<%BANNER%>

Validation of stresses caused by thermal gradients in segmental concrete bridges

HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Abstract
 Introduction
 Background
 Test beam
 Beam construction
 Instrumentation
 Methodology of imposing thermal...
 CTE tests
 Conclusions and recommendation...
 Appendices
 References
 Biographical sketch
University of Florida Institutional Repository

PAGE 1

VALIDATION OF STRESSES CAUSED BY THERMAL GRADIENTS IN SEGMENTAL CONCRETE BRIDGES: CO NSTRUCTION, TEST SETUP, AND COEFFICIENT OF THERMAL EXPANSION TESTS By DAVID CLANCY WALTER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 By David Clancy Walter

PAGE 3

iii ACKNOWLEDGMENTS I thank my supervisory committee chairs, Dr. H.R. Hamilton III and Dr. Gary R. Consolazio, for their guidance th roughout this research and in my development as an engineer. I would like to tha nk Dr. Ronald A. Cook for his help and support on my thesis committee. I would also like to thank my research partne r, Farouque Mahama, for his help with and extensive knowle dge on this research project. Very special thanks go to the Florida Department of Transportation (FDOT) Structures Lab personnel, especially Marc Ansley and Frank Cobb. Without their help with the construction of the T-beam, this re search project would not be possible. The same thanks go to the University of Florida Structures Lab personnel, especially Charles Broward, Hubert “Nard” Martin and John Gam ache. Not only did they devote countless hours to helping with the research project, they also provided gui dance, relief and an enjoyable work environment. I would also like to thank the FDOT Stat e Materials Office, especially Richard DeLorenzo, for assisting with and performi ng concrete cylinder tests. I thank DYWIDAG Systems International for donating all of the prestressing bars, nuts and plates. Most importantly, I thank my mom, dad, brot her, sisters, nephew and niece, and my college friends along the way. Their support, love and humor brightened my life during my years as a student.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT....................................................................................................................... xi CHAPTER 1 INTRODUCTION........................................................................................................1 2 BACKGROUND..........................................................................................................3 Thermal Gradients........................................................................................................3 Design Thermal Gradients............................................................................................5 Concrete Beam Behavior under Thermal Loading.......................................................5 3 TEST BEAM..............................................................................................................10 Selection of Cross-Section Geometry.........................................................................10 Segmental Construction Details.................................................................................12 Prestressing.................................................................................................................13 Steel Reinforcement....................................................................................................14 Load Test Setup..........................................................................................................14 Segment Heating System............................................................................................15 4 BEAM CONSTRUCTION.........................................................................................28 Forms.......................................................................................................................... 28 Concrete......................................................................................................................2 8 Delivery, Storage and Prestressing.............................................................................29 5 INSTRUMENTATION..............................................................................................37 Load Cells...................................................................................................................37 Deflections..................................................................................................................37 Joint Opening..............................................................................................................38 Strain Rings.........................................................................................................38

PAGE 5

v Foil Gages............................................................................................................39 Displacement Devices.........................................................................................40 Thermocouples and Inline Temperature Sensors.......................................................40 Final Instrumentation Setup........................................................................................41 6 METHODOLOGY OF IMPOSING THERMAL PROFILES...................................54 Uniform Profile...........................................................................................................54 Linear Profile..............................................................................................................55 AASHTO Positive Thermal Gradient.........................................................................56 AASHTO Negative Thermal Gradient.......................................................................57 7 CTE TESTS................................................................................................................65 AASHTO Standard Test Method................................................................................65 In-situ CTE Test Method............................................................................................67 Positive Uniform Temperature Profile................................................................68 Linearly Increasing Temperature Profile.............................................................68 Comparison of CTE data.....................................................................................69 8 CONCLUSIONS AND RECOMMENDATIONS.....................................................74 APPENDIX A TEST BEAM SHOP DRAWINGS............................................................................76 B FRAME DRAWINGS................................................................................................80 C CONCRETE MIX TICKETS.....................................................................................84 LIST OF REFERENCES...................................................................................................87 BIOGRAPHICAL SKETCH.............................................................................................88

PAGE 6

vi LIST OF TABLES Table page 1 Basis for temperature gradients..............................................................................6 2 Approximate service-I stresses in SRB Bridge (psi)............................................19 3 Dates of concrete pours.........................................................................................33 4 Concrete pump mix proportions...........................................................................34 5 Results from compression tests.............................................................................34 6 Moduli of elasticity for each segment...................................................................34 7 Prestress load increments for tests performed on July 2, 2006.............................35 8 Prestress load increments for tests performed on July 5, 2006.............................35 9 Inventory for available data acquisition channels.................................................50 10 Absolute temperatures to achieve +41 F profile based on initial temperatures....60 11 Absolute temperatures to achieve linear profile...................................................61 12 Absolute temperatures to achieve AASHTO positive gradient............................63 13 Absolute temperatures to achieve AASHTO negative gradient...........................64 14 Cylinder lengths at room temperature...................................................................70 15 Absolute temperatures and temperat ure changes for AASHTO CTE test............71 16 Recorded displacements from LVDT's.................................................................71 17 Calculated strains..................................................................................................71 18 Calculated CTE values..........................................................................................71 19 Comparison of CTE values...................................................................................73

PAGE 7

vii LIST OF FIGURES Figure page 1 Positive vertical temp erature gradient....................................................................6 2 Solar radiation zones for the United States............................................................7 3 Positive thermal gradient for Florida......................................................................7 4 Negative thermal grad ient for Florida.....................................................................8 5 Decomposition of a nonlinear thermal gradient......................................................8 6 Strain difference that leads to self-equilibrating stresses........................................9 7 Cross section of SRB bridge................................................................................17 8 I-section representation of SR B bridge cross section...........................................18 9 Self-equilibrating stresses due to AASHTO positive thermal gradient................18 10 Self-equilibrating stresses due to AASHTO negative thermal gradient...............18 11 Chosen shape for final test beam with self-equilibra ting stresses........................19 12 Schematic of segmental beam test specimen........................................................19 13 Prestress assembly cross section view..................................................................19 14 View of prestess assembly....................................................................................20 15 Prestress assembly in elevation view....................................................................20 16 Elevation views of prestress assembly..................................................................21 17 Mild steel reinforcements in Segments 1 and 4....................................................21 18 Mild steel reinforcement.......................................................................................21 19 Frame locations.....................................................................................................22 20 Mid-support frame a.) drawing, b.) picture...........................................................22

PAGE 8

viii 21 Piping schematics for a) protot ype beam, b) test beam........................................23 22 Pipe spacing and thermal gradients for positive thermal grad ient and negative thermal gradient....................................................................................................23 23 Typical manifold in flange...................................................................................24 24 Typical manifold in web......................................................................................24 25 Flow rates for web manifold.................................................................................24 26 Flow rates for flange manifold s with straight inlet pipe.......................................25 27 Modified manifold for the flange..........................................................................25 28 Flow rates for modifi ed web manifold..................................................................25 29 Heating systems....................................................................................................26 30 Piping schematics for uniform temperature rise...................................................26 31 Piping schematics for AASHTO positive thermal gradient..................................26 32 Piping schematics for AASHTO negative thermal gradient.................................27 33 Beam layout with casting sequence......................................................................30 34 Open form with steel reinforcement.....................................................................31 35 Closed form with reinforcing and lifting hooks....................................................31 36 Form for shear keys..............................................................................................32 37 Thermal segment with copper piping and thermocouples....................................32 38 Finished concrete pours for a) Segment 1, b) Segment 2, c) Segment 3, d) Segment 4..............................................................................................................33 39 Positioning a hydraulic cylinder on the top bar....................................................35 40 Prestressed beam...................................................................................................36 41 Load cell layout for final beam.............................................................................42 42 Load cell layout for prestressing...........................................................................42 43 Cross section view of prestress load cells.............................................................42 44 Cross section of rectangular test beam..................................................................43

PAGE 9

ix 45 Load setup for small beam tests............................................................................43 46 Strain ring on mounting blocks across joint.........................................................43 47 Modified mounts for strain rings..........................................................................44 48 Load vs. longitudinal strain (strain ring)..............................................................44 49 Incremental loading of strain rings.......................................................................45 50 Strain vs. loading for strain rings..........................................................................45 51 Foil strain gauges at segment interface.................................................................46 52 Load vs. longitudinal strain (foil gauges).............................................................46 53 Load vs. tip deflection...........................................................................................46 54 Non-contact displacement trans ducer, LVDT and linear POT...........................47 55 Load vs. relative jo int displacement.....................................................................47 56 Thermal gradients and thermocouple la youts for a) positive thermal gradient, b) negative thermal gradient.................................................................................47 57 Thermocouple grid with attached thermocouples.................................................48 58 Elevation view of segments 2 and 3 al ong with thermocouple grid locations......48 59 Thermocouple labels in Segment 2 and Segment 3..............................................49 60 Elevation view of instrument ation layout (north side).........................................50 61 Elevation view of instrume ntation layout (south side).........................................51 62 Detail view of instrumentation on segments 2 and 3 (north side)........................51 63 Detail view of instrumentation on segments 2 and 3 (south side)........................52 64 Plan view of instrumentation on segments 2 and 3...............................................52 65 Typical instrumentation labe l for displacement device........................................52 66 Typical instrumentation la bel for strain device....................................................53 67 Sections at which thermo couples were embedded................................................58 68 Pipe layers used for thermal profile tests..............................................................58 69 Thermocouples used in each section.....................................................................59

PAGE 10

x 70 Piping schematic for tap water flow.....................................................................59 71 Piping schematic for +41 F uniform profile........................................................60 72 Uniform profile for segment 2.............................................................................60 73 Uniform profile for Segment 3.............................................................................61 74 Piping schematic for linear profile........................................................................61 75 Linear profile for Segment 2.................................................................................62 76 Linear profile for Segment 3.................................................................................62 77 Piping schematic for AASHTO positive gradient................................................62 78 Experimental AASHTO positive gradient............................................................63 79 Piping schematic to heat beam..............................................................................63 80 Piping schematic for AASHTO negative gradient................................................64 81 Experimental AASHTO negative gradient...........................................................64 82 Measuring device for AASHTO CTE tests..........................................................70 83 CTE test setup.......................................................................................................71 84 Longitudinal displacements from LVDT's for Segment 2 for positive uniform temperature profile................................................................................................72 85 Longitudinal displacements from LVDT's for Segment 3 for positive uniform temperature profile................................................................................................72 86 Longitudinal displacements from LVDT's for Segment 2 for linearly increasing temperature profile................................................................................................73 87 Longitudinal displacments from LVDT's for Segment 3 for linearly increasing temperature profile................................................................................................73

PAGE 11

xi Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering VALIDATION OF STRESSES CAUSED BY THERMAL GRADIENTS IN SEGMENTAL CONCRETE BRIDGES: CO NSTRUCTION, TEST SETUP, AND COEFFICIENT OF THERMAL EXPANSION TESTS By David Clancy Walter December 2006 Chair: H. R. Hamilton III Cochair: Gary R. Consolazio Major Department: Civil and Coastal Engineering American Association of State Highway and Transportation ( AASHTO) provisions for checking service limit states require that the stresses from dead and live loads be combined with those caused by heating and cool ing. Of particular concern are the “selfequilibrating” stresses caused by non-linear thermal gradients. The Florida Department of Transportation has found that these stresses in combinati on with dead and live load stresses can control the design and rating of se gmental concrete bridge s. Yet cracking or other indications of distress th at might be associated with these stress combinations have not been noted in regular inspecti ons of bridges in service. This thesis presents the laboratory setup and methodologies implemented to evaluate stresses resulting from thermal gradie nts in segmental concrete bridges. A 20foot long segmental concrete T-beam was c onstructed to determine the actual thermal stresses that occur under a non-linear thermal gradient. The T-shaped cross section was

PAGE 12

xii chosen to simulate the web-fl ange portion of a box girder. Th e test specimen consisted of four 5 ft. segments that were match-cast and then prestressed together. Two of the sections contained multiple layers of copper pipes that carried heated water through the section. By varying the water temperature in each layer of piping, thermal gradients were imposed on the cross section that closel y matched the thermal gradients found in AASHTO guidelines for the State of Florida. Load cells, displacement transducers, strain gauges, strain rings, and thermocouples were us ed to collect data during heating and load testing. In future work, the data from lo ad tests will be analy zed to determine the magnitude of the stresses caused by a combination of heating and loading. An important variable in the project was the concrete’s coefficient of thermal expansion (CTE). This thesis also presents the setup, procedures and results of tests conducted to determine the CTE of the test b eam. The results are compared to CTE tests conducted on concrete cylinders taken from the test beam pour.

PAGE 13

1 CHAPTER 1 INTRODUCTION Stresses caused by heating and cooling mu st be considered in the design of segmental concrete bridges. In general, st resses are generated when the temperature of all or part of the superstructure varies signi ficantly from the temper ature at which it was constructed. Seasonal and diurna l variations in temperature ar e usually the cause of such stress. The nature of the stresses that develop in the superstructure depends on the magnitude of the overall temperature change as well as the magnitude of the temperature gradient over the height of the cross-section. Thermal gradients are created when the cross-section is heated or cool ed unevenly over its height. Stresses are generated when the superstruc ture is restrained by redundant supports or when the thermal gradient is non-linear, which creates “s elf-equilibrating” stresses. Non-linear thermal gradients cause cross-section distortion that is not compatible with the fundamental kinematic assumption that plane sec tions remain plane. Internal stresses are required to enforce this basic kinematic assump tion. Lack of external forces ensures that, to maintain static equilibrium, the stresses are self-equilibrati ng. Of particular interest are the tensile stresses that form when the top of the cross-section is at a lower temperature than the bottom. There are no data in the literature showi ng that the self-equilibr ating stresses have been experimentally verified. Determining the state of stress unde r the application of thermal and mechanical loading is made di fficult because stresses cannot be measured directly.

PAGE 14

2 Additionally, the design analys is of thermal gradients does not account for possible effects of creep. Heating of bridge concrete does not occu r instantaneously, but rather, over a period of hours. The response of the co ncrete to the thermal loading also occurs over hours. During this long period, the high stresses predicted to occur from thermal loading may not actually occur due to creep effects. In an attempt to experimentally meas ure the effect of the non-linear thermal stresses, a 20 ft. segmental T-beam was constructed. The beam will be subjected to thermal loading and mechanical “tip” loading. Loads, deflections and strains will be measured on the beam, and will be analyzed to determine the effects of thermal gradients on segmental concrete bridges. Recommendations for future bridge analysis and design will then be made. This thesis documents the test beam de sign, laboratory setup and instrumentation setup used in this study. Cr oss-section geometry, span length and the plumbing used inside the beam to impose the thermal gradie nts are also document ed. Laboratory setup documentation includes loading frames, reacti on frames, pumps and heaters. The types of load, displacement and strain measuring devices are discussed in the instrumentation section. The thesis also covers in-situ test s performed on the test beam to determine a coefficient of thermal expansion.

PAGE 15

3 CHAPTER 2 BACKGROUND Thermal effects, particularly of thermal gradients, are of great concern to bridge designers. They are believed to cause stre sses in the bridge deck that may lead to unwanted cracks. Current AASHTO guidelines se t forth a procedure to calculate thermal stresses and design for them. These codes ma y be overly-conservative in the calculations of actual thermal stresses. A better understand ing of actual thermal st resses is needed for a more economical design of bridges that will maintain safety and serviceability. This chapter covers the thermal gradient s as set forth by the AASHTO guidelines. Mechanical and thermal loading of the test beam are also addressed. Thermal Gradients Design for thermal effects in concrete segmental bridges is based on AASHTO specifications. The specifications discuss tim e-dependent fluctuations in the bridge temperature and also the temperature differentia ls within the superstructure. All concrete bridges are designed for time-dependent temp erature fluctuations (AASHTO 1989). For prestressed concrete bridges, stresses and movements caused by temperature differentials must also be considered. Stresses and movements do not cause a loss of strength; however, they are significant because they can cause cracks in the bridge which may lead to deterioration or corrosion of the bridge superstructure. The two temperature differentials, or gradients, discussed in the AASHTO provisions are the positive thermal gradient and the negative thermal gradient (AASHTO 2001). These gradients arise through the radi ation, convection and conduction that a box-

PAGE 16

4 girder bridge undergoes throughout the day and throughout the year. Because concrete is a poor thermal conductor, the box-girder crosssection does not heat or cool uniformly over its height, creating thermal gradients. A positive thermal gradient occurs as the sun is rising and heating the top of the bridge cross-section more than the lower porti on. Compressive stresses are formed in the bridge deck due to the higher temperatures in that region. The stresses are due to nonlinearities of the temperature profile and are self-equilibra ting, meaning that no external reaction is required. They will be disc ussed in further detail in this chapter. As the sun sets, the bridge begins to cool. The bridge deck cools to lo wer temperatures than the rest of the superstructure due to outward ra diation of the stored heat. This creates the negative gradient in which the top of the cr oss-section is cooler than the bottom. Designers are concerned with th is gradient because it causes tensile stresses, particularly during the winter season, and possi bly cracks in the bridge deck. Tables and charts are available (AASHT O 2001) that give minimum and maximum effective bridge temperatures and maximum solar radiation based on surface type and geographical location for both the positive and the negative thermal gradients. The figures and tables shown below are used in the analysis procedure to predict expansion and stresses caused from thermal loading. Th e figures and tables are taken from site studies conducted throughout the United States on box-girder bridges. Figure 1, Table 1, and Figure 2 are used collectively to estab lish the positive thermal gradient in a boxgirder bridge for a particular ge ographical locatio n. The values, T1 through T3, are relative temperature changes; they do not re present absolute temperatures in F. The value T3 is taken as 0.0 unless a si te specific study is conduc ted to determine another

PAGE 17

5 value. For the negative thermal gradient, T values are taken as -0.30 or -0.20 times the T values of Table 1. The multiplier of -0.30 is used for plain concrete decks; the multiplier of -0.20 is used for decks with asphalt. The difference accounts for the insulating ability of the asphalt on the bridge deck. Design Thermal Gradients The AASHTO guide (AASHTO 2001) specifies temperature differentials for both the positive and negative gradient based on the zone of the count ry and the type of surface used. The resulting design thermal gr adients for zone 3 (Florida) are shown in Figure 3 and Figure 4. Concrete Beam Behavior under Thermal Loading To facilitate analysis, nonlinear ther mal gradients are divided into three components: uniform temperature, linear thermal gradient and non-linear selfequilibrating temperature grad ient (see Figure 5). The uni form temperature component causes uniform expansion or contraction of th e structure; axial fo rces and moments are formed if the structure is restrained agains t movement. The linear temperature gradient component causes a curvature that forms seconda ry moments if the structure is restrained against bending. The third component of the analysis is the calculation of self-e quilibrating stresses. If an arbitrary unrestrained cross section is subjected to a nonlinear temperature profile, the section fibers will deform in a way that the section does not remain plane. Typically, however, the cross-section is assumed to conform to the Navier-Bernoulli kinematic assumption for beams in whichi plane secti ons remain plane. Enforcement of this kinematic assumption generates strains. The differences between the thermal strains that

PAGE 18

6 would result from free expansion of the fibe rs and the developed strains of the plane section lead to internal self-equi librating stresses (see Figure 6). T1 4" A T28" T3 D epth of S uperS tructure Figure 1. Positive vertical temperature gradient. Table 1. Basis for temperature gradients. Zone T1 (F)T2 (F) 15414 24612 34111 4389

PAGE 19

7 Figure 2. Solar radiation zones for the United States. +41 F +11 F 0 F 4" 12" 20" Figure 3. Positive thermal gradient for Florida.

PAGE 20

8 -12.3 F -3.3 F 0 F 4" 12" 20" Figure 4. Negative thermal gradient for Florida. y zCross-Section Thermal Gradient=++Non-Linear Self-Equilibrating Temperature Distribution Linear Temperature Gradient Uniform Temperature T3T2T1TTopT1 + T2 + T3 = TTopNeutral Axis Figure 5. Decomposition of a nonlinear thermal gradient.

PAGE 21

9 Final (linear) strain distribution Temperature induced strain distribution assuming that the section's fibers do not influence one another T Figure 6. Strain difference that lead s to self-equilibrating stresses.

PAGE 22

10 CHAPTER 3 TEST BEAM This chapter covers the setup of the main te st beam. It addresses the size and shape of the beam, the reaction frame, the loading frame and the mid-support frame. The objective of the test program was to determine whether the stresses formed from the AASHTO positive and negative thermal gradients are as severe as the stresses predicted by AASHTO recommended procedures. To accomplish this objective, a segmental beam was designed that could be test ed in the UF structures laboratory, yet be of sufficient scale and geometry so that its behavior was sim ilar to that of a typical fullscale prestressed segmental concrete bridge. Selection of Cross-Section Geometry The Santa Rosa Bay (SRB) Bridge, located near Pensacola, Florida, was used as a prototype for the design of the length and cros s-section of the test beam (Figure 7). A scaled version of the box girder section was co nsidered for use in laboratory tests, which would also require scaling of the thermal gradient. It was not practical to implement a scaled version of the thermal gradient in lab conditions. The cro ss section of the SRB Bridge can be simplified to that of an I-s ection that has the same area and weak axis flexural stiffness as the box-girder sec tion (see Figure 8). AAS HTO does not stipulate that the codes and analysis procedures are rest ricted to box girders, nor does it restrict the shape to which a gradient is applied. Figure 9 and Figure 10 summarize the results of an analysis on the simplified crosssections for self-equilibrati ng stresses (Mahama). Both of the design positive and

PAGE 23

11 negative thermal gradients for Florida were considered. The negative gradient creates tensile stress in the top of the cross-section, which, when combined with dead and live load stresses from negative bending, could exceed the sum of the precompression and concrete tensile strength, causing cracking. Th e tensile stresses, however, are restricted to the top 4-in. of the crosssection, forming a relatively sharp gradient compared to the cross-section depth. Space and equipment constraints did not allow the full-scale section to be constructed and tested. The primary focus of this research was the tensile stresses created in the top of the section from the negative gradient. Consequently, it was decided to construct a T-beam matching the geometry of top portion of the modified section as illustrated by the hatching in Figure 9 and Figure 10. The figures also show the resulting stress state in the proposed section when the AASHTO temperature gradients are applied to the T-beam section. The important aspects of the stress profile ar e well approximated, including the stress magnitudes and gradient in the top of the section. For both the SRB Bridge and the T-beam, the maximum compre ssive stresses (shown as negative) occurred at the top for the AASHTO positive thermal gr adient. Likewise, the maximum tensile stresses (shown as positive) occurred at the top for the AASHTO negative thermal gradient. This unique approach of using a T-beam is permitted by the fact that the design temperature profile is applied to the top 16-in. of the crosssection and is independent of depth. The width of the T-beam flange was chosen on the basis of recommendations in the ACI Committee 318, Building Code Requirements fo r Structural Concrete and Commentary (ACI 2002) and on the basis of the limited thermal energy available to

PAGE 24

12 impose the gradients on the test beam. In ve ry wide flanged prestressed beams, shear deformations tend to relieve extreme fibers of the flange of longitudinal compressive stress, which leads to a non-uni form distribution of stress. Therefore, the ACI code suggests for design purposes that the width effective as a T-b eam flange shall not exceed one-quarter of the span length of the beam, and the flange width on each side of the web shall not exceed eight times the slab thickness. A width of 2 ft. was chosen to meet these standards and to ensure a uniform longitudi nal stress distribution in the flange. The width was also chosen on the basis of being able to supply enough thermal energy (heated water) to impose the thermal gradients on the test beam. Details of the final test beam cross-section geometry are shown in Figure 11. Segmental Construction Details Imposing the temperature gradients and m easurement of stress were two of the critical considerations when de signing the test setup. To facilitate both of these, the test beam was constructed segmentally using ma tch cast construction techniques as was done on the SRB Bridge. Segment lengths were chosen to match the 5-foot segments of the SRB Bridge and to fit the tie-down spaci ng in the laboratory. Figure 12 shows a schematic of the specimen in the load test setup with prestress force (PT) and tip load (Q). The specimen was composed of four segm ents that were externally prestressed to form a single 20-ft long beam. Segmental construction allowed the piping used to heat and cool the concrete to be easily restricted to segments 2 and 3, which were adjacent to the joint with the largest moment. Furthermore, the joints were left dry, which allowed the joint between heated segments (2 and 3) to open during loading. Incipient joint opening indicates a zero stress state to which strain measurements and anal ytical models can be calibrated. Prestress

PAGE 25

13 force, tip load and span were adjusted to fo rm stress states similar to those found at the mid-span and supports of the SRB Bridge. Prestressing An external post-tensioning (PT) system wa s used to apply pres tressing to the test specimen. The anchorages were suspended from the end segments to allow the PT force location to be adjusted vertical ly relative to the cross-section centroid. The PT force was positioned to develop stresses similar to the st ress conditions found at the mid-span and at the support of the SRB Bridge (see Table 2). The SRB Bridge was designed using AASHTO HS20-44 live load stresses; however, fo r the laboratory tests, live loads taken from the AASHTO HL-93 vehicle loading (LRFD) were used. In Table 2, M/St and M/Sb represent the stresses at the extrem e top and bottom fibers of the section, respectively. An external post-tensioning system was used to avoid problems with concrete void space and interference with the heating syst em and instrumentation. Four 1 3/8 in. diameter high-strength DYWIDAG thread bars were configured as shown in Figure 13 and Figure 14. The PT anchorage was fabricated from stru ctural steel shapes and plates as shown in Figure 15. Steel channels were placed back to back with sufficient space to allow passage of the PT bars. Stiffeners were added to the channels under the bar bearing plates. A set of back-to-back channels was us ed for the top two bars and another pair of channels was used for the bottom bars. Thes e channel systems evenly distributed the PT force over the web of the T-beam. The PT force within the concrete was gradually developed over a few feet from each end of the beam.

PAGE 26

14 Post-tensioning forces were monitored with load cells on each of the four bars (see Figure 15 and Figure 16). Tandem sixty-ton Enerpac Hollow Core single-acting jacks were used to apply the PT force. The jack s were pressurized with a manifold system attached to a single pump, allowing the two bars on opposite sides of the web to be stressed equally avoiding eccentric loading of the prestressing. Steel Reinforcement Steel reinforcement was required to resist th e high PT forces in segments 1 and 4. Number 3 vertical stirrups were placed at 1.75 in. o.c. to resist the principal tensile stresses that developed in th e general anchorage zone. Th is reinforcement was placed within 27 in. from the ends of the beam in Segments 1 and 4. The reinforcement design was based on the approximate method permitte d by LRFD specifications. Outside of the anchorage zone, vertical reinforcement was placed every 12” on center. It was also necessary to design for local zone for ces. AASHTO guidelines (AASHTO 2001) were utilized to determine the local zone stresses that were being develope d in the area that the prestressing channels transferred the force into the concrete. A set of three confinement spirals was used to resist the high anchorage zone stresses. The c onfinement steel for the ends of the beam is shown in Figure 17. Load Test Setup The load test setup was designed so the mechanical loading was reversible; both positive and negative moments could be applied at the joint between segments 2 and 3. Figure 19 shows the test set up layout of the support system. The beam was supported by a reaction frame at one end and a load ce ll support in the middl e. The beam was mechanically loaded at the opposite end (tip ) through the use of a MTS hydraulic jack. The beam was prestressed and then loaded at the tip until a joint opening occurred

PAGE 27

15 between segments 2 and 3. Tests were perf ormed on three different set-ups: tip load only, AASHTO positive thermal gradient wi th tip loading, and AASHTO negative thermal gradient with tip loading. Details of the loading frame, reaction frame and mid-support frame are shown in Appendix B. The loading frame consisted of steel W-section columns and deep channel beams that supported the MTS jack. It wa s designed to handle loads up to 400 kips, although the tip load would not exceed 55 kips. The reaction frame carried the equal and opposite load that was transferred from the jack to the beam. It consisted of a load cell, neoprene bearing pads and back-to-back channe ls that were supported by four all-thread bars; the system was capable of resisting up to 200 kips. It represented a roller because little or no moment or long itudinal resistance was developed from the frame. The mid-support frame is shown in Figure 20; it represents a roller because it allowed for longitudinal movement. As the beam was loaded, the segment interface at the support frame was expected to open up or unpeel from the top of the beam downward. The joint was predicted to open wh en the stress at the top became zero. The beam interface would continue to open, as th e tensile stresses created from the loading moment were greater than the compressive for ces created from the prestress force. The support frame was subjected to loads that were approximately twice the tip load. Segment Heating System Heated water was used to achieve the desired positive and negative thermal gradients. The water was passed through laye rs of copper pipes that were embedded in the beam. The number of pipes in each la yer was minimized to reduce the reinforcing effect and reduction in cross-se ctional properties. The necessa ry number of pipes in each layer was found from experiments conducted on a separate 5-foot prototype beam, which

PAGE 28

16 cross section is shown in Figure 21(a). This beam was heated to match the temperature profile of the AASHTO positive thermal gradient and to match that of the negative thermal gradient. Several comb inations of pipe layers were tested to determine the minimum number of pipes necessary to impose the thermal gradients. The piping schematics shown in Figure 21(b) allowed fo r minimal piping without sacrificing the ability to achieve the AASHTO thermal gradients. Pipe layers 1, 2, and 4 were located near the slope changes of the AASHTO thermal gradients (see Figure 22). The other pipe layers were positioned to aid in heating the entire beam and in shaping the AASHTO gradients. The top layer of pipes was placed as close to the top of the beam as possible while retaining adequate concrete cover, which resulted in a slight non-linearity of the thermal gradient near the surface. Results from thermal gradients tests indica ted that this behavi or was negligible. Manifold systems were designe d to distribute equal flow rates of heated water to each pipe (see Figure 23 and Figure 24) to achieve a uniform temperature distribution across the width of the beam. The smaller web manifolds consisted of a straight inlet and outlet pipes, which allowed for similar flows. The flow rates were checked before the manifolds were cast in the beam (see Figure 25). The same strai ght inlet/outlet pipe approach did not work for the larger flange manifolds, as the greatest flow was about three times the flow rate of the least flow (see Figure 26). Modified manifolds were fabricated to ensure more similarity in flow rates. Figure 28 shows that the tested flow rates were more similar for the modified manifo lds than the flow rates in the straight-inlet manifolds. In addition to testing the flow rates for the manifolds, temperatures were tested on top and bottom of each pipe at the inlet side and outlet side. No temperature

PAGE 29

17 changes were discovered between top and bo ttom and end-to-end. This was important because equal temperature distribution on both top and bottom of the pipe and longitudinally was the desired outcome. The water heaters and pumps were very important components of this project because they supplied the heat necessary to impose the thermal grad ients. Heaters and pumps that could handle temperatures up to 135F were required. Seisco S-H-7 electric heaters allowed high intake temperatures a nd discharged fully heated water nearly instantaneously. The pumps, HP Depco submersible pumps, were able to handle temperatures up to 200F and were able to supply the necessary flow rates to the manifolds. Hoses used in the system were plastic braided flexib le tubing, which could handle large flow rates and extreme temperatures. The pump, heater, and pipe schematic necessary to uniformly increase the crosssection temperature is shown in Figur e 30. Schematics for imposing the AASHTO positive and negative thermal gradient are shown in Figure 31 and Figure 32, respectively. Heaters are labe led 1 through 3. Heaters 1 and 2 are the main heaters that provided high temperatures between 105 F and 140 F. Heater 3 was a supplemental heater that was utilized for temperat ures in the 86 F to 100 F range. 8" 24' 1'-3" 9'-61 2" 8' 14'-111 4" 16' 100 100 53.21 59.26 7" 1'-3" 9'-61 2" Figure 7. Cross section of SRB bridge.

PAGE 30

18 81" 5 4 7 196.8" 26.4" 8" 7" 3 1.5" N.A. Figure 8. I-section representation of SRB bridge cross section. -106 221 330 -622 117 +41 +11 0 4" 12" 60" 547" 196.8" 26.4" 8" 7" 20" -568 118 172 -202 TEMP. GRADIENT (F) SRB (psi) TEST BEAM (psi) 31.5" N.A. Figure 9. Self-equilibrating stresses due to AASHTO positive thermal gradient. 32 -66 -99 1874" 12" 60" 547" 196.8" 26.4" 8" 7" 20" -35 -12.3 -3.3 0 TEMP. GRADIENT (F) SRB (psi) 170 -35 -52 61 TEST BEAM (psi) 31.5" N.A. Figure 10. Self-equilibrating st resses due to AASHTO negative thermal gradient.

PAGE 31

19 24"61 -52 -35 170 20" 8" 4" 12" 10" -99 -66 187 SRB Bridge (truncated at 3 ft.) Proposed Section 36" Self-Equilibrating Stresses for Negative Thermal Gradient (psi) Self-Equilibrating Stresses for Positive Thermal Gradient (psi) -568 -622 118 117 172 330 -202 221 Figure 11. Chosen shape for final test beam with self-equi librating stresses. 3' 5' 5' 5' 3' 2' 2' Q SEGMENT 1SEGMENT 2 (HEATED) SEGMENT 3 (HEATED) SEGMENT 4 PT PT LOCATION OF INTEREST Figure 12. Schematic of segmental beam test specimen. Table 2. Approximate service-I st resses in SRB Bridge (psi). Effective Prestress Dead + Live Load Support Mid-span Support Mid-span P/A M/St M/Sb P/A M/St M/Sb M/St M/Sb M/St M/Sb HS20-44 -700 -175 350 -700 460 -940 810 -1630 -610 1230 HL-93 -700 -175 350 -700 460 -940 810 -1735 -684 1380 PRESTRESS BAR BACK TO BACK C8X18.75 CHANNELSy=14.67" 3 3/8" 3 3/8" 1 1/2" Figure 13. Prestress assembly cross section view.

PAGE 32

20 Figure 14. View of prestess assembly. CHAIR (NOT SHOWN) HYDRAULIC CYLINDER LOAD CELL 3" 1 2" EMBED PLATE 4" 21" C8X18.75 CHANNEL 20' 3' 7' 13 8" DIA. DYWIDAG THREADBAR 1-3 4" PLATE AND NUT FROM DSI 3" PLATE Figure 15. Prestress assembly in elevation view.

PAGE 33

21 Figure 16. Elevation views of prestress assembly. 27" 3 6"# 3 STIRRUPS @1.75" # 3 STIRRUPS @ 12" 5' 24" 8" 28" 10" 9 3 4 12"# 3 BARS 2 1 4 5 1 2 #3 CONFINEMENT SPIRALS 2 1 4 Figure 17. Mild steel reinforcements in Segments 1 and 4. Figure 18. Mild steel reinforcement.

PAGE 34

22 1'-81 4" 2' JACK 16' 20' STRONG FLOOR BOLT LOCATIONS 4' SEGMENT 1SEGMENT 2SEGMENT 3SEGMENT 4 Figure 19. Frame locations. 10" C15x33.9 JOINT OPENING INTERFACE SEGMENT 2SEGMENT 3 (a) (b) Figure 20. Mid-support frame a.) drawing, b.) picture.

PAGE 35

23 24" 6" 10" 3" 24" 8" 10" 36" 3"28" 2" 9" 2" 3" 1" 3" 3" 2" 9" 9" 31 2" 11 2" 11 2" 13 4" 11 2" 18" 3" 11 2" 24" (a) (b) Figure 21. Piping schematics for a) prototype beam, b) test beam. +41 F +11 F 0 F 4" 12" 20" 24" 8" 10" 36" 28" -12.3 F -3.3 F 0 F 4" 12" 20" LAYER 1 LAYER 2 LAYER 3 LAYER 4 LAYER 5 LAYER 6 Figure 22. Pipe spacing and thermal gradients for positive thermal gradient and negative thermal gradient.

PAGE 36

24 5' 2' 13 8" WATER IN FROM HEATERS WATER OUT OF SEGMENT 2 INTO SEGMENT 3 THROUGH FLEXIBLE PLASTIC TUBING WATER RETURN TO TANKS 5' 1 4" I.D.COPPER TUBES SEGMENT 2 SEGMENT 3 Figure 23. Typical manifold in flange. 10" WATER OUT OF SEGMENT 2 INTO SEGMENT 3 THROUGH FLEXIBLE PLASTIC TUBING 5' WATER IN WATER RETURN TO TANKS 5' FROM HEATERS 13 8" 1 4" I.D.COPPER TUBES SEGMENT 2 SEGMENT 3 Figure 24. Typical manifold in web. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 123Flow Rate (GPM)Incoming Flow Figure 25. Flow rates for web manifold.

PAGE 37

25 0 0.1 0.2 0.3 0.4 0.5 0.6 1234567Flow (GPM)Incoming Flow Figure 26. Flow rates for flange manifolds with straight inlet pipe. 5 8" 3" 8 Figure 27. Modified manifold for the flange. 0 0.1 0.2 0.3 0.4 0.5 0.6 1234567Flow (GPM)Incoming H2O Figure 28. Flow rates for modified web manifold.

PAGE 38

26 Figure 29. Heating systems. H1 P1 H2 P2 H3 P3 HEATER 1 IN HEATER 2 IN HEATER 1 OUT HEATER 2 OUT ** *VALVE CLOSED Figure 30. Piping schematics for uniform temperature rise. HEATER 1 IN MIXED HEATER 3 IN H1 P1 H2 P2 H3 P3HEATER 1 OUT MIXED OUT HEATER 3 OUT *VALVE PARTIALLY OPEN TO MIX WATER Figure 31. Piping schematics for AASHTO positive thermal gradient.

PAGE 39

27 HEATER 1 IN MIXED HEATER 3 IN H1 P1 H2 P2 H3 P3HEATER 1 OUT MIXED OUT HEATER 3 OUT HEATER 2 IN HEATER 2 OUT *VALVE PARTIALLY OPEN TO MIX WATER* Figure 32. Piping schematics for AASHTO negative thermal gradient.

PAGE 40

28 CHAPTER 4 BEAM CONSTRUCTION The test beam was constructed at the FDOT Structures Re search Center in Tallahassee. Beam design plans (see Appendi x A) from the UF research team were submitted to the FDOT and construction began during the summer of 2006. All formwork, rebar setup and thermocouple grid construction was performed by the FDOT employees, particularly Frank Cobb. The conc rete pours for all the beam segments also took place in Tallahassee. When the segments were cast and cured, th ey were shipped to the University of Florida Structures Lab. Forms A single set of forms was used to indi vidually cast each segment. The casting sequence is shown in Figure 33. Segmen ts 1 and 4 contained the mild steel reinforcement for the prestress anchorage zone and a steel bearing pl ate to distribute the anchorage force into the concrete (Figur e 34 and Figure 35). The segment joints contained two shear keys to assist in ali gnment when post-tensioning (Figure 36). After Segment 1 and 4 were cast, the thermal segmen ts (Segment 2 and 3) were cast. These segments contained a grid consisting of coppe r pipes, thermocouples and mild reinforcing (see Figure 37). Concrete Figure 38 shows the finished concrete pours fo r all four segments. The dates of the concrete pours are shown in Ta ble 3. A 7000-psi pump mix wa s used to cast each of the four segments. The design mix proportions ar e shown in Table 4 a nd the delivery tickets

PAGE 41

29 for each mix are provided in Appendix C. Alth ough the slump of the concrete is listed as 5 inches, the mix was delivered with a lower water content and slump to allow the slump to be adjusted just prior to placement. Wh en Segment 3 was cast, the weather was rainy and the concrete in the first truck had a slum p that exceeded the specifications. The first truck was rejected, and a second truck with a lower slump concrete was delivered. The segments were cast and cured inside the FDOT research lab. Each segment was cured for a week before it was removed from the form. Fifteen 6-in. diameter cylinders were ta ken from each concrete pour. Compression tests and modulus of elasticity (MOE) tests we re performed on the cylinders at the FDOT State Materials Office (SMO). For each segmen t, three cylinders were tested. The first cylinder of each segment was tested for breaki ng load and stress. The next two cylinders of each segment were instrumented with disp lacement transducers and were loaded to 40% of the breaking load. The recorded disp lacements were converted to strains and an MOE was found. After the cylinders were loaded to 40%, the instrumentation was removed and the cylinder was loaded to failu re. The failure loads are shown in Table 5 and the values of the MOE for each segment are shown in Table 6. Delivery, Storage and Prestressing After all the segments were cast and allo wed to cure in Tallahassee, they were delivered via a large flat bed tr uck to the structures lab at the University of Florida. When they arrived on February 23, 2006, the segments were placed by fork lift on their sides in the lab. They were then instrument ed and prepared for CTE and thermal gradient tests, which are reported in later chapters. When the CTE and thermal gradient tests were completed, the segments were aligned in the load test position shown in Figure 40. Each segment was placed on two

PAGE 42

30 wood frame blocks to ensure stability during tensioning. Th e segments were placed as close to each other as possible to avoid exces sive movement as the joints closed during tensioning. The anchorages and bars were positioned at 1.4 inches below the beam centroid after segment alignment, and the hydr aulic cylinders were positioned to tension the PT bars. During the first prestressing, which occurred on June 2, 2006, the prestress loads were applied slowly to ensure that al l the segments were br ought together and lined up properly. The beam was prestressed in several stages. Firs t, the two hydraulic cylinders were positioned on th e top bars as shown in Figure 39. Through the use of a piping manifold system, the top two bars were prestressed simultaneously until a pressure developed in the hand jack, which indicated a stressing of the bars. The hydraulic cylinders were then positioned on the botto m two bars, and the same procedure was followed. The beam segments came together and aligned successfully with the help of the shear keys. After the beam came together and was aligne d, the bars were loaded in increments that are shown in Table 7. The beam was prestressed ag ain on July 5, 2006; Table 8 shows the load increments for the prestressing procedure. Figure 40 shows the beam after the initial prestressing. SEGMENT 1 December 1, 2005 SEGMENT 2 January 19, 2006 SEGMENT 3 February 2, 2006 SEGMENT 4 December 12, 2005 Figure 33. Beam layout with casting sequence.

PAGE 43

31 Figure 34. Open form with steel reinforcement. Figure 35. Closed form with reinforcing and lifting hooks.

PAGE 44

32 Figure 36. Form for shear keys. Figure 37. Thermal segment with copper piping and thermocouples.

PAGE 45

33 Table 3. Dates of concrete pours. Segment Number Date Cast Segment 1 1 December 2005 Segment 4 12 December 2005 Segment 2 19 January 2006 Segment 3 2 February 2006 (a) (b) (c) (d) Figure 38. Finished concrete pours for a) Segment 1, b) Segment 2, c) Segment 3, d) Segment 4.

PAGE 46

34 Table 4. Concrete pump mix proportions. Mix Number FC82JC Strength (psi) 7000 W/C Ratio 0.31 Slump (in) 5 +/1” Air Content (%) 4.5 +/1.5% Plastic Unit Weight (lbs/cf) 140.1 +/1.5 Material ASTM Type Cement C 150 I/II 820 Cement C 618 F. Ash 160 Water --304 Fine Aggregate C 33 Sand 1095 Aggregate C 33 #89STONE 1400 Admixture C 260 AIR Admixture C 494 W/Reducer Dosage rates vary with manufacturers recommendations Table 5. Results from compression tests. Segment Number Age of Specimen at Time of Test (days) Test Number Breaking Load (lbs) Breaking Stress (psi) 1 263,000 9302 2 266,160 9413 3 255,880 9050 1 77 Average 261,680 9,255 1 200,520 7092 2 196,120 6936 3 199,690 7063 2 28 Average 198,777 7,030 1 224,670 7946 2 209,000 7392 3 215,910 7636 3 28 Average 216,527 7,658 1 257,440 9105 2 248,300 8782 3 243,550 8614 4 66 Average 249,763 8,834 Table 6. Moduli of elastic ity for each segment. Segment Number Modulus of Elasticity (psi) Segment 1 4,862,282 Segment 4 4,575,301 Segment 2 3,835,573 Segment 3 4,000,000

PAGE 47

35 Figure 39. Positioning a hydraulic cylinder on the top bar. Table 7. Prestress load increments for tests performed on July 2, 2006. Load Cell Readings(kips) Load Step L1 L2 L3 L4 1 --9.1 13.9 2 19.3 32.8 9.1 13.9 3 19.3 32.8 53.3 69.8 4 31.4 42.0 53.3 69.8 Final Load 24.7 28.3 51.5 62.6 Table 8. Prestress load increments for tests performed on July 5, 2006 Load Cell Readings(kips) Load Step L1 L2 L3 L4 1 29.2 18 --2 29.2 18 43.8 46.7 3 48.3 32.2 43.8 46.7 4 48.3 32.2 76.3 68.8 5 76.7 68.8 76.3 68.8 6 76.7 68.8 97.6 93.2 7 94.6 100.8 97.6 93.2 8 94.6 100.8 107.7 100.9 9 106.5 112.1 107.7 100.9 10 106.5 112.1 111.8 106.1 Final Loads 97.2 89.2 93.3 92.4

PAGE 48

36 Figure 40. Prestressed beam.

PAGE 49

37 CHAPTER 5 INSTRUMENTATION Thermocouples, strain rings, foil gauges, linear variable displacement transducers (LVDTs) and load cells were used to measur e the response of the beam to mechanical and thermal loading. Details of instrument ation and the data ac quisition (DAQ) system are covered in this chapter and include th e selection process, methods, mounts, and calibrations used for the final beam instrumentation. Load Cells Load cells were used to measure the tip load and prestress forces. The layout of the load cells are shown in Figure 41. Load cells L1, L2 and L3 were used to measure the reaction forces; load cells L4 through L7 were used for the prestress forces. The applied tip load was a maximum of 55 kips, which wa s measured by a built-in load cell in the hydraulic jack. Load cells L2 and L3 were pl aced to measure the forces at the load cell support frame under segment 2 and at the r eaction frame over segment 1, respectively. The expected maximum load at L1 was 55 ki ps, which would result in a maximum load of 110 kips at L2. A 75 kip Geokon load cell was used at L3, and a 150 kip Geokon load cell was used at L2. Four 200 kip hollow core load cells were placed as shown in Figure 42. Figure 43 shows the layout of the PT load cells. Deflections DCTH Series LVDTs from RDP Electrosense were used in the final test beam to measure all deflections. They were placed ve rtically to measure tip deflection, deflection at the mid support and deflection at the reacti on frame. Horizontally, they were placed

PAGE 50

38 throughout the depth of the beam across the in terface of Segments 2 and 3. They were also placed at the top of the beam across th e interface of Segments 2 and 3. The LVDTs were positioned across the interface to determine if there was joint opening. Joint Opening Tests were conducted on a small rectangular beam to determine if joint opening could be detected using strain gages, strain rings, or linear po tentiometers. The test beam consisted of two 30 in. long rectangular b eams that were match-cast and prestressed together. Prestressing was applied with two 0.5-in. diameter threaded bars inside PVC conduits that were cast into the segments at 1 in. above and 1 in. below the centroid of the cross-section (Figure 44). The beam was pl aced in the loading frame shown in Figure 45 and the gage to be tested was installed on the beam at the segment. The beam was loaded, and data were recorded and plotted vs. tip load to determine the sensitivity of the gage to crack opening. Opening of the joint was confirmed with vi sual observations. Joint opening was important because it indicated that there was zero flexural stress in the concrete at the top of the segment interface. The following sections detail the results of the tests conducted on each of the gages evaluated. Strain Rings Strainstall Type-5745 sealed strain rings were used to measure longitudinal strains. Figure 46 and Figure 47 show two setups of th e strain rings across the beam interface. Figure 48 is representative stra in data of the mounts across the joint interface. After several tests using either m ounting condition, no conclusive data that would confirm a joint opening was found because the curves were non-linear. The interface between the two segments was not perfectly matched due to shrinkage effects during the concrete curing. This could have added an effect wh ere the segments rotate d about high points in

PAGE 51

39 the interface during lo ading and not allowed for full cross sectional properties. Additionally, it was believed that the gauges were adding an ex tra variable of stiffness across the joint. Therefore, in the final test beam, these gauges were not used across the joint. The strain rings were used on the final te st beam to record strains at the centroid of the beam close to the segment interface betw een the two thermal segments. They were placed on either side of the beam, and were used to supplement the foil gauge data. The strain rings were tested for linearity to ensure that the signal sent from the strain rings did not vary nonlin early. The strain rings were loaded incrementally with 1 kilogram weights (approximately 2.2 pounds) and strains were measured (see Figure 49). Figure 50 shows that the lo ading and unloading of the strain ring was linear. Foil Gages Strain foil gauges were used to measure strains through the depth of the beam. The gauges were selected based upon the necessity to measure concrete strains, to handle temperatures higher than 140 F, to be shor t in contact length, and to be able to compensate for temperature. A 3-wired 60 mm “PL” series foil gauge from Texas Measurements was selected based on these stipulations. The gauges were placed throughout the depth of the test beam near the joint interface (Figure 51). Strain data was plotted vs. tip load to determine any crack opening (Figure 50). On the final test beam, the strain gauges were also placed throughout the depth of the beam on the two thermal segments. They were placed near the segment interfaces between segments 1, 2, 3 and 4 to determine the behavior at the joints under loading conditions. They were also placed at the cen ter of Segment2 and the center of Segment 3 to determine the strain behavior there and the differences, if any, from the strain behavior at the joints.

PAGE 52

40 Displacement Devices A linear potentiometer (POT) was used in the small beam tests to measure the deflection of the tip of the beam. The tip de flection was plotted vers us tip load. It was found that the beam would deflect linearly up to a certain load, gradually change slope, and then deflect linearly again (Figure 53). An Omega LD-700 Series non-contact di splacement transducer (NCDT), an Omega LD400 Series LVDT, and a linear POT (LP) were used to attempt to detect joint opening (see Figure 54). The instruments were able to accurately measure in the 1/1000 of an inch range, which was necessary to m easure joint opening. The data collected from these devices were plotted against the tip load (see Figure 55). Data collected from these instruments followed a similar trend to the da ta from the load vs. tip deflection data. Relative joint displacement was linear, grad ually changed slope and then became linear again. Thermocouples and Inline Temperature Sensors Thermocouples were fabricated in the lab and embedded in the c oncrete to measure temperatures within the beam. A Teflon neof lon type T thermocouple wire was selected due to it ability to handle up to 392 F temp eratures. Thermocouples were positioned to measure temperatures at the heights corre sponding to slope cha nges in the thermal gradients, see Figure 56. They were also placed at intermediate points between slope changes to verify linear thermal gradients. Thermocouple cages, consisting of small diameter steel, were fabricated to ensure that the thermocouples were placed at exact depths of the beam. Figure 57 shows the thermocouple grids placed in the form. There were six sets of thermocouple grid s; three placed in Segment 2 and three placed in Segment 3 (see Figure 58). The gr ids close to the segment interfaces were

PAGE 53

41 placed three inches from the e nds to accurately measure temp eratures near the interface. In both segments 2 and 3, thermocouples were placed in the center of the segment to record data that would demonstrate that the temperatures throughout the beam length were uniform. Final Instrumentation Setup. Final instrumentation setup was based on tests conducted on the small rectangular beam and on available channels in the DAQ system. The system consisted of three chassis that contained modules that read input signals from terminal blocks and converted and output the signal to a lapt op computer. The channel inve ntory is shown in Table 9. The three chassis that were us ed provided 12 available slot s for modules. Seven of the slots were used for voltage (thermocouple) te rminal blocks, and 5 slots were used for strain module terminal blocks. An overview of the instrumentation is shown in Figure 60 and Figure 61. Details of the gages on Segments 2 and 3 are shown in Figure 62 and Figure 63. The instruments were given the following labe ling designation: L for load a nd S for strain. Load cells were designated with a label such as L3, in which the L indicates load cell and the 3 is the load cell number. Linear potentiometers a nd strain gages were labeled as shown in Figure 65 and Figure 66.

PAGE 54

42 LOAD CELL SEG 1 SEG 2 SEG 3 SEG 4 J1 J2 J3 L4 L5 L6 L7 L1 L2 L3 Figure 41. Load cell layout for final beam. ANGLE : 2 X 2 X 0.25" CHAIR (NOT SHOWN) HYDRAULIC CYLINDER LOAD CELL 1 2" INBED PLATE C8X18.75 CHANNEL 13 8" \U+2205 DYWIDAG THREADBAR (150 ksi) 1-3 4" PLATE AND NUT FROM DSI Figure 42. Load cell layout for prestressing. LOAD CELL PRESTRESS BAR L4L5 L6L7 Figure 43. Cross section view of prestress load cells.

PAGE 55

43 4 1 2 4" 1" 1" PVC PIPE 8" ALLTHREAD BAR 1 2 Figure 44. Cross section of rectangular test beam. 30 "30 "3-1/4 Q2-5/8 1-1/4 Figure 45. Load setup for small beam tests. Figure 46. Strain ring on moun ting blocks across joint.

PAGE 56

44 Figure 47. Modified mounts for strain rings. 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 02004006008001000120014001600Strain (microstrain)Load (lbs) Figure 48. Load vs. longitudi nal strain (strain ring).

PAGE 57

45 Figure 49. Incremental loading of strain rings. 0 5 10 15 20 250200400600800Strain (Microstrain)Load (lb) LOADING UNLOADING Linear (LOADING) Linear (UNLOADING) Figure 50. Strain vs. loading for strain rings.

PAGE 58

46 Figure 51. Foil strain gauges at segment interface. 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300-100-80-60-40-20020406080100Strain (microstrain)Load (lbs) fg1 fg2 fg3 fg4 fg5 fg6 fg7 fg8 Figure 52. Load vs. longitudinal strain (foil gauges). 0 200 400 600 800 1000 1200 1400 1600 1800 00.050.10.150.20.250.30.350.4Deflection (in)Load (lbs) Figure 53. Load vs. tip deflection.

PAGE 59

47 Figure 54. Non-contact displacement tr ansducer, LVDT and linear POT. 0 200 400 600 800 1000 1200 1400 1600 00.010.020.030.040.050.060.07Relative Joint Displacement (in)Load (lbs) LP NCDT LVDT LP NCDT LVDT Figure 55. Load vs. relative joint displacement. +41 F +11 F 0 F -12.3 F -3.3 F 0 F 4" 12" 20" 4" 12" 20" 24" 10" 4 @ 2" 4 @ 5" 3 4 1 2 3 4 4" 1 2 1 2 24" 10" 4 @ 2" 4 @ 5" 3 4 1 2 3 4 4" 1 2 1 2 (a) (b) Figure 56. Thermal gradients and thermocoupl e layouts for a) positive thermal gradient, b) negative thermal gradient.

PAGE 60

48 Figure 57. Thermocouple grid with attached thermocouples. THERMOCOUPLE GRID 3" SECTION ASECTION BSECTION C SECTION DSECTION ESECTION F 3" SEGMENT 2SEGMENT 3 Figure 58. Elevation view of segments 2 a nd 3 along with thermocouple grid locations.

PAGE 61

49 A-T-1 A-T-6 A-T-11 A-T-2 A-T-7 A-T-12 A-T-3 A-T-8 A-T-13 A-T-4 A-T-9 A-T-14 A-T-5 A-T-10 A-T-15 A-T-16A-T-17A-T-18 A-T-19A-T-20A-T-21 A-T-22A-T-23A-T-24 A-T-25A-T-26A-T-27 A-T-28A-T-29A-T-30 A-T-31A-T-32A-T-33 A-T-34A-T-35A-T-36 A-T-37A-T-38A-T-39 B-T-1 B-T-6 B-T-11 B-T-2 B-T-7 B-T-12 B-T-3 B-T-8 B-T-13 B-T-4 B-T-9 B-T-14 B-T-5 B-T-10 B-T-15 B-T-16B-T-17B-T-18 B-T-19B-T-20B-T-21 B-T-22B-T-23B-T-24 B-T-25B-T-26B-T-27 B-T-28B-T-29B-T-30 B-T-31B-T-32B-T-33 B-T-34B-T-35B-T-36 B-T-37B-T-38B-T-39 C-T-6C-T-7C-T-8C-T-9C-T-10 C-T-1 C-T-11 C-T-2 C-T-12 C-T-3 C-T-13 C-T-4 C-T-14 C-T-5 C-T-15 C-T-16C-T-17C-T-18 C-T-19C-T-20C-T-21 C-T-22C-T-23C-T-24 C-T-25C-T-26C-T-27 C-T-28C-T-29C-T-30 C-T-31C-T-32C-T-33 C-T-34C-T-35C-T-36 C-T-37C-T-38C-T-39 E-T-1 E-T-6 E-T-11 E-T-2 E-T-7 E-T-12 E-T-3 E-T-8 E-T-13 E-T-4 E-T-9 E-T-14 E-T-5 E-T-10 E-T-15 E-T-16E-T-17E-T-18 E-T-19E-T-20E-T-21 E-T-22E-T-23E-T-24 E-T-25E-T-26E-T-27 E-T-28E-T-29E-T-30 E-T-31E-T-32E-T-33 E-T-34E-T-35E-T-36 E-T-37E-T-38E-T-39 F-T-6F-T-7F-T-8F-T-9F-T-10 F-T-1 F-T-11 F-T-2 F-T-12 F-T-3 F-T-13 F-T-4 F-T-14 F-T-5 F-T-15 F-T-16F-T-17F-T-18 F-T-19F-T-20F-T-21 F-T-22F-T-23F-T-24 F-T-25F-T-26F-T-27 F-T-28F-T-29F-T-30 F-T-31F-T-32F-T-33 F-T-34F-T-35F-T-36 F-T-37F-T-38F-T-39 D-T-6D-T-7D-T-8D-T-9D-T-10 D-T-1 D-T-11 D-T-2 D-T-12 D-T-3 D-T-13 D-T-4 D-T-14 D-T-5 D-T-15 D-T-16D-T-17D-T-18 D-T-19D-T-20D-T-21 D-T-22D-T-23D-T-24 D-T-25D-T-26D-T-27 D-T-28D-T-29D-T-30 D-T-31D-T-32D-T-33 D-T-34D-T-35D-T-36 D-T-37D-T-38D-T-39SECTION ASECTION BSECTION C SECTION DSECTION ESECTION F Figure 59. Thermocouple labels in Segment 2 and Segment 3.

PAGE 62

50 Table 9. Inventory for availabl e data acquisition channels CHASSIS # OF CHASSIS SLOTS PER CHASSIS TOTAL # OF SLOTS 3 4 12 MODULES MODEL DESCRIPTION COUNT TOTAL CHANNELS SCXI 1520 8 CHANNEL STRAIN MODULE 5 40 STRAIN SCXI 1102 32 CHANNEL T-COUPLE 4 128 T-COUPLE SCXI 1102C 32 CHANNEL T-COUPLE 3 96 T-COUPLE TERMINAL BLOCKS MODEL DESCRIPTION COUNT TOTAL CHANNELS SCXI 1314 8 CHANNEL STRAIN BLOCK 5 40 STRAIN SCXI 1303 32 CHANNEL T-COUPLE BLOCK 4 128 T-COUPLE SCXI 1300 32 CHANNEL T-COUPLE BLOCK 3 96 T-COUPLE CAPABILITY AVAILABLE SLOTS WITH 3 CHASSIS : 12 NUMBER OF T-COUPLES : 224 USING : 7 SLOTS LEAVING : 5 SLOTS AVAILABLE NUMBER OF STRAIN DEVICES : 40 USING : 5 SLOTS STRAIN RING FOIL GAUGE LOAD CELL LVDT SEG 1 SEG 2 SEG 3 SEG 4 J1 J2 J3 L4 L5 L6 L7 L1 L2 L3 S2-T-D1 S3-T-D1 S4-T-D1 S1-T-D1 Figure 60. Elevation view of instru mentation layout (north side).

PAGE 63

51 SEG 1 SEG 2 SEG 3 SEG 4 J3 J2 J1 L2 L1 L3 S2-T-D1 S3-T-D1 S4-T-D1 S1-T-D1 STRAIN RING FOIL GAUGE LOAD CELL LVDT L4 L5 L6 L7 Figure 61. Elevation view of instru mentation layout (south side). STRAIN RING FOIL GAUGE LOAD CELL LVDT SEGMENT 3SEGMENT 2C.G.C.G.8" 28" L2 J1 J2 J3S3-N-S-02-33.5 S3-N-S-02-30.3 S3-N-S-02-27.4 S3-N-S-02-14.6 S3-N-S-02-7.8 S3-N-S-02-1.1 S3-N-R-03.7-21.3S2-T-D1S3-T-D1S3-N-S-30-35.1 S3-N-S-30-28.5 S3-N-S-30-14.6 S3-N-S-30-7.7 S3-N-S-30-1.1 S3-N-S-30-21.4 S3-N-S-58-35.5 S3-N-S-58-28.5 S3-N-S-58-14.6 S3-N-S-58-7.8 S3-N-S-58-1.1 S3-N-S-58-21.3 S2-N-S-02-33.4 S2-N-S-02-30.4 S2-N-S-02-27.5 S2-N-S-02-14.5 S2-N-S-02-7.7 S2-N-S-02-1.0 S2-N-R-03.7-21.3 S2-N-S-30-35.5 S2-N-S-30-28.5 S2-N-S-30-14.6 S2-N-S-30-7.9 S2-N-S-30-1.0 S2-N-S-30-21.2 S2-N-S-58-35.5 S2-N-S-58-28.6 S2-N-S-58-14.6 S2-N-S-58-8.5 S2-N-S-58-1.2 S2-N-S-58-21.2 S3-N-S-3.7-21.3 S2-N-S-3.7-21.3 Figure 62. Detail view of instrumentati on on segments 2 and 3 (north side).

PAGE 64

52 14.67" C.G.C.G. SEGMENT 2 SEGMENT 3 S3-T-D1 L2 J3 J2 J1 J2-S-D-33.75 J2-S-D-30.25 J2-S-D-25.0 J2-S-D-4.375S2-T-D1S3-S-S-02-35.3 S3-S-S-02-32 S3-S-S-02-28.6 S3-S-S-02-27.5 S3-S-S-02-14.6 S3-S-S-02-7.9 S3-S-S-02-0.9 S3-S-R-3.7-21.3 S3-S-S-30-35.4 S3-S-S-30-28.6 S3-S-S-30-14.9 S3-S-S-30-7.8 S3-S-S-30-1.0 S3-S-S-30-21.4 S3-S-S-58-33.4 S3-S-S-58-30.4 S3-S-S-58-14.6 S3-S-S-58-7.8 S3-S-S-58-1.0 S3-S-S-58-21.3 S2-S-S-02-35.5 S2-S-S-02-32 S2-S-S-02-28.5 S2-S-S-02-27.3 S2-S-S-02-14.5 S2-S-S-02-7.7 S2-S-S-02-0.9 S2-S-R-3.7-21.2 S2-S-S-30-35.5 S2-S-S-30-28.5 S2-S-S-30-14.6 S2-S-S-30-8.0 S2-S-S-30-0.9 S2-S-S-30-21.2 S2-S-S-58-33.6 S2-S-S-58-30.4 S2-S-S-58-14.5 S2-S-S-58-7.7 S2-S-S-58-0.9 S2-S-S-58-21.2 STRAIN RING FOIL GAUGE LOAD CELL LVDT 8" 28" S3-S-S-3.7-21.3S2-S-S-3.7-21.2 J2-S-D-11.375 J2-S-D-18.25 Figure 63. Detail view of instrumenta tion on segments 2 and 3 (south side). LVDT J2-TS-D-9.5 S3-TS-S-5.5-7.25 FOIL GAUGE J2-TN-D-9.5 S3-T-S-5.5-0.75 S3-TN-S-5.5-7.25 SOUTH N ORTHBEAM CENTERLINE 12 12 S2-TS-S-5.5-8.25 S2-TN-S-5.5-7.75 J2-T-D Figure 64. Plan view of instrumentation on segments 2 and 3. Segment Number Top Face Displacement Device NumberS2 T D1 Figure 65. Typical instrumentation label for displacement device.

PAGE 65

53 Segment Number North (N) or South (S) Face Strain Gauge Distance from Joint 2 Distance Above Bottom of BeamS2 N S 580.6 Figure 66. Typical instrumentati on label for strain device.

PAGE 66

54 CHAPTER 6 METHODOLOGY OF IMPOSING THERMAL PROFILES The thermal profiles were imposed on the beam by pumping heated water through the copper pipes that were embedded in the concrete. Four different profiles were imposed on the beam. A uniform profile and a linear profile were imposed individually on beam Segments 2 and 3 for coefficient of thermal expansion (CTE) tests. The other two profiles, the AASHTO positive and negative thermal gradient, were imposed with the segments together. The methods used to impose the gradient s are discussed in this chapter. Throughout the chapter, pipe la yers, heaters and thermocouple sections are referenced. Heaters are labeled as described in Chapter 3. Figure 68 through Figure 69 show the pipe labelings, thermocouple section labeling and the thermocouples used in each section. The X markings in Figure 69 show the ther mocouples that were not read due to limitations on the number of instrumentation ch annels that could be simultaneously read by the data acquisition system. Uniform Profile CTE tests were conducted on Segments 2 and 3 by imposing a uniform increase in temperature of 41 F and measuring the resul ting longitudinal displacement. Tap water, with a temperature of approximately 80 F, was passed through the segment overnight to ensure a uniform starting temperature (see Figure 70). The starting temperatures were recorded and used to determine the necessary absolute temperatures to impose the 41 F uniform profile.

PAGE 67

55 Heated water was then passed through the beam as illustrated in Figure 71. The heaters were initially set to about 10 F higher than the re quired +41 F temperature to expedite the beam heating process. Throughout the test, the temperatures in the concrete were monitored. When the temperature change s approached +41 F, the heater settings were adjusted to match the required temperat ures. When the desired temperatures were achieved, they were held in steady state by allowing the pumps to continue passing the heated water through the pipes. Figure 72 a nd Figure 73 show the experimental profiles for beam Segment 2 and 3, respectively. In Segment 2, the temperature changes exceeded the target temperature change of + 41 F due to overheating the segment. The actual average uniform temperature change was + 42 F. The temperature change in Segment 3 matched the target temperature ch ange of 41 F. Throughout the depth of the beam, the temperatures fall within 1 degree of the target temperatur e, which is about a 2% difference from the target temperature ch ange. Due to constraints of the heating systems and laboratory conditions, this difference was found to be acceptable. Linear Profile CTE tests were also conducted on segments 2 and 3 by imposing a linear increase in temperature over the height of the section that was varied from 0 F at the bottom of the beam to 41 F at the top of the beam Tap water was passed through the beam to provide a uniform starting temperature. Heated water was then passed through th e beam as shown in Figure 74 to impose the profile. Table 11 shows th e required absolute temperatur es in each pipe layer to achieve the linear profile based from the st arting ambient temperatures. Initially the heaters were set higher than the required temperatures to expedite the beam heating process. Temperatures in the concrete we re closely monitored throughout the test, and

PAGE 68

56 the heaters were adjusted to match the required temperatures. When the desired temperatures were achieved, they were held in steady state by allowing the pumps to continue running. Figure 75 and Figure 76 s how the experimental profile along with the desired linear profile for beam Segment 2 and 3, respectively. AASHTO Positive Thermal Gradient The AASHTO positive thermal gradient was im posed in two steps. The first step was to bring the sections to a uniform temper ature similar to the CTE tests. The second step was to impose the gradient. Figure 80 shows the piping setup used to impose the temperature change necessary to create the gradient. Tap water continued to run through the pipe layers in the web to ensure that the original beam temperature wa s maintained at depths greater than 16 in. from the top. Temperatures of the water that passed through layers 1 thru 3 were set to the absolute temperatures shown in Table 12. Initially the temperatur es were actually set higher than the required temperature to expedite the beam heating process. Temperatures in the concrete were closely monitored througho ut the test, and the he aters were adjusted to match the required temperatures. When th e desired temperatures were achieved, they were held in steady state. Figure 78 shows the experiment al profile along with the theoretical AASHTO positive gradient. The temp eratures matched within 0 to 2 F of the target temperatures, with th e exception of one layer in Section A that had an average temperature within 3 F of the target temperat ure. These temperatures matched closely to the target profile and were found to be accep table with the given laboratory and heating systems.

PAGE 69

57 AASHTO Negative Thermal Gradient The AASHTO negative thermal gradient was im posed in two steps. The first step was to uniformly heat the beam overnight to a temperature ar ound 122 F. Having the beam elevated to this temperature allowed fo r the heaters to be in their working range when backing down temperature changes to -12.3 F. After the beam was uniformly heated, the negative gradient was imposed. Figure 80 shows the piping setup used to impose the temperature change necessary to create the gradient. Th e high temperature water continued to run through the pipe layers in the web to ensure that 0 F temperat ures were maintained at depths greater than 16 in. from the top. The temperatures of th e water that passed through layers 1 through 3 were set to the absolute temperatures th at are shown in Table 13. Initially the temperatures were set lower than the required temperatures to expedite the beam cooling process. Temperatures in the concrete we re closely monitored throughout the test, and the heaters were adjusted to match the required temperatures. When the desired temperatures were achieved, they were he ld in steady state. Figure 81 shows the experimental profile along with the th eoretical AASHTO negative gradient. The temperatures are within 0.5 F, with the excep tion of one temperature average in Section A that is within 1 F.

PAGE 70

58 THERMOCOUPLE GRID 3" SECTION ASECTION BSECTION C SECTION DSECTION ESECTION F 3" SEGMENT 2SEGMENT 3 Figure 67. Sections at which thermocouples were embedded. 24" 8" 10" 36" 28" LAYER 1 LAYER 2 LAYER 3 LAYER 4 LAYER 5 LAYER 6 Figure 68. Pipe layers used for thermal profile tests.

PAGE 71

59 SECTION ASECTION BSECTION C SECTION DSECTION ESECTION F Figure 69. Thermocouples used in each section. TAP WATER IN TAP WATER IN TAP WATER OUT TAP WATER OUT Figure 70. Piping schematic for tap water flow.

PAGE 72

60 HEATER 1 IN HEATER 2 IN HEATER 1 OUT HEATER 2 OUT Figure 71. Piping schematic for +41 F uniform profile. Table 10. Absolute temperatures to achieve +41 F profile ba sed on initial temperatures. Initial Ambient Temperatures (F) 79 80 81 82 83 84 85 86 Destination Temp ( F) 120 121 122 123 124 125 126 127 0 4 8 12 16 20 24 28 32 36 051015202530354045Temperature Difference (deg. F)Elevation (inches) Section A Section B Section C Target Profile Figure 72. Uniform profile for segment 2.

PAGE 73

61 0 4 8 12 16 20 24 28 32 36051015202530354045Temperature Difference (deg. F)Elevation (inches) Section D Section E Section F Target Profile Figure 73. Uniform profile for Segment 3. HEATER 1 IN HEATER 2 IN MIXED HEATER 3 IN HEATER 1 OUT HEATER 2 OUT MIXED OUT HEATER 3 OUT Figure 74. Piping schematic for linear profile. Table 11. Absolute temperatures to achieve linear profile. Target Temperatures of Pipe Layers Based on Initial Ambient Temperatures (F) Initial Temp ( F) 79 80 81 82 83 84 85 86 Layer 1 119.4 120.4 121.4122.4123.4124.4 125.4 126.4 Layer 2 No Flow Layer 3 111.5 112.5 113.5114.5115.5116.5 117.5 118.5 Layer 4 102.3 103.3 104.3105.3106.3107.3 108.3 109.3 Layer 5 91.0 92.0 93.0 94.0 95.0 96.0 97.0 98.0 Layer 6 No Flow

PAGE 74

62 0 4 8 12 16 20 24 28 32 36 051015202530354045Temperature Difference (deg. F)Elevation (inches) Section A Section B Section C Target Profile Figure 75. Linear profile for Segment 2. 0 4 8 12 16 20 24 28 32 36051015202530354045Temperature Difference (deg F)Elevation (inches) Section D Section E Section F Target Profile Figure 76. Linear profile for Segment 3. HEATER 1 IN MIXED HEATER 3 IN HEATER 1 OUT MIXED OUT HEATER 3 OUT TAP WATER IN TAP WATER OUT Figure 77. Piping schematic for AASHTO positive gradient.

PAGE 75

63 Table 12. Absolute temperatures to achieve AASHTO positive gradient. Target Temperatures of Pipe Layers Based on Initial Ambient Temperatures (F) Initial Temp ( F) 79 80 81 82 83 84 85 86 Layer 1 120.0 121.0 122.0 123.0 124.0 125.0 126.0 127.0 Layer 2 90.0 91.0 92.0 93.0 94.0 95.0 96.0 97.0 Layer 3 86.8 87.8 88.8 89.8 90.8 91.8 92.8 93.8 Layer 4 79.0 80.0 81.0 82.0 83.0 84.0 85.0 86.0 Layer 5 79.0 80.0 81.0 82.0 83.0 84.0 85.0 86.0 Layer 6 79.0 80.0 81.0 82.0 83.0 84.0 85.0 86.0 0 4 8 12 16 20 24 28 32 36 -505101520253035404550Temperature Difference (deg F)Elevation (inches) Section A Section B Section C Section D Section E Section F AASHTO Figure 78. Experimental AASHTO positive gradient. HEATER 1 IN HEATER 2 IN HEATER 1 OUT HEATER 2 OUT Figure 79. Piping schema tic to heat beam.

PAGE 76

64 HEATER 1 IN MIXED HEATER 3 IN HEATER 1 OUT MIXED OUT HEATER 3 OUT HEATER 2 IN HEATER 2 OUT Figure 80. Piping schematic for AASHTO negative gradient. Table 13. Absolute temperatures to achieve AASHTO negative gradient. Target Temperatures of Pipe Layers Based on Initial Ambient Temperatures (F) Initial Temp ( F) 79 80 81 82 83 84 85 86 Layer 1 105.7 106.7 107.7 108.7 109.7 110.7 111.7 112.7 Layer 2 114.7 115.7 116.7 117.7 118.7 119.7 120.7 121.7 Layer 3 115.7 116.7 117.7 118.7 119.7 120.7 121.7 122.7 Layer 4 118.0 119.0 120.0 121.0 122.0 123.0 124.0 125.0 Layer 5 118.0 119.0 120.0 121.0 122.0 123.0 124.0 125.0 Layer 6 118.0 119.0 120.0 121.0 122.0 123.0 124.0 125.0 0 4 8 12 16 20 24 28 32 36 -14-12-10-8-6-4-202Temperature Difference (deg F)Elevation (inches) Section A Section B Section C Section D Section E Section F AASHTO Figure 81. Experimental AAS HTO negative gradient.

PAGE 77

65 CHAPTER 7 CTE TESTS The coefficient of thermal expansion of the concrete, CTE, was necessary for computer modeling and stress calculations It was found using two methods: the AASHTO standard test method and in-situ tests. This chapter will discuss the two methods for determining the CTE value. AASHTO Standard Test Method A standard procedure for determining the CTE value for concrete is outlined in the AASHTO Designation: TP 60-00 (2004). This designation is the Standard Method of Test for Coefficient of Thermal Expansion of Hydraulic Cement Concrete. The scope of the procedure involves fully sa turating a concrete cylinder, increasing and decreasing the temperature of the cylinder and measuring the length changes. The CTE is then calculated as the change in le ngth divided by the cylinder length and temperature change. The value is useful in determining the pot ential for length and volume changes of concrete due to either a uniform temperat ure change or a temperature gradient. AASHTO test specimens are concrete cylinders 7.0 0.1 inches in length and 4 inches in diameter. They are submersed in saturated limewater at 73 4 F for at least two days. After the cylinders are fully saturated, they are removed from the water, and their lengths are measured at room temperat ure to the nearest 0.004 in. They are then placed in a measuring apparatus that consists of a frame and LVDTs (see Figure 82). The cylinder and apparatus are placed into a prepar ed water bath, and the temperature of the water is then set to 50 2 F. Once thermal equilibrium of the specimen has been

PAGE 78

66 reached, the LVDT will consistently read the same deflection to the nearest 0.00001 inches. LVDT readings shall be taken ev ery 10 minutes over a one-half hour time period. Both the temperature and LVDT readings will be recorded as the initial readings. The temperature of the water will then be set to 122 2 F, and thermal equilibrium of the specimen will be reached similarly to the in itial stage. The temperature and LVDT readings will be recorded as the second read ing. The water temperat ure is then set back to the initial 50 2 F, thermal equilibrium is reached, and the third and final set of readings is recorded. The next step is to calculate the CTE value according to the equation: CTE T L La / ) / (0. The actual length change of the specimen, La, is calculated by taking the sum of the measured length change of the specimen (increase is positive, decrease is negative) and the length change of the measuring apparatus. The length change of the measuring apparatus is taken as the multipli cation of the CTE of stainless steel (17.3 C / 106), the change in temperature and the original specimen length, L0. CTE values shall be computed for both the expansion and contraction stages. The final CTE value is calculated by averaging the two CTE values The two values must be within 0.5 F / 106; additional tests are necessa ry if they are not. The AASHTO CTE tests were performed by the FDOT SMO on concrete cylinders taken from the casting of beam segment 3. Ideally, concrete cylinders from beam Segment 2 would also have been tested because it was a thermal section. However, 4 in. concrete cylinders were not taken from segment 2. The AASHTO tests were performed on the Segment 3 cylinders at the FDOT Stat e Materials Lab, while in-situ tests were being performed on segment 3 (described later). The proximity of the time that the two

PAGE 79

67 tests were run ensured that the concrete was at the same material properties (i.e. moisture content) between the cylinders and the beam segment. Data from the AASHTO CTE tests are show n in the following tables. Table 14 shows the initial lengths of each tested cyli nder after the cylinders had soaked for more than two days. These lengths were used to de termine the final strains. Table 15 displays the absolute temperatures a nd the temperature changes ( t) recorded during the test. The displacements recorded from the LVDTs are sh own in Table 16. The final strains were calculated by dividing the displacements by the initial lengths (Table 17). The strains were divided by the temperature changes and CTE values were recorded for the heating and cooling phases. These values were th en averaged to find a CTE value for each cylinder (Table 18). A representative CTE value for beam segment 3 was taken as the average of the three cylinders’ CTE va lues. The value was found to be 7.84 F / 106, which is in the high end of t ypical CTE values (4.1 to 7.3 F / 106). In-situ CTE Test Method Methods were developed to determine CTE va lues for the actual beam in laboratory conditions. The general scope of the tests was to heat a beam segment to a given thermal profile and record th e longitudinal elongation of the beam Deflections were recorded using LVDT’s, and temperatures were reco rded with thermocouples. Figure 83 shows the positioning of LVDTs on the beam in the la b. LVDTs were pla ced vertically along the cross sectional centerline on each end of the beam at the top, the centroi d, the bottom and halfway between the centroid and bottom. The CTE value was then computed by dividing the average longitudinal displacements by the average temperature differentials and the total beam length, according to the following equations:

PAGE 80

68 c c total avgA dA (average total displacement of segment) segment Lavg avg_ (average engineering strain of segment) c c grad avgA dA T TG (average temperature differential) avg avgTG (coefficient of thermal expansion) Where: total total axial elongation of segment L_segment total axial elongation of segment Tgrad total axial elongation of segment Ac cross sectional area of segment Positive Uniform Temperature Profile A positive uniform temperature profile of + 41 F was imposed individually on beam segments 2 and 3. Data from thermo couples, LVDTs, foil ga uges and strain rings were recorded. The desired and imposed th ermal profiles for segment 2 and 3 are shown in Figure 72 and Figure 73 in Chapter 6. The LVDT data are shown in Figure 84 and Figure 85. The calculated CTE values for Segment 2 and Segment 3 from LVDT data were found to be 6.3610/ F and 7.8610/ F, respectively. Linearly Increasing Temperature Profile A linear profile as described in Chapter 5 was imposed individually on beam segments 2 and 3. Data from thermocouples LVDT’s, foil gauges and strain rings were recorded. The desired and imposed therma l profiles are shown in Figure 75 and Figure

PAGE 81

69 76 in Chapter 6. The LVDT data is show n in Figure 86 and Figure 87. The calculated CTE values for Segment 2 and Segment 3 from LVDT data were found to be 7.6610/ F and 8.3610 / F, respectively. Comparison of CTE data Table 19 compares CTE values found from the AASHTO test and the uniform and linear profile tests on segment 2 and 3. The CTE value was lower for Segment 2 than it was for Segment 3, which can most likely be attributed to the moisture content being lower in segment 2 than in segment 3. The CTE values for the linear gradient profile tests were found to be higher th an the values for the uniform tests. The reason for this occurrence is not known. The CTE value for the uniform profile test on Segment 3 matched the average CTE value found from the AASHTO CTE tests.

PAGE 82

70 Figure 82. Measuring device for AASHTO CTE tests. Table 14. Cylinder lengths at room temperature. Cylinder 1 Cylinder 2 Cylinder 3 L1 (inches) 7.065 7.098 7.056 L2 7.078 7.108 7.064 L3 7.061 7.092 7.056 L4 7.054 7.104 7.054 Average Length 7.064 7.100 7.056

PAGE 83

71 Table 15. Absolute temperatures and te mperature changes for AASHTO CTE test. Cylinder 1 Cylinder 2 Cylinder 3 Initial Temp (F) 50.2 50.2 50.2 Maximum Temp (F) 123.4 123.4 123.4 Final Temp (F) 50.0 50.0 50.0 t : Initial to Max (F) 73.2 73.2 73.2 t : Max to Final (F) -73.2 -73.2 -73.2 Table 16. Recorded displacements from LVDT's. Cylinder 1 Cylinder 2 Cylinder 3 L : Initial to Max (in.) 0.00404 0.00412 0.00411 L : Initial to Max (in.) -0.00406 -0.00396 -0.00407 Table 17. Calculated strains. Cylinder 1 Cylinder 2 Cylinder 3 L/L : Initial to Ma x (in.) 0.00058 0.00059 0.00059 L/L : Initial to Max (i n.) -0.00058 -0.00057 -0.00058 Table 18. Calculated CTE values. Cylinder 1 Cylinder 2 Cylinder 3 CTE intial max (1/F) 7.82 E-06 7.92 E-06 7.97 E-06 CTE max final (1/F) 7.84 E-06 7.61 E-06 7.87 E-06 Average CTE (1/F) 7.83 E-06 7.77 E-06 7.92 E-06 HEATED SEGMENT C.G.C.G. LVDT 4"X4" STEEL COLUMN VERY RIGID SHORING 1 2 5 6"STRAIN RING FOIL GAUGE LVDT 1 2 10.9" 14.67" 1 2 Figure 83. CTE test setup.

PAGE 84

72 0 5 10 15 20 25 30 35 40 00.0050.010.0150.02Longitudinal Displa cements (inches)Elevation (inches) East West Sum Figure 84. Longitudinal displacements from LVDT's for Segment 2 for positive uniform temperature profile. 0 5 10 15 20 25 30 35 40 00.0050.010.0150.020.025Longitudinal Displa cements (inches)Elevation (inches) East West Sum Figure 85. Longitudinal displacements from LVDT's for Segment 3 for positive uniform temperature profile.

PAGE 85

73 0 5 10 15 20 25 30 35 40 -0.00500.0050.010.0150.02Longitudinal Displacements (inches)Elevation (inches) East LVDT's West LVDT's SUM Figure 86. Longitudinal displacements fr om LVDT's for Segment 2 for linearly increasing temperature profile. 0 5 10 15 20 25 30 35 40 00.0050.010.0150.02Longitudinal Displa cements (inches)Elevation (inches) East West Sum Figure 87. Longitudinal displacments from L VDT's for Segment 3 for linearly increasing temperature profile. Table 19. Comparison of CTE values. Segment 2 CTE Value (micro strain / F) Segment 3 CTE Value (micro strain / F) AASHTO N.A. 7.8 Uniform Profile 6.3 7.8 Linear Profile 7.6 8.3

PAGE 86

74 CHAPTER 8 CONCLUSIONS AND RECOMMENDATIONS The main objective of this research proj ect was to quantify the stresses formed from the AASHTO nonlinear thermal gradient s. Instrumentation was necessary to measure the loads, strains and deflections to quantify the thermal stresses in a 20 ft. long segmental concrete T-beam. Before the in strumentation was placed on the T-beam, many tests were performed on small rectangular pr estressed concrete beams to determine the characteristics of the instrumentation under m echanical loading. The beams were loaded until a joint opening occurred between the tw o segments of the prestressed beams. Data taken from tests performed on the rectangular beams were analyzed to determine the load required to cause a joint opening, and associated with zero flexural stress state in the concrete, at the top of the segment interface. Data were taken from strain rings, foil gauges, crack gauges, LVDT s, linear potentiometer s and load cells. Deflections and strains were pl otted against tip load to characterize the beams behavior. The graphs of tip deflections and deflections across the joints s howed the beam would deflect linearly until the joint began to open, would start to deflect nonlinearly, and then would deflect linearly again. Until the joint opened, the stiffness was constant; while the joint was opening, the stiffness was cha nging. Joint opening caused a reduction in stiffness that remained constant with continue d loading. It was recommended that the tip deflections and deflections across the joints ar e further researched on the T-beam. It was also recommended that strains throughout the de pth of the beam at the joints and away from the joints be recorded and analyzed.

PAGE 87

75 One noticeable concern on the small rect angular test beams was the interface between the two match-cast segments. Ev en though much care was taken to form a smooth, even interface, some shrinkage occurr ed during the concrete curing process. The shrinkage resulted in reduced cross sectional prop erties and therefore, flexural stiffness. There may be some similar effects from shri nkage on the test T-beam. However, it is believed that the T-beam section and prestre ss forces are large enough that shrinkage will not affect the behavior of the test beam under loading conditions. A very important aspect of the research project thus far, is the creation of an in-situ test for the coefficient of thermal expansion, CTE, value on a given concrete beam. CTE values were found using AASHTO procedures that included taking 4 in. concrete cylinders and testing them in a bath of wate r. It was felt that testing concrete in a completely soaked environment was not truly representative of the concrete in the field. Therefore, Segments 2 and 3 were individua lly heated and longit udinal movements were recorded. CTE values were then calculate d from the change in length divided by the original length and the temp erature change. It was found that the average CTE value from the AASHTO test for concrete cylinders taken from Segment 3 and CTE value from the in-situ uniform heating test performed on Segment 3 were both 7.8610/ F. Ongoing research being conducted by the University of Florida will focus on performing tests on the segmental T-beam usi ng the instrumentation systems described in this thesis. Through data analysis and comput er modeling, stresses wi ll be quantified and recommendations will be made to the FDOT. In the future, this data will be used to help develop a more economical bridge design.

PAGE 88

76 APPENDIX A TEST BEAM SHOP DRAWINGS This appendix provides shop drawings given to the Florida Department of Transportation for building of the main test beam.

PAGE 89

77

PAGE 90

78

PAGE 91

79

PAGE 92

80 APPENDIX B FRAME DRAWINGS This appendix provides the drawings and details for the frames. W16x89 PL 2"x11.25"x17 3 4 11 2" 1 5 BOLT HOLES 1" DIA @ 15" o.c. 4 1 2 1 5 'PL 2"x1'-1"X1'-7" 1'-67 8" MC18x45.8 5 16 11 2"

PAGE 93

81 MC18x45.8 W16x89 COPE FLANGES @ COLUMN (TYP.) 71 4" 1 13 4" 71 4" 1" DIA BOLT HOLES 1 9 13" 4' STRONG-FLOOR BOLT HOLES LOCATED IN LAB 7 1 2 PL 2"x1'-1"x1'-7" W16x89 4' 1 9 13" W16x89 23 4" 71 2" 23 4" 5 3 4 71 2 53 4 5 16 2" DIA BOLT HOLES

PAGE 94

82 10" C15x33.9 JOINT OPENING INTERFACE 3' LOAD CELL 3'-67 8 67 8" 11 2" X 10" X 10" PLATE 10" DRILLED AND TAPPED FOR 0.5" DIA ALLTHREAD 1 2" NUT DRILLED 5 8" DIA 6 7 8 SEGMENT INTERFACE 8 3 4 3 7 8 1 2 3 7 8

PAGE 95

83 CHANNELS: C15x33.9 LOAD CELL 11 2" X 9" X 9" PLATE 5'-7 7 8 SHORING LOAD CELL 3'-6 7 8 6 7 8 1'-81 4 SHORING

PAGE 96

84 APPENDIX C CONCRETE MIX TICKETS This appendix provides the mix tickets for each concrete pour.

PAGE 97

85

PAGE 98

86

PAGE 99

87 LIST OF REFERENCES American Association of State Highway a nd Transportation Offi cials AASHTO (1989). “AASHTO Guide Specifications, Therma l Effects in Concrete Bridge Superstructures.” Washington D.C. American Association of State Highway a nd Transportation Offici als (AASHTO) (2001). “AASHTO LRFD Bridge Desi gn Specifications – U.S. Units.” Washington D.C. American Association of State Highway a nd Transportation Offici als (AASHTO) (2004). “Coefficient of Thermal Expansion of H ydraulic Cement Concrete.” Washington D.C., TP 60-1 – TP60-7. American Concrete Institute (ACI) Committ ee 318 (2002). “Building Code Requirements for Structural Concrete ( 318-02) and Commentary (318R02).” American Concrete Institute, Farmington Hills, Michigan. Mahama, Farouque (2006). “Validation of St resses Caused by Thermal Gradients in Segmental Concrete Bridges.” PhD Proposal, University of Florida, Department of Civil and Coastal Engineeri ng, Gainesville, FL.

PAGE 100

88 BIOGRAPHICAL SKETCH David Clancy Walter was born on April 17, 198 2, in Alexandria, Virginia. In the same year, he moved to South Florida. After high school, he successfully completed his undergraduate studies at the Univ ersity of Florida and received a Bachelor of Science in civil engineering in May of 2004. The author then worked in a structural design firm in his hometown, Stuart, Florida. He then pur sued his master’s degree in the field of structural engineering at the University of Florida. U pon completion of his graduate school, the author plans to continue his prof essional engineering career with Atlantic Engineering Services in Jacksonville, Florida.


Permanent Link: http://ufdc.ufl.edu/UFE0016221/00001

Material Information

Title: Validation of stresses caused by thermal gradients in segmental concrete bridges : construction, test setup, and coefficient of thermal expansion tests
Physical Description: xii, 88 p. ; ill. , tables
Language: English
Creator: Walter, David Clancy ( Dissertant )
Hamilton, Homer R. ( Thesis advisor )
Consolazio, Gary R. ( Thesis advisor )
Cook, Ronald A. ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2006
Copyright Date: 2006

Subjects

Subjects / Keywords: Civil Engineering thesis, M.E
Dissertations, Academic -- UF -- Civil and Coastal Engineering

Notes

Abstract: American Association of State Highway and Transportation (AASHTO) provisions for checking service limit states require that the stresses from dead and live loads be combined with those caused by heating and cooling. Of particular concern are the "self-equilibrating" stresses caused by non-linear thermal gradients. The Florida Department of Transportation has found that these stresses in combination with dead and live load stresses can control the design and rating of segmental concrete bridges. Yet cracking or other indications of distress that might be associated with these stress combinations have not been noted in regular inspections of bridges in service. This thesis presents the laboratory setup and methodologies implemented to evaluate stresses resulting from thermal gradients in segmental concrete bridges. A 20-foot long segmental concrete T-beam was constructed to determine the actual thermal stresses that occur under a non-linear thermal gradient. The T-shaped cross section was chosen to simulate the web-flange portion of a box girder. The test specimen consisted of four 5 ft. segments that were match-cast and then prestressed together. Two of the sections contained multiple layers of copper pipes that carried heated water through the section. By varying the water temperature in each layer of piping, thermal gradients were imposed on the cross section that closely matched the thermal gradients found in AASHTO guidelines for the State of Florida. Load cells, displacement transducers, strain gauges, strain rings, and thermocouples were used to collect data during heating and load testing. In future work, the data from load tests will be analyzed to determine the magnitude of the stresses caused by a combination of heating and loading. An important variable in the project was the concrete's coefficient of thermal expansion (CTE). This thesis also presents the setup, procedures and results of tests conducted to determine the CTE of the test beam. The results are compared to CTE tests conducted on concrete cylinders taken from the test beam pour.
Subject: bridges, concrete, expansion, gradients, segmental, thermal
General Note: Title from title page of source document.
General Note: Includes vita.
General Note: Document formatted into pages; contains 100 pages.
Thesis: Thesis (M.E.)--University of Florida, 2006.
Bibliography: Includes bibliographical references.
Original Version: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0016221:00001

Permanent Link: http://ufdc.ufl.edu/UFE0016221/00001

Material Information

Title: Validation of stresses caused by thermal gradients in segmental concrete bridges : construction, test setup, and coefficient of thermal expansion tests
Physical Description: xii, 88 p. ; ill. , tables
Language: English
Creator: Walter, David Clancy ( Dissertant )
Hamilton, Homer R. ( Thesis advisor )
Consolazio, Gary R. ( Thesis advisor )
Cook, Ronald A. ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2006
Copyright Date: 2006

Subjects

Subjects / Keywords: Civil Engineering thesis, M.E
Dissertations, Academic -- UF -- Civil and Coastal Engineering

Notes

Abstract: American Association of State Highway and Transportation (AASHTO) provisions for checking service limit states require that the stresses from dead and live loads be combined with those caused by heating and cooling. Of particular concern are the "self-equilibrating" stresses caused by non-linear thermal gradients. The Florida Department of Transportation has found that these stresses in combination with dead and live load stresses can control the design and rating of segmental concrete bridges. Yet cracking or other indications of distress that might be associated with these stress combinations have not been noted in regular inspections of bridges in service. This thesis presents the laboratory setup and methodologies implemented to evaluate stresses resulting from thermal gradients in segmental concrete bridges. A 20-foot long segmental concrete T-beam was constructed to determine the actual thermal stresses that occur under a non-linear thermal gradient. The T-shaped cross section was chosen to simulate the web-flange portion of a box girder. The test specimen consisted of four 5 ft. segments that were match-cast and then prestressed together. Two of the sections contained multiple layers of copper pipes that carried heated water through the section. By varying the water temperature in each layer of piping, thermal gradients were imposed on the cross section that closely matched the thermal gradients found in AASHTO guidelines for the State of Florida. Load cells, displacement transducers, strain gauges, strain rings, and thermocouples were used to collect data during heating and load testing. In future work, the data from load tests will be analyzed to determine the magnitude of the stresses caused by a combination of heating and loading. An important variable in the project was the concrete's coefficient of thermal expansion (CTE). This thesis also presents the setup, procedures and results of tests conducted to determine the CTE of the test beam. The results are compared to CTE tests conducted on concrete cylinders taken from the test beam pour.
Subject: bridges, concrete, expansion, gradients, segmental, thermal
General Note: Title from title page of source document.
General Note: Includes vita.
General Note: Document formatted into pages; contains 100 pages.
Thesis: Thesis (M.E.)--University of Florida, 2006.
Bibliography: Includes bibliographical references.
Original Version: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0016221:00001


This item has the following downloads:


Table of Contents
    Title Page
        Page i
        Page ii
    Acknowledgement
        Page iii
    Table of Contents
        Page iv
        Page v
    List of Tables
        Page vi
    List of Figures
        Page vii
        Page viii
        Page ix
        Page x
    Abstract
        Page xi
        Page xii
    Introduction
        Page 1
        Page 2
    Background
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
    Test beam
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
    Beam construction
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
    Instrumentation
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
    Methodology of imposing thermal profiles
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
    CTE tests
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
    Conclusions and recommendations
        Page 74
        Page 75
    Appendices
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
    References
        Page 87
    Biographical sketch
        Page 88
Full Text











VALIDATION OF STRESSES CAUSED BY THERMAL GRADIENTS IN
SEGMENTAL CONCRETE BRIDGES: CONSTRUCTION, TEST SETUP, AND
COEFFICIENT OF THERMAL EXPANSION TESTS















By

DAVID CLANCY WALTER


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING

UNIVERSITY OF FLORIDA


2006



































Copyright 2006

By

David Clancy Walter















ACKNOWLEDGMENTS

I thank my supervisory committee chairs, Dr. H.R. Hamilton III and Dr. Gary R.

Consolazio, for their guidance throughout this research and in my development as an

engineer. I would like to thank Dr. Ronald A. Cook for his help and support on my thesis

committee. I would also like to thank my research partner, Farouque Mahama, for his

help with and extensive knowledge on this research project.

Very special thanks go to the Florida Department of Transportation (FDOT)

Structures Lab personnel, especially Marc Ansley and Frank Cobb. Without their help

with the construction of the T-beam, this research project would not be possible. The

same thanks go to the University of Florida Structures Lab personnel, especially Charles

Broward, Hubert "Nard" Martin and John Gamache. Not only did they devote countless

hours to helping with the research project, they also provided guidance, relief and an

enjoyable work environment.

I would also like to thank the FDOT State Materials Office, especially Richard

DeLorenzo, for assisting with and performing concrete cylinder tests. I thank

DYWIDAG Systems International for donating all of the prestressing bars, nuts and

plates.

Most importantly, I thank my mom, dad, brother, sisters, nephew and niece, and my

college friends along the way. Their support, love and humor brightened my life during

my years as a student.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iii

LIST OF TA BLES .................................................................... ............ .. vi

LIST OF FIGURE S ......... ....................... ............. ........... vii

ABSTRACT .............. ......................................... xi

CHAPTER

1 IN TRODU CTION ................................................. ...... .................

2 B A C K G R O U N D .................... .... .............................. .......... ........ ......... .. ....

Therm al Gradients ....................... .................. .... .. ............. .... .3
D design Therm al Gradients ........................................................ ... ...............
Concrete Beam Behavior under Thermal Loading ...................................................5

3 TEST BEAM ................................... ................................ ........... 10

Selection of Cross-Section Geom etry...................................................................... 10
Segm ental C construction D details ........................................... ........... ...............12
P re store ssin g .................................................. ................... ................ 1 3
Steel R einforcem ent....... .................................................................. .. .. .... .. .. .... 14
L oad T est Setup .............. ................................................................. .......... 14
Segm ent H eating System ............................ ..................................... ............... 15

4 BEAM CON STRU CTION ............................ ................. ................. ............... 28

F o rm s .................................................................................................................... 2 8
C concrete ............... .......... ..................................................................................28
Delivery, Storage and Prestressing ............................. ........... .................. 29

5 IN STR U M E N TA TIO N ................................................................... .....................37

L o a d C e lls ............................................................................................................. 3 7
D efle ctio n s .......................................................................................3 7
Jo in t O p e n in g .................................................................................................... 3 8
S tra in R in g s ................................................................................................... 3 8


iv










F o il G ag e s.....................................................3 9
D isplacem ent D evices ........................................ ................... ............... 40
Thermocouples and Inline Temperature Sensors ........................................... 40
F inal Instrum entation Setup ................................................ ..................................4 1

6 METHODOLOGY OF IMPOSING THERMAL PROFILES.................................54

U n form P rofi le ................................. ................................................... .. 54
L inear Profile .................................. ..... ......................... ..... .........55
AA SH TO Positive Therm al Gradient..................................... ........................56
AASHTO Negative Thermal Gradient ........... ................................ ...............57

7 C T E T E ST S ........................................................................... 65

A A SH TO Standard Test M ethod.................................................................. ........ 65
In-situ CTE Test M ethod ...................................................... ..... ............ ............... 67
Positive Uniform Temperature Profile ..................................... ...............68
Linearly Increasing Temperature Profile................................. ...............68
C om prison of C T E data ......................................................................... ..... 69

8 CONCLUSIONS AND RECOMMENDATIONS ............................................... 74

APPENDIX

A TEST BEAM SHOP DRAWINGS ........................................ ........................ 76

B F R A M E D R A W IN G S ..................................................................... .....................80

C CON CRETE M IX TICKETS ......................................................... ............... 84

L IST O F R E F E R E N C E S ........................................................................ .....................87

B IO G R A PH IC A L SK E TCH ..................................................................... ..................88




















v
















LIST OF TABLES


Table p

1 Basis for tem perature gradients. ......................... ......... ................................ 6

2 Approximate service-I stresses in SRB Bridge (psi). ........................................ 19

3 D ates of con create p ou rs............................................................... .................... 33

4 Concrete pump mix proportions. ................... ...................................... 34

5 Results from com pression tests................................................................. ...... 34

6 Moduli of elasticity for each segment.......................................................... 34

7 Prestress load increments for tests performed on July 2, 2006............................. 35

8 Prestress load increments for tests performed on July 5, 2006............................. 35

9 Inventory for available data acquisition channels......................................... 50

10 Absolute temperatures to achieve +41 F profile based on initial temperatures.... 60

11 Absolute temperatures to achieve linear profile. ............................................... 61

12 Absolute temperatures to achieve AASHTO positive gradient........................ 63

13 Absolute temperatures to achieve AASHTO negative gradient ....................... 64

14 Cylinder lengths at room tem perature......................................... .................... 70

15 Absolute temperatures and temperature changes for AASHTO CTE test.......... 71

16 Recorded displacem ents from LVD T's ............................................. .............. 71

17 Calculated strains...................... .... .. .......... ............. 71

18 C alcu lated C T E v alu es ........................................................................... .... ......... 7 1

19 C om prison of C T E values ......................................................................... .... 73
















LIST OF FIGURES


Figure page

1 Positive vertical temperature gradient. ........................................ .............. 6

2 Solar radiation zones for the United States....................... ............................... 7

3 Positive therm al gradient for Florida. ........................................... .............. 7

4 Negative thermal gradient for Florida...... ..................... .............. 8

5 Decomposition of a nonlinear thermal gradient ............. ..... ................. 8

6 Strain difference that leads to self-equilibrating stresses........... ............... 9

7 Cross section of SRB bridge. ........................................ .......................... 17

8 I-section representation of SRB bridge cross section. ........................................ 18

9 Self-equilibrating stresses due to AASHTO positive thermal gradient ............... 18

10 Self-equilibrating stresses due to AASHTO negative thermal gradient .............. 18

11 Chosen shape for final test beam with self-equilibrating stresses .................... 19

12 Schematic of segmental beam test specimen. ............................................... 19

13 Prestress assembly cross section view. .............. ................................... ......... 19

14 View of prestess assembly......................................... ............. 20

15 Prestress assembly in elevation view ....... ...................... ............. 20

16 Elevation views of prestress assembly............... ..................................... ........ 21

17 Mild steel reinforcements in Segments 1 and 4 ................................................. 21

18 M ild steel reinforcem ent. ........................ .............................. 21

19 Fram e locations .................. ..................................... .. .......... 22

20 M id-support fram e a.) drawing, b.) picture................................. .................... 22









21 Piping schematics for a) prototype beam, b) test beam. ....................................... 23

22 Pipe spacing and thermal gradients for positive thermal gradient and negative
thermal gradient. ................ ......................... ............... 23

23 T typical m anifold in flange ......................................................................... .... 24

24 Typical manifold in web....................... ................................. 24

25 Flow rates for w eb m anifold. ............................................ ......................... 24

26 Flow rates for flange manifolds with straight inlet pipe ..................................... 25

27 M odified m anifold for the flange.................................... ................................... 25

28 Flow rates for m odified w eb m anifold............................................ ... ... .............. 25

29 H eating system s. ................................. .................. 26

30 Piping schematics for uniform temperature rise ................................................ 26

31 Piping schematics for AASHTO positive thermal gradient............................... 26

32 Piping schematics for AASHTO negative thermal gradient............................. 27

33 Beam layout with casting sequence. ....................................................... 30

34 Open form with steel reinforcement. ...................................................... 31

35 Closed form with reinforcing and lifting hooks .............................................. 31

36 Form for shear keys ............. ...................................... .............. ........... 32

37 Thermal segment with copper piping and thermocouples. .................................. 32

38 Finished concrete pours for a) Segment 1, b) Segment 2, c) Segment 3, d)
S eg m en t 4 ..................... ................................ ...... .......... ..... 3 3

39 Positioning a hydraulic cylinder on the top bar. ................................................ 35

40 Prestressed beam .................. ...................... ............ ................. 36

41 Load cell layout for final beam ................................... ........... ..... ......... 42

42 Load cell layout for prestressing. ..... ................................ ......................... .............. 42

43 Cross section view of prestress load cells......................... ... .......... ........ 42

44 Cross section of rectangular test beam ............................................ ... ... .............. 43









45 L oad setup for sm all beam tests.................................................. ... ................. 43

46 Strain ring on mounting blocks across joint. .................................................... 43

47 Modified mounts for strain rings. ................... ..................................... 44

48 Load vs. longitudinal strain (strain ring).................................... .................... 44

49 Incremental loading of strain rings. ....................................................... 45

50 Strain vs. loading for strain rings ................................................................... 45

51 Foil strain gauges at segm ent interface ............................................................ 46

52 Load vs. longitudinal strain (foil gauges). .................................. .............. 46

53 L oad v s. tip deflection .................... .......................................... .. ................ 46

54 Non-contact displacement transducer, LVDT and linear POT........................... 47

55 Load vs. relative joint displacem ent. .................................. ............. ......... 47

56 Thermal gradients and thermocouple layouts for a) positive thermal gradient,
b) negative therm al gradient. ........................................... ................... ...... 47

57 Thermocouple grid with attached thermocouples......................................... 48

58 Elevation view of segments 2 and 3 along with thermocouple grid locations...... 48

59 Thermocouple labels in Segment 2 and Segment 3........................ ............ 49

60 Elevation view of instrumentation layout (north side). ......... .................. 50

61 Elevation view of instrumentation layout (south side). .......... ................ 51

62 Detail view of instrumentation on segments 2 and 3 (north side). ....................... 51

63 Detail view of instrumentation on segments 2 and 3 (south side) ....................... 52

64 Plan view of instrumentation on segments 2 and 3....... .................................. 52

65 Typical instrumentation label for displacement device. ..................................... 52

66 Typical instrumentation label for strain device ............................... ................. 53

67 Sections at which thermocouples were embedded...................... .............. 58

68 Pipe layers used for thermal profile tests.................................. .................. .... 58

69 Therm couples used in each section................................................ .. .................. 59









70 Piping schematic for tap water flow. ............................................. ......... 59

71 Piping schematic for +41 OF uniform profile ................................................. 60

72 Uniform profile for segm ent 2. ........................................ ................. ...... 60

73 Uniform profile for Segment 3 .............. ............... ....................... .... 61

74 Piping schematic for linear profile....................... ........... ............ 61

75 Linear profile for Segment 2 ........ .............. .......................................... 62

76 Linear profile for Segm ent 3 ...................... .... ......... .................. .............. 62

77 Piping schematic for AASHTO positive gradient. ............................................. 62

78 Experimental AASHTO positive gradient .................................................... 63

79 Piping schem atic to heat beam ................... ... ............................. .............. 63

80 Piping schematic for AASHTO negative gradient............ ......................... 64

81 Experimental AASHTO negative gradient. ................................. .............. 64

82 Measuring device for AASHTO CTE tests. ................................ ............. 70

83 C TE test setup. ...................... .................................................. 71

84 Longitudinal displacements from LVDT's for Segment 2 for positive uniform
tem perature profile ........................... .................................. .... ........... .. 72

85 Longitudinal displacements from LVDT's for Segment 3 for positive uniform
tem perature profile ........................... .................................. .... ........... .. 72

86 Longitudinal displacements from LVDT's for Segment 2 for linearly increasing
tem perature profile ........................... .................................. .... ........... .. 73

87 Longitudinal displacments from LVDT's for Segment 3 for linearly increasing
tem perature profile ........................... .................................. .... ........... .. 73















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering

VALIDATION OF STRESSES CAUSED BY THERMAL GRADIENTS IN
SEGMENTAL CONCRETE BRIDGES: CONSTRUCTION, TEST SETUP, AND
COEFFICIENT OF THERMAL EXPANSION TESTS

By

David Clancy Walter

December 2006

Chair: H. R. Hamilton III
Cochair: Gary R. Consolazio
Major Department: Civil and Coastal Engineering

American Association of State Highway and Transportation (AASHTO) provisions

for checking service limit states require that the stresses from dead and live loads be

combined with those caused by heating and cooling. Of particular concern are the "self-

equilibrating" stresses caused by non-linear thermal gradients. The Florida Department

of Transportation has found that these stresses in combination with dead and live load

stresses can control the design and rating of segmental concrete bridges. Yet cracking or

other indications of distress that might be associated with these stress combinations have

not been noted in regular inspections of bridges in service.

This thesis presents the laboratory setup and methodologies implemented to

evaluate stresses resulting from thermal gradients in segmental concrete bridges. A 20-

foot long segmental concrete T-beam was constructed to determine the actual thermal

stresses that occur under a non-linear thermal gradient. The T-shaped cross section was









chosen to simulate the web-flange portion of a box girder. The test specimen consisted of

four 5 ft. segments that were match-cast and then prestressed together. Two of the

sections contained multiple layers of copper pipes that carried heated water through the

section. By varying the water temperature in each layer of piping, thermal gradients were

imposed on the cross section that closely matched the thermal gradients found in

AASHTO guidelines for the State of Florida. Load cells, displacement transducers, strain

gauges, strain rings, and thermocouples were used to collect data during heating and load

testing. In future work, the data from load tests will be analyzed to determine the

magnitude of the stresses caused by a combination of heating and loading.

An important variable in the project was the concrete's coefficient of thermal

expansion (CTE). This thesis also presents the setup, procedures and results of tests

conducted to determine the CTE of the test beam. The results are compared to CTE tests

conducted on concrete cylinders taken from the test beam pour.














CHAPTER 1
INTRODUCTION

Stresses caused by heating and cooling must be considered in the design of

segmental concrete bridges. In general, stresses are generated when the temperature of

all or part of the superstructure varies significantly from the temperature at which it was

constructed. Seasonal and diurnal variations in temperature are usually the cause of such

stress. The nature of the stresses that develop in the superstructure depends on the

magnitude of the overall temperature change as well as the magnitude of the temperature

gradient over the height of the cross-section. Thermal gradients are created when the

cross-section is heated or cooled unevenly over its height.

Stresses are generated when the superstructure is restrained by redundant supports

or when the thermal gradient is non-linear, which creates "self-equilibrating" stresses.

Non-linear thermal gradients cause cross-section distortion that is not compatible with the

fundamental kinematic assumption that plane sections remain plane. Internal stresses are

required to enforce this basic kinematic assumption. Lack of external forces ensures that,

to maintain static equilibrium, the stresses are self-equilibrating. Of particular interest are

the tensile stresses that form when the top of the cross-section is at a lower temperature

than the bottom.

There are no data in the literature showing that the self-equilibrating stresses have

been experimentally verified. Determining the state of stress under the application of

thermal and mechanical loading is made difficult because stresses cannot be measured

directly.









Additionally, the design analysis of thermal gradients does not account for possible

effects of creep. Heating of bridge concrete does not occur instantaneously, but rather,

over a period of hours. The response of the concrete to the thermal loading also occurs

over hours. During this long period, the high stresses predicted to occur from thermal

loading may not actually occur due to creep effects.

In an attempt to experimentally measure the effect of the non-linear thermal

stresses, a 20 ft. segmental T-beam was constructed. The beam will be subjected to

thermal loading and mechanical "tip" loading. Loads, deflections and strains will be

measured on the beam, and will be analyzed to determine the effects of thermal gradients

on segmental concrete bridges. Recommendations for future bridge analysis and design

will then be made.

This thesis documents the test beam design, laboratory setup and instrumentation

setup used in this study. Cross-section geometry, span length and the plumbing used

inside the beam to impose the thermal gradients are also documented. Laboratory setup

documentation includes loading frames, reaction frames, pumps and heaters. The types

of load, displacement and strain measuring devices are discussed in the instrumentation

section. The thesis also covers in-situ tests performed on the test beam to determine a

coefficient of thermal expansion.














CHAPTER 2
BACKGROUND

Thermal effects, particularly of thermal gradients, are of great concern to bridge

designers. They are believed to cause stresses in the bridge deck that may lead to

unwanted cracks. Current AASHTO guidelines set forth a procedure to calculate thermal

stresses and design for them. These codes may be overly-conservative in the calculations

of actual thermal stresses. A better understanding of actual thermal stresses is needed for

a more economical design of bridges that will maintain safety and serviceability.

This chapter covers the thermal gradients as set forth by the AASHTO guidelines.

Mechanical and thermal loading of the test beam are also addressed.

Thermal Gradients

Design for thermal effects in concrete segmental bridges is based on AASHTO

specifications. The specifications discuss time-dependent fluctuations in the bridge

temperature and also the temperature differentials within the superstructure. All concrete

bridges are designed for time-dependent temperature fluctuations (AASHTO 1989). For

prestressed concrete bridges, stresses and movements caused by temperature differentials

must also be considered. Stresses and movements do not cause a loss of strength;

however, they are significant because they can cause cracks in the bridge which may lead

to deterioration or corrosion of the bridge superstructure.

The two temperature differentials, or gradients, discussed in the AASHTO

provisions are the positive thermal gradient and the negative thermal gradient (AASHTO

2001). These gradients arise through the radiation, convection and conduction that a box-









girder bridge undergoes throughout the day and throughout the year. Because concrete is

a poor thermal conductor, the box-girder cross-section does not heat or cool uniformly

over its height, creating thermal gradients.

A positive thermal gradient occurs as the sun is rising and heating the top of the

bridge cross-section more than the lower portion. Compressive stresses are formed in the

bridge deck due to the higher temperatures in that region. The stresses are due to

nonlinearities of the temperature profile and are self-equilibrating, meaning that no

external reaction is required. They will be discussed in further detail in this chapter. As

the sun sets, the bridge begins to cool. The bridge deck cools to lower temperatures than

the rest of the superstructure due to outward radiation of the stored heat. This creates the

negative gradient in which the top of the cross-section is cooler than the bottom.

Designers are concerned with this gradient because it causes tensile stresses, particularly

during the winter season, and possibly cracks in the bridge deck.

Tables and charts are available (AASHTO 2001) that give minimum and maximum

effective bridge temperatures and maximum solar radiation based on surface type and

geographical location for both the positive and the negative thermal gradients. The

figures and tables shown below are used in the analysis procedure to predict expansion

and stresses caused from thermal loading. The figures and tables are taken from site

studies conducted throughout the United States on box-girder bridges. Figure 1, Table 1,

and Figure 2 are used collectively to establish the positive thermal gradient in a box-

girder bridge for a particular geographical location. The values, T1 through T3, are

relative temperature changes; they do not represent absolute temperatures in F. The

value T3 is taken as 0.0 unless a site specific study is conducted to determine another









value. For the negative thermal gradient, T values are taken as -0.30 or -0.20 times the T

values of Table 1. The multiplier of -0.30 is used for plain concrete decks; the multiplier

of -0.20 is used for decks with asphalt. The difference accounts for the insulating ability

of the asphalt on the bridge deck.

Design Thermal Gradients

The AASHTO guide (AASHTO 2001) specifies temperature differentials for both

the positive and negative gradient based on the zone of the country and the type of

surface used. The resulting design thermal gradients for zone 3 (Florida) are shown in

Figure 3 and Figure 4.

Concrete Beam Behavior under Thermal Loading

To facilitate analysis, nonlinear thermal gradients are divided into three

components: uniform temperature, linear thermal gradient and non-linear self-

equilibrating temperature gradient (see Figure 5). The uniform temperature component

causes uniform expansion or contraction of the structure; axial forces and moments are

formed if the structure is restrained against movement. The linear temperature gradient

component causes a curvature that forms secondary moments if the structure is restrained

against bending.

The third component of the analysis is the calculation of self-equilibrating stresses.

If an arbitrary unrestrained cross section is subjected to a nonlinear temperature profile,

the section fibers will deform in a way that the section does not remain plane. Typically,

however, the cross-section is assumed to conform to the Navier-Bemoulli kinematic

assumption for beams in which plane sections remain plane. Enforcement of this

kinematic assumption generates strains. The differences between the thermal strains that









would result from free expansion of the fibers and the developed strains of the plane

section lead to internal self-equilibrating stresses (see Figure 6).

T,


Depth of
S u p e r-
S tru ctu re


T,

Figure 1. Positive vertical temperature gradient.

Table 1. Basis for temperature gradients.


Zone T1 (F) T2 (F)
1 54 14
2 46 12
3 41 11
4 38 9

























Figure 2. Solar radiation zones for the United States.

+41 F


Figure 3. Positive thermal gradient for Florida.











-12.3 F


-3.3 F


OF


4"


12"





20"


Figure 4. Negative thermal gradient for Florida.

T, + T2 + T3 = TTo


Cross-Section Thermal Uniform Linear
Gradient Temperature Temperature
Gradient


Figure 5. Decomposition of a nonlinear thermal gradient.


Neutral Axis


Non-Linear
Self-Equilibrating
Temperature
Distribution


















STemperature induced
strain distribution assuming
that the section's fibers do
not influence one another


Final (linear) strain
distribution


Figure 6. Strain difference that leads to self-equilibrating stresses.














CHAPTER 3
TEST BEAM

This chapter covers the setup of the main test beam. It addresses the size and shape

of the beam, the reaction frame, the loading frame and the mid-support frame.

The objective of the test program was to determine whether the stresses formed

from the AASHTO positive and negative thermal gradients are as severe as the stresses

predicted by AASHTO recommended procedures. To accomplish this objective, a

segmental beam was designed that could be tested in the UF structures laboratory, yet be

of sufficient scale and geometry so that its behavior was similar to that of a typical full-

scale prestressed segmental concrete bridge.

Selection of Cross-Section Geometry

The Santa Rosa Bay (SRB) Bridge, located near Pensacola, Florida, was used as a

prototype for the design of the length and cross-section of the test beam (Figure 7). A

scaled version of the box girder section was considered for use in laboratory tests, which

would also require scaling of the thermal gradient. It was not practical to implement a

scaled version of the thermal gradient in lab conditions. The cross section of the SRB

Bridge can be simplified to that of an I-section that has the same area and weak axis

flexural stiffness as the box-girder section (see Figure 8). AASHTO does not stipulate

that the codes and analysis procedures are restricted to box girders, nor does it restrict the

shape to which a gradient is applied.

Figure 9 and Figure 10 summarize the results of an analysis on the simplified cross-

sections for self-equilibrating stresses (Mahama). Both of the design positive and









negative thermal gradients for Florida were considered. The negative gradient creates

tensile stress in the top of the cross-section, which, when combined with dead and live

load stresses from negative bending, could exceed the sum of the precompression and

concrete tensile strength, causing cracking. The tensile stresses, however, are restricted

to the top 4-in. of the cross-section, forming a relatively sharp gradient compared to the

cross-section depth.

Space and equipment constraints did not allow the full-scale section to be

constructed and tested. The primary focus of this research was the tensile stresses created

in the top of the section from the negative gradient. Consequently, it was decided to

construct a T-beam matching the geometry of top portion of the modified section as

illustrated by the hatching in Figure 9 and Figure 10. The figures also show the resulting

stress state in the proposed section when the AASHTO temperature gradients are applied

to the T-beam section. The important aspects of the stress profile are well approximated,

including the stress magnitudes and gradient in the top of the section. For both the SRB

Bridge and the T-beam, the maximum compressive stresses (shown as negative) occurred

at the top for the AASHTO positive thermal gradient. Likewise, the maximum tensile

stresses (shown as positive) occurred at the top for the AASHTO negative thermal

gradient. This unique approach of using a T-beam is permitted by the fact that the design

temperature profile is applied to the top 16-in. of the cross-section and is independent of

depth.

The width of the T-beam flange was chosen on the basis of recommendations in the

ACI Committee 318, Building Code Requirementsfor Structural Concrete and

Commentary (ACI 2002) and on the basis of the limited thermal energy available to









impose the gradients on the test beam. In very wide flanged prestressed beams, shear

deformations tend to relieve extreme fibers of the flange of longitudinal compressive

stress, which leads to a non-uniform distribution of stress. Therefore, the ACI code

suggests for design purposes that the width effective as a T-beam flange shall not exceed

one-quarter of the span length of the beam, and the flange width on each side of the web

shall not exceed eight times the slab thickness. A width of 2 ft. was chosen to meet these

standards and to ensure a uniform longitudinal stress distribution in the flange. The

width was also chosen on the basis of being able to supply enough thermal energy

(heated water) to impose the thermal gradients on the test beam. Details of the final test

beam cross-section geometry are shown in Figure 11.

Segmental Construction Details

Imposing the temperature gradients and measurement of stress were two of the

critical considerations when designing the test setup. To facilitate both of these, the test

beam was constructed segmentally using match cast construction techniques as was done

on the SRB Bridge. Segment lengths were chosen to match the 5-foot segments of the

SRB Bridge and to fit the tie-down spacing in the laboratory. Figure 12 shows a

schematic of the specimen in the load test setup with prestress force (PT) and tip load

(Q). The specimen was composed of four segments that were externally prestressed to

form a single 20-ft long beam.

Segmental construction allowed the piping used to heat and cool the concrete to be

easily restricted to segments 2 and 3, which were adjacent to the joint with the largest

moment. Furthermore, the joints were left dry, which allowed the joint between heated

segments (2 and 3) to open during loading. Incipient joint opening indicates a zero stress

state to which strain measurements and analytical models can be calibrated. Prestress









force, tip load and span were adjusted to form stress states similar to those found at the

mid-span and supports of the SRB Bridge.

Prestressing

An external post-tensioning (PT) system was used to apply prestressing to the test

specimen. The anchorages were suspended from the end segments to allow the PT force

location to be adjusted vertically relative to the cross-section centroid. The PT force was

positioned to develop stresses similar to the stress conditions found at the mid-span and at

the support of the SRB Bridge (see Table 2). The SRB Bridge was designed using

AASHTO HS20-44 live load stresses; however, for the laboratory tests, live loads taken

from the AASHTO HL-93 vehicle loading (LRFD) were used. In Table 2, M/St and

M/Sb represent the stresses at the extreme top and bottom fibers of the section,

respectively.

An external post-tensioning system was used to avoid problems with concrete void

space and interference with the heating system and instrumentation. Four 1 3/8 in.

diameter high-strength DYWIDAG thread bars were configured as shown in Figure 13

and Figure 14.

The PT anchorage was fabricated from structural steel shapes and plates as shown

in Figure 15. Steel channels were placed back to back with sufficient space to allow

passage of the PT bars. Stiffeners were added to the channels under the bar bearing

plates. A set of back-to-back channels was used for the top two bars and another pair of

channels was used for the bottom bars. These channel systems evenly distributed the PT

force over the web of the T-beam. The PT force within the concrete was gradually

developed over a few feet from each end of the beam.









Post-tensioning forces were monitored with load cells on each of the four bars (see

Figure 15 and Figure 16). Tandem sixty-ton Enerpac Hollow Core single-acting jacks

were used to apply the PT force. The jacks were pressurized with a manifold system

attached to a single pump, allowing the two bars on opposite sides of the web to be

stressed equally avoiding eccentric loading of the prestressing.

Steel Reinforcement

Steel reinforcement was required to resist the high PT forces in segments 1 and 4.

Number 3 vertical stirrups were placed at 1.75 in. o.c. to resist the principal tensile

stresses that developed in the general anchorage zone. This reinforcement was placed

within 27 in. from the ends of the beam in Segments 1 and 4. The reinforcement design

was based on the approximate method permitted by LRFD specifications. Outside of the

anchorage zone, vertical reinforcement was placed every 12" on center. It was also

necessary to design for local zone forces. AASHTO guidelines (AASHTO 2001) were

utilized to determine the local zone stresses that were being developed in the area that the

prestressing channels transferred the force into the concrete. A set of three confinement

spirals was used to resist the high anchorage zone stresses. The confinement steel for the

ends of the beam is shown in Figure 17.

Load Test Setup

The load test setup was designed so the mechanical loading was reversible; both

positive and negative moments could be applied at the joint between segments 2 and 3.

Figure 19 shows the test setup layout of the support system. The beam was supported by

a reaction frame at one end and a load cell support in the middle. The beam was

mechanically loaded at the opposite end (tip) through the use of a MTS hydraulic jack.

The beam was prestressed and then loaded at the tip until a joint opening occurred









between segments 2 and 3. Tests were performed on three different set-ups: tip load

only, AASHTO positive thermal gradient with tip loading, and AASHTO negative

thermal gradient with tip loading.

Details of the loading frame, reaction frame and mid-support frame are shown in

Appendix B. The loading frame consisted of steel W-section columns and deep channel

beams that supported the MTS jack. It was designed to handle loads up to 400 kips,

although the tip load would not exceed 55 kips. The reaction frame carried the equal and

opposite load that was transferred from the jack to the beam. It consisted of a load cell,

neoprene bearing pads and back-to-back channels that were supported by four all-thread

bars; the system was capable of resisting up to 200 kips. It represented a roller because

little or no moment or longitudinal resistance was developed from the frame.

The mid-support frame is shown in Figure 20; it represents a roller because it

allowed for longitudinal movement. As the beam was loaded, the segment interface at

the support frame was expected to open up or unpeel from the top of the beam

downward. The joint was predicted to open when the stress at the top became zero. The

beam interface would continue to open, as the tensile stresses created from the loading

moment were greater than the compressive forces created from the prestress force. The

support frame was subjected to loads that were approximately twice the tip load.

Segment Heating System

Heated water was used to achieve the desired positive and negative thermal

gradients. The water was passed through layers of copper pipes that were embedded in

the beam. The number of pipes in each layer was minimized to reduce the reinforcing

effect and reduction in cross-sectional properties. The necessary number of pipes in each

layer was found from experiments conducted on a separate 5-foot prototype beam, which









cross section is shown in Figure 21(a). This beam was heated to match the temperature

profile of the AASHTO positive thermal gradient and to match that of the negative

thermal gradient. Several combinations of pipe layers were tested to determine the

minimum number of pipes necessary to impose the thermal gradients. The piping

schematics shown in Figure 21(b) allowed for minimal piping without sacrificing the

ability to achieve the AASHTO thermal gradients.

Pipe layers 1, 2, and 4 were located near the slope changes of the AASHTO

thermal gradients (see Figure 22). The other pipe layers were positioned to aid in heating

the entire beam and in shaping the AASHTO gradients. The top layer of pipes was

placed as close to the top of the beam as possible while retaining adequate concrete

cover, which resulted in a slight non-linearity of the thermal gradient near the surface.

Results from thermal gradients tests indicated that this behavior was negligible.

Manifold systems were designed to distribute equal flow rates of heated water to

each pipe (see Figure 23 and Figure 24) to achieve a uniform temperature distribution

across the width of the beam. The smaller web manifolds consisted of a straight inlet and

outlet pipes, which allowed for similar flows. The flow rates were checked before the

manifolds were cast in the beam (see Figure 25). The same straight inlet/outlet pipe

approach did not work for the larger flange manifolds, as the greatest flow was about

three times the flow rate of the least flow (see Figure 26). Modified manifolds were

fabricated to ensure more similarity in flow rates. Figure 28 shows that the tested flow

rates were more similar for the modified manifolds than the flow rates in the straight-inlet

manifolds. In addition to testing the flow rates for the manifolds, temperatures were

tested on top and bottom of each pipe at the inlet side and outlet side. No temperature









changes were discovered between top and bottom and end-to-end. This was important

because equal temperature distribution on both top and bottom of the pipe and

longitudinally was the desired outcome.

The water heaters and pumps were very important components of this project

because they supplied the heat necessary to impose the thermal gradients. Heaters and

pumps that could handle temperatures up to 1350F were required. Seisco S-H-7 electric

heaters allowed high intake temperatures and discharged fully heated water nearly

instantaneously. The pumps, 12 HP Depco submersible pumps, were able to handle

temperatures up to 2000F and were able to supply the necessary flow rates to the

manifolds. Hoses used in the system were plastic braided flexible tubing, which could

handle large flow rates and extreme temperatures.

The pump, heater, and pipe schematic necessary to uniformly increase the cross-

section temperature is shown in Figure 30. Schematics for imposing the AASHTO

positive and negative thermal gradient are shown in Figure 31 and Figure 32,

respectively. Heaters are labeled 1 through 3. Heaters 1 and 2 are the main heaters that

provided high temperatures between 105 F and 140 F. Heater 3 was a supplemental

heater that was utilized for temperatures in the 86 F to 100 OF range.

9'-61" 24' 9'-61"


-I-
8" \\ 53.21
100100 14'-11" 8
59.267 1 7"
I \ ,I


Figure 7. Cross section of SRB bridge.








18


547"


31.5" 1


26.4"


196.8"


Figure 8. I-section representation of SRB bridge cross section.

547"
S1 +41 -622


S330
221


-568
118

172

202


I-1Ub
196 8"
TEMP GRADIENT SRB TEST BEAM
(F) (psi) (psi)


Figure 9. Self-equilibrating stresses due to AASHTO positive thermal gradient.


170
-35

52
61


196 8" TEMP GRADIENT SRB TEST BEAM
(F) (psi) (psi)


Figure 10. Self-equilibrating stresses due to AASHTO negative thermal gradient.














170 187 -622-568

118
117


24"


8"



10"


ndge (truncated at 3ft)
ed Section


172 ,330


-202 _-_-_-_ 221
Self-Equilibrating Stresses
for Positive Thermal Gradient
(psi)


Figure 11. Chosen shape for final test beam with self-equilibrating stresses.


Figure 12. Schematic of segmental beam test specimen.


m 11 i A


iapie z. A proximate service-i stresses in K nrinage (psi).

Effective Prestress Dead + Live Load

Support Mid-span Support Mid-span

P/A M/St M/Sb P/A M/St M/Sb M/St M/Sb M/St M/Sb

HS20-44 -700 -175 350 -700 460 -940 810 -1630 -610 1230
HL-93 -700 -175 350 -700 460 -940 810 -1735 -684 1380


PRESTRESS BAR



1 1/2"- -------


y=14.67"

3 3/8"
T3 3/8"


BACK TO BACK
C8X18.75 CHANNELS


Figure 13. Prestress assembly cross section view.


4"
-35

12"


9 1 -52. ----------SRB B
9--- Propos






-66 61
Self-Equilibrating Stresses
for Negative Thermal Gradient
(psi)


r nr\T\ T\ 1 \




































Figure 14. View of prestess assembly.


7' 20'
CHAIR (NOT SHOWN) LOAD CELL 3" 4"
HYDRAULIC CYLINDER


C8X18.75
CHANNEL


Figure 15. Prestress assembly in elevation view.

























Figure 16. Elevation views of prestress assembly.


12" -
h -- ^ 9-4?


# 3 STIRRUPS @1.75" # 3 STIRRUPS @12"


24"





#3 CONFINEMENT
SPIRALS


# 3 BARS

2 0"
10"


Figure 17. Mild steel reinforcements in Segments 1 and 4.


Figure 18. Mild steel reinforcement.


27"
























Figure 19. Frame locations.


(a) (b)


Figure 20. Mid-support frame a.) drawing, b.) picture.











24"

-,, 1,,
O1

oX 0 X 0 0 X o o X o0 -
3"
oXoXoXoXo Xo




oX oX o

17"- 3"
X X 1X

10"


24"

h4 1"


Figure 21. Piping schematics for a) prototype beam, b) test beam.

24"


. . LAYER 1
So 0" LAYER 2
S.o 0 0 0 LAYER 3


-12.3 F


-3.3 F


LAYER 4


LAYER 5


LAYER 6


10"

Figure 22. Pipe spacing and thermal gradients for positive thermal gradient and negative
thermal gradient.


+41 F


+11 F



OF -
F


8"





28"


o o
. o 0


0 0 0











WATER IN
FROM HEATERS


WATER RETURN
TO TANKS


SEGMENT SEGMENT


VU)
WATER OUT OF SEGMENT 2
INTO SEGMENT 3 THROUGH
FLEXIBLE PLASTIC TUBING


Figure 23. Typical manifold in flange.


WATER IN
FROM HEATERS


WATER RETURN
TO TANKS


5' 5'
SEGMENT 2 SEGMENT 3

10" E "


1 I.D. COPPER
TUBES


Figure 24. Typical manifold in web.


0.8
-

z 0.6
-
._ 0.4
LIL
0.2

0
Incoming
Flow


WATER OUT OF SEGMENT 2
INTO SEGMENT 3 THROUGH
FLEXIBLE PLASTIC TUBING


Figure 25. Flow rates for web manifold.


2'





8


1"I.D. COPPER
TUBES


I I* *


m












0.6

0.5

S0.4
a.
- 0.3
0
EL 0.2

0.1

0
Incoming
Flow -


Figure 26. Flow rates for flange manifolds with straight inlet pipe.





8 3"
8








Figure 27. Modified manifold for the flange.


0.6

0.5

0.4-
a.



0. -




0 --

Incoming H2* 1


2 3 4 5 6 7


Figure 28. Flow rates for modified web manifold.


2 3 4 5 6 7




































Figure 29. Heating systems.


o o HEATER1

HEATER 1 OUT o -. o-


-fe o.


-HEATER 2


* VALVE CLOSED


Figure 30. Piping schematics for uniform temperature rise.



HEATER OUT o o o o o 0 HEATER 1 IN
HEATER 3 OUT o H o o o HEATER 3 IN
MIXED OUT o. ... ,--MIXED









0 o 0 VALVE PARTIALLY OPEN
TO MIX WATER


Figure 31. Piping schematics for AASHTO positive thermal gradient.


HEATER 2 OUT---


o *


ffH3 ^P3
















HEATER1OUT, o o O --HEATER 1 IN
HEATER3OUT o o o --HEATER 3 IN
MIXEDOUT o MIXED--


-HEATER 2


* VALVE PARTIALLY OPEN
TO MIX WATER


Figure 32. Piping schematics for AASHTO negative thermal gradient.


HEATER 2 OUT *


-* *














CHAPTER 4
BEAM CONSTRUCTION

The test beam was constructed at the FDOT Structures Research Center in

Tallahassee. Beam design plans (see Appendix A) from the UF research team were

submitted to the FDOT and construction began during the summer of 2006. All

formwork, rebar setup and thermocouple grid construction was performed by the FDOT

employees, particularly Frank Cobb. The concrete pours for all the beam segments also

took place in Tallahassee. When the segments were cast and cured, they were shipped to

the University of Florida Structures Lab.

Forms

A single set of forms was used to individually cast each segment. The casting

sequence is shown in Figure 33. Segments 1 and 4 contained the mild steel

reinforcement for the prestress anchorage zone and a steel bearing plate to distribute the

anchorage force into the concrete (Figure 34 and Figure 35). The segment joints

contained two shear keys to assist in alignment when post-tensioning (Figure 36). After

Segment 1 and 4 were cast, the thermal segments (Segment 2 and 3) were cast. These

segments contained a grid consisting of copper pipes, thermocouples and mild reinforcing

(see Figure 37).

Concrete

Figure 38 shows the finished concrete pours for all four segments. The dates of the

concrete pours are shown in Table 3. A 7000-psi pump mix was used to cast each of the

four segments. The design mix proportions are shown in Table 4 and the delivery tickets









for each mix are provided in Appendix C. Although the slump of the concrete is listed as

5 inches, the mix was delivered with a lower water content and slump to allow the slump

to be adjusted just prior to placement. When Segment 3 was cast, the weather was rainy

and the concrete in the first truck had a slump that exceeded the specifications. The first

truck was rejected, and a second truck with a lower slump concrete was delivered. The

segments were cast and cured inside the FDOT research lab. Each segment was cured for

a week before it was removed from the form.

Fifteen 6-in. diameter cylinders were taken from each concrete pour. Compression

tests and modulus of elasticity (MOE) tests were performed on the cylinders at the FDOT

State Materials Office (SMO). For each segment, three cylinders were tested. The first

cylinder of each segment was tested for breaking load and stress. The next two cylinders

of each segment were instrumented with displacement transducers and were loaded to

40% of the breaking load. The recorded displacements were converted to strains and an

MOE was found. After the cylinders were loaded to 40%, the instrumentation was

removed and the cylinder was loaded to failure. The failure loads are shown in Table 5

and the values of the MOE for each segment are shown in Table 6.

Delivery, Storage and Prestressing

After all the segments were cast and allowed to cure in Tallahassee, they were

delivered via a large flat bed truck to the structures lab at the University of Florida.

When they arrived on February 23, 2006, the segments were placed by fork lift on their

sides in the lab. They were then instrumented and prepared for CTE and thermal gradient

tests, which are reported in later chapters.

When the CTE and thermal gradient tests were completed, the segments were

aligned in the load test position shown in Figure 40. Each segment was placed on two









wood frame blocks to ensure stability during tensioning. The segments were placed as

close to each other as possible to avoid excessive movement as the joints closed during

tensioning. The anchorages and bars were positioned at 1.4 inches below the beam

centroid after segment alignment, and the hydraulic cylinders were positioned to tension

the PT bars. During the first prestressing, which occurred on June 2, 2006, the prestress

loads were applied slowly to ensure that all the segments were brought together and lined

up properly. The beam was prestressed in several stages. First, the two hydraulic

cylinders were positioned on the top bars as shown in Figure 39. Through the use of a

piping manifold system, the top two bars were prestressed simultaneously until a pressure

developed in the hand jack, which indicated a stressing of the bars. The hydraulic

cylinders were then positioned on the bottom two bars, and the same procedure was

followed. The beam segments came together and aligned successfully with the help of

the shear keys. After the beam came together and was aligned, the bars were loaded in

increments that are shown in Table 7. The beam was prestressed again on July 5, 2006;

Table 8 shows the load increments for the prestressing procedure. Figure 40 shows the

beam after the initial prestressing.


Figure 33. Beam layout with casting sequence.

























Figure 34. Open form with steel reinforcement.


Figure 35. Closed form with reinforcing and lifting hooks.































Figure 36. Form for shear keys.


Figure 37. Thermal segment with copper piping and thermocouples.









Table 3. Dates of concrete pours.
Segment Number Date Cast

Segment 1 1 December 2005

Segment 4 12 December 2005

Segment 2 19 January 2006

Segment 3 2 February 2006


/
(b)


(c) (d)

Figure 38. Finished concrete pours for a) Segment 1, b) Segment 2, c) Segment 3, d)
Segment 4.










Table 4. Concrete pump mix proportions.
Mix Number FC82JC
Strength (psi) 7000
W/C Ratio 0.31
Slump (in) 5 +/- 1"
Air Content (%) 4.5 +/- 1.5%
Plastic Unit Weight (lbs/cf) 140.1 +/- 1.5

Material ASTM Type
Cement C 150 I/II 820
Cement C 618 F. Ash 160
Water -- -- 304
Fine Aggregate C 33 Sand 1095
Aggregate C 33 #89STONE 1400
Admixture C 260 AIR Dosage rates vary
Admixture C 494 W/Reducer with manufacturers
recommendations

Table 5. Results from compression tests.
Age of Specimen
Segment Age of Specimen Breaking Load Breaking Stress
Number at Time of Test Test Number bs) (psi)
Number (Ibs) (psi)
1 263,000 9302
772 266,160 9413
77
3 255,880 9050
Average 261,680 9,255
1 200,520 7092
2 22 196,120 6936
2 28
3 199,690 7063
Average 198,777 7,030
1 224,670 7946
32 209,000 7392
3 28
3 215,910 7636
Average 216,527 7,658
1 257,440 9105
42 248,300 8782
4 66
3 243,550 8614
Average 249,763 8,834

Table 6. Moduli of elasticity for each segment.
Segment Number Modulus of Elasticity (psi)
Segment 1 4,862,282
Segment 4 4,575,301
Segment 2 3,835,573
Segment 3 4,000,000

































Figure 39. Positioning a hydraulic cylinder on the top bar.


Table 7. Prestress load increments for tests performed on July 2, 2006.
Load Load Cell Readings(kips)
Step L1 L2 L3 L4
1 -- -- 9.1 13.9
2 19.3 32.8 9.1 13.9
3 19.3 32.8 53.3 69.8
4 31.4 42.0 53.3 69.8
Final Load 24.7 28.3 51.5 62.6

Table 8. Prestress load increments for tests performed on July 5, 2006
Load Load Cell Readings(kips)
Step L1 L2 L3 L4
1 29.2 18 -- --
2 29.2 18 43.8 46.7
3 48.3 32.2 43.8 46.7
4 48.3 32.2 76.3 68.8
5 76.7 68.8 76.3 68.8
6 76.7 68.8 97.6 93.2
7 94.6 100.8 97.6 93.2
8 94.6 100.8 107.7 100.9
9 106.5 112.1 107.7 100.9
10 106.5 112.1 111.8 106.1
Final Loads 97.2 89.2 93.3 92.4






36














Figure 40. Prestressed beam...

Figure 40. Prestressed beam.













CHAPTER 5
INSTRUMENTATION

Thermocouples, strain rings, foil gauges, linear variable displacement transducers

(LVDTs) and load cells were used to measure the response of the beam to mechanical

and thermal loading. Details of instrumentation and the data acquisition (DAQ) system

are covered in this chapter and include the selection process, methods, mounts, and

calibrations used for the final beam instrumentation.

Load Cells

Load cells were used to measure the tip load and prestress forces. The layout of the

load cells are shown in Figure 41. Load cells L1, L2 and L3 were used to measure the

reaction forces; load cells L4 through L7 were used for the prestress forces. The applied

tip load was a maximum of 55 kips, which was measured by a built-in load cell in the

hydraulic jack. Load cells L2 and L3 were placed to measure the forces at the load cell

support frame under segment 2 and at the reaction frame over segment 1, respectively.

The expected maximum load at L1 was 55 kips, which would result in a maximum load

of 110 kips at L2. A 75 kip Geokon load cell was used at L3, and a 150 kip Geokon load

cell was used at L2. Four 200 kip hollow core load cells were placed as shown in Figure

42. Figure 43 shows the layout of the PT load cells.

Deflections

DCTH Series LVDTs from RDP Electrosense were used in the final test beam to

measure all deflections. They were placed vertically to measure tip deflection, deflection

at the mid support and deflection at the reaction frame. Horizontally, they were placed









throughout the depth of the beam across the interface of Segments 2 and 3. They were

also placed at the top of the beam across the interface of Segments 2 and 3. The LVDTs

were positioned across the interface to determine if there was joint opening.

Joint Opening

Tests were conducted on a small rectangular beam to determine if joint opening

could be detected using strain gages, strain rings, or linear potentiometers. The test beam

consisted of two 30 in. long rectangular beams that were match-cast and prestressed

together. Prestressing was applied with two 0.5-in. diameter threaded bars inside PVC

conduits that were cast into the segments at 1 in. above and 1 in. below the centroid of the

cross-section (Figure 44). The beam was placed in the loading frame shown in Figure 45

and the gage to be tested was installed on the beam at the segment. The beam was

loaded, and data were recorded and plotted vs. tip load to determine the sensitivity of the

gage to crack opening. Opening of the joint was confirmed with visual observations.

Joint opening was important because it indicated that there was zero flexural stress in the

concrete at the top of the segment interface. The following sections detail the results of

the tests conducted on each of the gages evaluated.

Strain Rings

Strainstall Type-5745 sealed strain rings were used to measure longitudinal strains.

Figure 46 and Figure 47 show two setups of the strain rings across the beam interface.

Figure 48 is representative strain data of the mounts across the joint interface. After

several tests using either mounting condition, no conclusive data that would confirm a

joint opening was found because the curves were non-linear. The interface between the

two segments was not perfectly matched due to shrinkage effects during the concrete

curing. This could have added an effect where the segments rotated about high points in









the interface during loading and not allowed for full cross sectional properties.

Additionally, it was believed that the gauges were adding an extra variable of stiffness

across the joint. Therefore, in the final test beam, these gauges were not used across the

joint. The strain rings were used on the final test beam to record strains at the centroid of

the beam close to the segment interface between the two thermal segments. They were

placed on either side of the beam, and were used to supplement the foil gauge data.

The strain rings were tested for linearity to ensure that the signal sent from the

strain rings did not vary nonlinearly. The strain rings were loaded incrementally with 1

kilogram weights (approximately 2.2 pounds) and strains were measured (see Figure 49).

Figure 50 shows that the loading and unloading of the strain ring was linear.

Foil Gages

Strain foil gauges were used to measure strains through the depth of the beam. The

gauges were selected based upon the necessity to measure concrete strains, to handle

temperatures higher than 140 F, to be short in contact length, and to be able to

compensate for temperature. A 3-wired 60 mm "PL" series foil gauge from Texas

Measurements was selected based on these stipulations. The gauges were placed

throughout the depth of the test beam near the joint interface (Figure 51). Strain data was

plotted vs. tip load to determine any crack opening (Figure 50).

On the final test beam, the strain gauges were also placed throughout the depth of

the beam on the two thermal segments. They were placed near the segment interfaces

between segments 1, 2, 3 and 4 to determine the behavior at the joints under loading

conditions. They were also placed at the center of Segment2 and the center of Segment 3

to determine the strain behavior there and the differences, if any, from the strain behavior

at the joints.









Displacement Devices

A linear potentiometer (POT) was used in the small beam tests to measure the

deflection of the tip of the beam. The tip deflection was plotted versus tip load. It was

found that the beam would deflect linearly up to a certain load, gradually change slope,

and then deflect linearly again (Figure 53).

An Omega LD-700 Series non-contact displacement transducer (NCDT), an

Omega LD400 Series LVDT, and a linear POT (LP) were used to attempt to detect joint

opening (see Figure 54). The instruments were able to accurately measure in the 1/1000

of an inch range, which was necessary to measure joint opening. The data collected from

these devices were plotted against the tip load (see Figure 55). Data collected from these

instruments followed a similar trend to the data from the load vs. tip deflection data.

Relative joint displacement was linear, gradually changed slope and then became linear

again.

Thermocouples and Inline Temperature Sensors

Thermocouples were fabricated in the lab and embedded in the concrete to measure

temperatures within the beam. A Teflon neoflon type T thermocouple wire was selected

due to it ability to handle up to 392 F temperatures. Thermocouples were positioned to

measure temperatures at the heights corresponding to slope changes in the thermal

gradients, see Figure 56. They were also placed at intermediate points between slope

changes to verify linear thermal gradients. Thermocouple cages, consisting of small

diameter steel, were fabricated to ensure that the thermocouples were placed at exact

depths of the beam. Figure 57 shows the thermocouple grids placed in the form.

There were six sets of thermocouple grids; three placed in Segment 2 and three

placed in Segment 3 (see Figure 58). The grids close to the segment interfaces were









placed three inches from the ends to accurately measure temperatures near the interface.

In both segments 2 and 3, thermocouples were placed in the center of the segment to

record data that would demonstrate that the temperatures throughout the beam length

were uniform.

Final Instrumentation Setup.

Final instrumentation setup was based on tests conducted on the small rectangular

beam and on available channels in the DAQ system. The system consisted of three

chassis that contained modules that read input signals from terminal blocks and converted

and output the signal to a laptop computer. The channel inventory is shown in Table 9.

The three chassis that were used provided 12 available slots for modules. Seven of the

slots were used for voltage thermocouplee) terminal blocks, and 5 slots were used for

strain module terminal blocks.

An overview of the instrumentation is shown in Figure 60 and Figure 61. Details

of the gages on Segments 2 and 3 are shown in Figure 62 and Figure 63. The instruments

were given the following labeling designation: L for load and S for strain. Load cells

were designated with a label such as L3, in which the L indicates load cell and the 3 is the

load cell number. Linear potentiometers and strain gages were labeled as shown in

Figure 65 and Figure 66.











O LOAD CELL


L4
L5
SL6
SEG 1 SEG 2 SEG 3 SEG 4 L7



L2

Figure 41. Load cell layout for final beam.

ANGLE:
2 X 2 X 0.25"

CHAIR (NOT SHOWN) LOAD CEL

HYDRAULIC CYLINDER


1-3" PLATE AND
NUT FROM DSI


C8X18.75
CHANNEL


1- 3" \U+2205 DYWIDAG
8
THREADBAR (150 ksi)
1/" INBED PLATE
/2


Figure 42. Load cell layout for prestressing.


PRESTRESS BAR


O LOAD CELL


Figure 43. Cross section view of prestress load cells.


.11 1 II
















O PVC PIPE
1" ALLTHREAD BAR


Figure 44. Cross section of rectangular test beam.
301" 30"
3-1/4"


Figure 45. Load setup for small beam tests.


Figure 46. Strain ring on mounting blocks across joint.


S2-5/8"






























Figure 47. Modified mounts for strain rings.


1200
1100
1000
900
800
700
600
500
400
300
200
100
0
0 200 400 600 800 1000 1200 1400 1600
Strain (microstrain)


Figure 48. Load vs. longitudinal strain (strain ring).











































Figure 49. Incremental loading of strain rings.


25


20


S15
LOADING

O 10 UNLOADING
.-J
5 Linear (LOADING)

Linear
(UNLOADING)
0
0 200 400 600 800
Strain (Microstrain)


Figure 50. Strain vs. loading for strain rings.





























Figure 51. Foil strain gauges at segment interface.


1300
1200
1100
1000
900
S800
700
S600
0 500
400
300
200
100


0 O. I
-100 -80 -60 -40 -20 0 20 40 60
Strain (microstrain)


Figure 52. Load vs. longitudinal strain (foil gauges).


1800
1600
1400
S1200
S1000
800
-I 600
400
200
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Deflection (in)


Figure 53. Load vs. tip deflection.


fgl
fg2
fg3
fg4
fg5
fg6
fg7
fg8


80 100


AT







47





















Figure 54. Non-contact displacement transducer, LVDT and linear POT.


1600 NcuD
1400 LVDT', LP

1200
1000
800 -LP
0o NCDT
-. 600 -- LVDT
400
200
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Relative Joint Displacement (in)


Figure 55. Load vs. relative joint displacement.

24" 24"
+41 F 1 -12.3 F 1 2
r -r
,, _4" 4 4" 4 "
12" T *. 12"
4 @ 2" 4 @ 2" 2
O F-I- OF
4 @5" 4 @5"
20" 20" *
I 1,, 1,,

4 4 4 4
10" 10.
(a) (b)


Figure 56. Thermal gradients and thermocouple layouts for a) positive thermal gradient,
b) negative thermal gradient.







48

























Figure 57. Thermocouple grid with attached thermocouples.

THERMOCOUPLE GRID



SEGMENT 2 j i SEGMENT 3


SECTION A SECTION B SECTION C SECTION D SECTION E SECTION F

Figure 58. Elevation view of segments 2 and 3 along with thermocouple grid locations.














A-T-1

A-T-6


B-T-1 B-T-2 B-T-3 B-T-4 B-T-5

B-T-6 B-T-7 B-T-8 B-T-9 B-T-10
B-T-11 B-T-12 B-T-13 B-T-14 B-T-15


B-T-16 B-T-17 B-T-18
B-T-19 B-T-20 B-T-21
B-T-22 B-T-23 B-T-24
B-T-25 B-T-26 B-T-27



B-T-28 B-T-29 B-T-30


A-T-2 A-T-3 A-T-4 A-T-5

A-T-7 A-T-8 A-T-9 A-T-10
A-T-12 A-T-13 A-T-14 A-T-15

A-T-16 A-T-17" A-T-18
AA-T-19 AA-T-20 A-T-21
A-T-22 A-T-23 A-T-24
*A-T-25 A-T-26i A-T-27



A-T-28 A-T-29 A-T-30


A-T-31 A-T-32" A-T-33


A-T-34 A-T-35 A-T-36

A-T-37 A-T-38 A-T-39


SECTION A


C-T-1 C-T-2 C-T-3 C-T-4 C-T-5

C-T-6 C-T-7 C-T-8 C-T-9 C-T-10
C-T-11 C-T-12 C-T-13 C-T-14 C-T-15

C-T-16 C-T-17" C-T-18
TC-T-19 C-T-20 T C-T-21
TC-T-22 C-T-23 T C-T-24
C-T-25 C-T-26 T C-T-27



C-T-28 C-T-29 C-T-30


C-T-31 C-T-32 C-T-33


C-T-34 C-T-35 C-T-36

C-T-37 C-T-38 C-T-39


SECTION C


D-T-1 D-T-2 D-T-3 D-T-4 D-T-5

D-T-6 D-T-7 D-T-8 D-T-9 D-T-10
D-T-11 D-T-12 D-T-13 D-T-14 D-T-15

D-T-16 i-T-17 D-T-18
D-T-19 D-T-20 D-T-21
D-T-22 D-T-23i D-T-24
SD-T-25 D-T-26i D-T-27



D-T-28 D-T-29 D-T-30


D-T-31 D-T-32 D-T-33


D-T-34 D-T-35i D-T-36

D-T-37 D-T-38 D-T-39


SECTION D


E-T-1 E-T-2 E-T-3 E-T-4 E-T-5

E-T-6 E-T-7 E-T-8 E-T-9 E-T-10
E-T-11 E-T-12 E-T-13 E-T-14 E-T-15

E-T-16 E-T-17 E-T-18
E-T-19 E-T-20 E-T-21
E-T-22 E-T-23 E-T-24
E-T-25 E-T-26 E-T-27



E-T-28 E-T-29 E-T-30


E-T-31 E-T-32 E-T-33


E-T-34 E-T-35 E-T-36


E-T-37 E-T-38


E-T-39


SECTION E


F-T-1 F-T-2 F-T-3 F-T-4 F-T-5

F-T-6 F-T-7 F-T-8 F-T-9 F-T-10
F-T-11 F-T-12 F-T-13 F-T-14 F-T-15

F-T-16 F-T-17 F-T-18
F-T-19 F-T-20 F-T-21
F-T-22 F-T-23" F-T-24
TF-T-25 'F-T-26' F-T-27



F-T-28 F-T-29 F-T-30


F-T-31 F-T-32 F-T-33


-T-34 F-T-35 F-T-36

F-T-37 F-T-38 F-T-39


SECTION F


Figure 59. Thermocouple labels in Segment 2 and Segment 3.


B-T-31 B-T-32 B-T-33


B-T-34 B-T-35 B-T-36

B-T-37 B-T-38 B-T-39


SECTION B







50


Table 9. Inventory for available data acquisition channels
CHASSIS
SLOTS TOTAL#
# OF
PER OF
CHASSIS
CHASSIS SLOTS
3 4 12
MODULES
MODEL DESCRIPTION COUNT TOTAL CHANNELS
SCXI 1520 8 CHANNEL STRAIN MODULE 5 40 STRAIN
SCXI 1102 32 CHANNEL T-COUPLE 4 128 T-COUPLE
SCXI 1102C 32 CHANNEL T-COUPLE 3 96 T-COUPLE

TERMINAL BLOCKS
MODEL DESCRIPTION COUNT TOTAL CHANNELS
SCXI 1314 8 CHANNEL STRAIN BLOCK 5 40 STRAIN
SCXI 1303 32 CHANNEL T-COUPLE BLOCK 4 128 T-COUPLE
SCXI 1300 32 CHANNEL T-COUPLE BLOCK 3 96 T-COUPLE

CAPABILITY
AVAILABLE SLOTS WITH 3 CHASSIS : 12

NUMBER OF T-COUPLES : 224
USING : 7 SLOTS
LEAVING : 5 SLOTS AVAILABLE

NUMBER OF STRAIN DEVICES : 40
USING : 5 SLOTS


LVDT
H LOAD CELL
FOIL GAUGE
-e- STRAIN RING


S1-T-D1


SEG 1


S2-T-D1 S3-T-D1


L1
Hq


SEG2 SE


S4-T-D1


Figure 60. Elevation view of instrumentation layout (north side).


.. __ .


L4
L5
L6
L7


SEG 3


SEG 4









51



LVDT
H LOAD CELL
FOIL GAUGE
-e- STRAIN RING


S S1-T-D1

4


L2 I



Figure 61. Elevation view of instrumentation layout (south side).


-4 LVDT

D LOAD CELL

J1 s- FOIL GAUGE

S-- STRAIN RING


S2-T-D1 S3-T-D1



2-N-S-30-35 3SSN-S-30-35
S2-N-S-30355 S2-N-S-02-3340 H 53N-S-02-335
S2-N-S-30-285 S2-N-S-02-304 0 -0 S3-N-S-02-303 SEGMENTS
0 D 53N-S-30-28 5


S2-N-S-02-275 H

5S2-N-S-30-21 2 S2-N-R-03 7-21
S2-N-S-3 7-21 3 0-

-GS2-N-S-30-146 2-N-S-02-145 H

S2-N-S-02-77
- S2-N-S-30-79

S2-N-S-30-1 0 S2-N-S-02-1 0


S3-N-S-02-274

S3-N-R-03 7-21 3
-0-a
- 53-N-S-3 7-21 3

SS3-N-S-02-146

H S3N-S-02-78


S3-N-S-02-1 1


C.G. 3-N-S-30-21 4


- S3-N-S-30-14 6


- S3-N-S-30-7 7

3-N-S-30-1 1


S3-N-S-58-35 5

S3-N-S-58-28 5
53N5582


J3


S3-N-S-58-21 3



S3-N-S-58-14 6-1


S3-N-S-58-7 809-

S3-N-S-58-1 1


L2 LI



Figure 62. Detail view of instrumentation on segments 2 and 3 (north side).


S2-N-S-58-35 5
SEGMENT 2
S2-N-S-58-28 6


S2-N-S-58-21 2



W--S2-N-S-58-14 6


--S2-N-S-58-8 5

5S2-N-S-58-1 2


---t-


I













LVDT

D LOAD CELL

. FOIL GAUGE

-0- STRAIN RING


J2 -



S3-T-D1 S2-T-D1


Ji


S-58-3334 S S-30-354 53-SS-02-353 S2-S--02-355 S2-S-S-30-355 -S-S- 6
$53-S5-58-33 4 I F J2-SD-3375 525558336
0 SEGMENT 3 3-SS&02-32 H -H 2-S-S02-32 SEGMENT2
SS-58-304 -- J2-SD-3025 S2-S-S-58-304
$3-5-S-30-28 6a 14.67 3-S-S-02-28 6 S2-S--02-285 S2-S-S-30-285 0


3-S-S-58-21 3 3-S-S-30-21 4 C.G.


S3-SS-30-14 9-


S3-S-S-30-7 8H-

S3-S-S-30-1 O


S3-S--02-27 5
I-
53-S-R-3 7-21 3-
S3-SS37-2133
I-


S3-S-S-02-7 9

-SS02-09
S3-S-5-02-09


S2-S-S-02-27 3
S J2-SD-250
2S2-SR-3 7-21 2
S2-SS-3 7-21 2
1 J2-SD-1825
SS2-S-S-02-14 5
S J2-SD-11 375
SS2-S-S-02-7 7
S J2-S-D-4375
S2-S-S-02-0 9


C.G. S2-S-S-30-21 2 S2-S-S-58-21 2
pz --


S2-S-S-30-146 2-S-S-58-14 5 -
o

S2-S-S-30-8 0 S2-S-S-58-7 7

SS2-S-S-30-0 9 S2-S-S-58-0 9


Figure 63. Detail view of instrumentation on segments 2 and 3 (south side).



FOIL GAUGE

= i LVDT


SOUTH





NORTH


12"


S2-TS-S-5 5-8 25


-I J2-TS-D-9 5
S3-TS-S-5 5-7 25


J2-T-D J e


S2-TN-S-5 5-7 75 0-
E -


BEAM
S3-T-S-5 5-0 75 BEAM
CENTERLINE


S3-TN-S-5 5-7 25
-I J2-TN-D-9 5


Figure 64. Plan view of instrumentation on segments 2 and 3.


Segment Number

Top Face -I

Displacement Device Number C




Figure 65. Typical instrumentation label for displacement device.


J3


- 53-SS-58-14 6
o

-- S3-SS-58-7 8

S3-S-S-58-1 0


4L2






53


Segment Number C

North (N) or South (S) Face z
Strain Gauge on
Distance from Joint 2 g

Distance Above Bottom of Beam o
Figure 66. Typical instrumentation label for strain device.
Figure 66. Typical instrumentation label for strain device.














CHAPTER 6
METHODOLOGY OF IMPOSING THERMAL PROFILES

The thermal profiles were imposed on the beam by pumping heated water through

the copper pipes that were embedded in the concrete. Four different profiles were

imposed on the beam. A uniform profile and a linear profile were imposed individually

on beam Segments 2 and 3 for coefficient of thermal expansion (CTE) tests. The other

two profiles, the AASHTO positive and negative thermal gradient, were imposed with the

segments together.

The methods used to impose the gradients are discussed in this chapter.

Throughout the chapter, pipe layers, heaters and thermocouple sections are referenced.

Heaters are labeled as described in Chapter 3. Figure 68 through Figure 69 show the pipe

labelings, thermocouple section labeling and the thermocouples used in each section.

The X markings in Figure 69 show the thermocouples that were not read due to

limitations on the number of instrumentation channels that could be simultaneously read

by the data acquisition system.

Uniform Profile

CTE tests were conducted on Segments 2 and 3 by imposing a uniform increase in

temperature of 41 F and measuring the resulting longitudinal displacement. Tap water,

with a temperature of approximately 80 F, was passed through the segment overnight to

ensure a uniform starting temperature (see Figure 70). The starting temperatures were

recorded and used to determine the necessary absolute temperatures to impose the 41 F

uniform profile.









Heated water was then passed through the beam as illustrated in Figure 71. The

heaters were initially set to about 10 F higher than the required +41 F temperature to

expedite the beam heating process. Throughout the test, the temperatures in the concrete

were monitored. When the temperature changes approached +41 F, the heater settings

were adjusted to match the required temperatures. When the desired temperatures were

achieved, they were held in steady state by allowing the pumps to continue passing the

heated water through the pipes. Figure 72 and Figure 73 show the experimental profiles

for beam Segment 2 and 3, respectively. In Segment 2, the temperature changes

exceeded the target temperature change of +41 F due to overheating the segment. The

actual average uniform temperature change was + 42 F. The temperature change in

Segment 3 matched the target temperature change of 41 F. Throughout the depth of the

beam, the temperatures fall within 1 degree of the target temperature, which is about a

2% difference from the target temperature change. Due to constraints of the heating

systems and laboratory conditions, this difference was found to be acceptable.

Linear Profile

CTE tests were also conducted on segments 2 and 3 by imposing a linear increase

in temperature over the height of the section that was varied from 0 F at the bottom of

the beam to 41 F at the top of the beam. Tap water was passed through the beam to

provide a uniform starting temperature.

Heated water was then passed through the beam as shown in Figure 74 to impose

the profile. Table 11 shows the required absolute temperatures in each pipe layer to

achieve the linear profile based from the starting ambient temperatures. Initially the

heaters were set higher than the required temperatures to expedite the beam heating

process. Temperatures in the concrete were closely monitored throughout the test, and









the heaters were adjusted to match the required temperatures. When the desired

temperatures were achieved, they were held in steady state by allowing the pumps to

continue running. Figure 75 and Figure 76 show the experimental profile along with the

desired linear profile for beam Segment 2 and 3, respectively.

AASHTO Positive Thermal Gradient

The AASHTO positive thermal gradient was imposed in two steps. The first step

was to bring the sections to a uniform temperature similar to the CTE tests. The second

step was to impose the gradient.

Figure 80 shows the piping setup used to impose the temperature change necessary

to create the gradient. Tap water continued to run through the pipe layers in the web to

ensure that the original beam temperature was maintained at depths greater than 16 in.

from the top. Temperatures of the water that passed through layers 1 thru 3 were set to

the absolute temperatures shown in Table 12. Initially the temperatures were actually set

higher than the required temperature to expedite the beam heating process. Temperatures

in the concrete were closely monitored throughout the test, and the heaters were adjusted

to match the required temperatures. When the desired temperatures were achieved, they

were held in steady state. Figure 78 shows the experimental profile along with the

theoretical AASHTO positive gradient. The temperatures matched within 0 to 2 OF of the

target temperatures, with the exception of one layer in Section A that had an average

temperature within 3 F of the target temperature. These temperatures matched closely to

the target profile and were found to be acceptable with the given laboratory and heating

systems.









AASHTO Negative Thermal Gradient

The AASHTO negative thermal gradient was imposed in two steps. The first step

was to uniformly heat the beam overnight to a temperature around 122 F. Having the

beam elevated to this temperature allowed for the heaters to be in their working range

when backing down temperature changes to -12.3 F. After the beam was uniformly

heated, the negative gradient was imposed.

Figure 80 shows the piping setup used to impose the temperature change necessary

to create the gradient. The high temperature water continued to run through the pipe

layers in the web to ensure that 0 F temperatures were maintained at depths greater than

16 in. from the top. The temperatures of the water that passed through layers 1 through 3

were set to the absolute temperatures that are shown in Table 13. Initially the

temperatures were set lower than the required temperatures to expedite the beam cooling

process. Temperatures in the concrete were closely monitored throughout the test, and

the heaters were adjusted to match the required temperatures. When the desired

temperatures were achieved, they were held in steady state. Figure 81 shows the

experimental profile along with the theoretical AASHTO negative gradient. The

temperatures are within 0.5 F, with the exception of one temperature average in Section

A that is within 1 F.







58


THERMOCOUPLE GRID



SEGMENT 2 SEGMENT 3


SECTION A SECTION B SECTION C SECTION D SECTION E SECTION F

Figure 67. Sections at which thermocouples were embedded.

24"

S o o... LAYER 1
8" ..... LAYER 2
0... 0 LAYER3


36" o 0 LAYER 4

28"
S LAYER 5


o LAYER 6


10"

Figure 68. Pipe layers used for thermal profile tests.







59


















SECTION A SECTION B SECTION C


















SECTION D SECTION E SECTION F


Figure 69. Thermocouples used in each section.

o O O0 0 -TAP WATER
IN
TAP WATEROUT o a o o a



a ** -TAP WATER IN
-- ; ; -- -- ; ; -- ----- ; -----


























.-.o o + --TAP WATER IN


-a0 o 0


TAP WATER OUT o L o


Figure 70. Piping schematic for tap water flow.







60


o o o o o HEATER 1 IN

HEATER1UT o a o a



-o ---HEATER 2 IN


-ft 0 0-


HEATER 2 OUT --to


Figure 71. Piping schematic for +41 OF uniform profile.


Table 10. Absolute temperatures to achieve +41 F profile based on initial temperatures.
Initial Ambient Temperatures (F)
79 80 81 82 83 84 85 86


Destination Temp (F) 120 121 122 123 124 125 126 127


36
32
" 28
S 24
20
0 16
> 12
8
4
0


0 5 10 15 20 25 30 35 40 45
Temperature Difference (deg. F)


Figure 72. Uniform profile for segment 2.


-Section A
- -Section B
Section C _
--Target Profile










36
32
'" 28
U 24
20
0 16
> 12
0 8
4
0


0 5 10 15 20 25 30 35
Temperature Difference (deg. F)


40 45


Figure 73. Uniform profile for Segment 3.


HEATER 1 OUT O 0 0 0 0 0 0- HEATER 1 IN

MIXEDOUT o o o o -MIXED


HEATER 2 OUT


HEATER 3 OUT


o o ---HEATER 2 IN


0 0


o--HEATER 3 IN


Figure 74. Piping schematic for linear profile.

Table 11. Absolute temperatures to achieve linear profile.
Target Temperatures of Pipe Layers Based on Initial
Ambient Temperatures (oF)
Initial Temp
(0F) 79 80 81 82 83 84 85 86

Layer 1 119.4 120.4 121.4 122.4 123.4 124.4 125.4 126.4
Layer 2 No Flow
Layer 3 111.5 112.5 113.5 114.5 115.5 116.5 117.5 118.5
Layer 4 102.3 103.3 104.3 105.3 106.3 107.3 108.3 109.3
Layer 5 91.0 92.0 93.0 94.0 95.0 96.0 97.0 98.0
Layer 6 No Flow


--Section D
-
-Section E
-*-Section F
-Target Profile

^^^


0 0 0












36 ,
Section A
32 -Section B
t' 28 -- Section C
2 24 -- Target Profile
S20
0 16
> 12 -
,, 8
4
0
0 5 10 15 20 25 30 35 40 45
Temperature Difference (deg. F)


Figure 75. Linear profile for Segment 2.


36
-Section D
32 -- Section E
28 -Section F
i -rTarget Profile
24
20
16
1 12
8
4
0








HEATER OUT 0 0 a 0 0 0 HEATER 3 IN





-TAP WATER IN
^ o ^ -TAP WATER IN


TAP WATER OUT o o


-.0 0 0-


Figure 77. Piping schematic for AASHTO positive gradient.







63


Table 12. Absolute temperatures to achieve AASHTO positive gradient.
Target Temperatures of Pipe Layers Based on Initial
Ambient Temperatures (oF)
Initial Temp
(0F) 79 80 81 82 83 84 85 86

Layer 1 120.0 121.0 122.0 123.0 124.0 125.0 126.0 127.0
Layer 2 90.0 91.0 92.0 93.0 94.0 95.0 96.0 97.0
Layer 3 86.8 87.8 88.8 89.8 90.8 91.8 92.8 93.8
Layer 4 79.0 80.0 81.0 82.0 83.0 84.0 85.0 86.0
Layer 5 79.0 80.0 81.0 82.0 83.0 84.0 85.0 86.0
Layer 6 79.0 80.0 81.0 82.0 83.0 84.0 85.0 86.0


36 R
32
2 28
S24-
- 20 --*-Section A
20 T//
c, --- Section B
.o 16 --Section C
12- ---Section D
2 8 _-___ -Section E
J -Section F
4 --- AASHTO
0 I--
-5 0 5 10 15 20 25 30 35 40 45 50
Temperature Difference (deg F)


Figure 78. Experimental AASHTO positive gradient.

o o o o -HEATER 1 IN

HEATER OUT o


HEATER 2 OUT--oa -


S--1HEATER 2 IN


- a-0


Figure 79. Piping schematic to heat beam.


- O








64


HEATER 1 OUT 0 0 0 0 0 0 -HEATER1IN
HEATER30UT o o o *o HEATER31N
MIXED OUTo o 0 o o *o O MIXED



So o--HEATER 2 IN


HEATER 2 OUT o D-


-f a-


Figure 80. Piping schematic for AASHTO negative gradient.


Table 13. Absolute temperatures to achieve AASHTO negative gradient.
Target Temperatures of Pipe Layers Based on Initial
Ambient Temperatures (oF)
Initial Temp
(0F) 79 80 81 82 83 84 85 86


Layer 1 105.7 106.7 107.7 108.7 109.7 110.7 111.7 112.7
Layer 2 114.7 115.7 116.7 117.7 118.7 119.7 120.7 121.7
Layer 3 115.7 116.7 117.7 118.7 119.7 120.7 121.7 122.7
Layer 4 118.0 119.0 120.0 121.0 122.0 123.0 124.0 125.0
Layer 5 118.0 119.0 120.0 121.0 122.0 123.0 124.0 125.0
Layer 6 118.0 119.0 120.0 121.0 122.0 123.0 124.0 125.0


36
32
S28

U
S20
C
O 16
> 12
w 8
4
0


-Section A
--Section B
- Section C
--- Section D
-- Section E
-Section F
--AASHTO


-14 -12 -10 -8 -6


-4 -2 0 2


Temperature Difference (deg F)


Figure 81. Experimental AASHTO negative gradient.














CHAPTER 7
CTE TESTS

The coefficient of thermal expansion of the concrete, CTE, was necessary for

computer modeling and stress calculations. It was found using two methods: the

AASHTO standard test method and in-situ tests. This chapter will discuss the two

methods for determining the CTE value.

AASHTO Standard Test Method

A standard procedure for determining the CTE value for concrete is outlined in the

AASHTO Designation: TP 60-00 (2004). This designation is the Standard Method of

Test for Coefficient of Thermal Expansion of Hydraulic Cement Concrete. The scope of

the procedure involves fully saturating a concrete cylinder, increasing and decreasing the

temperature of the cylinder and measuring the length changes. The CTE is then

calculated as the change in length divided by the cylinder length and temperature change.

The value is useful in determining the potential for length and volume changes of

concrete due to either a uniform temperature change or a temperature gradient.

AASHTO test specimens are concrete cylinders 7.0 0.1 inches in length and 4

inches in diameter. They are submersed in saturated limewater at 73 40 F for at least

two days. After the cylinders are fully saturated, they are removed from the water, and

their lengths are measured at room temperature to the nearest 0.004 in. They are then

placed in a measuring apparatus that consists of a frame and LVDTs (see Figure 82). The

cylinder and apparatus are placed into a prepared water bath, and the temperature of the

water is then set to 50 20 F. Once thermal equilibrium of the specimen has been









reached, the LVDT will consistently read the same deflection to the nearest 0.00001

inches. LVDT readings shall be taken every 10 minutes over a one-half hour time period.

Both the temperature and LVDT readings will be recorded as the initial readings. The

temperature of the water will then be set to 122 20 F, and thermal equilibrium of the

specimen will be reached similarly to the initial stage. The temperature and LVDT

readings will be recorded as the second reading. The water temperature is then set back

to the initial 50 20 F, thermal equilibrium is reached, and the third and final set of

readings is recorded.

The next step is to calculate the CTE value according to the equation: CTE

= (AL, / L)/AT. The actual length change of the specimen, ALa, is calculated by taking

the sum of the measured length change of the specimen (increase is positive, decrease is

negative) and the length change of the measuring apparatus. The length change of the

measuring apparatus is taken as the multiplication of the CTE of stainless steel (17.3

x 10 6 /C), the change in temperature and the original specimen length, Lo. CTE values

shall be computed for both the expansion and contraction stages. The final CTE value is

calculated by averaging the two CTE values. The two values must be within 0.5

x 10-6 /F ; additional tests are necessary if they are not.

The AASHTO CTE tests were performed by the FDOT SMO on concrete cylinders

taken from the casting of beam segment 3. Ideally, concrete cylinders from beam

Segment 2 would also have been tested because it was a thermal section. However, 4 in.

concrete cylinders were not taken from segment 2. The AASHTO tests were performed

on the Segment 3 cylinders at the FDOT State Materials Lab, while in-situ tests were

being performed on segment 3 (described later). The proximity of the time that the two









tests were run ensured that the concrete was at the same material properties (i.e. moisture

content) between the cylinders and the beam segment.

Data from the AASHTO CTE tests are shown in the following tables. Table 14

shows the initial lengths of each tested cylinder after the cylinders had soaked for more

than two days. These lengths were used to determine the final strains. Table 15 displays

the absolute temperatures and the temperature changes (At) recorded during the test. The

displacements recorded from the LVDTs are shown in Table 16. The final strains were

calculated by dividing the displacements by the initial lengths (Table 17). The strains

were divided by the temperature changes and CTE values were recorded for the heating

and cooling phases. These values were then averaged to find a CTE value for each

cylinder (Table 18). A representative CTE value for beam segment 3 was taken as the

average of the three cylinders' CTE values. The value was found to be 7.84x 106 /F ,

which is in the high end of typical CTE values (4.1 to 7.3 x 106 /F ).

In-situ CTE Test Method

Methods were developed to determine CTE values for the actual beam in laboratory

conditions. The general scope of the tests was to heat a beam segment to a given thermal

profile and record the longitudinal elongation of the beam. Deflections were recorded

using LVDT's, and temperatures were recorded with thermocouples. Figure 83 shows

the positioning of LVDTs on the beam in the lab. LVDTs were placed vertically along

the cross sectional centerline on each end of the beam at the top, the centroid, the bottom

and halfway between the centroid and bottom. The CTE value was then computed by

dividing the average longitudinal displacements by the average temperature differentials

and the total beam length, according to the following equations:










Jt total dA
avg (average total displacement of segment)
A


av = a"g (average engineering strain of segment)
g L segment


Sra dA A
TGag = dA (average temperature differential)


a = ag (coefficient of thermal expansion)
TG g

Where:

tota total axial elongation of segment

L_segment total axial elongation of segment

Tgrad total axial elongation of segment

A, cross sectional area of segment

Positive Uniform Temperature Profile

A positive uniform temperature profile of + 41 F was imposed individually on

beam segments 2 and 3. Data from thermocouples, LVDTs, foil gauges and strain rings

were recorded. The desired and imposed thermal profiles for segment 2 and 3 are shown

in Figure 72 and Figure 73 in Chapter 6. The LVDT data are shown in Figure 84 and

Figure 85. The calculated CTE values for Segment 2 and Segment 3 from LVDT data

were found to be 6.3 x 10 6 / oF and 7.8 x 10 6 / F, respectively.

Linearly Increasing Temperature Profile

A linear profile as described in Chapter 5 was imposed individually on beam

segments 2 and 3. Data from thermocouples, LVDT's, foil gauges and strain rings were

recorded. The desired and imposed thermal profiles are shown in Figure 75 and Figure









76 in Chapter 6. The LVDT data is shown in Figure 86 and Figure 87. The calculated

CTE values for Segment 2 and Segment 3 from LVDT data were found to be 7.6 x 106 /

F and 8.3 x 10-6 / F, respectively.

Comparison of CTE data

Table 19 compares CTE values found from the AASHTO test and the uniform and

linear profile tests on segment 2 and 3. The CTE value was lower for Segment 2 than it

was for Segment 3, which can most likely be attributed to the moisture content being

lower in segment 2 than in segment 3. The CTE values for the linear gradient profile

tests were found to be higher than the values for the uniform tests. The reason for this

occurrence is not known. The CTE value for the uniform profile test on Segment 3

matched the average CTE value found from the AASHTO CTE tests.









Bae Plato Die.= 10"


Top View


,- Spring Loaded LVDT


Frame Heigt = 10"


4" Die. Concrete
- --- Core shown


3 Sarnl-phamcal muprr
-- Buttons equally spaced
about 2" De. Circle


Front View

Figure 82. Measuring device for AASHTO CTE tests.


Table 14. Cylinder lengths at room temperature.
Cylinder 1 Cylinder 2 Cylinder 3
L1 (inches) 7.065 7.098 7.056
L2 7.078 7.108 7.064
L3 7.061 7.092 7.056
L4 7.054 7.104 7.054
Average Length 7.064 7.100 7.056


.n PT~.o.










Table 15. Absolute temperatures and temperature changes for AASHTO CTE test.
Cylinder 1 Cylinder 2 Cylinder 3
Initial Temp (F) 50.2 50.2 50.2
Maximum Temp (F) 123.4 123.4 123.4
Final Temp (F) 50.0 50.0 50.0

At: Initial to Max (F) 73.2 73.2 73.2
At: Max to Final (F) -73.2 -73.2 -73.2

Table 16. Recorded displacements from LVDT's.
Cylinder 1 Cylinder 2 Cylinder 3
AL : Initial to Max (in.) 0.00404 0.00412 0.00411
AL : Initial to Max (in.) -0.00406 -0.00396 -0.00407

Table 17. Calculated strains.
Cylinder 1 Cylinder 2 Cylinder 3
AL/L : Initial to Max (in.) 0.00058 0.00059 0.00059
AL/L : Initial to Max (in.) -0.00058 -0.00057 -0.00058

Table 18. Calculated CTE values.
Cylinder 1 Cylinder 2 Cylinder 3
CTE initial max (1/F) 7.82 E-06 7.92 E-06 7.97 E-06
CTE max final (1/F) 7.84 E-06 7.61 E-06 7.87 E-06

Average CTE (1/F) 7.83 E-06 7.77 E-06 7.92 E-06


LVDT
FOIL GAUGE
-- STRAIN RING


Figure 83. CTE test setup.







72



40
4O35- -East I
"- West
a 30 ---Sum
& 25
20
15-
S10-- ------- \-
0 10
5

0 0.005 0.01 0.015 0.02
Longitudinal Displacements (inches)


Figure 84. Longitudinal displacements from LVDT's for Segment 2 for positive uniform
temperature profile.


40
East
S35 -West
S30 -- Sum
= 25
20
0
15--
C 10


0-
0 0.005 0.01 0.015 0.02 0.025
Longitudinal Displacements (inches)


Figure 85. Longitudinal displacements from LVDT's for Segment 3 for positive uniform
temperature profile.











40
35
30
25
20
15
10
5
0
-0


.005


0 0.005 0.01 0.015
Longitudinal Displacements (inches)


Figure 86. Longitudinal displacements from LVDT's for Segment 2 for linearly
increasing temperature profile.


40
35
a 30
S25
- 20
0
*I 15
c 10
w
5
0


- East
-West
-Sum










0 0.005 0.01 0.015 0.C
Longitudinal Displacements (inches)


Figure 87. Longitudinal displacments from LVDT's for Segment 3 for linearly increasing
temperature profile.

Table 19. Comparison of CTE values.
Segment 2 Segment 3
CTE Value CTE Value
(micro strain / F) (micro strain / F)
AASHTO N.A. 7.8
Uniform Profile 6.3 7.8
Linear Profile 7.6 8.3


S-East LVDT's
West LVDT's
--SUM





//














CHAPTER 8
CONCLUSIONS AND RECOMMENDATIONS

The main objective of this research project was to quantify the stresses formed

from the AASHTO nonlinear thermal gradients. Instrumentation was necessary to

measure the loads, strains and deflections to quantify the thermal stresses in a 20 ft. long

segmental concrete T-beam. Before the instrumentation was placed on the T-beam, many

tests were performed on small rectangular prestressed concrete beams to determine the

characteristics of the instrumentation under mechanical loading. The beams were loaded

until a joint opening occurred between the two segments of the prestressed beams.

Data taken from tests performed on the rectangular beams were analyzed to

determine the load required to cause a joint opening, and associated with zero flexural

stress state in the concrete, at the top of the segment interface. Data were taken from

strain rings, foil gauges, crack gauges, LVDTs, linear potentiometers and load cells.

Deflections and strains were plotted against tip load to characterize the beams behavior.

The graphs of tip deflections and deflections across the joints showed the beam would

deflect linearly until the joint began to open, would start to deflect nonlinearly, and then

would deflect linearly again. Until the joint opened, the stiffness was constant; while the

joint was opening, the stiffness was changing. Joint opening caused a reduction in

stiffness that remained constant with continued loading. It was recommended that the tip

deflections and deflections across the joints are further researched on the T-beam. It was

also recommended that strains throughout the depth of the beam at the joints and away

from the joints be recorded and analyzed.









One noticeable concern on the small rectangular test beams was the interface

between the two match-cast segments. Even though much care was taken to form a

smooth, even interface, some shrinkage occurred during the concrete curing process. The

shrinkage resulted in reduced cross sectional properties and therefore, flexural stiffness.

There may be some similar effects from shrinkage on the test T-beam. However, it is

believed that the T-beam section and prestress forces are large enough that shrinkage will

not affect the behavior of the test beam under loading conditions.

A very important aspect of the research project thus far, is the creation of an in-situ

test for the coefficient of thermal expansion, CTE, value on a given concrete beam. CTE

values were found using AASHTO procedures that included taking 4 in. concrete

cylinders and testing them in a bath of water. It was felt that testing concrete in a

completely soaked environment was not truly representative of the concrete in the field.

Therefore, Segments 2 and 3 were individually heated and longitudinal movements were

recorded. CTE values were then calculated from the change in length divided by the

original length and the temperature change. It was found that the average CTE value

from the AASHTO test for concrete cylinders taken from Segment 3 and CTE value from

the in-situ uniform heating test performed on Segment 3 were both 7.8 x 10 6 / F.

Ongoing research being conducted by the University of Florida will focus on

performing tests on the segmental T-beam using the instrumentation systems described in

this thesis. Through data analysis and computer modeling, stresses will be quantified and

recommendations will be made to the FDOT. In the future, this data will be used to help

develop a more economical bridge design.














APPENDIX A
TEST BEAM SHOP DRAWINGS

This appendix provides shop drawings given to the Florida Department of

Transportation for building of the main test beam.













:w *= ---- EB 1y99 HNH: 1 ld I 3f3Ot1d HMV1V VdnSO S:
onAnr : bOd Vt IS 1iS31 03SOdOkud tViNlOl M 0 AJSS1AInl


I b I


3l
a

L:1










C, I






Z Hl

* w

-U^ U
---s |





C






7


H
z

0
O


ci,
H

C



-41
H

z
0




H
>
a
z


Qd
0
h

z











l-o I Ma %fl' ---- "a -:
;- :-- ---- HaY3vSa 1NlIfNVI9 Io wY3H1 iN l vda ISNnina ;S
F ,Annr : bOd flVaa 1S31 03SOdOt1d VaINOld AO ASHyAINrfl


cj~


I I


- IEI.-


I ,,
~


1-' il







aa **

















Orfi :gL v M0O "J'I= :Vfl HOV3S3N M '
O:tI V HWaH : Sld 3rACOyd sun,'.nt
:'' s,*,t H-3Svf3ts 1N3IC9o -iVW3H-1 NcnitV4dc3 so ni ~Cn"lC ;S
.r0C ;b iOd lwV3a9 ISa1 a3SOdOid VUQIrO:Id AIISHAAINn


I '-I' I


I -" -

1--^-


0 1

i2~

[ 3~


"















APPENDIX B
FRAME DRAWINGS

This appendix provides the drawings and details for the frames.











COPE FLANGES @
COLUMN (TYP.)


W16x89


SSTRONG-FLOOR BOLT HOLES
LOCATED IN LAB







4' O.


-PL 2"x1'-1"x1'-7"


2" DIA BOLT
HOLES


W\AI fivQ


0-


VV I'







































INTERFACE


FOR 0.5" DIA











CHANNELS:
C15x33.9


LOAD CELL


12" X 9" X 9"
PLATE


SHORING


1,,
1'-84


LOAD CELL


SHORING


3'-6 8























APPENDIX C

CONCRETE MIX TICKETS


This appendix provides the mix tickets for each concrete pour.


RECEIVE-D Br'


T1he nddHiaenl water nu rdded pl m~ reqIar


WVir Addcd .G(i
44,,- -B4' ID ifI, l[RER t
I IhP~N CODFB
'.1 4a-66i


Slump
FLORIDAi RUI IMND AinltFSL, MC
P*J BDY Io 2
2Ji4 OUTJH n IU, .31 PEET
I3 L.N~tFWLL Fl- ?lirv


al ..Fi -Iu;

F|0 1, LKi" 'fCu


lluL' Jr,,' l.r i | ; "litB\% n)m
Min.-L,


,;. I' II-.j CF:4IIl II I "
Da1 P1 it17,11 Ir.W.. iii ;, Tr P r 1-11,, .rL Lt-''C iQ 111 i
r r "'.ll :'r. -' l"r h *;l"'. r-, F,- r Ti'Cf L -' TiI I T


I i Uinou Tirm C icL.a mnfi.d

Nm Jtob Addit Inr



L e llr .t. I .nrl
MJp C'ildlji ill' Suiz 1 .,"rflr ,


I S1rni
Ir ,
-mu.-
C' urn.


I L. -T .Inl*(- .l
ll,. ", :-.I
'l!'i' i 1.2 I I .i




\ "'


I I. ,IE .
.; i 1EL -
Nir' ri-F ,L F,


Tlra DuIi

LIiIl Fri,


- i' "


I T1-r Lou.dler










I 1- 1
1h" i b-


*' / 9I


S: LuTTiT.L


CpN.J Qq.

La~
Qiirinii,


- ; I.-


111111111111111111U111111
Silllllll li L llll lll
a D [] 'I 2 i "3 b: 3 L. i


----












RECJEIVED B ,, llllllll lllll
Tm i.dddlaul nai mu aided lI m.ay q4 IIImI
9 O% n a L 2 2 3 B 0
taref A.4t4h Gi, Slump ___
--*,r.'' .-l.TER UF1 ,. '.-RIOR D1t' S NUSTi.ES. 1N : "
-.TE COBB .O. '. '. :- r !,: '.,
. -- I :" ',-OUt.H WIN'STREE ; -. .


:;.._:'-_ J _. -J. Gt: .L____.____--
ppm AJ, e U= r A No Dta r Na D- harm t D rimr Veiftes



.L- .. u. ... [ e .....-
i dr.u MIv jotrb Mdrpa Ll.r

I .arNwitn LUoNtatjld No. Exti PRodlij
L;Cc W.0RK CROS FREOM Fi AS OFF F LEVY ---BO TO E YND w00 ..
7-#;:* ; L '?r 'f ---6 INTO 197 EiTRtWaCE ON THE RIGHT MIpCGorfint snws'eiM








y.-I L1 5l
LrWd bRm U s Jlnlptio" n U m11111 II I ritml 7I a
Quridty Q| i ly j /?
I. X0 1.00 .* s PSI PPp.ix 0 exJ YDB %9.00 91.
10.0 1 E-/D EN i iPEt'k. FaE Ep A- .00 M.I00 onIE
,00 F BLUFiaISE RS g FUELS EA 15.00 [15.00
:.V ; xCl Lmu a asRE LL lW 1. S 610,00 IS 0.


O -e r

















ru NL. D-T No Nilr.-I Nh Link,, ik" ru.,,ir.4',-,-d
RECEIVED B l__I' ,
D- -.l ..*.t-7..L. 'lHIlEilIIIIIII II II II

Wala l..iL L,_ded ia ;l.mp YT_ i0 0 I 3 a 3 3 a L
-PA----0AVID MfLlEa (4F1 FLLORIDA K0C; INWASTRIVES; INC
r are COBB P. 0. BOx 102a
1,-t4EEF 94 i-: f Fa' rn STRET
BLnRlneSUIL.LL, FL. Wiso
Mr .1'. TlE- ID T- Ir 1 L' rm No. ft1t4 r ax*- Ne llu,. i c, .d








P kIt..4 ",' Sluiriy I5 .IrIc oJ ML. toI Ft'l N~d '" l lInti r'l.. Thn I.,.. lni
- I ., ',U .ilL ._ .'.. .
PuAl 13] N L[ k lm ITw C..jnr- V


r iL 'U L s ; -Ti.,L I .:- i.-1 ,





2Dd QImuJ l,. 16 ,r.kk, 1. LE t.l'oh
1.[. TFii | F- -. i -F f U -'. -- S TO F'Jf i_ % eL c.f.



J.I.i i
I*r'? ''-F TIW LEf-FT M-- f IFli l .l i dir ... J.' or... r Mi '.,.. Tr a




rrq.
Ii'iWr..il',lt; Shlil i llw Inid '. f.t B l*iri ^ __ lime blur r I -J .




. \.\ Ir Fl i F, -' LE EF Ern I ,.

'^r i il;:? t Ihe r F1 -: EFr [. ((i(, 1 ._ ,.



i T lr*h














RECEIVED B Y


Tke addhlMaol wditr ,x


Wn.er Wdcd .____ J.
-AIF- -DAVID iALTER (IF)
FRAf14 CCBB
*-il44Cfl


adk iq'rrwa II 111111 I IN 11111111r1
r I., 11____ 1 13 2 1 L I
ELORTIOP ROCK INDMUSTWREB, |N
P. O. BOx t0R-a
9e4 SaQTH"i. ar T 5-tE
GrifESILLCE. FL VS2OE


PF.i Lur D ef T.-Tck t'Jr. Dlrr r, 1 ln-r Iatmr
... Ji. ,. -[ rrB ,' --" ... 7'.:,',.* B-,j.i ,n. rT. ,ig [.
PCO E'LJI3N,. F T
"-.

DEIst .!tal Fl.-nPll jl;T[-i
Dd Ir anldi.ss
[NNil 'iJ iIN 'A WC rh I 1 ,L. I !'11 Irf.r LAE. r]- 1: w LL. I ---UO1 EL JNT
1? i.', ? 'LI LEFT -BO JiNrn I I-Nip'JrF- 5[rt THF RII3HT


.rle NJ L';l...I. Ten, rl,.r -r.n




I f.r ri' ll FI ,lr P'.in1 r
\.nrilr
1 l JiiMdY EL. sIIIl,, rri-i


T.Iwiep
;. IV1


C 'ur
r)Qunaiu


Frili IUPCiia:.- R


,,Us, ,iu l Itl d | tb


...;r IEE .. |.

isAF921 tMTDtsi


w! I't Do.

VI



I, '
-Jnl tw
A.# O,,


lio L. dni





.I rAw r ii'
I I, I.,

.IJ!1


Fiwh |
SPI.. ..... ,

Plrtr nrnin


U.dlcn'. (q
I Iii i

Ii i


I. 'jfl
1. -A
-WOOL-


r t4rli.rd


I _


i~ra~i -r















LIST OF REFERENCES


American Association of State Highway and Transportation Officials AASHTO (1989).
"AASHTO Guide Specifications, Thermal Effects in Concrete Bridge
Superstructures." Washington D.C.

American Association of State Highway and Transportation Officials (AASHTO) (2001).
"AASHTO LRFD Bridge Design Specifications U.S. Units." Washington D.C.

American Association of State Highway and Transportation Officials (AASHTO) (2004).
"Coefficient of Thermal Expansion of Hydraulic Cement Concrete." Washington
D.C., TP 60-1 TP60-7.

American Concrete Institute (ACI) Committee 318 (2002). "Building Code Requirements
for Structural Concrete (318-02) and Commentary (318R-02)." American Concrete
Institute, Farmington Hills, Michigan.

Mahama, Farouque (2006). "Validation of Stresses Caused by Thermal Gradients in
Segmental Concrete Bridges." PhD Proposal, University of Florida, Department of
Civil and Coastal Engineering, Gainesville, FL.















BIOGRAPHICAL SKETCH

David Clancy Walter was born on April 17, 1982, in Alexandria, Virginia. In the

same year, he moved to South Florida. After high school, he successfully completed his

undergraduate studies at the University of Florida and received a Bachelor of Science in

civil engineering in May of 2004. The author then worked in a structural design firm in

his hometown, Stuart, Florida. He then pursued his master's degree in the field of

structural engineering at the University of Florida. Upon completion of his graduate

school, the author plans to continue his professional engineering career with Atlantic

Engineering Services in Jacksonville, Florida.