Prestressed Pavement Rehabilitation

Prestressed Pavement Rehabilitation FINAL REPORT June 23, 2009 By Shelley M. Stoffels, Andrea J. Schokker, Hao Yin, Vishal Singh, Lin Yeh and Maria Lopez de Murphy The Thomas D. Larson Pennsylvania Transportation Institute COMMONWEALTH OF PENNSYLVANIA DEPARTMENT OF TRANSPORTATION CONTRACT No. 510602 PROJECT No. PSU-015

Technical Report Documentation Page 1. Report No. 2. Government Accession No. 3. Recipient s Catalog No. FHWA-PA-20-012-PSU-015 4. Title and Subtitle Prestressed Pavement Rehabilitation 5. Report Date June 23, 2009 6. Performing Organization Code 7. Author(s) Shelley M. Stoffels, Andrea J. Schokker, Hao Yin, Vishal Singh, Lin Yeh and Maria Lopez de Murphy 8. Performing Organization Report No. PTI 2009-20 9. Performing Organization Name and Address The Thomas D. Larson Pennsylvania Transportation Institute The Pennsylvania State University 201 Transportation Research Building University Park, PA 16802-4710 12. Sponsoring Agency Name and Address The Pennsylvania Department of Transportation Bureau of Planning and Research Commonwealth Keystone Building 400 North Street, 6 th Floor Harrisburg, PA 17120-0064 10. Work Unit No. (TRAIS) 11. Contract or Grant No. 510602, PSU 015 13. Type of Report and Period Covered Final Report 1/7/20 11/7/20 14. Sponsoring Agency Code 15. Supplementary Notes COTR: Ed Stoltz, 814-696-7204 16. Abstract In 1989, a landmark pavement project was opened to traffic in Blair County, Pennsylvania, that received national attention. The pavement was a two-mile section of prestressed concrete pavement that was constructed on the northbound lanes of what is now Interstate 99. The pavement has now started to show signs of distress, including transverse cracks, longitudinal cracks, spalling, and seal failure at the armored contraction joints. In addition, there are signs of shoulder separation and some localized subgrade failure. However, the ride quality of the pavement is still very good. Due to the uniqueness of this pavement section, and the lack nationally of any other similar sections in terms of design and age, traditional rehabilitation and overlay options are not necessarily directly applicable. The objective of this project was to develop recommendations for feasible rehabilitation strategies for the prestressed concrete pavement on I-99 in Blair County. As part of the work, the expected design lives and cost-effectiveness of the strategies were evaluated. 17. Key Words Pavement, prestressed concrete, distress, rehabilitation, design life 18. Distribution Statement No restrictions. This document is available from the National Technical Information Service, Springfield, VA 22161 19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price Unclassified Unclassified 240 Form DOT F 1700.7 (8-72) Reproduction of completed page authorized

This work was sponsored by the Pennsylvania Department of Transportation and the U.S. Department of Transportation, Federal Highway Administration. The contents of this report reflect the views of the authors, who are responsible for the facts and the accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies of either the Federal Highway Administration, U.S. Department of Transportation, or the Commonwealth of Pennsylvania at the time of publication. This report does not constitute a standard, specification, or regulation.

TABLE OF CONTENTS 1. INTRODUCTION...1 2. FIELD DATA COLLECTION AND ANALYSIS...5 3. FORMULATION OF ALTERNATIVES...31 4. LIFE-CYCLE COST ANALYSIS...42 5. RESULTS, DISCUSSION AND RECOMMENDATIONS...48 6. REFERENCES AND SOURCES CONSULTED...50 APPENDIX A: DISTRESS MAPS APPENDIX B: CONDITION SURVEY TABULATIONS APPENDIX C: VISUAL CONDITION PHOTOGRAPHS APPENDIX D: BACKCALCULATED MODULUS VALUES APPENDIX E: LOAD TRANSFER RESULTS APPENDIX F: PRELIMINARY PAVEMENT THICKNESS CALCULATIONS APPENDIX G: LCCA PRINTOUTS APPENDIX H: ORIGINAL CONSTRUCTION PLAN DETAILS v

1. INTRODUCTION Scope In 1989, a landmark pavement project was opened to traffic in Blair County, Pennsylvania, that received national attention. The pavement was a two-mile section of prestressed concrete pavement that was constructed on the northbound lanes of what is now Interstate 99. The pavement has now started to show signs of distress, including transverse cracks, longitudinal cracks, spalling, and seal failure at the armored contraction joints. In addition, there are signs of shoulder separation and some localized subgrade failure. However, the ride quality of the pavement is still very good. Due to the uniqueness of this pavement section, and the lack nationally of any other similar sections in terms of design and age, traditional rehabilitation and overlay options are not necessarily directly applicable. The objective of this project was to develop recommendations for feasible rehabilitation strategies for the prestressed concrete pavement on I-99 in Blair County. As part of the work, the expected design lives and costeffectiveness of the strategies were evaluated. Pavement History and Characteristics Design Features The two-mile section of prestressed pavement was constructed in November 1988 along a section of U.S. 220 in Blair County in Pennsylvania. That pavement section is now part of I-99 and is in the northbound lanes of a four-lane divided highway. The 31,411-sq-yd prestressed pavement is 24 ft wide and 7 inches thick. Active joints are spaced at 400 ft, with 396 ft of main slab and 4 ft of gap slab. Details about the design and construction were obtained from the project design documents as well as from a research report resulting from extensive testing and instrumentation done near the time of construction (Okamoto, 19). 1

The outside tied concrete shoulder is 7 inches thick and 10 ft wide. The inside tied concrete shoulder is 7 inches thick and 4 ft wide. The cross section of the prestressed pavement consists of the following (from top down): 7-in-thick prestressed pavement 2 layers of 6-mil polyethylene sheets 5-in-thick lean concrete base (LCB) course 3-in-thick open-graded subbase 3-in-thick dense-graded subbase Natural compacted subgrade The minimum compressive strength for class AA concrete for the prestressed pavement was specified to be 3,000 psi at 7 days and 3,750 psi at 28 days. The following prestressing details for the encased tendons were used: 7-wire strand, grade 270 ksi (stress relieved or low-relaxation steel) Strand diameter: 0.6 inch Strand spacing: 18 inch No of tendons: 15 Ultimate strength of strand: 58.6 kip Stressing load (80% of ultimate): 46.9 kip Tendon placement: 0.5 inch below mid-depth To guard against corrosion, the tendons were made with a special extrusion coating process. Each seven-wire steel tendon was encased in grease and then coated with seamless polypropylene sheath 0.06 inch thick. To accommodate large horizontal movements, steel channels were used to protect the slab ends, with stainless steel dowels crossing the joints to transfer vertical loads. To avoid early shrinkage cracks, each main slab was stressed in three stages. One-third of the design stress was applied when test cylinders showed compressive strength of 1,000 psi. Two-thirds stress was applied when the concrete reached 2,500 psi, and full stress was applied at 3,750 psi. 2

Problems During Construction As might be anticipated during such a unique project, several problems developed during construction. The polyethylene sheets developed wrinkles as concrete was spread across them. One of the 480 tendons failed during stressing, possibly because of misaligned wedges at the tendon anchor. Stress was transferred smoothly to the gap slabs at most tendons, but some of the permanent anchors failed to hold. This may have been because of concrete leaking into anchors. The problem was compounded when workers burned away several anchors in an attempt to relieve stress at the temporary face. The anchors were repaired by cutting away part of the gap slab. Broken tendons were replaced with 5/8-inch tendons pulled through the existing sheathing with the old tendon. Dowel expansion caps were damaged, and the ends were drilled after concrete placement to allow for the expansion. Early Condition Assessments and Observations Falling Weight Deflectometer Test Results Falling weight deflectometer (FWD) data were used to calculate deflection load transfer efficiency and estimate modulus of subgrade reaction. The joint efficiency was calculated at each joint at the center of the truck lane. The effective modulus of subgrade reaction was backcalculated for each slab at stations 2+00 and 3+00. The full results are listed in the 19 research report (Okamoto, 19). A brief summary from that report is provided in Table 1. Condition Surveys Condition surveys were performed in April and August 1989, and again in 1990 and 19. The 1989 and 1990 surveys were summarized by Okamoto (19). The 19 survey was provided 3

by PennDOT District 9. The 19 survey was used as a reference to observe subsequent changes in condition, as discussed in a later section and shown in Appendices A and B. The predominant observed distresses were longitudinal cracks, generally over prestressing tendons. Some transverse cracks were also recorded. Table 1. Summary of falling weight deflectometer testing in September 1989 (Okamoto, 19). Subgrade Support (pci) Joint Efficiency (%) Minimum 160 59 Maximum 380 94 Range 220 35 Average 286 77 Standard Deviation 50 9 Effective Prestress at One Year Analysis of instrumentation data at 8 months indicated a long-term creep and drying shrinkage prestress loss of 27 psi. This produces an effective midslab prestress loss of 27 psi, resulting in an effective midslab prestress of 74 and 202 psi for slab 1 and slab 2, respectively. This assumes that no loss due to steel relaxation had yet taken place. Joint Width Measurements indicated that for the prestressed pavement the joint width changes were approximately 83 percent of those calculated. The measured change in width (movement) between August 1989 and March 1990 measurements averaged approximately 1.2 inches. 4

2. FIELD DATA COLLECTION AND ANALYSIS The field data collection included a visual distress survey, falling weight deflectometer testing, and Portable Seismic Pavement Analyzer (PSPA) testing. All three activities were conducted in April 20. Field Investigation Location Referencing The prestressed pavement section is on the northbound lanes of I-99. For convenience, the inspection and testing procedures were conducted from south to north. The slabs were also numbered from south to north, and designated as slab 1 to slab 26. To assist in identification of testing and analysis locations within each slab, testing was referenced by distance from south to north, from 0 ft at the south joint to 400 ft at the north joint. Testing was also referenced by transverse position in the slab, such as corner or wheelpath positions. Distress Survey The distress survey included the preparation of a detailed map of cracks and other distresses. Measurements of width and faulting were taken at cracks and joints, as appropriate. The detailed distress maps are included in Appendix A, Figure A (key) and Figures A-1 through A-26. These figures include the distresses shown on the maps from the 19 distress survey, as well as the additional/different distresses observed in 20. Appendix B provides tabulated details of the distress and faulting measurements. The distresses in Appendix B were referenced to the original construction stationing, in order to correspond to the distress surveys conducted soon after construction. The starting and ending stations for each slab are cross-referenced with the distances used for other testing. Appendix C presents additional photographs from the visual survey. While some new cracks have appeared and propagated, the majority of the cracking has simply experienced some spalling and deterioration since the early surveys. The photographs in Figures 5

1 and 2 illustrate fairly typical high-severity longitudinal cracking. Although the tendon coating is visible, and the sealant out in some locations, the tendons appear to be undamaged. Localized subsidence near slab 17 had been previously observed and repaired by PennDOT District 9 engineers. Subsealing has improved the ride, but subsidence is still evident. In any rehabilitation alternative, special repairs and attention would be needed for this section of the prestressed pavement. Figure 1. Longitudinal cracking that extends across a number of slabs. This crack was present at the time of the early distress surveys. 6

Figure 2. Deteriorated longitudinal crack, with tendon coating visible. Also observed during the distress survey were significant amounts of vertical displacement of the transverse shoulder joints. The lane-shoulder joints at and near the mainline transverse joints were principally affected. It is inferred that the cause of the vertical displacement is not the typical faulting mechanism, but rather an interaction between the plain shoulder slabs and the prestressed mainline slabs, resulting in the apparent faulting or heaving. There is also separation of the lane-shoulder joints near the transverse joints. Figure 3 shows an occurrence of such vertical displacement on the shoulder joint between slabs 8 and 9. 7

Figure 3. Severe vertical displacement of the shoulder joint between mainline slabs 8 and 9. Falling Weight Deflectometer Testing The FWD testing was conducted by the PennDOT Bureau of Maintenance and Operations, with project staff on site. The Dynatest falling weight deflectometer was utilized for the testing, and time history data were collected. Temperature measurements were also recorded during the course of the testing. The following approximate load sequence was performed for each test position, for a total of nine drops at each test position: 8

Seating load 15,000 lb 1 drop First load level 6,000 lb 2 drops Second load level 9,000 lb 2 drops Third load level 12,000 lb 2 drops Fourth load level 15,000 lb 2 drops Testing was conducted at three interior locations per slab, at 100 ft and 300 ft from the slab approach joint, and at two transverse positions, as illustrated in Figure 4. Load transfer testing was performed at all transverse joints, and on the longitudinal lane/shoulder joints. Testing at the slab corners was also performed. Travel Direction 400 ft Both joints Wheel Path CL Travel Lane FWD test for backcalculation FWD test for load transfer Figure 4. FWD test layout for each prestressed slab. FWD Data Examination For each FWD test, the data were plotted and examined, as shown in Figure 5. By examining the full history of the load pulse and the deflection responses, anomalies can be detected. This can 9

prevent erroneous data from being included in the analysis. The slight buffering during the increase of the load application is typical of the Dynatest FWD and was observed for most locations. In general, the data from this site were well behaved. The few drops with problems were not used in the subsequent analysis. 6.00 12000.00 5.00 10000.00 4.00 8000.00 3.00 2.00 6000.00 4000.00 D1 D2 D3 D4 D5 D6 D7 Load 1.00 2000.00 0.00 0.00-10.00 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00-1.00-2000.00 Figure 5. Example of time series graph, showing load pulse and deflection sensor responses, slab 1. Backcalculation of Layer Moduli Backcalculation was performed for all center slab and wheel path locations, considering five pavement layers. Several potential layer configurations were tested to achieve the best consistent convergence and modulus stability. The layers assumed for the reported backcalculated moduli were: Prestressed concrete layer 7 inches 10

Lean concrete base 5 inches Subbase 6 inches (two subbase layers were combined together) Upper subgrade layer 6 inches (The top 6 inches of the subgrade was taken as a separate layer.) Subgrade layer The program used for backcalculation was BAKFAA, developed and distributed by the Federal Aviation Administration. The program is based upon layered elastic analysis and is designed to work effectively for both flexible and rigid pavements. The user interface for the program is shown in Figure 6. The program allows for seven sensors to be used in the backcalculation. Backcalculated results from all layered-elastic programs can be sensitive to the number and spacing of sensors, depending upon the thickness and stiffness of the pavement layers. Therefore, various combinations of the available sensors were utilized to determine the most consistent assumptions for backcalculation. In addition, the interface friction parameter between layers was varied, again to find the values that provided the best convergence. Figure 6. BAKFAA user interface and typical inputs. 11

The results were evaluated on the basis of the backcalculated modulus values of the prestressed concrete, upper subgrade and subgrade layers. Backcalculated values were categorized on the basis of minimum modulus values across all load levels for the test location. Potential areas of concern included the following: Backcalculated values for the prestressed concrete were less than 4 million psi for four test locations in slab 3, slab 8, slab 15, and slab 17. Of these test locations, those in slabs 3 and 15 had backcalculated values of approximately 2.5 million psi. These locations should be noted for localized investigation. Slab 4, slab 14, slab 16, slab 18, and slab 24 had locations with subgrade backcalculated values less than 20,000 psi. Slab 2, slab 4, slab 15, slab 22, and slab 26 had locations with upper subgrade backcalculated values less than 10,000 psi. Upper subgrade backcalculated values for other locations were typically between 20,000 psi and 40,000 psi. Slab 4 had low modulus values for both the subgrade and upper subgrade. Areas with the highest backcalculated modulus values were as follows: Slab 8, slab 9, slab 10, slab 11, slab 12, slab 13, slab 20, and slab 21 had locations with subgrade backcalculated values more than 50,000 psi. Slab 21 had a subgrade backcalculated value of more than 70,000 psi. Slab 1, slab 4, slab 5, slab 21, slab 22, slab 23, and slab 25 had prestressed concrete backcalculated moduli of more than 8 million psi. Observational correlation with the distress survey showed that these test locations were not in highly cracked areas. Overall, it was found that the backcalculated moduli provided a general indication that substantial tendon prestress remained in most tested locations. Locations with lower moduli 12

corresponded with observed distresses. With the exception of only 10 tested locations, the upper subgrade backcalculated moduli exceeded 20,000 psi, and should provide good support for rehabilitation. More detailed summary tabulations of the backcalculated moduli are included in Appendix D. Load Transfer Calculations The falling weight deflectometer data were also utilized to examine the load transfer across joints. The load transfer value is very important to the consideration of the deterioration of the pavement and to the progression of roughness. Load transfer was calculated using the deflection from the sensor on the unloaded side of the joint divided by the deflection at the sensor on the loaded side of the joint. In addition, corrected load transfer was calculated, considering the deflection sensor further away from the joint on the loaded side, but equidistant from the load. Corrected load transfer values are higher, and on some pavements with very stiff joints may exceed 100 percent. Brief summaries of the load transfer values are included here. Further tabulations are provided in Appendix E. Load Transfer at Mainline Transverse Joint Corners Load transfer calculations were performed for the corner FWD drops, for the transverse mainline joints. Corner load transfer calculations often produce the lowest values, as slab corners typically are the most subject to curling, warping, sealant damage, erosion, etc. The corner load transfer values were categorized on the basis of average normal load transfer values of all load levels, with key summarized observations as follows. (Joint 2 3 means that LT is being calculated from slab 2 to slab 3, for example. The first slab number represents the location of the load plate.) 13

Only the test locations at slab 1 (joint 1 0), slab 2 (joint 2 3), slab 17 (joint 17 18), slab 21 (joint 21 22), slab 22 (joint 22 21), and slab 24 (joint 24 25) had calculated deflection load transfer values greater than 70 percent. The load transfer values for slab 1 (joint 1 2), slab 2 (joint 2 1), slab 3 (joint 3 4), slab 4 (joint 4 3), slab 7 (joint 7 8), and slab 11 (joint 11 10) were less than 20 percent. The low load transfer values were not anticipated on the mainline transverse joints. The transfer through the lean concrete base was expected to provide a significant level of load transfer. However, no differential performance problems or faulting were observed at the transverse joints with poor load transfer. Lane/Shoulder Joint Load Transfer Load transfer testing and calculations were performed for a number of lane/shoulder joint locations, sometimes exceeding the number of locations shown in Figure 3. The results are summarized as follows. Slab 3, slab 4, slab 18, slab 20, slab 21, slab 22, slab 23, slab 25, and slab 26 had tested locations with lane/shoulder load transfer values greater than 80 percent. Slab 2, slab 4, slab 7, slab 8, slab 9, slab 10, slab 11, slab 12, and slab 13 had tested locations with lane/shoulder load transfer values less than 20 percent. The remainder of the tested locations had lane/shoulder load transfer values between 20 and 80 percent. The locations of poor load transfer do indicate a greater potential for initiation of reflection cracking for overlay rehabilitation options. However, no surface distress correlations were currently observed. 14

Void Detection Calculations When a slab is fully supported, with no curling, warping or erosion, the magnitude of deflection typically increases linearly with load, from a load of zero. If gaps are present between the slab and underlying layer, a regression line through different load levels will indicate a positive intercept at zero load. This indicates that much less load is needed to cause deflection under such poorly supported circumstances. Examination of the load-deflection plots can also show nonlinear behavior of the pavement structure. The results for all corner tests are provided in Table 2. With the possible exception of the initial terminal joint, no voids were detected. Table 2. Analysis of corner deflections for possible voids. Slab Intercept (mils) Joint Void/ Potential Stress Softening Behavior 1 2.74 1 0 Possible Void 2-1.01 2 1 -- 4-3.03 4 3 Stress Softening 5-0. 5 6 -- 6-0.58 6 7 -- 7-1.24 7 6 -- 8-1.24 8 7 -- 9-2.37 9 8 Stress Softening 10-0.20 10 9 -- 11-0.66 11 10 -- 13-1.09 13 12 -- 14-0.67 14 13 -- 15-0.10 15 14 -- 17-0.02 17 16 -- 18-0.27 18 19 -- 19-0.27 19 18 -- 19-0.17 19 20 -- 20-0.60 20 19 -- 20-0.10 20 21 -- 21-0.36 21 20 -- 21-0.42 21 22 -- 22-0.53 22 21 -- 22-0.40 22 23 -- 23-0.37 23 22 -- 15

Slab Intercept (mils) Joint Void/ Potential Stress Softening Behavior 23-0.43 23 24 -- 24-0.34 24 23 -- 25-1.04 25 24 -- 25-0.46 25 26 -- 26-1.04 26 25 -- 26-0.24 26 27 -- Portable Seismic Pavement Analyzer Testing Within the last decade, nondestructive test (NDT) methods have become popular for the assessment of existing pavement conditions. Because of the effects of temperature, moisture, and traffic on pavement materials, knowledge of the in-situ material properties of pavement layers is essential for evaluating the effective structural capacity of the pavement and selecting an appropriate rehabilitation strategy. The Portable Seismic Pavement Analyzer, as shown in Figure 7, is a device designed to determine the modulus of the top pavement layer in real-time. The PSPA consists of two receivers (accelerometers) and source packaged into a hand-portable system that can perform high-frequency seismic tests. Figure 7. Portable Seismic Pavement Analyzer (PSPA). 16

The analysis method implemented in the PSPA is called the ultrasonic surface waves (USW) method. As surface waves contain most of the seismic energy, the USW method utilizes the surface wave energy to determine the variation in modulus with wavelength. At wavelengths less than or equal to the thickness of the uppermost layer, the velocity of propagation is independent of wavelength. The relationship amongst velocity, V R, travel time, Δt, and receiver spacing, ΔX, can be written in the following form: V R ΔX = (1) Δt Therefore, if one simply generates high-frequency (short-wavelength) waves, and if one assumes that the properties of the uppermost layer are uniform, the shear wave velocity of the upper layer, V S, can be calculated from surface wave velocity, V R, and Poisson's ratio, υ. Then the elastic modulus of the top layer, E, can be determined from mass density, ρ. V S = V R ( 1.13 0.16υ ) E [(1.13 0.16υ ) V ] 2 2ρ(1 + υ) s (2) = (3) To collect data with a PSPA, the operator initiates the testing sequence through the computer. The high-frequency source is activated four to six times. The outputs of the two receivers from the last three impacts are saved and averaged (stacked). The other (pre-recording) impacts are used to adjust the gains of the amplifiers. The gains are set in a manner that optimizes the dynamic range. The PSPA testing was conducted by the Penn State research team at the same time that FWD measurements were conducted. A prior study found that the variation of PSPA data collected within a small area (3 ft 2 ) can be up to 10 percent. Therefore, it was desirable to perform PSPA tests at three locations that were within a reasonable range (less than 1 ft) from each FWD testing location. The PSPA measurements were repeated at least three times for each PSPA testing location. 17

The PSPA results were analyzed to obtain the effective seismic moduli at all testing locations. An overall modulus was determined for each specific location. Those moduli are shown in Tables 3 through 5, for the lane-shoulder testing locations, the wheelpath testing locations, and the center slab locations, respectively. Table 3. PSPA results from lane-shoulder testing locations. Slab No Testing Location (ft) 1 1 29875 1 230 21930 1 397 22033 2 1 25012 2 25 23943 2 53 23865 2 77 25203 2 103 24885 2 125 28714 2 150 236 2 183 28064 2 203 24127 2 225 22875 2 251 23486 2 275 21390 2 303 24187 2 325 224 2 350 21300 2 375 24548 2 399 21751 3 1 24403 3 203 21061 3 400 23435 4 1 25276 4 25 22939 4 50 23045 4 75 22849 4 105 27800 4 128 22233 Seismic Modulus of Prestressed Concrete (MPa) 18

Slab No Testing Location (ft) Seismic Modulus of Prestressed Concrete (MPa) 4 150 23681 4 176 25490 4 200 24934 4 225 21923 4 250 21967 4 275 25016 4 300 243 4 325 25770 4 350 23700 4 375 23307 4 400 23782 5 1 28337 5 200 25858 5 400 26945 6 1 271 6 205 25214 6 400 25016 7 1 27312 7 200 28946 7 400 23732 8 1 25639 8 200 22654 8 400 28418 9 1 30663 9 200 26538 9 400 29277 10 1 29751 10 200 25398 10 400 31309 11 1 26520 11 200 27984 11 400 28162 12 1 25278 12 200 24883 12 400 24690 13 1 23794 19

Slab No Testing Location (ft) Seismic Modulus of Prestressed Concrete (MPa) 13 200 23162 13 400 27165 Table 4. PSPA results from wheelpath testing locations. Slab No Testing Location (ft) Seismic Modulus of Prestressed Concrete (MPa) 14 100 23821 14 200 27092 14 300 20654 15 100 20346 15 200 25855 15 300 17773 16 100 19583 16 200 21565 16 300 20126 17 100 27450 17 200 210 17 300 23681 18 100 26692 18 200 24168 18 300 20356 19 100 23279 19 200 20767 19 300 24293 20 100 30578 20 200 26713 20 300 28942 21 100 20656 21 200 22654 21 300 23759 22 100 21511 22 200 21856 22 300 23759 23 100 25885 23 200 20022 20

Slab No Testing Location (ft) Seismic Modulus of Prestressed Concrete (MPa) 23 300 19213 24 100 17336 24 200 18379 24 300 17823 25 100 21767 25 200 25118 25 300 27747 26 100 304 26 200 28880 26 300 26287 Table 5. PSPA Results from Center Slab Testing Locations Slab No Testing Location (ft) Seismic Modulus of Prestressed Concrete (MPa) 14 100 22185 14 300 19367 15 100 23999 15 300 25704 16 100 17784 16 300 23711 17 100 24095 17 300 23755 18 100 18538 18 300 253 19 100 23725 19 300 21436 20 100 16382 20 300 19668 21 100 17520 21 300 188 22 100 18857 22 300 21804 23 100 25798 23 300 28593 24 100 27255 21

Slab No Testing Location (ft) Seismic Modulus of Prestressed Concrete (MPa) 24 300 25042 25 100 27676 25 300 30739 26 100 238 26 300 171 The seismic modulus values obtained along the length of the prestressed pavement section are shown in Figure 8. The PSPA modulus values are given in MPa, as those are the output units of the device s output software. The values of the FWD backcalculated moduli and the seismic moduli have not been converted to compatible units, as the underlying assumptions under each are different. The seismic moduli do provide an independent means of assessing the variability of the concrete condition. Examination of Figure 8 shows that when lower values of seismic modulus occurred, the occurrences were in the wheelpath and center positions. The lower moduli may correlate to damage, but also may correlate to interference from the tendon channels. Figures 9 through 11 show the variation of seismic modulus with depth in the pavement. The interface between the prestressed slab and the underlying concrete is clearly discernible in all three plots. In Figure 9, the results of testing near the lane-shoulder joint on slabs 1 and 4 are illustrated. The seismic modulus profile for the prestressed slab is similar for both locations. There is a greater variation in the modulus of the lean concrete. Similarly, in Figure 10, greater variation is apparent for the lean concrete base. In Figure 11, the results are plotted for a center slab location in slab 25. 22