2.5 mm/min and 0.25 mm/min on s pecimens with inch diameter semicircular groove. Tests perfo rmed on three specimens at a loading rate of 2.5 mm/min showed that all composite specimens broke from the open-graded mixture side; but 2 of them broke near the loading he ads while only one of them broke at the center. The cracks in composite specimen are shown in Figure 4-16. Cr acks near loading head Cracks at the center Figure 4-16. Cracks in the specimen at loading rate of 2.5 mm/min 85

PAGE 86

Typical dis placements at loading rate of 2.5 mm/min are shown in Figures 4-17 and 4-18. Gage Displacement for OGFC and Dense-Graded 0 0.00005 0.0001 0.00015 0.0002 0.00025 0.0003 0.00035 0.0004 0.00045 0.0005 456789 t(sec)Displacement(inch) 1 0 OGFC Dense-Graded Figure 4-17. Strain gage displacement at lo ading rate 2.5 mm/min-broke near the end Gage Displacemnet for OGFC and Dense-Graded 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0510152025 t(sec)Displacement(inch) OGFC Dense-Graded Figure 4-18. Strain gage displacement at loading rate 2.5 mm/min-broke at the center 86

PAGE 87

Figures 4-17 and 4-18 clearly indicate that composite specimens broke from the open-graded mixture side. Howev e r, it took around twice longer for the specimen that broke at the center to fail t han the specimen that broke near the loadi ng heads to fail. This is because the rough surface of open-graded mixture forms natural stress concentrators; in some cases, these natural stress concentrators work more effectively than the cored groove stress concentrators, especially near the loading heads where end-effects further concentrated stresses. Loading rates were further reduced to 0.25 mm/min on specimens with cored groove. Interestingly, test results showed th at all composite specimens broke from the dense-graded mixture side near the loading heads. The cracks in composite specimen are shown in Figure 4-19. Typical disp lacements at loading rate 0.25 mm/min are shown in Figure 4-20. Figure 4-19. Cracks in the specim en at loading rate of 0.25mm/min The reason that the specimens broke at the dense-gr aded side instead of the open-graded side is that at this slow loading rate, t he stiffer, lower compliance dense87

PAGE 88

graded mixtures attract more stress, while at the same time dissi pate less stress during testing, while open-graded mixtur es with the lower stiffness and higher strain tolerance have higher capacity to release part of the input energy used to damage the material. Therefore, the compos ite specimens failed once the stra in limit of dense-graded mixture was reached. Gage Displacement for OGFC and Dense-Graded 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0255075100125150175200t(sec)Displacement(inch) OGFC Dense-Graded Figure 4-20. Strain gages displacement at loading rate of 0.25mm/min 4.3.4 Conclusion for Asymmetric Monotonic Strength Tests Tests performed on composite specimens at the loading rate of 25 mm/min, 2.5 mm/min, and 0.25 mm/min show that this m onotonic strength test is not a good choice for interface evaluation because either the testing time is too short to identify the crack initiation point or because composite s pecimen broke from t he dense-graded mixture side, even though stress concentrators were introduced on the open-graded mixture side. 88

PAGE 89

4.4 Repeated Loading Test on Asy mmetrical Composite Specimens Based on the results presented above, it seems clear that a simple monotonic strength test, which is appropriate to determi ne strength and fracture energy in single mixture specimens, is not suitable for co mposite specimens. Therefore, repeated loading tests on asymmetrical composite s pecimen were attempted. All tests were performed at 10C. 4.4.1 Repeated Loading Test on Asymmetr ical Composite Specimen with 1 inch Dense-graded Mixture Layer Repeated loading approach involved a repeated haversine load of one-tenth second duration followed by nine-tenth second rest period. Based on previous research experience, a peak load of 1200-lb peak load was selected for preliminary tests. The repeated loading schematic diagram is shown in Figure 4-21. 0.1 sec loading 0.9 sec rest time load peak load Figure 4-21. Repeated loading schematic diagram 4.4.1.1 Materials The dense-graded mixture for composite s pecimen tests was the same as that used in section 3.2.1. The open-graded mixture used for the t op layer was the same as 89

PAGE 90

that used in section 4.3.1. The composite specimen configuration was the same as that in section 4.3.1 except that a inch diameter semi-circular groove was used instead of diamond s aw cut groove. Neither c onventional tack coat nor Novabond was introduced to the interface between open-graded mixture and dense-graded mixture. 4.4.1.2 Analysis of results Initial trials showed the composite s pecimen broke near the loading head instead of through the groove, as shown in Figure 4-22. The failure occurred near the loading head partly because of end effects, and partly because of low compaction density near the end and a large amount of air voids on the rough open-graded mixture surface forming natural stress concentrators. It appears that these natural stress concentrators were more effective than the drilled groove to initiate cracks. The comparison between newly compacted open-graded mixt ure surfaces and cut surfaces are shown in Figures 4-23 and 4-24. Since composite specimens tend to fail near the loading head because the higher air voids on open-graded mixture surface, which might more effectively concentrate stresses than the groove, slicing off the rough surface to have a uniform surface like in Figure 4-24 was attempted. Tests on compos ite specimen with the top 3/8 inch sliced off show the crack initiated from the groove bottom. Failed specimen is shown in Figure 4-25. Typical total recoverable deformations for OGFC and dense-graded mixture are shown in Figure 4-26; total or instantaneous recoverable deformation from repeated loading test is a measure of resilient m odulus and increase in recoverable deformation is a measure of damage induced in the specimen. 90

PAGE 91

Figure 4-22. Composite specimen broke near the end Figure 4-23. Compacted open-graded mixture surface with rough surface 91

PAGE 92

Figure 4-24. Compacted opengraded mixture surface after rough surface sliced Figure 4-25. Composite spec imen with top 3/8 inch rough surface sliced off 92

PAGE 93

During this repeated loading proc ess, dam age started to accumulate in opengraded mixture right after loading; but littl e or no damage was induced in dense-graded mixture until cracks prorogated in to it. This damage feature can be used to evaluate the effect of different types of interface materials, like conventional tack coat and Novabond. For specimens with same geometry and external load but with different interface conditions, difference in the number of loading cycles to failure is direct measurement of interface condition effect on cracking performance. Total Recoverable Deformation for OGFC and Dense-Graded Mixture 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0200400600800100012001400 t(sec)Total Recoverable Deformation(inch ) Microdamage Macrodamagee Little or no damage Figure 4-26. Total recoverable deformation for OGFC and dense-graded mixture 4.4.2 Repeated Loading Test on Asymmetr ical Composite Specimen with 3 inch Dense-graded Mixture Layer The success of composite specimen with inch diameter semi-circular groove introduced on OGFC surface under repeated loading make it possible to evaluate the effect of interface conditions on stress redistribution in OGFC and dense-graded layer near the interface and on cracking retar dation from OGFC layer into dense-graded 93

PAGE 94

layer. Considering the fracture process involves crack initiation, stable crack propagation and unstable cra ck propagation, dense-graded mi xture thickness in composite specimen was increased to 3 inch from the original 1 inch to make sure that the cracking is still in stable crack propagation range in order to evaluate the effect of interface; otherwise the effect of interf ace may be overshadowed by the unstable crack propagation period. In order to identify the crack tip once the crack propagates from the OGFC through the interface to the dense-graded mixture, strain gages in dense-graded mixture were placed near the interface as shown in Figure 4-27. Gage H1(H2) Gage V1(V2) OGFC Dense-Graded Figure 4-27. Strain gage distributions on specimen with 3 inch dense-graded mixture 4.4.2.1 Materials The dense-graded mixture used for composit e specimen was the same as that used in section 3.2.1. The open-graded mixture used for the t op layer was the same as that used in section 4.3.1. The composite specimen configuration was the same as that in section 4.4.2 except that dense-graded mi xture thickness was increased to 3 inch. 94

PAGE 95

Tack coat and Novabond were applied on the dense-graded surface at the application rate of 0.045 gal/sy and 0.3 gal/sy, respectively. 4.4.2.2 Analysis of results Since the composite specimen thickness was increased to 4 inches, external load was increased up to 2850 lbs (from 1200 lbs fo r 2 inch thick specimen) to reduce the number of loading cycles to fail the specimens. Typical total recoverable deformations for OGFC and dense-graded mixture are shown in Figures 4-28 and 4-29 for tack coat and Novabond interface, respectively. Both Figures 4-28 and 4-29 indicate that damage was accumulating in densegraded mixture even though the damage rate is lower than that in OGFC for thick specimens; the reason is that 3 inch thi ck dense-graded mixture is carrying a larger portion of the external load than the 1 inch thick dense-graded mixture composite specimen. It took more loading cycles fo r specimen with tack coat interface than specimen with Novabond interface to fail; actua lly, specimens with Novabond interface were expected to fail after many more loading cycles. This idea of increasing dense-graded mixture thickness to keep inte rface in stable cr ack propagation region was proved to be inappropriate because OGFC and dense-graded are accumulating damage at the same time; premature failu re in the underlying HMA layer can be addressed by introducing a more effectiv e stress concentrator, like a rectangular groove, to concentrate high enough stress in OGFC so that dense-graded damage is caused by crack propagation rather than exte rnal load during the loading process. 95

PAGE 96

Total Recoverable Deformation for OGFC and Dense-Graded0 0.001 0.002 0.003 0.004 0.005 0.006 04008001200160020002400 t(sec)Total Recoverable Deformation(inch) OGFC Dense-Graded Figure 4-28. Total recoverable deformation for OGFC and dense-graded mixture with tack coat interface (3 inch dense-graded mixture layer) Total Recoverable Deformation for OGFC and Dense-Graded 0 0.001 0.002 0.003 0.004 0.005 0.006 020040060080010001200140016001800200022002400 t(sec)Total Recoverable Deformation(inch) OGFC Dense-Graded Figure 4-29. Total recoverable deformation for OGFC and dense-graded mixture with Novabond interface (3 inch dense-graded mixture layer) 96

PAGE 97

4.4.3 Repeated Loading Test on Asy mmet rical Composite Specimen with Rectangular Groove Rectangular groove was introduced on OGFC su rface in composite specimen with dense-graded mixture thickness reduced back to 1 inch, as shown in Figure 4-30. The groove is 3/8 inch deep and 5 diamond-tip saw cuts wide along the external load pulling direction. 4.4.3.1 Materials The dense-graded mixture used for compos ite specimen was the same as in section 3.2.1. Nova Scotia-granite FC-5 (open-graded friction cour ses) instead of oolitic limestone FC-5 was used for top layer of the composite specimen. Its aggregate gradation is shown in Table 5-4. The aspha lt binder used is ARB-12. The compacted composite specimen consisted of 1inc h dense-graded mixture and 1 3/8 inch opengraded mixture with 20% air voids. The top 3/8 inch OGFC was sliced off. The amount of each component of open-graded mixtur e needed was 83.8 grams of ARB-12 and 1313.2 grams of aggregate with 6.0% binde r content. Conventional tack coat and Novabond were applied on dense-graded surface at the application rate of 0.045 gal/sy and 0.3 gal/sy, respectively. Figure 4-30. Composite spec imen with rectangular groove 97

PAGE 98

Table 4-2. Nova scotia-granite FC-5 mixture aggregate gradation Percent Passing Blend 77% 12% 10% 1% 100% Number 1 2 3 4 JMF 3/4" 100 100 100 100 100.0 1/2" 95 100 100 100 96.2 3/8" 64 92 100 100 75.0 # 4 11 20 97 100 21.6 # 8 3 5 68 100 10.7 # 16 2 3 43 100 7.2 # 30 2 3 28 100 5.7 # 50 2 3 18 100 5.0 # 100 2 3 11 100 4.0 Sieve Size # 200 1.1 2.5 8.0 100 3.1 4.4.3.2 Analysis of results Tests on 4 samples of composite specim en with tack coat interface and 3 samples of composite spec imen with Novabond interface were performed. Because the rectangular groove was used, external load was reduced to 720lbs. Tests results are shown in Figures 4-31 through 4-34. Samples are denoted as: CST1, CST2, CST3, CST4, CSN1, CSN2, CSN3, with letters CS denoting composite specimen, the letter T denoting tack coat interface, and N denoting Novabond interface. Figures 4-31 to 4-34 indicate huge variat ions from sample to sample for both composite specimens with ta ck coat interface and Novabond interface. Specimen CST1 and CST3 failed shortly after loading; but specimen CST2 and CST4 failed by slowly accumulated damage. It took relatively t he same amount of time for CST2, and CST4 and CN3 to fail; this indica tes that the effect of Novabond interface was overshadowed somehow. 98

PAGE 99

Total Recoverable Deformation for OGFC 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01 0250050007500100001250015000 t(sec)Total Recoverable Deformation(inch) CST1-OGFC CST2-OGFC CST3-OGFC CST4-OGFC Figure 4-31. Total recoverable deformation for OGFC with tack coat interface (rectangular groove) Total Recoverable Deformation for Dense-Graded 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0250050007500100001250015000 t(sec)Total Recoverable Deformation(inch) CST1-Dense-Graded CST2-Dense-Graded CST3-Dense-Graded CST4-Dense-Graded Figure 4-32. Total recoverable deformation for dense-graded with tack coat interface (rectangular groove) 99

PAGE 100

Total Recoverable Deformation for OGFC0 0.002 0.004 0.006 0.008 0.01 0.012 025005000750010000125001500017500t(sec)Total Recoverable Deformation(inch) CSN1-OGFC CSN2-OGFC CSN3-OGFC Figure 4-33. Total recoverable def ormation for OGFC with Novabond interface (rectangular groove) Total Recoverable Deformation for Dense-Graded 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 025005000750010000125001500017500 t(sec)Total Recoverable Deformation(inch) CSN1-Dense-Graded CSN2-Dense-Graded CSN3-Dense-Graded Figure 4-34. Total recoverable deformation for dense-graded with Novabond interface (rectangular groove) Jones (1980) pointed out that it is impo ssible to pull on an asymmetric composite specimen without at the same time bending and/or twisting the composite specimen. 100

PAGE 101

This huge variation was determined to be caused by rectangular stress concentration and the bending of composite specimen. 4.5 Repeated Loading Test on Sy mme trical Composite Specimen In order to mitigate bendi ng induced during pulling on asymmetric composite specimen, two single composite specim ens were bonded together at the open-graded mixture surface to create a symmetrical co mposite specimen as shown in Figure 4-35. All tests were performed at 10C. 4.5.1 Materials and Testing Method The dense-graded mixture used for composit e specimen was the same as that used in section 3.2.1. Nova Scotia-granite FC-5 (open-graded friction courses) was used for the top layer of the composite s pecimen. The asphalt binder used was ARB-12. The compacted composite specimen consisted of 1inch dense-graded mixture and 1cinch open-graded mixture with 23% air voids. The natural voids on the compacted OGFC surface were used as stress concentrators to initiate crack. The amount of each component of open-graded mixture needed was 50.6 grams of ARB-12 and 792.3 grams of aggregate with 6.0% bi nder content. Tack coat and Novabond were applied on dense-graded specimen surface at an applicati on rate of 0.045 gal/sy and 0.3 gal/sy, respectively. The composite specimens were bonded together with one spacer throughout the whole symmetric plane except 1 inch wide at the center. This 1 inch open space created a gap between two asymmetric composite specimens and functioned as a stress concentrator. Repeated loading (Figure 421) was applied using a peak load of 2500 lbs. 101

PAGE 102

Both fiber glass and cardboard were tried as spacers. The cracking surfaces are presented in Figure 4-36. Figure 4-35. Symmetrical composite specimen Because of the brittleness of the fiber glass, the composite specimen broke right along the edge of the fiber glass; while the composite specimen with cardboard spacer broke along the natural weak path. Thus cardboard was selected as the spacer material. Figure 4-36. Cracking surfaces for fiber glass (left) and cardboard (right) spacer 102

PAGE 103

4.5.2 Alignment Sy stem for Composite Specimen Preparation One of the issues encount ered in composite spec imen preparation was the alignment of the composite specimen and loading heads. Poor alignment not only makes the testing setup difficult but also introduces bending in the specimen. Two pieces of steel angles were mounted on two diagonal corners of t he loading heads to enforce good alignment. Loading heads assembl ed with two steel angles are shown in Figure 4-37. Figure 4-37. Loading heads with two steel angles 4.5.3 Results of Symmetrical Composite Specimen Two symmetrical composite specimens for tack coat and Novabond interface each were tested and the results are show n in Figure 4-38. Samples are denoted as: CSST1, CSST2, CSSN1, and CSSN2 with letters CSS denoting composite specimen with symmetrical configuration, the letter T denoting tack coat interface, and N denoting Novabond interface. Four extensometers were pos itioned at the center of each of the 103

PAGE 104

symmetrical composite specimens four su rfaces with gage H1 and H2 mounted on the interface and gage V1 and V2 mounted on densegraded mixture surface. The results shown in Figure 4-38 are the average results of gage H1 and H2, and gage V1 and V2. 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005 02000400060008000100001200014000t(sec)Total Recoverable Deformation (inch) cssn1-h cssn1-v csst1-h csst1-v cssn2-h cssn2-v csst2-h csst2-v Figure 4-38. Total recoverable deformation for symmetrical composite specimen The results clearly show that the crack di d initiate from the stress concentrator in the open-graded mixture, propagat ed through the interface and eventually failed the specimen. The repeatability of the test re sults was greatly improved for composite specimens with both conventional tack coat and Novabond interface with the help of this alignment system and symmetrical s pecimen configurati on as compared to asymmetric composite specimen. However, t he effect of the interface conditions was not identified because the ent ire composite specimen was being subjected to tensile load with the loading head glued to bot h the open-graded and dens e-graded mixture. This loading configuration led to most of the external load being applied to the dense104

PAGE 105

graded mixture because of its higher stiffness. It can be conc luded that the e ffect of the interface conditions was overwhelmed by this loading configuration. 4.6 Monotonic Partial Loading on Sy mmetric Composite Specimen The results from secti on 4.3 to 4.5 indicate that the effect of interface conditions on the composite specimen cracking performa nce is overshadowed when the composite specimen behavior is dominated by the dens e-graded mixture. One possible way to reduce this dense-graded mixture dominance behavior is to decrease the load carried by the dense-graded mixture by loading the OGFC mixture part only. The following two sections deal with the optimum composite specimen geometry and test results of the new loading configuration. All tests were performed at 10C. 4.6.1 Optimum Geometry of Symmetric Composite Specimen for Partial Loading Three dimensional finite element me thod (FEM) analyses were performed to determine the optimum diameter and constraint conditions of the symmetric composite specimens. The material properties of eac h component of the composite specimen, including OGFC, dense-graded mixture, and load ing head are shown in Table 4-3. The sketch of half composite specimen (symmetric to OGFC surface) is shown in Figure 4-39. Since the desirable stress distribution is that the stress is dominant on the OGFC mixture rather than on the dense-graded mixture under tensil e loading conditions, the stresses in the direction of ex ternal loading (x direction in the model) were checked. Table 4-3. Material properti es of composite specimen Type Loading head OGFC Dense-Graded Modulus(psi) 1.00E7 1.01E6 1.57E6 Poisson's ratio 0.2 0.35 0.35 105

PAGE 106

Composite specimen diameter s in the range of 3 inch to 6 inch were analyzed. The 3 inch diameter was selected for practical reasons, such as easy handling and the smallest available strain gage length. The 6 inch diameter was selected because this is typically the largest composite specimen that can be fabricated or cored. Three constraint conditions were considered: 1) only OGFC is loaded; 2) OGFC plus half dense-graded mixture are l oaded; 3) both OGFC and densegraded mixture are loaded. The composite specimen is 1.5 inch wide with 1 inch thick OGFC and 1 inch densegraded mixture. The diameter effect of the composit e specimen on the stress distribution in external loading (x direction in the model ) under 1000 psi tensile stress is shown in Figure 4-40, whereas the effect of constrai nt conditions is shown in Figure 4-41. It can be concluded from this FEM anal ysis that the stresses carried by OGFC increase as compared with the stresses carried by dense-graded mixture when the composite specimen diameter and/or the loading area are reduced. Figure 4-39. Half composit e specimen modeling sketch 106

PAGE 107

-250 250 750 1250 1750 2250 00.511.52 Distance from OGFC Surface (inch)Stress in X Direction(psi) D=3 D=4 D=5 D=6 Figure 4-40. Effect of composite specimen diameter on stress distribution -250 250 750 1250 1750 2250 00.511.5 Distance from OGFC Surface(inch)Stress in X Direction(psi) 2 Only OGFC OGFC Plus Half Dense-Graded Both OGFC and Dense-Graded Figure 4-41. Effect of constraint condition on stress distribution 107

PAGE 108

4.6.2 Analysis of Results The materials used for this evaluation were the same as that used in section 4.5.1 except that the compacted OGFC layer was 1.5 inch thick and sliced down to 1 inch thick to remove the rough surface. Tack coat and Novabond were applied on densegraded surface at the applicat ion rate of 0.045 gal/sy and 0.3 gal/sy, respectively. Based on the results of the FEM analysis, the final specimen geometry was determined to be 1.5 inch wide by 2 in thick for 3 in diameter. This 3 in diameter composite specimen was obtained by coring of the 6 in laboratory prepared composite specimen, as in Figure 4-42. Following the composite spec imen preparation procedure st ated in section 4.2 and 4.5, the final specimen is shown in Figur e 4-43. One new set of loading heads with 3 inch diameter and 2 inch width was fabricated for the 3 inch diameter specimen, as shown in Figure 4-44. The di gital image correlation system (Birgisson et al. 2009) was tried for the strain field measurement instead of the tr aditional strain gage measurement. Because of the time consuming iss ue involved with repeated loading tests, monotonic loading at the rate of 0.025 mm/mi n was applied to reduce the testing time. This slow loading rate was expected to allo w the interface material to have enough time to dissipate the stress and/or reduce the st ress transmitted through the interface. The low quality of the pictures taken during this test run led to no valuable strain field information. The failed composite specimen is shown in Figure 4-45. The failure mode shown in Figure 4-45 indica ted that the crack did start from the bottom of the groove, but the shear stress induced in OGFC reached its failure limit before the tensile stress i nduced near the interface coul d drive the crack from OGFC 108

PAGE 109

through the interface and into the dense-graded mixture. It can be concluded that when the composite specimen was only loaded on OGFC section, specimens were failed because of the low shear strength of OGFC. Figure 4-42. Schematic diagram of t he 3 inch diameter specimen coring Figure 4-43. Prepared 3 in ch diameter symmetrical composite specimen 109

PAGE 110

Figure 4-44. Loading heads, alig nment bar, and shim blocks Figure 4-45. Failure mode for composite specimen under monotonic partial loading 110

PAGE 111

4.7 Monotonic Internal Loading on Sy mmetric Composite Specimen Based on the results from section 4.6, it was determined that on the composite specimen, from inside the stress concentration groove, even more favorable tensile stress distribution to initiate and drive the crack to propagate across the interface could be achieved by internally loading (expand ing). The loading system shown in Figure 446, involving a split cylinder placed inside the stress concentrating hole, was conceived and designed. All tests were performed at 10C. 4.7.1 Materials and Loading Assembly The materials used for this evaluation was t he same as that used in section 4.6. Conventional tack coat and Novabond were applied on densegraded specimen surface at an application ra te of 0.045 gal/sy and 0.3 gal /sy, respectively. The composite specimen geometry is the same as in secti on 4.6. Since the composite specimen was intended to be loaded from within the stress concentrator, a new of loading assembly was fabricated as show n in Figure 4-46. This loading assembly consists of one set of loading heads, two pi ns, two cross bars (split cylinders through which the load is applied) and four clevises that form two sets of loading yokes. The crossbar is a in diameter half cylinder bar; its size is the same as that of the stress concentration groove. 4.7.2 Analysis of Results Monotonic loading at a rate of 0.025 mm/ min was applied on the loading head to pull the composite specimen ap art from the stress concentra tion groove, as shown in Figure 4-47. Strain field measurements ar e not available because of the low image qualities for the digital image correlation syst em. The failed specim ens are presented in Figures 4-48 and 4-49 for conventional tack coat and Novabond interface, respectively. 111

PAGE 112

Figure 4-46. Loading assembly for internal loading The results shown in Figures 4-48 and 4-49 indicate that the crack initiated from the bottom of the groove and propagated towards the interface. It should be noted that the crack propagated into the dense-graded mixt ure for the specimen with conventional tack coat interface (Figure 4-48) whereas the crack propagated along the plane of maximum shear stress for the specimen with Novabond interface (Figure 4-49). For the specimen with conventional tack coat in terface, after the crack propagated through the interface, bending was introduced in the OGFC mixture near the composite specimen symmetrical plane because of the constraint from dense-graded mixture. The tensile stress induced by this bending behavio r reached the tensile strength limit of OGFC before the tensile stress induced in the dense-graded mixture exceeds its limit, which led to the final failure of the com posite specimen with conventional tack coat 112

PAGE 113

interface. On the other hand, for the specimen with Novabond interface, which introduced a higher fracture resi stant material at the interf ace, the crack also started from the bottom of the groov e but propagated along the plane of maximum shear stress. It can be concluded that the final failu re was caused by the bending behavior for the specimen with conventional tack coat in terface and by shear for the specimen with Novabond interface. Figure 4-47. Test setup for internal loading 113

PAGE 114

Figure 4-48. Failure mode for composit e specimen with tack coat interface Figure 4-49. Failure mode for co mposite specimen with Novabond interface 114

PAGE 115

4.8 Monotonic Internal Loading on Sy mmet ric Composite Specimen with Carbon Fiber Sheet Reinforcement The internal loading configur ation successfully forced the crack to initiate at the stress concentrator and OGFC. However, the secondary bending and shear stresses resulting after crack initiation caused the crack to change paths to go through the OGFC in conventional tack coat interface and shear failure with the Novabond interface. In order for the crack to continue propagat ing into the dense-graded mixture, the composite specimen curved end surface was rein forced with carbon fiber to minimize or eliminate the potential for failure in bending or shear through the OGFC. All tests were performed at 10C. 4.8.1 Materials and Specimen Preparation The materials used for this evaluation was t he same as that used in section 4.6. Tack coat and Novabond were applied on dense-graded specimen surface at an application rate of 0.045 gal/ sy and 0.3 gal/sy, respectively. The composite specimen geometry is the same as in section 4.6. Carbon fiber sheet was glued to the cu rved end surface of the composite specimen using Hardman double/bubble regular setting epoxy, as shown in Figure 450. This very low-viscosity epoxy cures to a light color (almost colorless) solid. The specimen curved end surface with glued carbon fi ber sheet is shown in Figure 4-51. 4.8.2 Analysis of Results Monotonic loading at the rate of 0. 025 mm/min was applied on curved end surface reinforced composite specimen fo r both tack coat and Novabond interface. The typical failure mode of the composite specimen is shown in Figure 4-52 and the loaddisplacement curve is shown in Figure 4-53. 115

PAGE 116

Figure 4-52 clearly showed that the cr ack propagated through the interface and into the dense-graded mixture when the reinforcement was used. Howev er, Figure 4-53 indicated that not much difference was seen between specimen with conventional tack coat interface and Novabond interface in terms of frac ture energy. For the specimen with tack coat interface, the load reached t he peak after the crack tip propagated into the dense-graded mixture. For the specimen with Novabond interface, even after the crack tip propagated into the dense-graded mixture, the OGFC part with Novabond materials still held together and the external effort was overcoming both the OGFC and dense-graded mixture, wh ich explained why the peak load for specimen with Novabond interface was lower than specim en with tack coat interface. Figure 4-50. Composite specimen curved end surface, epoxy, and carbon fiber sheet 116

PAGE 117

Figure 4-51. Composite specimen curved end surface with glued carbon fiber sheet Figure 4-52. Failure mode of composite s pecimen with curved end surface reinforced 117

PAGE 118

0 50 100 150 200 250 300 350 400 450 00.010.020.030.040.050.060.070.08 Displacement(inch)Load(lbs) Novabond Conventional Tack Figure 4-53. Load-displacement curve for spec imens with curved end surface reinforced 4.9 Repeated Internal Loading on Symmetric Composite Specimen The successful elimination of bending and /o r shear failure through the OGFC with the carbon fiber sheet reinforcement led to the desired composite specimen failure mode, which is characterized by t he crack propagating through the dense-graded mixture. However, the com posite specimen fracture energy was not sensitive to the interface conditions and failure occurred on one side of the specimen only. The reason appears to be that the Novabond interface materials can not release the stress accumulated near the interfac e under monotonic loading. 4.9.1 Materials and Lo ading Configuration The materials used for this evaluation was the same as that in section 4.6. Conventional tack coat and Novabond were applied on dense-gr aded surface at the 118

PAGE 119

application rate of 0.045 gal /sy and 0.3 gal/sy, respectively. Composite specimen curved end surfaces w ere reinforced with carbon fiber sheet. The 6-inch diameter composite specim en has the advantage of easy fabrication and long shear path as compared to 3 inch diameter specimen. Thus, the composite specimen geometry is the same as in sect ion 4.6 except that the specimen diameter was increased back to 6 inch, which can be eas ily obtained from gyratory compaction or field coring. 4.9.2 Analysis of Results The resting period in the repeated loading approach can provide the stress release time for interface materials and balance the damage accumulated on the two sides of symmetrical plane. The repeated loading mode shown in Figure 4-21 was applied. Based on previous research and FEM modeling results, 570 lbs peak load was tried. All tests were performed at 10C. Thr ee specimens with tack coat and Novabond interface each were prepared and tested. The typical fa ilure mode is shown in Figure 4-54 and the number of loading cycles for specimen failure is shown in Figure 4-55. Figure 4-54 shows that the crack initiated from the groove and propagated through the interface and into the dense-graded mixture, which caused the final failure on both sides of the composite specimen. The testi ng configuration successfully identified the benefit of Novabond, which enhanced the cracking resistance of the composite specimen as shown in Figure 4-55. The results presented in this section indica te that this testing configuration and mode of loading appears be appropriate to evaluate the interface conditions on composite specimen for top-down cracking. It can also be used for reflective cracking if 119

PAGE 120

the existing cracked pavement layer rather than overlay is glued together at the symmetrical plane. Figure 4-54. Failure mode for composite s pecimen under repeated internal loading Test1 Test1 Test2 Test2 Test3 Test3 Average Average0 10000 20000 30000 40000 50000 60000 70000 NovabondConventional TackNo. of Cycles to Failur e Figure 4-55. Number of cycl es to failure of Novabond and conventional tack 120

PAGE 121

CHA PTER 5 DATA COLLECTION AND IN TERPRETATION METHOD The test method developed in the Chapter 4 showed excellent potential for the evaluation of interface bond condition effe ct on top-down and reflective cracking. However, although the number of loading cycles to spec imen breaking can be used as a parameter to differentiate different interf ace conditions, it does not provide information regarding cracking initiation and propagation. Local deformation measurement, like strain gage measurement, was used for the identification of cra ck initiation and the crack propagation near the interface. 5.1 Data Collection Method In order to fully capture the characteristi cs of the repeated loading shown in Figure 4-21, previous research experience indicated that at least 500 points of data per second have to be acquired. A typical repeated loadi ng at the rate of 512 points per second versus time plot is shown in Figure 5-1. The fact that the applied load has to be low enough to allow the interface materials to relax during the resting period and thus dissipate the accumulated stress near the interface leads to considerably long testi ng time as compared to monotonic strength test. If the data, including four strain gages, external load, displacement and the corresponding time, are acquired at the ra te of 512 points per second as shown in Figure 5-1 for the entire testing period, which may be as long as 18 hours as shown in Figure 4-55, the amount of collected data will surpass the processing capability of the computer available in the l aboratory. Fortunately, not all of the data during the entire testing period are necessary for crack initia tion and propagation evaluation; data points 121

PAGE 122

where abrupt strain gage deformation occurs ar e sufficient to chara cterize the damage accumulation in the composite specimen. During the test, frequency data acquired at the rate of 5 points per second were plotted versus time to identify sudden changes in strain gage deformation as an indication of local damage evolvement. Once this sudden change occurs, or whenever desired, the operator re corded a burst of data for 6 cons ecutive loading cycles at a rate of 512 data points per second to calculate the specimens total recoverable deformation. The data points acquired for recoverable def ormation calculation was shown in Figure 5-2. Total recoverable deformation, includi ng both the instantaneous recoverable and the time-dependent recoverable deformation during the unloading and rest-period portion of each loading, as defined in Figure 5-3, was calculated. The results at each acquired data point are the four strain gage total recoverable deformations, and peak load as an average of 6 consecutive loading cycles. It should be noted that the data format of these 6 consecutive loading cycles is the same as that of the data used to calculate total resilient modul us in Superpave IDT test. 5.2 Data Interpretation Method 5.2.1 Data Interpretation Method As stated in section 4.9, t he total number of loading cycle it took for the composite specimen to break is a straightforward cr acking resistance com parison parameter for specimens with different interface conditi ons under the same magnitude of loading. However, this parameter provides only the fracture resistance of the whole specimen without any information regarding the damage evolution in the specimen. 122

PAGE 123

0 50 100 150 200 250 300 350 400 450 500 0123456Time ( sec ) Load(lbs) Figure 5-1. Typical repeat ed loading versus time 0 0 1 3 1 -0.013 -0.0129 -0.0128 -0.0127 -0.0126 -0.0125 -0.0124 -0.0123 -0.0122 050100150200 Time(sec)Strain Gage Deformation(inch) Acquired Data Points Figure 5-2. Data points recorded for recoverable deformation calculation 123

PAGE 124

-0.01295 -0.0129 -0.01285 -0.0128 -0.01275 -0.0127 -0.01265 -0.0126 0123456Time(sec)Strain Gage Deformation(inch) Total recoverable deformation Figure 5-3. Total recoverable deformation definition It has been well recognized that damage induced in the specimen can be measured by stiffness reduction in the spec imen. Because of the complicated stress distribution in composite s pecimen under repeated loading, t he stiffness calculation is not practical for routine evaluation. However, the easily obtained total recoverable deformation of strain gage measurement which is inversely related to the specimens stiffness was calculated to facilitate comparison of the specimens behavior and performance throughout the test. The typical total recoverable deformation versus time is presented in Figure 5-4. The total recoverable deformation versus time curve can be divided into three stages: t he initial stage, which is known to involve changes in temperature and local damage ad jacent to the loading yoke s; the second stage, which involves steady-state damage; and the final stage, when the crack propagates rapidly 124

PAGE 125

and the specimen breaks. Damage rate is de fined as the slope of the steady state of total recoverable deformation progre ssion curve shown in Figure 5-4. 0.00E+00 2.00E-04 4.00E-04 6.00E-04 8.00E-04 1.00E-03 1.20E-03 05000100001500020000250003000035000 Time ( sec ) Total Recoverable Deformation(inch) Steady state Slope = Damage rate Figure 5-4. Typical total recoverable deformation 5.2.2 Damage Rate as a Differentiation Parameter For the tests mentioned in section 4.9, strain gage deformations are measured for total recoverable deformation calculation. The distance between the gage point center and interface is inch for both gages in OGFC and in dense-graded mixture as shown in Figure 4-54. Because of the specimen to specimen variation, total recoverable deformation was normalized to the intercept of the steady state line to total recoverable deformation axis as shown in Figure 54 (cycled). Normalized total recoverable deformation for strain gage measurement in OGFC is shown in Figure 5-5. 125

PAGE 126

1 1.25 1.5 1.75 2 2.25 2.5 05000100001500020000250003000035000400004500050000 Time(sec)Normalized Total Recoverable Deformation Novabond 1 Tack coat 1 Novabond 2 Tack coat 2 Novabond 3 Tack coat 3 N o v a b o n d A v e r a g eC o n v e n t i o n a l T a c k A v e r a g e Figure 5-5. Normalized total recoverable deformation Figure 5-5 clearly indicates that damage rate can indentif y the effect of different interface material on the damage development inside composite specimen. The lower damage rate developed in composite specimen with Novabond interface is consistent with the fact that s pecimens with Novabond interface can resist more loading cycles to failure when compared to specimens with tack coat interface. It can be concluded that the damage rate as defined in this section can be used to evaluate the damage evolution for different interface materials. It is an additional evaluation parameter besides the number of loading cycles to failure that does not require long testing time for evaluation. 126

PAGE 127

CHA PTER 6 INTERFACE CRACKING PERFORMANCE EVALUATION A good bond at the interface can help dissipate the stresses near the interface and increase the resistance to fracture through the interface. Composite specimens with polymer modified asphalt emul sion applied at the interface are expected to have better cracking performance compared to specimens with conventi onal tack coat materials. Therefore, the effect of interface materials on both t op-down and reflective cracking performance will be evaluated in this chapter. Since OGFC mixtures may be the first front in resisting top-down cracki ng, it is also of great impor tance to evaluate the effect of OGFC on pavement top-down cracking resistance. All test s were performed at 10C. 6.1 Effects of Interf ace on Top-down Cracking In this section, the effects of three types of interface materials, conventional tack coat, Novabond from Road Science, LLC and trackless tack, on top-down cracking were evaluated. Except for interface materials, the asphalt and aggregate used for composite specimen production, specimen g eometry and loading configuration and magnitude are the same as stat ed in section 4.9. All damage rate results were obtained from strain gage measurements at the same location with respect to each composite specimen as shown in Figure 6-1. 6.1.1 Effects of Novabond on Top-down Cracking As part of the test method devel opment, the effect of Novabond interface on composite specimen cracking performance in te rms of the number of cycles to failure was reported in section 4.9 as shown in Figure 4-55. The effect of Novabond interface on damage rate as compared to conventional tack coat interface is shown in Figure 6-2. 127

PAGE 128

The typical cracking surfaces for specimens with Novab ond and conventional tack coat interface are presented in Figure 6-3. Figure 4-55 together with Figure 6-2 cl early indicated that specimens with Novabond interface outperformed the specimen wit h conventional tack coat in terms of cracking resistance. The reason appears to be the stress relief capability of the Novabond material and the higher polymer modified asphalt emulsion concentration near the interface as shown in Figure 6-3. Figure 6-1. Stain gage distribut ion on composite specimen 6.1.2 Effects of Trackless Tack on Top-down Cracking Since the trackless tack evaluated in this section is a non-emulsified trackless tack coat material, it has to be distributed at a temperature of ar ound 350F. The primary advantage of this product is that it cools to touch in 20 to 30 seconds and can be driven over by haul trucks and paver without tracking. It can be applied at very thick residue rates up to 0.18 gal/sy as compared to 0. 03 gal/sy residue rate for conventional tack coat used for current FC-5. 128

PAGE 129

Test1 Test1 Test2 Test2 Test3 Test3 Average Average0.00E+00 2.00E-09 4.00E-09 6.00E-09 8.00E-09 1.00E-08 1.20E-08 1.40E-08 1.60E-08 NovabondConventional TackDamage Rate Figure 6-2. Damage rate of Novabond and conventional tack coat interface Novabond Conventional Tack Figure 6-3. Cracking surfaces of specimens with Novabond and tack coat interface The effects of trackless tack on cracki ng performance were evaluated at two residue application rates, 0.2 gal/sy and 0.13 gal/sy. This 0.2 gal/sy residue was 129

PAGE 130

selected to match the residue ap plication rate of the Novabond interface used in section 6.1.1, whereas 0.13 gal/sy was the residue applicati on rate used by FDOT. The emulsion and residue application rates used in this section for Novabond, conventional tack coat and trackless tack are summarized in Table 6-1.Three composite specimens were produced for each application rate. Table 6-1. Novabond, conventional tack and trackless tack application rate Conventional Tack Type Novabond (AC-20 residue) Trackless Tack Emulsion Application Rate (gal/sy) 0.3 0.045 Not applicable Not applicable Residue Application Rate (gal/sy) 0.2 0.025 0.2 0.13 The effect of trackless tack inte rface on composite specimen cracking performance in terms of the number of cycles to failure is shown in Figure 6-4. The effect of trackless tack interface on damage ra te as compared to conventional tack coat interface is presented in Figure 6-5. The ty pical cracking surfaces of specimens with trackless tack interface are presented in Figure 6-6. The results presented in Figures 6-4 and 6-5 indicated that the trackless tack interface has a negative effect on the cracki ng resistance of the composite specimens as compared to the specimens with conventional tack coat interface. This finding is contrary to the presumption t hat this thick trackless tack application rate can increase the cracking performance of composite spec imen. The reason appears to be that the brittleness of trackless tack led to the reducti on of fracture resistance of specimens with 130

PAGE 131

trackless ta ck interface. The dry cracking surfaces shown in Figure 6-5 seem to confirm this conclusion when compared to Figure 6-3. Test1 Test1 Test1 Test2 Test2 Test2 Test3 Test3 Test3 Average Average Average0 5000 10000 15000 20000 25000 30000 35000 40000 Trackless Tack (0.2 gal/sy)Trackless Tack (0.13 gal/sy)Conventional TackNo. of Cycles to Failur e Figure 6-4. Number of cycles to failure of trackless tack and conventional tack Test1 Test1 Test1 Test2 Test2 Test2 Test3 Test3 Test3 Average Average Average0.00E+00 5.00E-09 1.00E-08 1.50E-08 2.00E-08 2.50E-08 3.00E-08 3.50E-08 4.00E-08 Trackless Tack (0.2 gal/sy)Trackless Tack (0.13 gal/sy)Conventional TackDamage Rate Figure 6-5. Damage rate of trackless tack and conventional tack 131

PAGE 132

Trackless Tack Trackless Tack Figure 6-6. Cracking surfaces of specimens with trackless tack interface 6.2 Effects of Interf ace on Reflective Cracking In this section, the effects of two types of interface materials, conventional tack coat and Novabond from Road Science, LLC, on reflective cracking were evaluated. Tests were performed on composite specim ens produced in the laboratory by Road Science and field cores obtained from I-70/ Broadway avenue exit project in St. Louis. All tests were performed at 10C. 6.2.1 Effects of Novabond on Reflective Cracking Six composite specimens for each of the tw o types of interface, 0.1 gal/sy diluted conventional tack coat and 0.2 gal/sy Novabond, were prepared by Road Science in their laboratory. The specimens consist ed of two dense-graded layers. The geometry and strain gage distribution of prepared testing specimen are shown in Figure 6-7. 132

PAGE 133

Figure 6-7. Composite specimen geometry and strain gage distribution Tests were performed under the loading mode stated in section 4.9 with 570lbs peak load. Tests results are presented in Fi gure 6-8 for number of loading cycles to failure and Figure 6-9 for damage rate. Test1 Test1 Test2 Test2 Test3 Test3 Average Average0 10000 20000 30000 40000 50000 60000 70000 80000 Novabond (0.2 gal/sy)Conventional Tack (0.1 gal/sy diluted)No. of Cycles to Failure Figure 6-8. Number of cycl es to failure of Novabond and diluted conventional tack 133

PAGE 134

Test1 Test1 Test2 Test2 Test3 Test3 Average Average0.00E+00 1.00E-09 2.00E-09 3.00E-09 4.00E-09 5.00E-09 6.00E-09 7.00E-09 8.00E-09 9.00E-09 1.00E-08 Novabond (0.2 gal/sy)Conventional Tack (0.1 gal/sy diluted)Damage Rate Figure 6-9. Damage rate of Novabond and diluted conventional tack Specimens with 0.2 gal/sy Novabond and 0.1 gal/sy diluted conventional tack interface exhibited almost the same fracture resistance. Afte r careful examination of the cracking surfaces as shown in Figure 610, the similarity in cracking performance between specimens with Novabond and conventional tack interface might be explained by the fact the low voids content of dense-graded mixture does not allow the migration of Novabond upwards into the mixture like in OGF C but is rather partly squeezed out from the interface. It shoul d also be noted that dense-graded mi xture has higher fracture energy than OGFC, which leads to shorter time for Novabond to have an effect (after crack propagated to the interface) for spec imen consisting of two layers of densegraded mixture. 134

PAGE 135

6.2.2 Effec ts of Novabond on Reflective Cracking on Specimens with Teflon Spacer Six specimens for each of the two types of interface, 0.1 gal/sy diluted conventional tack coat and 0.2 gal/sy Novabond, were prepared by Road Science in the same way as in section 6.2.1 except that a teflon sp acer was introduced to more effectively concentrate stress. The prepared composite specimen with teflon spacer is shown in Figure 6-11. The geom etry and strain gage distribution of test specimens were the same as in section 6.2.1. Novabond Conventional Tack Figure 6-10. Cracking surfaces of specimens with Novabond and diluted conventional tack interface Half the load used in section 6.2.1, 280 lbs, was applied bec ause of the teflon spacer stress concentrator. Specimens with Novabond and conventional tack coat failed in 103 to 125 hours. Careful exam ination of the strain gage deformations indicated that the specimens were not uniformly loaded, which made the results 135

PAGE 136

unreliable. Therefor e, the results with 280lbs peak load were not included in the following analysis. Figure 6-11. Composite spec imen with teflon spacer Peak load was increased to 430 and 520 lb s to reduce the testing time. Tests results are presented in Figure 6-12 for loadi ng cycles to failure and Figure 6-13 for damage rate. Results presented in Figures 6-12 and 6-13 indicate that specimens with Novabond interface exhibited higher frac ture resistance than specimens with conventional tack coat interface. These results also indicate that Novabond applied at the interface took effect ri ght from the loading moment with the introduction of teflon spacer as stress concentrator, which leads to better cracking performance for specimens with Novabond interface even with the possibility of being partially squeezed out. 136

PAGE 137

Novabond Novabond Conventional Tack Conventional Tack 0 20000 40000 60000 80000 100000 120000 140000 160000 Low StressHi g h StressNo. of Cycles to Failure Figure 6-12. Number of cycl es to failure of Novabond and diluted conventional tack for specimens with teflon spacer Novabond Novabond Conventional Tack Conventional Tack 0.00E+00 2.00E-09 4.00E-09 6.00E-09 8.00E-09 1.00E-08 1.20E-08 Low StressHigh StressDamage Rate Figure 6-13. Damage rate of Novabond and diluted conventional tack for specimens with teflon spacer 137

PAGE 138

6.2.3 Effec ts of Novabond on Double Novachip Reflective Cracking Six cored specimens, double Novachip on top of PCC base, were taken from 3 different sections, each with different Novabond emulsion application rate. As stated earlier, the specimens were cores from t he I-70/Broadway Avenue ex it project in St. Louis. Typical double Novachip composite specimen is shown in Figure 6-14. There were two passes of a gap graded Novachip mixture with approximately inch thick per pass. The details of Novabond application rates for each section are presented in Table 6-2. The actual lift thickness measurem ents are exhibited in Table 6-3, in which composite specimens are denoted by secti on number and test number (for instance s1t1 meaning section 1 test 1). The peak load used in section 6.1, 570lbs was first tried on composite specimen s1t2. The specimen failed very quickly after loading, around 700 seconds. The results of specimen s1t2 were not included in the fo llowing analysis. Around half of the load used in section 6.1, 270 lbs, was used on the rest of the specimens. For section 2 specimens, an unsuccessful attempt was made to introduce a crack in the concrete groove in specimen s2t2. Damage was introduced in the specimen while cracks were initiated by force with visible cr ushed concrete at the bottom of the groove. The results of specimen s2t2 were not included in the following analysis. For section 3, specimen s3t1 was broken by accident. Visibly le ss binder (drier) on top layer was observed on all three composit e specimens of section 3 as shown in Figure 6-15. This was probably caused by cons truction problems. Results of specimens from section 3 were therefore not included in the following analysis. 138

PAGE 139

Figure 6-14. Double Novachip composite specimen Table 6-2. Novabond application rate of double Novachip Number Lift 1(gal/sy) Lift 2(gal/sy) Total(gal/sy) Section 1 0.192 0.211 0.4 Section 2 0.308 0.329 0.64 Section 3 0.299 0.398 0.7 Tests results are presented in Figure 6-16 for number of loading cycles to failure and Figure 6-17 for damage rate. Damage rate results were obtained from strain gage measurement as shown in Figure 6-14. In Fi gures 6-16 and 6-17, low rate are results for section 1 and high rate are re sults for section 2. Test re sults indicated that section 1 with lower Novabond application rate underperformed rela tive to section 2 with higher Novabond application rate. 139

PAGE 140

Table 6-3. Double Novachip lift thickness Number Lift 1 (mm)Lift 2 ( mm)Concrete (mm)Total (mm) 1-1a 17.50 17.50 14.61 49.61 s1t1 1-1b 17.10 17.29 18.16 52.55 102.16 1-2a 19.32 21.02 15.46 55.80 s1t2 1-2b 20.31 21.51 12.45 54.27 110.07 1-3a 19.52 28.14 12.34 60.00 s1t3 1-3b 20.00 25.17 13.51 58.68 118.68 2-1a 19.54 28.86 11.98 60.38 s2t1 2-1b 17.17 29.29 13.05 59.51 119.89 2-2a 18.84 30.75 12.01 61.60 s2t2 2-2b 17.82 28.77 16.17 62.76 124.36 2-3a 18.51 33.15 11.21 62.87 s2t3 2-3b 17.77 31.40 13.46 62.63 125.50 3-1a 16.70 19.20 13.45 49.35 s3t1 3-1b 17.16 17.57 13.97 48.70 98.05 3-2a 15.69 16.83 13.26 45.78 s3t2 3-2b 15.15 17.12 12.77 45.04 90.82 3-3a 17.43 19.86 13.42 50.71 s3t3 3-3b 16.44 18.42 15.52 50.38 101.09 Figure 6-15. Cracking surfaces of double Novachip composite speciemn 140

PAGE 141

It should be noted that the inch thick concrete might cause local effects at the bottom of the concrete groove as shown in Fi gure 6-18. In the future concrete thickness should be increased in order to reduce this local effect. Test1 Test1 Test3 Test3 Average Average0 50000 100000 150000 200000 250000 300000 Low RateHigh RateNo. of Cycles to Failure Figure 6-16. Number of cycles to failure for double Novachip specimen Test1 Test1 Test3 Test3 Average Average0.00E+00 2.00E-09 4.00E-09 6.00E-09 8.00E-09 1.00E-08 1.20E-08 1.40E-08 Low RateHigh RateDamage Rate Figure 6-17. Damage rate of double Novachip specimen 141

PAGE 142

Figure 6-18. Uneven concrete thickness 6.3 Effects of OGF C on Top-down Cracking Because of its low fracture resistance, OGFC may increase the susceptibility to top-down cracking failure for pavements. Test specimens consist of only dense-graded mixture with no interface. The typical all dense-graded mixture specimen is shown in Figure 6-19. Tests were performed under the sa me condition as in section 6.1.1. Their results are compared with t hose from composite specimens (OGFC compacted on dense graded) with conventiona l tack coat interface as shown in section 6.1.1. Tests results are presented in Figure 6-20 for loading cycles to failure and Figure 6-21 for damage rate. Results indicate that to some extent OGFC reduce cracking resistance for pavements with OGFC overla y relative to pavements without OGFC. 142

PAGE 143

Figure 6-19. All dense-graded mixture specimen Test1 Test1 Test2 Test2 Test3 Test3 Average Average0 5000 10000 15000 20000 25000 30000 35000 40000 45000 OGFC on Dense with Conventional TackAll DenseNo. of Cycles to Failure Figure 6-20. Number of cycles to failu re of OGFC on dense with conventional tack interface and all dense-graded 143

PAGE 144

Test1 Test1 Test2 Test2 Test3 Test3 Average Average0.00E+00 2.00E-09 4.00E-09 6.00E-09 8.00E-09 1.00E-08 1.20E-08 1.40E-08 1.60E-08 OGFC on Dense with Conventional TackAll DenseDamage Rate Figure 6-21. Damage rate of OGFC on dense wit h conventional tack interface and all dense-graded 144

PAGE 145

CHA PTER 7 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 7.1 Summary and Conclusions A composite specimen interface cracking (CSIC) test system was developed in this research to evaluate the effects of pavement layer interface characteristics on cracking performance. During the develop ing process, a com posite specimen with interface preparation method was first conceived and evaluated; and the following variables were evaluated: loading mode incl uding monotonic and repeated; composite specimen thickness including 1 inch and 3 inch dense-graded mixture; stress concentrator type including rectangular and cy lindrical; composite specimen geometry including asymmetric and symmetric and specim en diameter; loading position including specimen curved end surface and within the st ress concentrator; composite specimen curved end surface reinforcement. Three types of interface conditions on top-down cracking, two types of interface conditions on reflective cracking, and the effect of OGFC on top-down cracking were evaluated. Some of the findings associated with this development and testing are as follows, The additional compaction of Superpave mi xture or OGFC had a negligible effect on the integrity the bottom dens e-graded mixture. If there is any, the effect would be the improvement in damage resistant. As compared with monotonic loading, repeated loading with resting period allows the interface materials to dissipate the accumulated stress and thus differentiate the effects of interface conditions on cracking. For asymmetric composite specimen under monotonic loading, the attempt to keep interface in stable crack propagat ion region by increasing lower layer thickness from 1 inch to 3 inch was proved to be ineffective because damage was accumulating in both composite specimen layers. Greater specimen to specimen variability was caused by rectangular stress concentrator and asymmetric specimen as compared to cylindrical stress concentrator to symmetric specimen, respectively. 145

PAGE 146

More favorable stress state was achieved by reducing the spec imen diameter and loading head width; however it led to specimen shear failure in OGFC. Shear failure was eliminated by loading within the stress concentrator; however this led to specimen bending failure, which was eliminated by use of carbon fiber reinforcement on specimen curved end surface. Repeated loading applied within the 6 inch diameter specimen stress concentrator with edge surface reinforcement successfully propagated cracks through the interface and effectively identified the effe cts of interface on cracking performance. For top-down cracking performance, as compared with conventional tack coat, Novabond increased the number of cycles to failure and reduced the damage rate; on the other hand, trackless tack r educed the number of cycles to failure and increased the damage rate. For reflective cracking performance, as compared with conventional tack coat, Novabond increased the number of cycles to failure and reduced the damage rate for composite specimen with teflon s pacer, which consisted of dense-graded mixture on dense-graded mi xture, whereas no difference was observed for specimens without teflon spacer. For reflective cracking performance, higher application rate of Novabond showed larger number of cycles to failure and lo wer damage rate for composite specimens consisting of double Novachip on Portland cement concrete. For the effects of OGFC on top-down cra cking, as compared with all dense-graded mixture specimens, composite specimens with OGFC on top of dense-graded mixture exhibited less number of cycles to failure and higher damage rate. After comprehensive evaluation of the effect s of different layer interface conditions on top-down and reflective cracking, and the effect of OGFC on topdown cracking, the following conclusions can be drawn, The CSIC test method developed in this study can be used to evaluate the relative effects of pavement layer interface c haracteristics on cracking performance. Polymer modified asphalt emulsion, like Novabond, increases the top-down cracking performance of composite specimen by increasing the cracking resistance of materials near the in terface and by dissipating the stress accumulated at the interface. It should be pointed out that the effectiveness of interface conditions is closely related to its brittleness. Trackless tack had a negative effect on the topdown cracking performance. Polymer modified asphalt emulsion, like Novabond, increases the reflective cracking performance of composite specimen when compared with conventional 146

PAGE 147

tack coat. Novabond interface application rate plays an important role on the cracking resistance. OGFC applied on top of dense-graded mixt ure to some extent reduces the cracking resistance of composite specimen with conventional tack coat interface as compared with specimens without OGFC. In order to maintain the cracking performance of pavements with OGFC over lay, polymer modified asphalt emulsion needs to be applied at the interface. 7.2 Recommendations Based on the studies completed, the fo llowing items are recommended for further research: Experimental road test on the effects of interface conditions on both top-down and reflective cracking performance needs to be conducted and evaluated. The effects of interface conditions on cr acking performance for different composite specimen combination, including both top and lower layer material type and gradation, need to be evaluated. Optimum interface condition, i.e., application rate, needs to be identified. A thorough analysis of the total recoverabl e deformation versus time curve should be performed to identify the crack in itiation and propagation stages in the specimen and thus the fracture resist ance of OGFC. It can be used as an approach to measure the properties of thin OGFC or other thin layers within the composite specimen. Besides the cracking performance of interfac e as stated in this study, the interface bonding condition between pavement layers al so has an important effect on the pavement performance. Literature review indicates that a new bond test (under repeated loading) needs to be developed ba sed on the test method identified in this study. Age hardening has a significant effect on the cracking performance of asphalt concrete mixture and the interface mate rials. An age conditioning procedure needs to be developed for composite s pecimen, including heat, ultraviolet and water. Laboratory aged composite specimen properties can be compared with field coring composite specimen. 147

PAGE 148

APPENDIX A THE AMOUNT OF INTERFAC E MATERIAL CALCULATION Cross section area of the SGC specimen: 227.5176.71Ar cm2 2 22 3 232318361.27;13785.41; 1/3785.41/8361.270.453/. yardcmgalloncm gallonydcmcmcmcm The amount of asphalt emulsion applied, 0.453176.71.bmassapplicationrate P Specific gravity of asphalt emuls ion is around 1. 0. For instance, the application rate is for open-graded friction course overlay according to 2010 FDOT Standard Specif ications for Road and Bridge Construction Section 300. bP2yd/045.0gallon The amount of asphalt emulsion applied, 80.050.0451.03.6. massgrams If paving grade binder is used, it should be adjusted from the amount of asphalt emulsion corresponding to the residue rate. 148

PAGE 149

149 APPENDIX B THE AMOUNT OF OVERLAY MATERIAL CALCULATION Assume compacted overlay materials like open-graded mixtures, height is air void content is maximum specific gravity is mass of overlay mixture is Gyratory compaction mold inner diam eter is D, base material height is overlayHbaseHoverlayACoverlayGmmoverlayM2(/2)(1).overlayasphalaggregate overlay overlay overlay overlayMVGmm DHACGmm This compaction is based on the compacted specimen height, but not the number of gyrations. overlaybaseHH

PAGE 150

LIST OF REFERENCES Abd El-Naby, R.M., and Easa, S.M. (2002). Evaluation of the bond characteristics between as phalt concrete layers Annual Conference of the Canadian Society for Civil Eng ineering Ameri-Gaznon, M., and Little, D. (1990). Octahedral shear stress analyses on an ACP overlay on a rigid base, Proceedings of the Association of Asphalt Paving Technologists 59, 443-479. ASTM (2003). ASTM D 8-02 Standard terminology relating to materials for roads and pavements, Annual Book of ASTM Standar ds 2003, Section 4: Construction, Volume 04.03, American Society for Test ing and Materials, West Conshohocken, PA. Bennert, T. (2010). Flexible overlays for ri gid pavements, Report No. FHWA-NJ-2009014, Federal Highway Administ ration, Washington, D.C. Birgisson B., Montepara A., Romeo E., Roque R., Roncella and Tebaldi G. (2009). An optical strain measurement system for asphalt mixtures, Materials and Structures 42(4), 427-441. Birgisson, B., Roque, R., Var adhan, A., Thai, T., and Jaiswal, L. (2006). Evaluation of Thick Open Graded and Bonded Fr iction Courses for Florida, Final Report of Florida Department of Transportation, Univ ersity of Florida, Gainesville, FL. Bozkurt, D., and Buttlar, W. (2002). Three dimensional fi nite element modeling to evaluate benefits of interlayer stress absorbing composite for reflective crack mitigation, Federal Aviation Administration Airport Technology Transfer Conference Brown S.F., Thom N.H., and Sanders P.J. (2001). A study of grid reinforced asphalt to combat reflection cracking, Annual Meeting of Associ ation of Asphalt Paving Technologists 543. Buchanan, M.S. and Woods, M. E. (2004). Field Tack Coat Evaluator (ATACKer), Report No. FHWA/MS-DOT-RD-04-168. Missi ssippi Department of Transportation, Jackson, MS. Canestrari F, Ferrotti G, Partl MN, and Sa ntagata F. (2005). Advanced testing and characterization of interlayer shear resistance, Transport Research Record 1929 68. Canestrari, F. and Santagata, E. (2005). Temperature effect s on the shear behaviour of tack coat emulsions used in flexible pavements, I nternational Journal of Pavement Engineering 6(1), 39-46. 150

PAGE 151

Chaignon F, and Roffe J-C. ( 2001). Characterisation tests on bond coats: worldwide study, impact, tests, recommendations, International Bit umen Emulsion Federation (IBEF) Chen, Jian Shiuh 1, Huang, and Chien Chung (2008). Mechanical properties of tack coat on asphalt concrete interfaces, Journal of the Chinese Institute of Civil and Hydraulic Engineering 20(1), 57-68. Chen, Dar-Hao. (2010). Slippage failure of a new hot-mix asphalt overlay, Journal of Performance of Constructed Facilities 24(3), 258-264. Choi YK, Sutanto MH, Collop AC, and Airey GD. (2005). Bond between Asphalt Layers, Project report to the UK high ways agency. Nottingham, UK: Scott Wilson Pavement Engineering Ltd. Chowdhury, A., J.W. Button and R.L. Lytton. (2009). Tes ts of HMA overlays using geosynthetics to reduce reflection cr acking, FHWA/TX-10/0-1777-3, Texas Transportation Institute, The Texas A&M University System. Cleveland, G. S., J. W. Button, and R. L. Lytton. (2002). G eosynthetics in flexible and rigid pavement overlay systems to reduc e reflection cracking, Report FHWA/TX02/1777-1, Texas Transportation Institute, The Texas A&M University System. Collop AC, Sutanto MH, Airey GD, and Elli ot AC. (2009). Shear bond strength between asphalt layers for prepared samples and field cores. Construction and Building Materials 23(6), 2251-2258. Collop, A.C, M.H., Sutanto G.D., Ai rey Richard C., and Elliott. (2010). Development of an automatic torque test to measur e the shear bond strength between asphalt Construction and Building Materials doi:10.1016/j.conbuildmat.2010.07.030 Collop, A.C.; Thom, N.H. and Sangiorgi, C. (2003). Assessment of bond condition using the leutner shear test, Proceedings of the Instit uti on of Civil Engineers: Transport 156(4), 211-217. Cooley, L.A. (1999). No-tack inlay on milled surface: project r eport, NCAT Report No. 99-2. Cooper, S. M. and L. N. M ohammad. (2004). NovachipTM surface treatment six year evaluation, State Projec t No. 407-04-0034, Louisiana Transportation Research Center. Cross, S.A., and P.P. Shrest ha. (2005). Guidelines for us ing prime and tack coats, Report FHWA-CFL/TD-02-002, Central Feder al Lands Highway Division, Federal Highway Administration Lakewood, Colorado. 151

PAGE 152

Dauzats, M. and Rampal, A. (1987). Mechanism of surface cracki ng in wearing courses, Proceedings, 6th International Conferenc e Structural Design of Asphalt Pavements 232-247. De Freitas, E. F., Pereira, P., and PicadoSantos, L. (2003). Asse ssement of top-down cracking causes in asphalt pavements, 3th International Symposium on Maintenance and Rehabilitation of Pavements and Technical Control Guimaraes, Portugal. Deysarkar, I. (2004). Test set-up to determine quality of tack coat, Masters Thesis, Department of Civil Engineer ing, The University of Texas at El Paso. Deysarkar, I. and Tandon, V. (2004). Developm ent of an objective field test to determine tack coat adequacy, Report No. 0-4129-1F, Center for Transportation Infrastructure Systems, The University of Texas at El Paso, El Paso, Texas. Deysarkar, I. and Tandon, V. (2005). Field ev aluation of tack coat quality measurement equipments, International Journal of Pavement Engineering 4(1-2), 25-37. Diakhate, M., Millien, A., Petit, C, Phe lipot, A., Pouteau, B. (2011). Experimental investigation of tack coat fatigue per formance: towards an improved lifetime assessment of pavement structure interfaces, Construction and Building Materials 25(2), 1123-1133. Diakh ate, M., Phelipot, A., M illien, A., Petit, C. (2006). S hear fatigue behaviour of tack coats in pavements, Road Materials and Pavement Design 7(2), 201. Emery, J.J. (2006). Evaluat ion and mitigation of asphalt pavement top-down cracking, Annual Conference of the Transportation Association of Canada Charlottetown, Prince Edward Island, Canada. Flexible Pavement of Ohio (2001). Proper Tack Coat App lication, Technical Bulletin, Flexible Pavement of Ohio, Columbus, OH. Gerritsen, A.H., van Gurp, C.A.P.M., van der Heide, J.P.J., Molenaar, A.A.A. and Pronk, A.C. (1987). Predicti on and prevention of surface cracking in asphaltic pavements, Proceedings, 6th International C onference Structural Design of Asphalt Pavements The University of Michigan. Ann Arbor, Michigan, 378-391. Gomba, S., Liddle, J., and Meht a, Y. A. (2005). Evaluation of interlayer bonding in hot mix asphalt pavements, The International Journal of Pavements 4(1-2),13-24. Groenendijk, J. (1998). Accelerated testing and surface cracking of asphaltic concrete pavements, A PhD dissertati on. Department of Civil Engi neering, Delft University of Technology, the Netherlands. 152

PAGE 153

Hachiy a, Y., and Sato, K. (1997). Effect of tack coat on bonding characteristics at interface between asphalt concrete layers, Proceedings of 8t h International Conference on Asphalt Pavements 1, 349-362. Harvey, J.T., du Plessis, L., Long, F. Deacon, J.A., Guada, I. Hung, D. and Scheffy, C. (1997). CAL/APT Program: Test result s from accelerated pavement test on pavement structure containing asphalt treated permeable base (ATPB) section 500RF, Report No. RTA65W485-3. Pavement Research Center, University of California at Berkeley, r eport for the California Department of Transportation, Sacramento, CA. Hot Mix Asphalt Paving Handbook. (2000). US Army Corps of Engineer. Hu, Xiaodi and Walubita, F. Lubinda. (2010). Effects of layer interfacial bonding conditions on the mechanistic responses in asphalt pavements, Journal of Transportation Engineering Hugo, F and Kennedy, T.W. (1985). Surface cracking of asphalt mixtures in southern africa, Proceedings of the Association of Asphalt Paving Technologists 54. Jacobs, M.M.J. (1995). Cracking in Asphalt Mixes, PhD, The Delft University of Technology, The Netherlands. Jones, R.M. (1980). Mechanics of Co mposite Materials, McGraw-Hill. Kandhal, P.S. and Lockett, L. (1997). Cons truction and Performance of Ultrathin Asphalt Friction Course, Report No. 975, The National Center for Asphalt Technology (NCAT). Kandhal, P. S., and Mallick, R. B. (1998). Open Graded Fricti on Course: State of the Practice, NCAT Report 98-7, National Center for Asphalt Technology, Auburn, AL. Khodaii, A., Fallah, S., Nej ad, F.M. (2009). Effects of geosynthetics on reduction of reflection cracking in asphalt overlays, Geotextiles and Geomembranes 27, 1-8. Kim, K. W., Doh, Y. S. and Lim, S. (1999). Mode I refl ection cracking resistance of strengthened asphalt concretes, Construction and Building Materials 13(5), 243 251. Koh, C. (2009). Tensile Proper ties of Open Graded Friction C ourse (OFGC) Mixture to Evaluate Top-Down Cracking Performance, Ph.D. Dissertation, University of Florida, Gainesville, FL. Kruntcheva, M., Collop, A.C. and Thom, N. (2005). Effect of bond condition on flexible pavement performance, Journal of Transportation Engineering 131(11), 880-888. 153

PAGE 154

Kruntcheva, M.R., Collop, A. C. and Thom, N.H. (2006). Pr operties of asphalt concrete layer interfaces, Journal of Materials in Civ il Engineering 18(3), 467-471. Leng, Z., Al-Qadi, I.L., Carpent er, S. and Ozer, H. (2009). Interface bonding between hot-mix asphalt and various portland cement concrete surfacesassessment of accelerated pavement testing and measur ement of interface strain, In Transportation Research Record: Journal of the Transport ation Research Board TRB, National Research Council, Washington, D.C. Leng, Z., Al-Qadi, I.L., Ozer, H. and Carpent er, S. (2008). Inter face bonding between hot-Mix asphalt and various portland cement concrete surfaces: laboratory assessment, In Transportation Research Record: Journal of the Transportation Research Board TRB, National Research Council, Washington, D.C. Leng, Z., Al-Qadi, I.L., Ozer, H. and Carpenter S. (2008). Tack coat application rate optimization and environmental effect on HMA overlay-PCC pavement bonding, Proceedings of the Inter national ISAP Symposium on Asphalt Pavements and Environment Zurich, Switzerland. 825-835. Mark A. Russell, Linda M. Pierce, Jeff S. Uhlmeyer, and Keith W. Anderson. (2008). NovaChip, Washington State Department of Transportation Materials Laboratory, WA-RD 697.1. Matsuno, S., and Nishizawa, T. (1992). Mechanism of longitudi nal surface cracking in asphalt pavement, 7th International Conference on Asphalt Pavements The University of Nottingham, 2, 277-291. McGhee, K.K. and T. Clark. (2009). Bond expec tations for milled surfaces and typical tack coat materials used in virginia, VTRC 09-R21. Virginia Transportation Research Council, Charlottesville, VA. Mohammad, L.N., Bae, A.I., Elseifi, M.A., Button, J. and Scherocman, J.A. (2009). Interface shear strength characte ristics of emulsified tack coats, Journal of the Association of Asphalt Paving Technologists 78, 249-278. Mohammad, L. N., Raqib, M. A., and Huang, B. (2002). Influence of asphalt tack coat materials on interface shear strength, Transportation Research Record 1789 5665. Mohammad, L., S. Saadeh, Yan Qi, J. Butto n,and J. Scherocman. (2008). Worldwide state of practice on the use of tack coats: a survey, Paper presented at Government and Industry Forum, Association of Asphalt Paving Technologists (AAPT) Meeting Philadelphia, PA. Mohammad, L.N., Zhong, W. an d Raqib, M.A. (2005). Inves tigation of the behavior of asphalt tack interface layer, FHWA /LA.04/394. Louisiana Transportation Research Center, Baton Rouge, LA. 154

PAGE 155

Molenaar, A.A.A. (1984). Fatigue and reflective cracking due to traffic (with discussion), Proceedings of the Association of Asphalt Paving Tec hnologists 53, 440-474. Mrawira, D., Damude, D.J. ( 1999). Revisiting the effectiveness of tack coats in HMA overlays: The shear strength of tack coats in young overlays, Proceedings 14th Annual Conference Canadian Technical Asphalt Association, 116-129. Mrawira, D. and D. Yin. (2006). Field evaluat ion of the effectiveness of tack coats in hot-mix asphalt paving, Transportation Research Board Annual Meeting Washington, D.C. MTAG (2007). Maintenance technical advis ory guide (MTAG), Volume 1 flexible pavement preservation, Second Edition Tr aining Modules. California Department of Transportation, Office of Pavement Preservation, Division of Maintenance. Muench, S.T. and Moomaw, T. 4 (2009). De-bonding of hot mix asphalt pavements in washington state: an initial investigation, WA-RD 712.1 and TNW 2008-10. Washington State Department of Trans portation (WSDOT) and Transportation Northwest Regional Center X (TransNow), Seattle, WA. Mukhtar, M. T., and Dempsey, B. J. ( 1996) Interlayer stress absorbing composite (ISAC) for mitigating reflection cracking in asphalt concrete overlays, Final Report Project IHR-533, Illinois Cooperativ e Highway Research Program. Myers, L. A. (2000). Development and propagat ion of surface-initiated longitudinal wheel path cracks in flexible highway pavements, PhD Thesis, University of Florida. NCHRP (2000). NCHRP Synthesis of Hig hway Practice 284, Performance characteristics of open-graded friction cour ses, Transportation Research Board, National Research Council, Washington D.C. Nesnas, K., Nunn, M. (2004). A model for top-down reflec tion cracking in composite pavements, Fifth International RILEM Confer ence on Reflective Cracking in Pavements Limoges, France. Nunn, M. (1998). Design of l ong-Life roads for heavy traffic, Australian Asphalt Pavement Association Industry Conference Surfers Paradise, Queensland, Australia. Pickett, D.L. and R.L. Lytton. (1983). Laboratory evaluation of selected fabrics for reinforcement of asphalt concrete overlays, Report No. FHWA-TX-84+261-1. Texas Transportation Institute, College Station, Texas. 155

PAGE 156

Ponniah J, Vacin O, Kennepohl G (2006). Effectiveness of tack coat to improve bonding between asphalt layers: laboratory evaluation, Proceedings of the FiftyFirst Annual Conferenc e of the Canadian Technical Asphalt Association (CTAA) 249-260. Raab C., and Partl M.N. (2004). Effect of tack coats on interlayer shear bond of pavements, Proceedings of the 8th Confer ence on As phalt Pavements for Southern Africa Raab C., and Partl M.N. (2008). Investigation into a long-te rm interlayer bonding of asphalt pavements, the Baltic Journal of Road and Bridge Engineering 3(2), 65 70. Raju, S. Kumar, S. S., Reddy, K. S., Bose S ., and Pandey, B. B. (2008). Analysis of top-down cracking behavior of asphalt pav ements, Transportation Research Board, National Research Council, Washington, D. C. Roberts, F., P. Kandhal, E. Brown, D. Lee, and T. Kennedy. (1991). Hot mix asphalt materials, mixture design, and cons truction, NAPA Education Foundation, Lanham, Maryland. Romanoschi, S.A. and Metcalf, J.B. (2001). Characterization of asphalt concrete layer interfaces, Transportation Research Record 1778, 56-65. Roque, R., Hardee, H., and Ruth, B. E. ( 1988). Thermal rippling of asphalt concrete pavements, Journal of the Association of Asphalt Paving Technologists 57,464483. Roque, R. and Ruth, B. E. (1990). Mechani sms and modeling of surface cracking in asphalt pavements, Journal of the Association of Asphalt Paving Technologists 59, 396-421. Rowe, G.M., Sauber, R., Fee, F. and Soliman, N. (2001). Development of long-life overlays for existing pavement infrastructu re projects with surface cracking in new jersey, Transportation Research Board/National Research Council, Transportation Research Circular 503, 96-107. Ruranika, M. M. and J. Geib. (2007). Per formance of ultra-thin bounded wearing course (UTBWC) surface treatment on US-169 Princeton, MN, Report No. MN/RC-2007-18, Minnesota Depart ment of Transportation. Santagata, F.A., Ferrotti, G., Partl, M. N., C anestrari, F. (2008). Sta tistical investigation of two different interlay er shear test methods, Materials and Structures 42(2), 705-714. Santagata, F.A., Partl, M. N. Ferrotti, G., Canestrari, F ., Flisch, A. (2008). Layer characteristics affecting interlayer s hear resistance in flexible pavements, Journal of the AAPT, 77, 221-254. 156

PAGE 157

Santucci, L. E. (2009). Recent findings on the use of tack coat between pavement layers, Pavem ent Technology Update Institute of Transportation Studies, University of California, Berkeley, CA; 1(1). Schorsch, M., Chnag, C-M., Baladi G.Y. (2003). Effects of s egregation of the initiation and propagation of top-down cracks, Paper presented in Transportation Research Board Annual Meet ing, January 12-16. Shahin, M. Y., Kirchner, K., Blackmon, E.W. and Tomita, H. (1986). Effect of layer slippage on performance of asphalt-concrete pavements, Transportation Research Record 1095 79-85. Sholar, G.A.; Page, G.C.; Muss elman, J.A.; Upshaw, P.B. and Moseley, H.L. (2002). Preliminary investigation of a test met hod to evaluate bond strength of bituminous tack coats, Research Report FL/DOT /SMO/02-459. Florida Department of Transportation, Gainesville, FL. Sutanto, Muslich. (2010). Assessment of bond between asphalt layers, PhD thesis, University of Nottingham. Svasdisant, T., Schorsch, M., Baladi G. Y. and Pinyosunun, S. (2002). Mechanistic analysis of top-down cracking in asphalt pavement, Transportation Research Record 1809 National Research Council, Nati onal Academy Press, Washington, D. C., 126-136. Tack Coat Guidelines (2006). Tack coat guidelines, California Department of Transportation, Division of Construction, Office of Construction Engineering, Sacramento, CA. Tashman, L., Kitae, N. and Papagiannakis, T. (2006). Evaluation of the influence of tack coat construction factors on the bond strength between pavement layers, WA-RD 645.1, Washington State Department of Trans portation, Olympia, WA. Tashman, L., K.Nam, and A.T.Papagiannakis et al. (2008) E valuation of construction practices that influence the bond strengt h at the interface between pavement layers, ASCE Journal of Performance of Constructed Facilities, 22(3), 154-161. Tayebali AA, Rahman MS, Kulkarni MB and Xu Q. (2004). A mechanistic approach to evaluate contribution of prime and tack coat in composite asphalt pavements, Research Report 2001-04, Department of Civil Engineer ing, North Carolina State University, USA. Texas Technical Advisory (2001). Proper use of tack coat, Technical Advisory 2001-1, Construction and Bridge Divisions, Texas Department of Transpo rtation, Austin, TX. 157

PAGE 158

Tschegg, K.E., Kroyer, G., Tan, D .M., Tschegg, S.E.S. and Litzka, J. (1995). investigation of bonding between asphal t layers on road construction, Journal of Transportation Engineering 309. Tschegg, E. K., Macht, J., Jamek, M., and Steigenberger, J. (2007). Mechanical and fracture-mechanical properties of asphalt-concrete interfaces, ACI Materials Journal, 104, 474. Uhlmeyer, J.S., Willoughby, K., Pierce, L. M. and Mahoney, J. P. (2000). Top-down cracking Washington state asphalt concrete wearing courses, Transportation Research Record 1730 National Research Council, National Academy Press, Washington, D. C., 110-116. UK Guidelines (2000). Guidelines document fo r the assessment and certification of thin surfacing systems for highways, Worki ng Draft SG3/98/173, British Board of Agreement, Watford, UK. Uzan, J., Livneh, M., and Eshed, Y. (1978) Investigation of adhesion properties between asphaltic-concrete layers, Proceedings of the Association of Asphalt Paving Technologists 47, 495-521. Von Quintus, Mallela, and Lytton. (2010). T echniques for mitigation of reflective cracks, FAA Worldwide Airport Technology Transfer Conference P10067. Atlantic City, New Jersey. Walsh, I.D., and J.T. William s, (2001). HAPAS certificates for procurement of thin surfacing, Highways and Transportation Wambura, J. H. Mania, J., and Smith, H. R. (1999). Ken ya bituminous materials study, Transportation Research Record 1681 National Research Council, National Academy Press, Washington, D.C., 129-137. Wheat, M. (2007). Evaluatio n of bond strength at asphalt interfaces, M.S Thesis, Kansas State University, Manhattan, KS. West, R.C.; Zhang, J. and Moore, J. (2005). Evaluati on of bond strength between pavement layers, NCAT Report 05-08. Nati onal Center for Asphalt Technology, Auburn, AL. Willis, J.R. and Timm, D.H. (2006) Forensic investigation of a rich-bottom pavement, NCAT Report 06-04. Nationa l Center for Asphalt Technology, Auburn, AL. Witczak, W.M., Kaloush, K., Pellinen, T., El-Basyouny, M., and Von Quintus, H. (2000). Simple performance test for Super pave mix design, NCHRP Report 465, National Research Council, Transportati on Research Board, Washington, DC. 158

PAGE 159

Yang, J.; P eng, D., Yu, L., Sumin, G., Liu, Q. and Qian, G. ( 2007). Bond condition between surface layer and immediate layer in porous asphalt pavement, Journal of Southeast University 23(4), 588-593. Zhou, F., and Scullion, T. (2004). Overlay tester: a rapid performance related crack resistance test, Report No. FHWA/T X-05/0-4467-2, Texas Transportation Institute. Ziari, H, and Khabiri.M.M. (2007) Interface condition influenc e on prediction of flexible pavement life, Journal of Civil Engi neering and Management 13(1), 71. 159

PAGE 160

160 BIOGRAPHICAL SKETCH Yu Chen was born in Siyang, Jiangsu Prov ince, Peoples Republic of China in 1981. He received a Bachelor of Science degree in civil engineering from ChangAn University in 2003. In August 2003, Yu Chen started a Master of Engineering program in civil engineering at ChangAn. After finishing his masters degr ee, Yu Chen came to the United States in 2006. He joined the Ph.D program of the ma terials group at the University of Florida and worked as a graduate research assistant with his doctoral advisor, Dr. Reynaldo Roque. After completi ng his Ph.D., he plans to work in academia, government agencies, or private companies in civil engineering.