Stage 5 Approved Version
March 2006
Research Report: UCPRC- RR- 2006- 02
Construction and Test Results on Dowel
Bar Retrofit HVS Test Sections 556FD,
557FD, 558FD, and 559FD:
State Route 14, Los Angeles
County at Palmdale
Authors:
Yi Bian, John Harvey, and Abdikarim Ali
This work was completed as part of Partnered Pavement Research Program Strategic Plan
Item 4.8 “ Dowel Bar Retrofit of Rigid Pavements”
PREPARED FOR:
California Department of Transportation
Sacramento
PREPARED BY:
University of California
Pavement Research Center
UC Davis and Berkeley
Stage 5 Approved Version
UCPRC- RR- 2006- 02
ii
DOCUMENT RETRIEVAL PAGE Report No: UCPRC- RR- 2006- 02
Title: Construction and Test Results on Dowel Bar Retrofit HVS Test Sections 556FD, 557FD, 558FD and
559FD: State Route 14, Los Angeles County at Palmdale
Authors: Y. Bian, J. Harvey, and A. Ali
Prepared for: Caltrans FHWA No.:
S/ CA/ RI- 2006/ 27b
Date: March 2006
Strategic Plan Element No: 4.8 Status: Final Status: Final, approved by
Caltrans
Abstract: This report presents the results of construction, Heavy Vehicle Simulator ( HVS) tests, deflection tests,
post- HVS forensic testing, and analysis on dowel bar retrofitted ( DBR) concrete pavement test sections at Palmdale,
California. This project was originally proposed in 2000 by the Caltrans Headquarters Division of Design. Benefits
expected from this research are to provide Caltrans with information about design and construction of DBR to help
determine where DBR may be a cost- effective strategy for rehabilitating rigid pavement and to help obtain best
performance where DBR is selected as the preferred rehabilitation strategy.
HVS Tests: Pavement sections include retrofitted joints and transverse cracks with three and four epoxy- coated steel
dowels, four hollow stainless steel dowels, and four fiber- reinforced polymer dowels per wheelpath. HVS and FWD
results at Palmdale also are compared with the results from previous HVS testing. HVS testing showed that joint
performance with four epoxy- coated steel dowels was generally the best of all the sections in terms of load transfer
efficiency ( LTE) and joint deflection. Three dowels per wheelpath was substantially worse than the other test
sections that had four dowels per wheelpath in terms of load transfer efficiency ( LTE), however, it was substantially
better than before DBR. Joint deflections were substantially better for four epoxy- coated steel dowels per wheelpath
than for the other sections. HVS results show that for each of the DBR alternatives, LTE was not substantially
affected by heavy HVS loading and that the slabs failed by fatigue cracking before LTE dropped substantially.
FWD Tests: The primary performance criteria are LTE and vertical deflection of the joints. Larger joint vertical
deflections and lower LTE are strongly correlated with increased rate of faulting and roughness development. FWD
deflection measurements agree with those under HVS wheel loading, showing that LTE was substantially improved
by DBR and was not substantially affected by HVS trafficking. Results are presented showing sensitivity of
deflections and LTE to dowel type, number of dowels per wheelpath, and slab temperature based on FWD
measurements.
Construction and materials: Observations about DBR construction and materials presented indicate variability in
depth of dowel bar placement, overall good condition of the slots and grout, and test results showing that the grout
met Caltrans specifications for flexural and compressive strength.
Keywords:
Dowel bar retrofit, concrete pavement, load transfer efficiency, fiber- reinforced polymer dowels, hollow stainless
steel dowels
Proposals for Implementation:
Related documents: UCPRC- RR- 2003- 03
Signatures:
J Harvey
1st Author
E. Kohler
Technical Review
D. Spinner
Editor
J Harvey
Principal Investigator
M. Samadian
Caltrans Contract
Manager
Stage 5 Approved Version
UCPRC- RR- 2006- 02
iii
DISCLAIMER
The contents of this report reflect the views of the authors who are responsible for the facts and
accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies
of the State of California or the Federal Highway Administration. This report does not constitute a
standard, specification, or regulation.
ACKNOWLEDGMENTS
The University of California Pavement Research Center would like to acknowledge the cooperation
and work devoted to this project by California Department of Transportation ( Caltrans) District 7,
especially Gary Laurent, Construction Resident Engineer, and the Maintenance Superintendents and
crews who supported this study.
The Caltrans Division of Research and Innovation ( DRI) Contract Monitor was Michael
Samadian, under the direction of Tom Hoover and Nick Burmas. David Bush was the Dynatest site
manager during HVS testing, and Peter Millar was the HVS operator.
Construction of the dowel bar retrofits was performed by PenHall, under the direction of Casey
Holloway. Field site sampling support was performed under the direction of Clark Scheffy. The
researchers would also like to thank the suppliers of the various types of dowels.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
iv
EXECUTIVE SUMMARY
This report presents the results of Heavy Vehicle Simulator ( HVS) tests, deflection tests, post- HVS
forensic testing, and analysis on dowel bar retrofitted ( DBR) concrete pavement test sections at Palmdale,
California. This project was originally proposed in 2000 by the Caltrans Headquarters Division of Design.
Other stakeholder Caltrans units included Headquarters METS Office of Rigid Pavement Materials and
Structural Concrete as well as Caltrans Districts 1 and 7. Benefits expected from this research are to
provide Caltrans with information needed for decisions about design and construction of DBR in order ( 1)
to help determine where DBR may be a cost- effective strategy for rehabilitating rigid pavement and ( 2) to
help obtain best performance where DBR is selected as the preferred rehabilitation strategy. This work
was completed as part of Partnered Pavement Research Program Strategic Plan Item 4.8, “ Dowel Bar
Retrofit of Rigid Pavements.”
Tasks for this project focus on four objectives agreed upon with Caltrans. This report completes
the requirements for the first objective. This report augments information provided in 2003 about HVS
and related tests at Ukiah ( 10).
1. Field Accelerated Pavement Testing with the HVS: Full- scale data tests of several types of
dowels to compare performance of retrofitted joints and cracks with those not retrofitted.
Observations and results are presented in this report.
2. Field Live Traffic Testing: Collecting field data on a long- term basis ( approximately two
years) under real loads will enable calibration of HVS and analysis results. Results will be
presented in a separate report.
3. Laboratory testing of materials: A report on completed corrosion tests was submitted to
Caltrans in 2005. A separate report on laboratory testing of FRP dowels will be provided.
4. Modeling ( future reports will present results for the following tasks):
• Finite element analysis of doweled concrete pavement joint
• Compilation of performance data from existing DBR projects throughout
the U. S.
• Life Cycle Cost analyses
Results for the first objective are presented in the report in sections that describe the HVS test
sections, results of HVS tests, results of FWD tests, and observations about construction and materials.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
v
OVERVIEW OF HVS TEST SECTIONS
Pavement sections tested include retrofitted joints and transverse cracks with three and four epoxy-coated
steel dowels, four hollow stainless steel dowels, and four fiber- reinforced polymer dowels per
wheelpath. The results of the HVS testing and other testing and analysis on the dowel bar retrofit sections
at Palmdale also are compared with the results from previous HVS testing of the following test sections:
• DBR sections at Ukiah,
• Sections of new pavement at Palmdale where dowels were installed during construction of
the slab, and
• Sections of new pavement at Palmdale that were constructed without dowels.
The dowel bar retrofit HVS tests were conducted on previously untrafficked portions of what is
referred to as “ Section 7” of the North Tangent at Palmdale. The pavement in Section 7 has cement- treated
base ( CTB), undoweled, jointed, plain concrete slabs with untied shoulders, and a standard width of 3.7 m.
Four sections on Section 7 were dowel bar retrofitted for HVS test sections as summarized below in Table A.
Trafficking of the four HVS test sections proceeded under similar temperature and rainfall to
enable the following comparisons:
• Four epoxy- coated steel dowels per wheelpath ( 556FD) versus three dowels per wheelpath
( 557FD), and
• Epoxy- coated steel dowels ( 559FD, Joint 32) versus hollow stainless steel dowels ( 559FD, Joint
33) versus fiber- reinforced polymer ( FRP) dowels ( 558FD), all with four dowels per wheelpath.
Trafficking consisted of 60 kN ( 13,500 lb) and 90 kN ( 20,250 lb) dual- wheel truck- tire loading and
150 kN ( 33,750 lb) aircraft, single- wheel loading.
Table A Summary of Palmdate Dowel Bar Retrofit Test Sections
HVS
Test
Section
Joint or
Crack
Number
Type of Dowels Number of
Dowels
Number of
HVS load
repetitions
Joint 41 Epoxy- coated steel
558FD Crack 2 Fiber reinforced polymer
Joint 42 Fiber reinforced polymer
Four per
wheelpath
2,208,578
556FD JJooiinntt 3398 EEppooxxyy-- ccooaatteedd sstteeeell
Four per
wheelpath 2,208,546
557FD JJooiinntt 3365 EEppooxxyy-- ccooaatteedd sstteeeell
Three per
wheelpath 1,121,600
Joint 32 Epoxy- coated steel
Crack 1 Hollow stainless steel in one wheelpath;
epoxy- coated steel in other wheelpath
559FD
Joint 33 Hollow stainless steel
Four per
wheelpath 2,001,497
Stage 5 Approved Version
UCPRC- RR- 2006- 02
vi
RESULTS OF HVS TESTS
HVS tests followed the three failure criteria shown below.
• Fatigue cracking of the concrete slab,
• Major damage to the DBR joints, or
• Loss of LTE of the loaded joint or crack.
The primary performance criteria for the HVS test sections is load transfer efficiency ( LTE) of
the joints and vertical deflection of the joints. Highlighted here are deflections and LTE results, from both
FWD and Joint Deflection Measurement Devices ( JDMDs).
Comparison of DBR alternatives: Fatigue life under HVS loads. All of the HVS test sections
failed by fatigue cracking of the concrete slab. Neither of the other two failure types occurred on any of
the test sections. Sections with four dowels per wheelpath withstood a similar number of HVS load
repetitions before failing by fatigue cracking.
HVS results show that for each of the DBR alternatives, load transfer efficiency was not
substantially damaged by heavy HVS loading and that the slabs failed by fatigue cracking before the LTE
dropped substantially. These results, where DBR outlasted the structural effectiveness of the concrete
slabs, are representative of the materials, quality, conditions, and workmanship in these test sections.
HVS trafficking was not applied to any of the Palmdale test sections until at least one month after
construction and after more than a year for some of the sections.
Comparison of DBR alternatives with three vs. four dowels per wheelpath: Fatigue life
under HVS loads. Fatigue life of the slab with three dowels per wheelpath was substantially shorter than
the other test sections that had four dowels per wheelpath. Longer fatigue life, higher LTE, and lower
deflections indicated better performance under these test conditions by four dowels per wheelpath when
compared to three dowels per wheelpath.
Comparison of DBR alternatives: Maximum joint deflections. Under HVS loading all of the
DBR joints but one showed an increase in joint maximum deflection after HVS trafficking. This
deflection is not attributed to temperature changes. The increases came under the 90 kN and 150 kN
loading. The only joint not showing an increase had FRP dowels in Section 558FD. The other joint with
FRP dowels in the same section behaved the same as all the other joints.
Comparison of DBR alternatives: Joint vertical deflections. The four epoxy- coated steel
dowels had much smaller joint vertical deflections than the others. Deflections for the alternatives ( four
FRP dowels per wheelpath, four hollow stainless steel dowels per wheelpath, and three epoxy- coated steel
Stage 5 Approved Version
UCPRC- RR- 2006- 02
vii
dowels per wheelpath) showed deflections similar to each other. At comparable temperatures, three
epoxy- coated steel dowels showed the largest deflections.
Comparison of DBR alternatives: LTE under HVS loads. All of the sections showed a slight
increase in initial LTE with increase in HVS wheel load ( from 60 to 90 to 150 kN). LTE is nearly always
higher than 90 percent on the DBR joints, regardless of temperature and load, under HVS loading ( using
the moving wheel definition of LTE described in this report).
Comparison of DBR alternatives: LTE after HVS trafficking ( JDMDs). All of the DBR
joints showed little or no decrease in LTE after HVS trafficking, based on measurements under the 60 kN
HVS wheel load.
Comparison of DBR with three vs. four dowels per wheelpath: Deflection and LTE results.
Under HVS loading, LTE values were lower with three epoxy- coated steel dowels per wheelpath than all
of the joints with four dowels per wheelpath, regardless of the type of dowel. Joint deflections were
higher with three dowels per wheelpath than with four dowels.
Comparison of DBR to undoweled joints: LTE sensitivity to temperature. All but one of the
joints showed a sensitivity of LTE to temperature, with LTE increasing with increased temperature. This
is expected, and is caused by slab expansion closing the joint and increasing aggregate interlock at the
joint. The temperature susceptibility was very low compared to undoweled joints. The one joint that did
not show temperature sensitivity of LTE had hollow stainless cylinder dowels.
Comparison of DBR at Ukiah vs. Palmdale: LTE results. Both locations showed similar lack
of damage to LTE under HVS trafficking. Comparison of results between the Palmdale DBR sections and
the Ukiah DBR sections show a tendency toward higher LTE at Palmdale than Ukiah. This trend
correlates with higher temperatures at Palmdale.
RESULTS OF FWD TESTS
As noted in the section “ Results of HVS Tests,” the primary performance criteria for the HVS test
sections are load transfer efficiency of the joints and vertical deflection of the joints. Larger joint vertical
deflections and lower LTE are strongly correlated with increased rate of faulting and roughness
development. Deflections were measured under the HVS loading ( discussed above) and with the FWD
before and after DBR as well as before and after HVS testing. Though the definitions of LTE are
somewhat different under HVS and FWD loading, correlation with performance is consistent in both
loading conditions. In addition to measuring joint deflections to calculate LTE, deflections were measured
at center slab to backcalculate concrete stiffness, support layer ( base + subgrade) stiffness, and subgrade
k- values.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
viii
Comparison of DBR alternatives: LTE after HVS trafficking. FWD deflection testing
performed after HVS testing showed that the HVS trafficking caused almost no change in LTE for all of
the DBR joints. This observation is the same as shown by LTE results with data from JDMD
measurements mentioned above.
Comparison of undoweled joints before DBR vs. joints with dowels originally installed:
Average LTE. Deflection tests prior to DBR showed that the average LTE was 91 percent in the
Palmdale test sections with originally installed dowels in the joints ( including a widened lane and others
with a tied shoulder). LTE values were consistent, with almost no difference between day and night
measurements. Measurements at different times of the year showed that the joints with originally installed
dowels never had less than 90 percent LTE. Many of these joints had been heavily trafficked during
previous HVS tests. The average LTE on the undoweled sections was 33 percent in the day at an average
surface temperature of 10° C, and 25 percent at night at an average surface temperature of 3° C.
Comparison of undoweled, DBR, and originally doweled joints: Temperature effects on
Average LTE. At higher temperatures, deflection testing after DBR on joints and cracks showed the LTE
of the undoweled and DBR joints was nearly the same as that of the originally installed dowel joints. The
average LTE was 97 percent at 34° C, with no difference between undoweled and DBR joints. At 23° C,
the average undoweled LTE was 85 percent, and the average DBR LTE was 82 percent.
The influence of temperature on LTE of undoweled joints was evident under cooler temperatures.
The average LTE for the undoweled joints was 73 percent at 13° C and dropped to 57 percent at 9° C. At
9° C, individual undoweled joint LTE values ranged between 27 and 93 percent, indicating the
inconsistency of relying on aggregate interlock for LTE.
The DBR strategy significantly improved the LTE of the previously undoweled joints and cracks,
and reduced the sensitivity of the LTE to different temperatures. The average LTE after DBR was 83
percent at 13° C and 79 percent at 9° C.
Comparison of DBR vs. originally installed dowels: Temperature effects on LTE.
Comparison of originally installed dowel performance with that of DBR showed that when the
temperature is above 33° C all joints have consistently high LTE, likely mostly carried by aggregate
interlock, making the presence and installation ( original or DBR) of dowels irrelevant. At lower
temperatures, the DBR joints had somewhat lower LTE than joints with originally installed dowels,
although it is still always greater than 80 percent and usually 85 to 90 percent when there are four dowels
per wheelpath. There is also greater variability in the LTE between different DBR joints than between
different joints with originally installed dowels.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
ix
Comparison of Undoweled vs. DBR with three dowels per wheelpath: LTE results. Three
dowels per wheelpath typically had much better LTE than undoweled joints. However, three dowels per
wheelpath had lower nighttime LTE than that of four dowels per wheelpath, regardless of the dowel type.
Comparison of DBR: Curling based on LTE. Built- in slab curling, which results from greater
concrete shrinkage on the surface of the slab than at the bottom, may have been reduced by DBR by about
2° C, although the results were not conclusive. This built- in curling was measured in terms of the
Equivalent Built- In Temperature Difference ( EBITD), which is backcalculated from FWD deflections
performed at different times of day in order to eliminate the effect of temperature gradients on curling.
High EBITD increases tensile stresses in the slab that cause longitudinal and corner cracking. It is
possible that any potential reduction in EBITD may only occur when the DBR backfill grout sets during
late afternoon and early evening when the temperature gradient in the slab is most positive ( hotter on top)
and the slab is the closest to being flat. Future measurement of FWD deflections before and after DBR
may provide data to check for similar results and to see if it is dependent on the slab temperature gradient
at the time that the backfill grout sets and more conclusively determine the significance of changes in
EBITD caused by DBR.
Comparison of Ukiah vs. Palmdale: Temperature effects on LTE and backcalculated
stiffness. Similar to results on the Ukiah test sections ( 10), FWD results show backcalculated stiffness
and LTE are highly dependent on temperature and temperature gradient. Low mid- slab temperature
causes the concrete slabs to contract, causing joints to open and greatly reducing LTE unless there are
dowels in the joint. Increasing mid- slab temperatures causes the slabs to expand, causing the joints to
close, which increases aggregate interlock, and may even place the joint faces in compression. At higher
temperatures, LTE is greater than 90 percent for all joints, regardless of their condition.
Deflections taken at center slab to backcalculate stiffness show the effects of temperature
gradients. Positive temperature gradients, with the slab hotter on top than on the bottom, result in lower
backcalculated moduli for the concrete slab and the underlying support layers. The change in the shape of
the slab caused by positive temperature gradients results in smaller deflections at the joints and corners.
Negative temperature gradients, with the slab cooler on the top than the bottom, result in greater
backcalculated moduli for the concrete slab and the underlying support layers. This shape also results in
greater deflections at the joints and corners.
OBSERVATIONS ABOUT CONSTRUCTION AND MATERIALS
Accuracy of placement. Dowel bar placement accuracy was measured on cores taken after all
testing was completed. The results showed that all dowels cored were above the mid- depth of the slab and
some were very near the top of the slab. Most of these cores were from the HVS wheelpath. These results
Stage 5 Approved Version
UCPRC- RR- 2006- 02
x
indicate that the test sections were not “ perfect” and resemble possible field construction in terms of
dowel placement variability.
Slot and grout condition. The condition of the grout backfill material in the dowel bar slots was
inspected periodically on both the wheelpath trafficked by the HVS and the wheelpath that was not
trafficked. Many of the dowel slots showed the grout to be in good condition, except for slightly out of
place foam back board on several of the joints.
Grout appearance. Some slots showed what appeared to be a lack of fine materials in the grout.
Inspection two years after construction and after all HVS testing was completed showed some transverse
cracking in the grout in the dowel slots. The grout did not come out of any of the slots, and the cracks
remained tightly interlocked.
Grout tests. Beam and cylinder specimens of the mixed backfill grout material were made from
material sampled at the site. The beams and cylinders were measured for 3rd point loading modulus of
rupture and compressive strength. The results showed that the grout met Caltrans specifications for
flexural strength and for compressive strength at twenty- four hours. The specification for compressive
strength at three hours ( 0.125 days) could not be checked because of a travel delay for the testing
contractor. The test at eight hours ( 0.33 days) indicated high early strength for the mix, but conformity
with the three- hour strength requirement can not be determined.
Early opening to traffic. Early opening on DBR projects was not included by Caltrans in the
scope of this study. Observations are not possible based on this project because trafficking was not
applied to any of the Palmdale test sections until at least one month after construction and after more than
a year for some sections.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xi
TABLE OF CONTENTS
Acknowledgments................................................................................................................ ....................... iii
Executive Summary ............................................................................................................................... ..... iv
List of Figures ............................................................................................................................... ............. xii
List of Tables ............................................................................................................................... ........... xviii
1 Introduction................................................................................................................... .................. 1
1.1 Purpose of This Report................................................................................................................ 1
1.2 Background ............................................................................................................................... . 2
1.2.1 Roughness on Rigid Pavements ....................................................................................... 2
1.2.2 Rehabilitation Strategies for Rigid Pavements................................................................. 5
1.2.3 Alternative DBR Strategies.............................................................................................. 6
1.3 Scope of This Report ................................................................................................................... 6
2 Design, Materials, And Construction of Original Pavement and Dowel Bar Retrofit Sections ...... 7
2.1 Design, Materials, and Construction of Original Pavement ........................................................ 7
2.1.1 Cross- Sections.................................................................................................................. 7
2.1.2 Materials...................................................................................................................... .... 9
2.1.3 Measured Concrete Properties ....................................................................................... 10
2.2 Condition after Original HVS Testing....................................................................................... 12
2.3 Design, Materials, and Construction of Dowel Bar Retrofit ..................................................... 13
2.3.1 Dowel Bar Retrofit Design and Layout.......................................................................... 13
2.3.2 Dowel Bar Types............................................................................................................ 13
2.3.3 Dowel Bar Retrofit Construction ................................................................................... 16
3 Overview of FWD and HVS Testing............................................................................................. 42
3.1 Measurement of Load Transfer Efficiency ( LTE)..................................................................... 42
3.1.1 Calculation Using JDMDs ............................................................................................. 42
3.1.2 Calculation Using the FWD........................................................................................... 44
3.2 Maximum Deflections ............................................................................................................... 45
3.3 Schedule and Conditions of FWD and HVS Tests.................................................................... 45
3.3.1 Chronology of Testing ................................................................................................... 45
3.3.2 FWD Test Conditions..................................................................................................... 45
3.3.3 HVS Test Conditions ..................................................................................................... 47
3.3.4 Use of the HVS to Evaluate Joint Performance ............................................................. 47
3.4 Expected Effects of Pavement Temperature on Deflections, Backcalculated Stiffnesses, and
Load Transfer Efficiency........................................................................................................... 48
4 FWD Test Data Prior to Dowel Bar Retrofit ................................................................................. 49
4.1 Backcalculated Stiffness............................................................................................................ 49
4.2 Load Transfer Efficiency........................................................................................................... 52
5 FWD Tests After DBR and Before HVS Testing .......................................................................... 54
5.1 April 2002 FWD Test................................................................................................................ 54
5.1.1 Backcalculated Stiffness ................................................................................................ 54
5.1.2 Load Transfer Efficiency ............................................................................................... 57
5.2 October 2002 FWD Test ........................................................................................................... 58
5.2.1 Backcalculated Stiffness ................................................................................................ 58
5.2.2 Load Transfer Efficiency ............................................................................................... 62
6 HVS Tests and FWD Tests After DBR and After HVS Testing ................................................... 64
6.1 HVS Results Analysis ............................................................................................................... 64
6.1.1 Failure Criteria for HVS Tests ....................................................................................... 64
6.1.2 Balancing of Environmental Conditions in Experiment Execution ............................... 64
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xii
6.1.3 556FD HVS Test............................................................................................................ 65
6.1.4 557FD HVS Test............................................................................................................ 85
6.1.5 558FD HVS Test.......................................................................................................... 104
6.1.6 559FD HVS Test.......................................................................................................... 124
6.2 FWD Results Analysis ............................................................................................................ 138
6.2.1 April 2003 .................................................................................................................... 138
6.2.2 June 2003 ..................................................................................................................... 143
6.2.3 February 2004 .............................................................................................................. 148
7 Comparison Between Palmdale and Ukiah DBR Sections, and Ukiah/ Palmdale DBR and
Palmdale LLPRS Sections Using FWD and HVS Results .......................................................... 154
7.1 Comparison between Palmdale DBR Sections........................................................................ 154
7.1.1 FWD Results ................................................................................................................ 154
7.1.2 Comparison of Palmdale DBR Section HVS Results .................................................. 164
7.2 Comparison between Palmdale DBR, Palmdale Originally Installed Dowels, and
Ukiah DBR Using FWD and HVS Results ............................................................................. 166
7.2.1 Comparison between Palmdale DBR and Originally Installed Dowel Sections .......... 166
7.2.2 Comparison between Palmdale DBR and Ukiah DBR Sections Using FWD Results. 174
8 Summary of Observations and Findings...................................................................................... 177
8.1 Observations and Findings ...................................................................................................... 177
8.1.1 Construction and Materials ( Chapter 2) ....................................................................... 177
8.1.2 FWD Results after Original HVS Testing and before DBR ( Chapter 4) ..................... 178
8.1.3 FWD Results after DBR and before HVS testing ( Chapter 5)..................................... 178
8.1.4 Results of HVS Tests ( Chapters 6 and 7)..................................................................... 179
8.1.5 Comparison of DBR Joints with Originally Installed Dowel Joints at Palmdale from
FWD Measurements ( Chapter 7) ................................................................................. 181
8.1.6 Reduction of Built- in Slab Curling from DBR ( Chapter 7) ......................................... 181
8.1.7 Comparison of DBR performance at Ukiah and Palmdale DBR HVS Test Sections
( Chapter 7) ................................................................................................................... 182
References..................................................................................................................... ........................... 183
Appendix A: Layout of DBR Test Sections............................................................................................. 186
Appendix B: Overhead Photographs of HVS Test Sections After Testing.............................................. 190
Appendix C: Grout Strength Data............................................................................................................ 198
Appendix D: Deflection Data .................................................................................................................. 200
Appendix E: Construction Specifications and Special Provisions........................................................... 226
LIST OF FIGURES
Figure 1.1. Upstream view showing faulting on an undoweled concrete pavement ( courtesy of
L. Khazanovich).................................................................................................................. 3
Figure 2.1. Location of Palmdale test sections. ........................................................................................... 8
Figure 2.2. Pavement Structure Diagram for North Tangent ( 23). .............................................................. 9
Figure 2.3. Slab and joint numbering and dimensions of slabs in Section 7 of North Tangent................... 9
Figure 2.4. Epoxy- coated steel dowel........................................................................................................ 15
Figure 2.5. Hollow stainless steel dowel.................................................................................................... 15
Figure 2.6. Glass fiber- reinforced dowel. .................................................................................................. 16
Figure 2.7. Dowel bar slot cutting machine ( rear view). ........................................................................... 17
Figure 2.8. Dowel bar slot cutting machine ( side view, blade arbor and vacuum between front and
back wheels). .................................................................................................................... 18
Figure 2.9. Removable saw blade arbor..................................................................................................... 18
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xiii
Figure 2.10. Quality control of dowel bar slot cut depth. .......................................................................... 19
Figure 2.11. Chipping out of dowel bar slots............................................................................................. 20
Figure 2.12. Removal of concrete from dowel bar slots. ........................................................................... 21
Figure 2.13. Cleaning dowel bar slots with air hose. ................................................................................. 22
Figure 2.14. Sealing joints with caulk........................................................................................................ 23
Figure 2.15. Putting together dowel bar retrofit assemblies. ..................................................................... 24
Figure 2.16. Dowel bar retrofit assembly. ................................................................................................. 24
Figure 2.17. Spraying bond- breaker material on dowel bar retrofit assemblies. ....................................... 25
Figure 2.18. Dowel bar retrofit assemblies in slots ( epoxy- coated steel dowels)...................................... 25
Figure 2.19. Dowel bar retrofit assemblies in slots ( hollow stainless steel dowels).................................. 26
Figure 2.20. Checking dowel bar depth and uniformity of depth. ............................................................. 26
Figure 2.21. Batching backfill grout material into mixer........................................................................... 27
Figure 2.22. Placement of backfill grout in dowel bar slots. ..................................................................... 28
Figure 2.23. Pulling backfill grout into slots with shovels. ....................................................................... 29
Figure 2.24. Vibration of backfill grout with small stinger. ...................................................................... 29
Figure 2.25. Sections 559FD ( Slabs 32, 33, and 34 in foreground), 557FD ( background), and
556FD ( far background under front of HVS) after dowel bar retrofit and grinding,
and before HVS testing..................................................................................................... 30
Figure 2.26. Close- up of surface texture of Section 557FD after dowel bar retrofit and grinding, and
before HVS testing ( blue lines painted on surface indicate future HVS wheeltrack)....... 30
Figure 2.27. Average compressive and flexural strengths from field- prepared backfill grout
specimens...................................................................................................................... ... 33
Figure 2.28. Comparison of long- term flexural beam strength for Palmdale and Ukiah DBR backfill
grout.......................................................................................................................... ....... 34
Figure 2.29. Comparison of long- term compressive strength for Palmdale and Ukiah DBR backfill
grout.......................................................................................................................... ....... 34
Figure 2.30. Core DBR33NE2 with hollow stainless steel dowel located 6 mm above mid- depth of
slab........................................................................................................................... ........ 37
Figure 2.31. Core DBR33NW with epoxy- coated steel dowel located 12 mm above mid- depth of
slab........................................................................................................................... ........ 37
Figure 2.32. Core DBR42NWC with FRP dowel located 62 mm above mid- depth of slab.................... 37
Figure 2.33. Core DBR42NNW with FRP dowel located 23 mm above mid- depth of slab. .................. 37
Figure 2.34. Photograph of Joint 38, Section 556FD, in March 2002, showing generally good
condition. ( Blue line painted on surface shows HVS wheelpath)..................................... 38
Figure 2.35. Photograph of Joint 36, Section 557FD, in March 2002, showing generally good
condition. .......................................................................................................................... 39
Figure 2.36. Photograph of Crack 1, Section 559FD, in March 2002, showing lack of fines,
transverse cracking, and separation at edges. ................................................................... 39
Figure 2.37. Photograph of Joint 33, Section 559FD, in March 2002, showing a small amount of
lack of fines and separation at the edges of the dowel slot closest to the slab center. ..... 40
Figure 2.38. Photograph of DBR slot, Section 557FD, in June 2003, showing tight transverse cracks
in grout and slab................................................................................................................ 40
Figure 2.39. Photograph of DBR slot, Section 559FD, Slab 39, in June 2002, showing some
separation of grout and slab. ............................................................................................. 41
Figure 2.40. Photograph of DBR slot, Section 558FD, Slab 42, in June 2002, showing some
transverse cracking in grout. ............................................................................................. 41
Figure 3.1. Example of LTE calculation from JDMD measurements. ...................................................... 43
Figure 3.2. LTE testing using the FWD..................................................................................................... 44
Figure 4.1. Backcalculated concrete stiffness from center slab deflections, Feb. 2001............................. 51
Figure 4.2. Backcalculated support layer stiffness from center slab deflections, Feb. 2001. ..................... 51
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xiv
Figure 4.3. Backcalculated subgrade k- value from center slab deflections, Feb. 2001. ............................ 52
Figure 4.4. Load Transfer Efficiency, Feb. 2001....................................................................................... 53
Figure 5.1. Backcalculated concrete stiffness from center- slab deflections, April 2002........................... 55
Figure 5.2. Backcalculated support layer stiffness from center- slab deflections, April 2002. .................. 56
Figure 5.3. Backcalculated subgrade k- value from center- slab deflections, April 2002. .......................... 56
Figure 5.4. Load Transfer Efficiency, April 2002 ( with joint numbers shown in DBR section)............... 57
Figure 5.5. Backcalculated concrete stiffness from center- slab deflections, October 2002. ..................... 60
Figure 5.6. Backcalculated support layer stiffness from center- slab deflections, October 2002. .............. 60
Figure 5.7. Backcalculated support layer k- value from center- slab deflections, October 2002. ............... 61
Figure 5.8. Load Transfer Efficiency, October 2002 ( joint numbers shown in DBR section). ................. 61
Figure 6.1. Air temperature and relative humidity across all Palmdale DBR HVS tests........................... 65
Figure 6.2. HVS Test Section 556FD prior to HVS testing....................................................................... 66
Figure 6.3. HVS Test Section 556FD, Joint 38, prior to HVS testing. ...................................................... 67
Figure 6.4. HVS Test Section 556FD, Joint 39, prior to HVS testing. ...................................................... 67
Figure 6.5. JDMD locations and numbering for Section 556FD. .............................................................. 68
Figure 6.6. Air Temperature and relative humidity on Section 556FD during HVS testing. .................... 69
Figure 6.7. Mid- slab temperatures for HVS Test Section 556FD.............................................................. 70
Figure 6.8. Slab temperature gradient for HVS Test Section 556FD. ....................................................... 70
Figure 6.9. Maximum joint deflections under HVS trafficking load on Section 556FD........................... 72
Figure 6.10. Maximum joint deflections under HVS loading with measurement load of 60 kN on
Section 556FD. ................................................................................................................. 72
Figure 6.11. JDMD 1 Peak Deflection vs. mid- slab temperature on Section 556FD under measuring
load of 60 kN ( Joint 39, epoxy- coated steel dowel, 4/ wheelpath). ................................... 73
Figure 6.12. JDMD 2 Peak deflection vs. mid- slab temperature on Section 556FD under measuring
load of 60 kN ( Joint 39, epoxy- coated steel dowel, 4/ wheelpath). ................................... 74
Figure 6.13. JDMD 4 peak deflection vs. mid- slab temperature on Section 556FD under measuring
load of 60 kN ( Joint 38, epoxy- coated steel dowel, 4/ wheelpath). ................................... 74
Figure 6.14. JDMD 5 Peak deflection vs. mid- slab temperature on Section 556FD under measuring
load of 60 kN ( Joint 38, epoxy- coated steel dowel, 4/ wheelpath). ................................... 75
Figure 6.15. LTE under HVS trafficking load on Section 556FD. ............................................................ 76
Figure 6.16. LTE under HVS loading with measuring load of 60 kN on Section 556FD. ........................ 76
Figure 6.17. JDMD 1 LTE vs. mid- slab temperature on Section 556FD ( with measuring load of
60 kN; Joint 39, epoxy- coated steel dowel, 4/ wheelpath). ............................................... 77
Figure 6.18. JDMD 2 LTE vs. mid- slab temperature on Section 556FD ( with measuring load of
60 kN; Joint 39, epoxy- coated steel dowel, 4/ wheelpath). ............................................... 78
Figure 6.19. JDMD 4 LTE vs. mid- slab temperature on Section 556FD with measuring load of
60 kN ( Joint 38, epoxy- coated steel dowel, 4/ wheelpath). ............................................... 78
Figure 6.20. JDMD 5 LTE vs. mid- slab temperature on Section 556FD with measuring load of
60 kN ( Joint 38, epoxy- coated steel dowel, 4/ wheelpath). ............................................... 79
Figure 6.21. Initial condition of slabs in Section 556FD with existing cracking in adjacent slabs
marked in red. ................................................................................................................... 80
Figure 6.22. Cracking in Slab 40 at end of wheelpath after approximately 1.291 million repetitions. ..... 81
Figure 6.23. Cracking in the center slab ( Slab 39) after 1.759 million repetitions. ................................... 82
Figure 6.24. Cracking in Slab 39 after 2.066 million repetitions............................................................... 83
Figure 6.25. Cracking in all slabs at end of HVS loading on Section 556FD after 2.209 million
repetitions.................................................................................................................... ..... 84
Figure 6.26. Close- up of Joint 39 at end of HVS loading after 2.209 million repetitions. ........................ 85
Figure 6.27. HVS Test Section 557FD prior to HVS testing..................................................................... 86
Figure 6.28. HVS Test Section 557FD, Joint 35, prior to HVS testing. .................................................... 87
Figure 6.29. HVS Test Section 557FD, Joint 36, prior to HVS testing. .................................................... 88
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xv
Figure 6.30. JDMD locations and numbering for Section 557FD. ............................................................ 89
Figure 6.31. Air temperature and relative humidity on Section 557FD during HVS testing..................... 90
Figure 6.32. Mid- slab temperature for HVS Test Section 557FD. ............................................................ 91
Figure 6.33. Slab temperature gradients for HVS Test Section 557FD..................................................... 91
Figure 6.34. Maximum joint deflections under HVS trafficking load on Section 557FD......................... 92
Figure 6.35. Maximum joint deflections with HVS test load of 60 kN on Section 557 FD. ..................... 93
Figure 6.36. JDMD 1 peak deflection vs. mid- slab temperature on Section 557FD with measuring
load of 60 kN ( Joint 36, epoxy- coated steel dowel, 3/ wheelpath). ................................... 94
Figure 6.37. JDMD 2 peak deflection vs. mid- slab temperature on Section 557FD with measuring
load of 60 kN ( Joint 36, epoxy- coated steel dowel, 3/ wheelpath). ................................... 94
Figure 6.38. JDMD 4 peak deflection vs. mid- slab temperature on Section 557FD with measuring
load of 60 kN ( Joint 35, epoxy- coated steel dowel, 3/ wheelpath). ................................... 95
Figure 6.39. JDMD 5 peak deflection vs. mid- slab temperature on Section 557FD with measuring
load of 60 kN ( Joint 35, epoxy- coated steel dowel, 3/ wheelpath). ................................... 95
Figure 6.40. LTE under HVS trafficking load on Section 557FD. ............................................................ 96
Figure 6.41. LTE under HVS loading with test load of 60 kN on Section 557FD. ................................... 97
Figure 6.42. JDMD 1 LTE vs. mid- slab temperature on Section 557FD ( with measuring load of
60 kN; Joint 36, epoxy- coated steel dowel, 3/ wheelpath). ............................................... 98
Figure 6.43. JDMD 2 LTE vs. mid- slab temperature on Section 557FD ( with measuring load of
60 kN; Joint 36, epoxy- coated steel dowel, 3/ wheelpath). ............................................... 98
Figure 6.44. JDMD 4 LTE vs. mid- slab temperature on Section 557FD with measuring load of
60 kN ( Joint 35, epoxy- coated steel dowel, 3/ wheelpath). ............................................... 99
Figure 6.45. JDMD 5 LTE vs. mid- slab temperature on Section 557FD with measuring load of
60 kN ( Joint 35, epoxy- coated steel dowel, 3/ wheelpath). ............................................... 99
Figure 6.46. Initial condition of slabs in Section 557FD showing no existing cracking. ........................ 100
Figure 6.47. Mid- slab transverse cracking in Slab 36 after 722,290 repetitions...................................... 101
Figure 6.48. Cracking at the end of the wheelpath in Slab 35 after 776,068 repetitions. ........................ 102
Figure 6.49. Final condition of Section 557FD after 1.122 million repetitions....................................... 103
Figure 6.50. Slab 42 near Joint 41 on Section 558FD after 2.002 million repetitions............................. 104
Figure 6.51. Slab 43 near Joint 42 on HVS Test Section 558FD after 2.002 million repetitions. ........... 105
Figure 6.52. HVS Test Section 558FD, Joint 42 after 2.002 million repetitions..................................... 106
Figure 6.53. HVS Test Section 558FD, Joint 41 after 2.002 million repetitions..................................... 106
Figure 6.54. HVS Test Section 558FD, Crack 2 after 2.002 million repetitions. .................................... 107
Figure 6.55. JDMD locations and numbering for Section 558FD. .......................................................... 107
Figure 6.56. Air temperature and relative humidity on Section 558FD during HVS testing................... 109
Figure 6.57. Mid- slab temperature for HVS Test Section 558FD. .......................................................... 110
Figure 6.58. Slab temperature gradient for HVS Test Section 558FD. ................................................... 110
Figure 6.59. Maximum JDMD Deflections on HVS Test Section 558FD under trafficking loads. ........ 111
Figure 6.60. Maximum JDMD deflections with HVS measuring load of 60 kN on Section 558FD. ..... 112
Figure 6.61. JDMD 1 Peak deflection vs. mid- slab temperature on Section 558FD ( with measuring
load of 60 kN; Joint 42, FRP dowel, 4/ wheelpath)......................................................... 113
Figure 6.62. JDMD 2 Peak deflection vs. mid- slab temperature on Section 558FD ( with measuring
load of 60 kN; Joint 42, FRP dowel, 4/ wheelpath)......................................................... 113
Figure 6.63. JDMD 4 peak deflection vs. mid- slab temperature on Section 558FD ( with measuring
load of 60 kN; Crack 2, FRP dowel, 4/ wheelpath). ........................................................ 114
Figure 6.64. JDMD 5 peak deflection vs. mid- slab temperature on Section 558FD ( with measuring
load of 60 kN; Crack 2, FRP dowel, 4/ wheelpath). ........................................................ 114
Figure 6.65. LTE under HVS trafficking loads on Section 558FD. ........................................................ 115
Figure 6.66. LTE under HVS loading with measuring load of 60 kN on Section 558FD. ...................... 116
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xvi
Figure 6.67. JDMD 1 LTE vs. mid- slab temperature on Section 558FD ( with measuring load of
60 kN; Joint 42, FRP dowel, 4/ wheelpath). .................................................................... 117
Figure 6.68. JDMD 2 LTE vs. mid- slab temperature on Section 558FD ( with measuring load of
60 kN; Joint 42, FRP dowel, 4/ wheelpath). .................................................................... 117
Figure 6.69. JDMD 4 LTE vs. mid- slab temperature on 558FD ( with measuring load of 60 kN;
crack 2, FRP dowel, 4/ wheelpath). ................................................................................. 118
Figure 6.70. JDMD 5 LTE vs. mid- slab temperature on Section 558FD ( with measuring load of
60 kN; crack 2, FRP dowel, 4/ wheelpath). ..................................................................... 118
Figure 6.71. First new crack on Section 558FD from slab edge to back of outside dowel bar slot on
Joint 41 at 1.782 million load repetitions........................................................................ 119
Figure 6.72. Close- up of first new crack after 2.002 million load repetitions. ........................................ 120
Figure 6.73 Mid- slab transverse cracking in Slab 42 near Joint 42 after 2.129 million repetitions. ....... 121
Figure 6.74. Final condition of mid- slab transverse crack and Crack 2 after 2.209 million repetitions... 122
Figure 6.75. Close- up of final condition of Crack 2 at end of HVS loading after 2.209 million
repetitions.................................................................................................................... ... 123
Figure 6.76. HVS Test wection 559FD prior to HVS testing. Blue lines painted on surface show
future HVS wheeltrack. .................................................................................................. 124
Figure 6.77. HVS Test Section 559FD, Joint 32, prior to HVS testing. .................................................. 125
Figure 6.78. HVS Test Section 559FD, Crack 1, prior to HVS testing. .................................................. 125
Figure 6.79. HVS Test Section 559FD, Joint 33, prior to HVS testing. .................................................. 126
Figure 6.80. JDMD locations and numbering for Section 559FD. .......................................................... 126
Figure 6.81. Air temperature and relative humidity during HVS Testing of Section 559FD. ................. 127
Figure 6.82. Mid- slab temperature for HVS Test Section 559FD. .......................................................... 128
Figure 6.83. Slab temperature gradient for HVS Test Section 559FD. ................................................... 128
Figure 6.84. Maximum JDMD deflections on HVS Test Section 559FD under trafficking loads. ......... 129
Figure 6.85. Maximum JDMD deflections with HVS measurement load of 60 kN on Section 559FD.. 130
Figure 6.86. JDMD 1 peak deflection vs. mid- slab temperature on Section 559FD ( with measuring
load of 60 kN; Joint 33, hollow stainless dowel, 4/ wheelpath). ..................................... 131
Figure 6.87. JDMD 2 peak deflection vs. mid- slab temperature on Section 559FD ( with measuring
load of 60 kN; Joint 33, hollow stainless dowel, 4/ wheelpath). ..................................... 131
Figure 6.88. JDMD 4 peak deflection vs. mid- slab temperature on Section 559FD ( with measuring
load of 60 kN; Joint 32, epoxy- coated steel dowel, 4/ wheelpath). ................................. 132
Figure 6.89. JDMD 5 peak deflection vs. mid- slab temperature on Section 559FD ( with measuring
load of 60 kN; Joint 32, epoxy- coated steel dowel, 4/ wheelpath). ................................. 132
Figure 6.90. LTE for all repetitions under trafficking load on Section 559FD........................................ 133
Figure 6.91. LTE under HVS measurement load of 60 kN on Section 559FD. ...................................... 134
Figure 6.92. JDMD 1 LTE vs. mid- slab temperature on Section 559FD ( with measuring load of
60 kN; Joint 33, hollow stainless steel dowel, 4/ wheelpath). ......................................... 135
Figure 6.93. JDMD 2 LTE vs. mid- slab temperature on Section 559FD ( with measuring load of
60 kN; Joint 33, hollow stainless steel dowel, 4/ wheelpath). ......................................... 135
Figure 6.94. JDMD 4 LTE vs. mid- slab temperature on Section 559FD ( with measuring load of
60 kN; Joint 32, epoxy- coated steel dowel, 4/ wheelpath). ............................................. 136
Figure 6.95. JDMD 5 LTE vs. mid- slab temperature on Section 559FD ( with measuring load of
60 kN; Joint 32, epoxy- coated steel dowel, 4/ wheelpath). ............................................. 136
Figure 6.96. Final condition of Slab 33 showing additional cracking around Crack 1 after 2.001
million repetitions. .......................................................................................................... 137
Figure 6.97. Close- up of final condition of Slab 33 and Crack 1. ........................................................... 138
Figure 6.98. Backcalculated concrete stiffness from center slab deflection, April 2003......................... 140
Figure 6.99. Backcalculated support layer stiffness from center slab deflections, April 2003................ 141
Figure 6.100. Backcalculated subgrade k- value from center- slab deflections, April 2003. .................... 141
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xvii
Figure 6.101. Load Transfer Efficiency, April 2003. .............................................................................. 142
Figure 6.102. Backcalculated concrete stiffness from center slab deflection, June 2003........................ 145
Figure 6.103. Backcalculated support layer stiffness from center slab deflections, June 2003............... 145
Figure 6.104. Backcalculated subgrade k- values from center slab deflections, June 2003. .................... 146
Figure 6.105. Load Transfer Efficiency, June 2003................................................................................. 147
Figure 6.106. Backcalculated concrete stiffness from center slab deflection, Feb. 2004. ....................... 150
Figure 6.107. Backcalculated support layer stiffness from center slab deflections, Feb. 2004. .............. 150
Figure 6.108. Backcalculated subgrade k- value from center slab deflections, Feb. 2004. ...................... 151
Figure 6.109. Load Transfer Efficiency, February 2004. ........................................................................ 152
Figure 7.1. Backcalculated concrete stiffness before and after DBR....................................................... 155
Figure 7.2. Backcalculated support layer stiffness before and after DBR. .............................................. 156
Figure 7.3. Backcalculated subgrade k- value before and after DBR. ...................................................... 156
Figure 7.4. LTE from FWD before DBR, after DBR but before HVS testing, and after DBR and
HVS testing for Joint 32 ( epoxy- coated steel dowel, 4/ wheelpath)................................ 157
Figure 7.5. LTE from FWD before DBR, after DBR but before HVS testing, and after DBR and
HVS testing for Joint 33 ( hollow stainless steel dowel, 4/ wheelpath)............................ 158
Figure 7.6. LTE from FWD before DBR, after DBR but before HVS testing, and after DBR and
HVS testing for Joint 35 ( epoxy- coated steel dowel, 3/ wheelpath)................................ 158
Figure 7.7. LTE from FWD before DBR, after DBR but before HVS testing , and after DBR and
HVS testing for Joint 36 ( epoxy- coated steel dowel, 3/ wheelpath)................................ 159
Figure 7.8. LTE from FWD before DBR, after DBR but before HVS testing , and after DBR and
HVS testing for Joint 38 ( epoxy- coated steel dowel, 4/ wheelpath)................................ 159
Figure 7.9. LTE from FWD before DBR, after DBR but before HVS testing , and after DBR and
HVS testing for Joint 39 ( epoxy- coated steel dowel, 4/ wheelpath)................................ 160
Figure 7.10. LTE from FWD before DBR, after DBR but before HVS testing , and after DBR and
HVS testing for Joint 41 ( epoxy- coated steel dowel, 4/ wheelpath)................................ 160
Figure 7.11. LTE from FWD before DBR, after DBR but before HVS testing , and after DBR and
HVS testing for Joint 42 ( FRP dowel, 4/ wheelpath). ..................................................... 161
Figure 7.12. LTE vs. surface temperature for joints with four epoxy- coated steel dowels per
wheelpath from FWD measurements.............................................................................. 162
Figure 7.13. LTE vs. surface temperature for joints with four hollow stainless steel dowels per
wheelpath from FWD measurements.............................................................................. 162
Figure 7.14. LTE vs. surface temperature for joints with four FRP dowels per wheelpath from FWD
measurements.................................................................................................................. 163
Figure 7.15. LTE vs. surface temperature for joints with three epoxy- coated steel dowels per
wheelpath from FWD measurements.............................................................................. 163
Figure 7.16. Comparison of JDMD vertical peak deflection regression lines under HVS loading,
60 kN testing load. .......................................................................................................... 164
Figure 7.17. Comparison of LTE regression lines under HVS loading, 60 kN testing load.................... 166
Figure 7.18 Example of matching of measured and calculated deflections to find EBITD...................... 170
Figure 7.19. Corner deflections for Section 535FD ( Section 7, DBR section before DBR) without
temperature control box. ................................................................................................. 172
Figure 7.20. Corner deflections for Section 536FD ( Section 9, with originally installed dowels)
without temperature control box..................................................................................... 173
Figure 7.21. Corner deflections for Section 539FD ( Section 11, with widened truck lane and
originally installed dowels) without temperature control box. ....................................... 173
Figure 7.22. Corner deflections for Section 556FD ( Section 7, DBR section after DBR) without
temperature control box. ................................................................................................. 174
Figure 7.23. Load Transfer Efficiency across time for Palmdale DBR Section 556FD ( four epoxy-coated
steel dowels per wheelpath)................................................................................. 176
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xviii
Figure 7.24. Daytime Load Transfer Efficiency across time for Ukiah DBR Sections 553FD and
554FD ( four epoxy- coated steel dowels per wheelpath). ............................................... 176
LIST OF TABLES
Table A Summary of Palmdate Dowel Bar Retrofit Test Sections................................................... v
Table 2.1. Summary of Flexural and Compressive Strengths from Specimens Prepared in the
Field during Construction of Section 7 of North Tangent ( 23)......................................... 11
Table 2.2. Summary of Thicknesses and Compressive Strengths from Cores from Section 7 of
North Tangent ( 23) Taken Several Weeks after Construction.......................................... 12
Table 2.3. HVS Dowel Bar Retrofit Test Sections, Test Durations, Slab Numbers and Joint
Numbers........................................................................................................................ ... 14
Table 2.4. Type and Number of Dowels Retrofitted on Each Joint or Crack .................................... 14
Table 2.5. Grout Flexural Beam ( Modulus of Rupture) and Compressive Strength Results from
Field Prepared Specimens at Palmdale ............................................................................. 32
Table 2.6. Grout Flexural Beam ( Modulus of Rupture) and Compressive Strength Results from
Field Prepared Specimens at Ukiah .................................................................................. 33
Table 2.7. Average Concrete Slab Thicknesses in DBR Sections .................................................... 35
Table 2.8. Distance from Top to Center of Dowel from Cores with Dowels .................................... 36
Table 3.1. Timetable of Testing on Palmdale Sections...................................................................... 46
Table 4.1. Summary of Pavement Surface Temperature from February 2001 .................................. 49
Table 4.2. Summary of Backcalculated Stiffness .............................................................................. 50
Table 4.3. LTE Summary from FWD Tests in February 2001 .......................................................... 52
Table 5.1. Summary of Pavement Surface Temperature from April 2002 ........................................ 54
Table 5.2. Summary of Backcalculated Stiffness .............................................................................. 55
Table 5.3. LTE Summary from FWD Tests in April 2002 ................................................................ 58
Table 5.4. Nighttime LTE of Section 7 ( DBR Section)..................................................................... 58
Table 5.5. DBR Joint Details ............................................................................................................. 58
Table 5.6. Summary of Pavement Surface Temperature from October 2002.................................... 59
Table 5.7. Summary of Backcalculated Stiffness .............................................................................. 59
Table 5.8. LTE Summary from FWD Tests in October 2002............................................................ 62
Table 5.9. DBR Joint Details ............................................................................................................. 62
Table 5.10. Average Daytime LTE of Section 7 ( DBR Section), October 2002 ................................. 62
Table 5.11. Average Nighttime LTE of Section 7 ( DBR Section), October 2002............................... 63
Table 6.1. Load History on Section 556FD ....................................................................................... 68
Table 6.2. Extreme Environmental Conditions During HVS Testing of Section 556FD .................. 69
Table 6.3. Load history on Section 557FD ........................................................................................ 89
Table 6.4. Extreme Environmental Conditions during HVS Testing of Section 557FD................... 90
Table 6.5. Load history on Section 558FD ...................................................................................... 108
Table 6.6. Extreme Environmental Conditions during HVS Testing of Section 558FD................. 108
Table 6.7. Load History on Section 559FD ..................................................................................... 127
Table 6.8. Extreme Environmental Conditions during HVS Testing of Section 559FD................. 127
Table 6.9. Temperature Summary from April 2003 ........................................................................ 139
Table 6.10. Summary of Backcalculated Stiffnesses......................................................................... 139
Table 6.11. LTE Summary from FWD Tests in April 2003 .............................................................. 142
Table 6.12. Average Nighttime LTE of Section 7 ( DBR Section), April.......................................... 143
Table 6.13. Temperature Summary from June 2003 ......................................................................... 144
Table 6.14. Summary of Backcalculated Stiffness, June 2003 .......................................................... 144
Table 6.15. LTE Summary from FWD Tests in June 2003 ............................................................... 147
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xix
Table 6.16. Average Nighttime LTE of Section 7 ( DBR Section), June 2003 .................................. 148
Table 6.17. Temperature Summary from February 2004 .................................................................. 149
Table 6.18. Summary of Backcalculated Stiffnesses......................................................................... 149
Table 6.19. LTE Summary from FWD Tests in February 2004 ........................................................ 153
Table 6.20. Average LTE of Section 7 ( DBR Section), February 2004 ............................................ 153
Table 7.1 Summary of Backcalculated Stiffness of DBR Section before and after DBR .............. 155
Table 7.2. HVS Testing Result Summary, under 60 kN testing load. ............................................. 165
Table 7.3. Comparison of LTE for between Palmdale Originally Installed Dowel ( OID) and
Dowel Bar Retrofit ( DBR) Joints ................................................................................... 168
Table 7.4. Calculated EBITD Before and After DBR ..................................................................... 171
Table 7.5. Backcalculated Stiffnesses for Palmdale DBR Sections................................................. 175
Table 7.6. Backcalculated Stiffnesses for Ukiah DBR Sections ( Feb. 2001) .................................. 175
Stage 5 Approved Version
UCPRC- RR- 2006- 02
xx
Stage 5 Approved Version
UCPRC- RR- 2006- 02
1
1 INTRODUCTION
1.1 Purpose of This Report
The purpose of this report is to present the results of a set of Heavy Vehicle Simulator ( HVS) tests,
deflection tests, post- HVS forensic testing and analysis on dowel bar retrofitted ( DBR) concrete pavement
test sections at Palmdale, California. The test sections included retrofitted joints and transverse cracks
with three and four epoxy- coated steel dowels, four hollow stainless steel dowels, and four fiber-reinforced
polymer dowels per wheelpath. The results of the HVS testing and other testing, and the
analysis on the dowel bar retrofit sections at Palmdale are also compared with the results from previous
HVS testing of the following test sections:
• DBR sections at Ukiah
• Sections of new pavement at Palmdale where dowels were installed during construction of
the slab and
• Section of new pavement at Palmdale that were constructed without dowels.
This project was originally proposed in 2000 by the Caltrans Headquarters Division of Design.
Other stakeholder Caltrans units included Headquarters METS Office of Rigid Pavement Materials and
Structural Concrete, as well as Caltrans Districts 1 and 7. Benefits expected from this research are to
provide Caltrans with information needed for decisions about design and construction of DBR in order
( 1) to help determine where DBR may be a cost- effective strategy for rehabilitating rigid pavement and
( 2) to help obtain best performance where DBR is selected as the preferred rehabilitation strategy. This
work was completed as part of Partnered Pavement Research Program Strategic Plan Item 4.8, “ Dowel
Bar Retrofit of Rigid Pavements.”
Tasks for this project focus on four objectives agreed upon with Caltrans. This report completes
the requirements for the first objective and presents observations and results only for the first objective.
This report augments information provided in 2003 about HVS and related tests at Ukiah ( 10).
5. Field Accelerated Pavement Testing with the HVS: To collect full- scale data quickly,
although with heavier loads than normally occur under real traffic. This will compare
performance of retrofitted joints and cracks with those not retrofitted. This testing also
includes measurement of load transfer efficiency ( LTE) and other pavement properties with
the Falling Weight Deflectometer ( FWD). Several generic types of dowels will be included in
the field test sections.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
2
6. Field Live Traffic Testing: To collect field data on a long- term basis ( approximately two
years) under real loads. This testing enables calibration of HVS and analysis results.
7. Laboratory testing of materials: Permits evaluation of additional variables that cannot be
included in the HVS testing, such as corrosion of the dowels and dowel types not included in
the field test sections. Laboratory testing is also used to characterize materials used in the
HVS test sections.
8. Modeling
• Finite element analysis of doweled concrete pavement joints: allows for performance
prediction of other options without testing; enables extrapolation of HVS results.
• Compilation of performance data from existing DBR projects throughout the U. S.: allows for
calibration of HVS and analysis results to field project results.
• Life Cycle Cost analyses.
1.2 Background
1.2.1 Roughness on Rigid Pavements
Rigid pavements, also referred to as concrete pavements or portland cement concrete ( PCC)
pavements, make up a large and important portion of the state highway network owned and maintained by
the California Department of Transportation ( Caltrans). They represent approximately 18 percent of the
centerline- kilometers in the state network and 32 percent of the lane- kilometers ( 1). The difference
between the centerline- kilometers and the lane- kilometers indicates the extent to which rigid pavements
have been used for multilane facilities in urban areas. Rigid pavements have also been used extensively
for the interstate system and other routes with heavy truck traffic in California.
Much of the rigid pavement network has performed well beyond its original twenty- year design
life. It has been estimated that approximately 90 percent of the states’ rigid pavement were constructed in
the fifteen years between 1959 and 1974 ( 2), and they are therefore now forty to fifty- five years old.
Because they have been subjected to many years of heavy truck traffic many of these pavements are now
in need of maintenance or rehabilitation. In 2003, rigid pavements made up 27 percent of the distressed
pavement lane- kilometers. ( 1)
Smoothness is the primary means by which the public evaluates pavement condition ( 3, 4, 5, 6,
7), and it is a significant variable controlling vehicle operating costs for both passenger and freight
vehicles ( 8, 9). Smoothness is also one of the three variables by which Caltrans prioritizes pavement
maintenance and rehabilitation, the others being the amount of traffic and the extent of cracking.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
3
The smoothness of newly constructed rigid pavements is controlled by the paving process. The
primary cause of increases in roughness ( or lack of smoothness) on rigid pavements after construction is
the development of faulting on transverse joints, and faulting of transverse cracks when they occur. The
smoothness of a concrete pavement through its life is therefore controlled by its initial smoothness and the
development of faulting under traffic loading. Faulting, sometimes referred to as “ step- faulting,” is the
difference in height between two concrete slabs at a transverse joint or crack, as shown in Figure 1.1. As
faulting develops, the edge of the “ upstream” slab at the joint becomes higher than the edge of the
“ downstream” slab.
Figure 1.1. Upstream view showing faulting on an undoweled concrete pavement
( courtesy of L. Khazanovich).
A previous report presented analyses using empirical models relating faulting to International
Roughness Index ( IRI, the measure of ride quality used by Caltrans) that showed that a relatively low
level of faulting ( 3 mm or more) can lead to levels of roughness that cause discomfort to drivers, increase
Stage 5 Approved Version
UCPRC- RR- 2006- 02
4
vehicle operating costs and freight damage, and that exceed Caltrans IRI criteria for triggering
maintenance or rehabilitation. ( 10)
The conclusion that roughness is primarily controlled by faulting is reinforced by other recent
work. The IRI model in the recently released draft report on the pavement design guide developed by
NCHRP Project 1- 37A ( 11) shows that faulting is the most important factor controlling ride quality in
jointed plain concrete pavement ( JPCP), the type of concrete pavement built by Caltrans. The report states
that:
By far the most sensitive factor affecting JPCP [ ride quality] is joint faulting and the most
critical factor affecting joint faulting is dowel diameter.
That report also states that:
The use of properly sized dowels is the most reliable and cost- effective way to control joint
faulting. Studies have shown that properly sized dowels with adequate consolidation will
reduce faulting dramatically.
These statements are based on analysis of new rigid pavement.
Historically, Caltrans has relied on improving the non- erodability of base materials through
cement stabilization as well as aggregate interlock between the slabs at the joint to control faulting
development. Empirical models developed from field data indicate that although the use of non- erodible
bases, usually meaning cement treated base or asphalt treated base, improves faulting performance, their
use is not as effective as the use of dowels. The field data and models indicate that the best faulting
performance is obtained by using dowels with non- erodible bases ( 12, 13, 14, 15).
Caltrans has not used dowels except for a few projects built since 1999. The decision not to use
dowels during initial construction of most of the rigid pavements in the state network was primarily based
on the results of test sections evaluated in the late 1940s ( 16). There were construction difficulties at that
time, primarily an inability to place dowels straight and level, which can lead to early failure of the
pavement.
Typical dowel practice at that time also included the use of small diameter dowels, typically 19 to
25 mm. Recent studies have shown that effectiveness of dowels placed in new highway pavements is
greatly improved when the diameters are at least 32 mm, although 25 mm dowels still perform better than
undoweled pavements ( 11, 13).
Small diameter dowels result in high bearing stresses in the surrounding concrete under traffic
loading, which causes the concrete to not hold the dowel as tightly as when originally constructed. As
dowels become loose in the concrete they lose their ability to perform their intended function, which is to
transfer loads from one slab to another across the transverse joint. Load transfer efficiency ( LTE) is
Stage 5 Approved Version
UCPRC- RR- 2006- 02
5
typically measured by deflection testing across joints and cracks, and has been found to be highly
correlated with the development of faulting ( 11, 13). The details of the mechanism of faulting
development were presented in detail in a previous report ( 10), and are also discussed in various other
reports ( 11, 13, 14, 15).
1.2.2 Rehabilitation Strategies for Rigid Pavements
Caltrans is interested in finding the most cost- effective rehabilitation strategies for concrete
pavements. In the five fiscal years, 1998/ 99 through 2002/ 03, Caltrans made a major effort to improve
smoothness on the rigid pavement network by performing diamond grinding on about 12 percent
( 3,220 ln- km) of its rigid pavement network, and placing asphalt concrete overlays on about 8 percent
( 2,089 ln- km), which reduced the lane- kilometers with poor ride quality but no major structural defects to
790 ln- km by the time of the 2003 State of the Pavement. Dowel bar retrofit ( DBR) has also been used by
Caltrans on some projects ( 1).
It was found that reflection cracking appears on the surface of asphalt concrete overlays of rigid
pavements approximately seven to eight years after construction, on average. This finding was based on
analysis of Caltrans PMS data for the years 1978 to 1993, and only considering initial overlays. There
was a wide variance of the reflection cracking lives around the average, with the variance likely
dependent on climate, traffic, and the distresses present at the time of overlay ( 17). Significant increases
in roughness typically do not occur until some time after the appearance of reflection cracking, after the
extent and severity of the reflection cracking has increased. The time to triggering of maintenance or
rehabilitation for asphalt concrete overlays of rigid pavements due to roughness, as measured by IRI, has
not been definitively established with Caltrans PMS data. It has also not been established for diamond
grinding projects on Caltrans pavements. The time to reach trigger values of roughness is also expected to
be dependent on climate, traffic, and existing pavement condition. The condition of cement- treated bases
and asphalt- treated permeable bases under older pavements and its effect on the lives of grinding and
asphalt concrete overlays is of concern because of the variability of the durability of some Caltrans bases,
particularly those designed and constructed before Caltrans began using lean concrete base. ( 15, 18,
19, 20).
While it is clear that the use of properly constructed dowels in new pavements significantly
reduces the rate of fault development, the long- term performance of DBR is not clear. Whether dowel bar
retrofit provides similar performance, as measured by Load Transfer Efficiency, is the purpose of this
research, as is how can dowel bar retrofit be made more cost- effective. The information presented in this
report compares the results of Heavy Vehicle Simulator ( HVS) testing and other testing, and analysis of
Stage 5 Approved Version
UCPRC- RR- 2006- 02
6
dowel bar retrofit sections with that of new doweled sections and undoweled sections. This information
will later be used with information presented in other reports that are part of this research project in order
to estimate the life of dowel bar retrofit sections and their life- cycle cost for comparison with alternative
rehabilitation strategies.
1.2.3 Alternative DBR Strategies
The information in this report also examines the performance of various dowel bar retrofit designs.
Corrosion of epoxy- coated dowels has been identified as a risk that could shorten the life of DBR projects
in locations where the pavement is exposed to high chloride contents, particularly where salt is used to
melt ice and snow. This risk has been identified through accelerated laboratory testing and observation of
some field sections ( 21). Alternative dowels have been proposed that are expected to have better
corrosion resistance. Alternatives tested in the HVS sections included in this study were dowels made of
grout– filled, hollow stainless steel cylinders and dowels made of fiber- reinforced polymer ( FRP).
Caltrans has used four dowels per wheelpath for DBR, which is the design originally used in
Washington State. The Washington State Department of Transportation ( WSDOT) and other states have
also used three dowels per wheelpath with the intention of reducing the cost of DBR. This study includes
a comparison under HVS loading of four and three dowels per wheelpath.
1.3 Scope of This Report
The original construction of the pavement sections is summarized from previous reports, and the
details of the DBR design, materials, and construction are presented in Chapter 2. The analysis and
comparison of HVS and Falling Weight Deflectometer ( FWD) testing data are described in Chapters 3
through 6. Chapter 3 presents an overview of the FWD and HVS testing performed. Chapter 4
summarizes FWD test data from the sections prior to dowel bar retrofit. Chapter 5 presents FWD test data
from the sections after DBR, but before HVS testing. Chapter 6 presents the HVS test data and FWD test
data from the sections after DBR and after HVS testing. Chapter 7 compares the FWD and HVS results
between the different Palmdale DBR test pavements, and between the Ukiah and Palmdale DBR and
Palmdale LLPRS test pavements. Chapter 8 presents the conclusions and recommendations drawn from
this study.
Appendix A presents the layout of the HVS test sections in detail. Appendix B includes overhead
photographs of all test sections with cracks marked on the pavement at the conclusion of HVS testing.
Appendix C contains details of grout strength testing. Appendix D includes detailed deflection test data.
Appendix E includes the test section construction specifications and special provisions.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
7
2 DESIGN, MATERIALS, AND CONSTRUCTION OF ORIGINAL
PAVEMENT AND DOWEL BAR RETROFIT SECTIONS
This chapter presents the details regarding the design, materials, and construction of the original
Palmdale test sections, and of the DBR sections.
Two sections of pavement were built in June 1998 for Heavy Vehicle Simulator testing to
investigate concrete pavement design features for Long- Life Pavement Rehabilitation Strategies
( LLPRS). Both sections are on State Route 14, approximately six kilometers south of Palmdale in
northeastern Los Angeles County, as shown in Figure 2.1. Each section is 210 m long. Both were built as
an outside lane tied to the existing traffic lanes, one in the southbound direction and one in the
northbound direction, and they are referred to as the South Tangent and North Tangent, respectively. All
of the sections are jointed plain concrete pavement ( JPCP).
The South Tangent sections were used to evaluate the fatigue performance of Fast- Setting
Hydraulic Cement Concrete ( FSHCC). The North Tangent sections were built to evaluate the
performance of undoweled pavements, doweled pavements with tied shoulders, and doweled pavements
with a wide lane. ( 22, 23)
2.1 Design, Materials, and Construction of Original Pavement
2.1.1 Cross- Sections
The dowel bar retrofit HVS tests were conducted on previously untrafficked portions of what is
referred to as “ Section 7” of the North Tangent. As can be seen in Figure 2.2, the pavement in Section 7
has cement- treated base ( CTB), undoweled jointed plain concrete slabs with untied shoulders, and a
standard width of 3.7 m.
The concrete slabs on Section 7 were nominally 200 mm thick. All slab joints were sawed at 90° with
spacing matching that of the adjacent slabs on SR 14. The joint spacing for the entire North Tangent
approximately follows the pattern of 3.7, 4.0, 5.5, and 5.8 m. Figure 2.3 shows the slab numbering for
Section 7 of the North Tangent, and the location and dimensions of the HVS test sections on the original
pavement.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
8
Figure 2.1. Location of Palmdale test sections.
Palmdale Test
Sections
Stage 5 Approved Version
UCPRC- RR- 2006- 02
9
Figure 2.2. Pavement Structure Diagram for North Tangent ( 23).
892+ 80 Station:
892+ 90 892+ 80 893+ 00 893+ 10 893+ 20 893+ 30 893+ 40 893+ 50
NORTH TANGENT
SECTION 7
( NO TIE BARS AND DOWELS)
30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
Previous HVS test sections are 8 m long, with HVS jacks placed on wooden sleepers 2.8 m off each end of the section.
Figure 2.3. Slab and joint numbering and dimensions of slabs in Section 7 of North Tangent.
2.1.2 Materials
Visual examination of the subgrade material indicated that it consists of uplifted alluvial deposits with
large stones (> 5 cm diameter) included and some weak- to- relatively strong cementing of the sand and
gravel. Dynamic Cone Penetrometer ( DCP) measurements were performed on the South Tangent at the
Stage 5 Approved Version
UCPRC- RR- 2006- 02
10
time of original construction. The results were highly variable, which was attributed to the presence of
stiff rock material at relatively shallow depths — sometimes as shallow as 0.3 m below the top of the
subgrade. The stiffness of the subgrade on the South Tangent was estimated to be between 118 and 447
MPa at those locations where the DCP did not reach the bedrock. It was assumed that results of the North
Tangent would be similar because both the North and South Tangents are in similar deep cuts and are
very close to each other. However, DCP measurements were not possible on the North Tangent. Subgrade
stiffnesses backcalculated from Falling Weight Deflectometer deflections measured seven to ninety days
after construction averaged for Section 7 were between 159 and 255 MPa. ( 23)
The North Tangent was constructed with 150 mm of Class 2 aggregate subbase ( ASB) placed on
subgrade compacted to Caltrans specifications. A 100- mm thick layer of Class A cement- treated base
( CTB) was placed on the aggregate subbase. The CTB was designed to have a seven- day compressive
strength of 1.895 MPa ( 275 psi) ± 0.345 MPa when tested with CT 312 to simulate material meeting the
pre- 1964 Caltrans specification. The CTB mix design is shown in Table 4.1 of Reference 23.
The mix design for the concrete is described in Section 4.3 of Reference 23 and Section 3.2 of
Reference 24. The specification called for a minimum cement content of 375 kg/ m3. The specification
called for flexural strengths of 2.8 MPa after eight hours and 4.1 MPa after seven days in accordance with
Caltrans Test 523 ( center- point loading). The specifications also called for a three- hour compressive
strength of 17.2 MPa and a three- day compressive strength of 34.5 MPa in accordance with ASTM C
109. Before the test sections were constructed, the contractor had to demonstrate through a trial slab that
the eight- hour and seven- day flexural strength specifications could be met with the proposed mix design.
The concrete mix design includes the following constituents: one coarse and fine aggregate, two
cement types ( Type I/ II portland cement and fast- setting hydraulic cement produced by Ultimax), water,
air entraining agent, DelvoTM liquid, or solid retarder. Table 4.2 of Reference 23 shows the proportion of
each mix constituent for one cubic meter. Ultimax is a proprietary cement with its main chemical
constituent being calcium sulfoaluminate. The contractor used a blend of the two cements to achieve the
required strength specifications. After trying several trial slabs with blends of cements ranging from
100 percent Ultimax and zero percent portland to 70 percent Ultimax and 30 percent portland, the
contractor finally chose a blend of 80 percent Ultimax and 20 percent portland. The 80/ 20 blend was used
on all of the test sections.
2.1.3 Measured Concrete Properties
The subgrade, subbase, and base materials were prepared in May and early June 1998. The slabs in
Section 7 were placed on June 18, 1998.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
11
Various fresh mix properties were tested at the time of construction, and specimens for strength
tests were prepared. Flexural and compressive strengths were measured from two different trucks on each
section at eight hours, seven days, ninety days, and long- term ( 575 days for flexural and 636 days for
compressive). Flexural strength tests were performed following ASTM C 78 ( third- point loading) with
some additional tests performed following CT 523 ( center- point loading). Compressive strength tests
were performed following ASTM C 39. The test information relevant to the DBR test sections is
summarized below from information in Reference 23.
Two specimens from each truck were tested at each age for each type of strength. Specimens
tested at eight hours were cured on site in a mold placed under wet burlap next to the pavement.
Specimens for later testing were demolded at forty- eight hours and sprayed with water, then wrapped in
wet burlap and plastic and cured at room temperature.
Tests were performed on material from the trucks that placed materials on Slabs 33 and 41 in
Section 7 of the North Tangent, which are on or are very close to the DBR sections, as can be seen in
Figures A1 through A3 in Appendix A.
Air entrainment test values were 2 percent and zero percent for the two trucks. Slump test results
were greater than six inches and four inches.
Flexural and compressive strength results averaged for the two trucks are shown in Table 2.1. The
results show that there was almost no long- term strength gain after ninety days.
Table 2.1. Summary of Flexural and Compressive Strengths from Specimens Prepared in the Field
during Construction of Section 7 of North Tangent ( 23)
Flexural Strength
( ASTM C 78)
Compressive Strength
Age Average ( ASTM C 39)
( MPa)
Coefficient of
Variation (%)
Average
( MPa)
Coefficient of
Variation (%)
8 hours 1.91 5 12.4 6
7 days 3.66 4 28.3 11
90 days 5.20 13 48.3 4
Long- term
575 days flexural
636 days compressive
5.20 5 50.1 18
Stage 5 Approved Version
UCPRC- RR- 2006- 02
12
Caltrans performed center- point flexural beam strength tests on beams cured at the site. The average
strength at eight hours was 1.17 MPa with a coefficient of variation of 44 percent. The average strength at
seven days was 3.87 MPa with a coefficient of variation of 9 percent.
Compressive strength tests were also performed on cores taken from the site from several slabs
on the opposite edge of the slab from the location of the future HVS wheel track. The cores were taken
and tested several weeks after construction. The results for Section 7 are summarized in Table 2.2.
Table 2.2. Summary of Thicknesses and Compressive Strengths from Cores from Section 7 of
North Tangent ( 23) Taken Several Weeks after Construction
Slab Length ( mm) Density ( g/ cm3)
Compressive Strength
Corrected for Dimensions
( MPa)
32 220 2.46 32.4
35 228 2.31 41.3
39 220 2.24 17.5
43 237 2.36 33.1
The results indicate that the slabs where typically somewhat thicker than the design thickness of 203
mm. The strengths indicate a fairly large degree of variation between different slabs.
Roesler and Rao reported an average elastic modulus of the concrete slabs backcalculated from
Falling Weight Deflectometer deflections on the North Tangent of approximately 42,500 MPa ( 24).
Coefficient of Thermal Expansion ( COTE) was measured on concrete specimens prepared in the
laboratory using raw materials collected at the site during construction ( 25). The average coefficient of
thermal expansion, in dry condition, was 8.03 ( 10- 6) ε/ º C.
Laboratory shrinkage tests indicated that the cement used for construction of the original HVS
test sections resulted in high concrete free shrinkage as measured using Caltrans and ASTM tests. Static
strain gauges cast in the concrete slabs also showed high differential shrinkage between the locations 50
mm above the bottom of the slabs and 50 mm below the top of the slabs. Very low humidity and low
rainfall greatly contributed to the differential shrinkage, which resulted in warped slabs with the edges
warped upward ( 25).
2.2 Condition after Original HVS Testing
All of the original slabs in the North Tangent that were 5.5 and 5.8 m in length had top- down
transverse cracking within three months of construction under shrinkage and curling stresses and before
the HVS was brought onto the site ( 25). HVS testing was performed on the shorter slabs. The slabs and
joints selected for DBR and subsequent HVS testing for this study were those that had not been
Stage 5 Approved Version
UCPRC- RR- 2006- 02
13
significantly damaged by the HVS testing during the previous experiment. Slabs tested in Section 7 as
part of the original experiment are shown in Figure 2.3.
2.3 Design, Materials, and Construction of Dowel Bar Retrofit
2.3.1 Dowel Bar Retrofit Design and Layout
Four sections on Section 7 were dowel bar retrofitted for later Heavy Vehicle Simulator ( HVS) test
sections. Details of the four HVS test sections are shown in Tables Table 2.3 and Table 2.4, and in
Appendix A. It can be seen in the two tables and in Figure 2.3 that two of the test sections, 558FD and
559FD, had long slabs with transverse cracks caused by shrinkage and temperature gradients and that the
cracks were dowel bar retrofitted.
2.3.2 Dowel Bar Types
Three types of dowels were included in the HVS test sections: epoxy- coated steel; grout- filled,
hollow, stainless steel cylinders; and fiber- reinforced polymer ( FRP).
2.3.2.1 Epoxy- coated steel
The epoxy- coated steel dowels were made of carbon steel coated with flexible epoxy ( green color
code). Epoxy- coated bars were also epoxy coated at the ends. The epoxy- coated dowels are 38 mm in
diameter and 457 mm long. One of the dowels is shown in Figure 2.4.
2.3.2.2 Hollow stainless steel dowel
The hollow stainless steel dowels consisted of a hollow- type A316 stainless steel cylinder
approximately 5 mm thick, filled with a cementitious grout. The hollow stainless steel dowels are 38 mm
in diameter and 457 mm long. One of the dowels is shown in Figure 2.5.
2.3.2.3 Fiber- reinforced polymer
Glass FRP ( GFRP) dowels were used for this project. The fiber- reinforced polymer consists of a
polyester matrix with 70 percent glass fibers by volume. The FRP dowels are 38 mm in diameter and 457
mm long. This type of dowel is shown in Figure 2.6.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
14
Table 2.3. HVS Dowel Bar Retrofit Test Sections, Test Durations, Slab Numbers and Joint
Numbers
HVS Test Section Slab Number Joint ( Crack) Number
38 38
556FD 39
( March 2002 – August 2002)
40
39
35 35
557FD 36
( August 2002 – October 2002)
37
36
41 41
42 Crack 2
558FD
( August 2001 – March 2002)
43
42
32
32
33 Crack 1
559FD
( October 2002 – March 2003)
34
33
Table 2.4. Type and Number of Dowels Retrofitted on Each Joint or Crack
Joint or Crack
Number
HVS Test
Section Type of Dowels Number of Dowels
Joint 32 559FD Epoxy- coated steel Four per wheelpath
Crack 1 559FD
Hollow stainless steel in one
wheelpath; epoxy- coated steel in
other wheelpath
Four per wheelpath
Joint 33 559FD Hollow stainless steel Four per wheelpath
Joint 35 557FD Epoxy- coated steel Three per wheelpath
Joint 36 557FD Epoxy- coated steel Three per wheelpath
Joint 38 556FD Epoxy- coated steel Four per wheelpath
Joint 39 556FD Epoxy- coated steel Four per wheelpath
Joint 41 558FD Epoxy- coated steel Four per wheelpath
Crack 2 558FD Fiber- reinforced polymer Four per wheelpath
Joint 42 558FD Fiber- reinforced polymer Four per wheelpath
Stage 5 Approved Version
UCPRC- RR- 2006- 02
15
Figure 2.4. Epoxy- coated steel dowel.
Figure 2.5. Hollow stainless steel dowel.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
16
Figure 2.6. Glass fiber- reinforced dowel.
2.3.3 Dowel Bar Retrofit Construction
The DBR construction was performed June 28– 30, 2001 by the PenHall Company. The construction
specification is included in Appendix E. The construction inspection was performed by the UCPRC based
on training provided earlier by the Washington State Department of Transportation for the construction of
DBR test sections at Ukiah ( 10).
2.3.3.1 Construction Process
A specially designed machine was used to cut the dowel bar slots ( see Figure 2.7 and Figure 2.8). The
machine has a system to vacuum up cutting water to prevent storm water contamination. The machine
also has a removable arbor on which the saw blades are mounted ( Figure 2.9). The blade configuration
was changed during construction to permit sawing of the sections with four dowels per wheelpath and
those with three dowels per wheelpath.
Quality control of saw cut depths was performed by the contractor as shown in Figure 2.10, and
some of the dowels were checked by UCPRC staff prior to chipping out of the dowel bar slots, as shown
in Figure 2.11 and Figure 2.12. A high- pressure air hose was then used to clean debris from the slots, as
Figure 2.13 shows. The joint was sealed with DAPTM 25- year Painter’s Acrylic Latex Caulk ( Figure
2.14).
Dowel bar retrofit assemblies were put together for each crack or joint ( Figure 2.15). The
assemblies consisted of the dowel, two end caps, two chairs, and a joint separator made of foam backer
board ( Figure 2.16). The assemblies were then sprayed with bond- breaker material, as shown in Figure
Stage 5 Approved Version
UCPRC- RR- 2006- 02
17
2.17, and placed into the slots, as shown in Figure 2.18 and Figure 2.19. Dowel bar depth and uniformity
of depth were then checked ( Figure 2.20).
Backfill grout was prepared on site using a small portable batch mixer, shown in Figure 2.21. The
backfill grout was dumped out of the mixer into a small front loader bucket and driven along the shoulder
to the dowel bar assembly. At the joint or crack the grout was pulled out of the loader into the slots with
shovels ( Figure 2.22 and Figure 2.23), and vibrated with a small stinger ( Figure 2.24). The next day a
grinding machine ground the grout flush with the surface of the pavement, resulting in the texture shown
in Figure 2.25 and Figure 2.26. The blue lines in the photographs indicate future HVS wheelpath center
over the dowel bar set.
Figure 2.7. Dowel bar slot cutting machine ( rear view).
Stage 5 Approved Version
UCPRC- RR- 2006- 02
18
Figure 2.8. Dowel bar slot cutting machine ( side view, blade arbor and vacuum
between front and back wheels).
Figure 2.9. Removable saw blade arbor.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
19
Figure 2.10. Quality control of dowel bar slot cut depth.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
20
Figure 2.11. Chipping out of dowel bar slots.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
21
Figure 2.12. Removal of concrete from dowel bar slots.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
22
Figure 2.13. Cleaning dowel bar slots with air hose.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
23
Figure 2.14. Sealing joints with caulk.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
24
Figure 2.15. Putting together dowel bar retrofit assemblies.
Figure 2.16. Dowel bar retrofit assembly.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
25
Figure 2.17. Spraying bond- breaker material on dowel bar retrofit assemblies.
Figure 2.18. Dowel bar retrofit assemblies in slots ( epoxy- coated steel dowels).
Stage 5 Approved Version
UCPRC- RR- 2006- 02
26
Figure 2.19. Dowel bar retrofit assemblies in slots ( hollow stainless steel dowels).
Figure 2.20. Checking dowel bar depth and uniformity of depth.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
27
Figure 2.21. Batching backfill grout material into mixer.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
28
Figure 2.22. Placement of backfill grout in dowel bar slots.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
29
Figure 2.23. Pulling backfill grout into slots with shovels.
Figure 2.24. Vibration of backfill grout with small stinger.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
30
Figure 2.25. Sections 559FD ( Slabs 32, 33, and 34 in foreground),
557FD ( background), and 556FD ( far background under front of HVS) after dowel
bar retrofit and grinding, and before HVS testing.
Figure 2.26. Close- up of surface texture of Section 557FD after dowel bar retrofit and grinding,
and before HVS testing ( blue lines painted on surface indicate future HVS wheeltrack).
Stage 5 Approved Version
UCPRC- RR- 2006- 02
31
2.3.3.2 Backfill Grout Material
The backfill material was prepared from aggregate, sacks of cement, and water brought to the site by
the contractor. The material was required to meet the specifications included in Appendix E of this report.
The contractor used Five Star Highway Patch for the cement.
Beam and cylinder specimens of the mixed backfill material were made from material sampled at
the site following ASTM C 31. The beam dimensions were 152 by 152 by 457 mm, and the cylinder
dimensions were 203 mm in height and 102 mm in diameter. The specimens were vibrated using a small
mechanical vibrating rod. Curing compound was placed on the surface of the specimens immediately
after finishing. The same curing compound used on the dowel bar retrofit locations was also used on the
specimens.
The first specimens were tested at the site by a commercial laboratory. One day after construction
the remaining specimens were transported to the UC Pavement Research Center laboratory in Richmond
and placed in a room with 97 percent humidity and a temperature of 20° C. The beams and cylinders were
measured for third- point loading modulus of rupture following ASTM C 78 and compressive strength
following ASTM C 39, respectively. The curing times and measured flexural and compressive strengths
are shown in Table 2.5 and are plotted in Figure 2.27.
From the results it can be seen that the grout met Caltrans specifications for flexural strength and
for compressive strength at twenty- four hours. The specification for compressive strength at three hours
( 0.125 days) could not be checked because of a travel delay for the testing contractor. The test at eight
hours ( 0.33 days) indicates high early strength for the mix, but it cannot be concluded as to whether it met
the three- hour strength requirement.
The strength for the backfill grout at the Ukiah DBR test sections is shown in Table 2.6 and
plotted for comparison against the Palmdale data in Figure 2.28 and Figure 2.29. It can be seen that the
Palmdale grout had greater compressive strength than the Ukiah grout. It can also be seen that the
Palmdale flexural strength was generally greater at each age, although the ultimate flexural strengths
appear to be approximately the same.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
32
Table 2.5. Grout Flexural Beam ( Modulus of Rupture) and Compressive Strength Results from
Field Prepared Specimens at Palmdale
Grout
Curing
Time
( days)
Flexural
Beam
Strength
( MPa)
Average
Caltrans
Specification
( minimum)
Compressive
strength
( MPa)
Average
Caltrans
Specification
( minimum)
0.33 3.16 3.4 0.33 37 21 ( at 0.125
days)
0.33 3.71 0.33
1 3.65 3.8 3.5 1 42 35
1 3.55 1
1 4.11 1
7 5.47 5.3 7 52
7 5.19 7
7 5.13 7
14 6.40 6.5 14 58
14 6.54 14
14 6.53 14
28 6.61 6.4 28 63
28 6.31 28
28 6.35 28
165 7.48 7.1
165 6.83
165 6.92
Stage 5 Approved Version
UCPRC- RR- 2006- 02
33
0
10
20
30
40
50
60
70
0 5 10 15 20 25 30
Time ( days)
Compressive Strength ( MPa)
0
1
2
3
4
5
6
7
8
9
10
Flexural Strength ( MPa)
Compressive Strength
Flexural Strength
Figure 2.27. Average compressive and flexural strengths
from field- prepared backfill grout specimens.
Table 2.6. Grout Flexural Beam ( Modulus of Rupture) and Compressive Strength Results from
Field Prepared Specimens at Ukiah
Grout
Curing
Time
( days)
Flexural
Beam
Strength
( MPa)
Average
Caltrans
Specification
( minimum)
Compressive
Strength
( MPa)
Average
Caltrans
Specification
( minimum)
1.125 3.9 3.8 21 21 21
0.125 3.8 21
1 2.2 2.5 3.5 35
1 2.7
8 3.5 3.7 39 39
8 3.9 39
14 47 44
14 41
37 6.8 6.7 53 54
37 6.5 55
Stage 5 Approved Version
UCPRC- RR- 2006- 02
34
0
1
2
3
4
5
6
7
8
9
10
0 20 40 60 80 100 120 140 160 180
Time ( days)
Flexural Strength ( MPa)
Palmdale Flexural
Strength
Ukiah Flexural
Strength
Figure 2.28. Comparison of long- term flexural beam strength for
Palmdale and Ukiah DBR backfill grout.
0
10
20
30
40
50
60
70
0 5 10 15 20 25 30 35 40
Time ( days)
Compressive Strength ( MPa)
Palmdale Compressive Strength
Ukiah Compressive Strength
Figure 2.29. Comparison of long- term compressive strength for Palmdale
and Ukiah DBR backfill grout.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
35
2.3.3.3 Other Materials
Other materials used on the project, including backer board, chairs, end caps, etc., appeared to meet
the specifications for the project.
2.3.3.4 Slab Thickness and Dowel Position After Construction
After HVS tests, twenty- six cores were obtained from the DBR sections in June 2003, including
fourteen with retrofitted dowels. All the cores were measured to determine the slab thicknesses; the
average results are shown in Table 2.7.
Table 2.7. Average Concrete Slab Thicknesses in DBR Sections
Slab No. 32 33 35 36 38 39 41 42
Thickness
( mm) 212.5 220.4 227.4 233.5 227 220.6 220.8 229.3
For the cores that had dowels, the distances from the slab surface to the dowel ( cover depths) are
shown in Table 2.8, as is the deviation from placement at the mid- depth of the slab. Some examples are
shown in Figure 2.30, Figure 2.31, Figure 2.32, and Figure 2.33. The dowels were cored between the
dowel end and the dowel center, as can be seen in the overhead photographs in Appendix B that are
marked with the dowel core locations. The results show that all of the dowels cored were above the mid-depth
of the slab, and some were very near the surface of the slab. These results indicate that the test
sections were not “ perfect” and probably resemble field construction in terms of dowel placement
variability. Most of the cores in Table 2.8 were taken from the HVS wheelpath.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
36
Table 2.8. Distance from Top to Center of Dowel from Cores with Dowels
Joint Description Core No. Actual
Cover Depth
a ( mm)
Slab Thickness
( mm)
Deviation
from slab
center b ( mm)
Joint 39, epoxy- coated
steel dowel, 4/ wheelpath DBR39NW 76 213 - 12
Joint 38, epoxy- coated
steel dowel, 4/ wheelpath DBR38NW 77 227 - 18
Joint 32, epoxy- coated
steel dowel, 4/ wheelpath DBR32NW 73 214 - 15
Joint 36, epoxy- coated
steel dowel, 3/ wheelpath DBR36NW 75 236 - 24
Joint 35, epoxy- coated
steel dowel, 3/ wheelpath DBR35NE 91 226 - 3
Joint 35, epoxy- coated
steel dowel, 3/ wheelpath DBR35NW 83 227 - 12
Joint 36, epoxy- coated
steel dowel, 3/ wheelpath DBR36NE 74 227 - 21
Joint 42, FRP dowel,
4/ wheelpath DBR42NWC 38 238 - 62
Joint 42, FRP dowel,
4/ wheelpath DBR42NNW 74 232 - 23
Joint 42, FRP dowel,
4/ wheelpath DBR42NNE 66 223 - 27
Joint 33, hollow stainless
steel dowel, 4/ wheelpath DBR33NW2 73 226 - 21
Joint 33, hollow stainless
steel dowel, 4/ wheelpath DBR33NE2 86 221 - 6
Joint 41, epoxy- coated
steel dowel, 4/ wheelpath DBR41NE 81 226 - 13
Joint 41, epoxy- coated
steel dowel, 4/ wheelpath DBR41NW 78 216 - 11
a Compares to expected cover depth of 83mm ( based on design slab thickness of 203mm)
b Vertical deviation of dowel center from slab center = cover depth + ( dowel diameter/ 2)- ( slab thickness/ 2);
positive value indicates dowel below slab center, negative value indicates dowel above slab center.
Stage 5 Approved Version
37
Figure 2.30. Core DBR33NE2 with
hollow stainless steel dowel located
6 mm above mid- depth of slab.
Figure 2.31. Core DBR33NW with
epoxy- coated steel dowel located
12 mm above mid- depth of slab.
Figure 2.32. Core DBR42NWC
with FRP dowel located 62 mm
above mid- depth of slab.
Figure 2.33. Core DBR42NNW
with FRP dowel located 23 mm
above mid- depth of slab.
Stage 5 Approved Version
38
2.3.3.5 Longer- Term Grout Condition
The condition of the grout backfill material in the dowel bar slots was inspected periodically on both
the wheelpath trafficked by the HVS and the wheelpath that was not trafficked. Many of the dowel slots
showed the grout to be in good condition. For example, Figure 2.34 and Figure 2.35 show a photographs
taken in March 2002, ten months after construction, and it can be seen that the only apparent problems
with the construction are slightly out- of- place foam backer boards on several of the joints.
However, some the slots showed what appear to be a lack of fine materials in the grout. This can
be seen in Figure 2.36, which shows the DBR slots on a set of transverse cracks. Some of the cracking on
this slot may be due to placement of the DBR on unconnected transverse cracks without a clear joint or
crack on which to place the foam backer board. A little of the apparent lack of fines can be seen in Figure
2.37, which shows a joint.
Photographs taken in June 2003, two years after construction and after HVS testing was
completed on all sections, showed some transverse cracking in the grout in the dowel slots. The grout did
not come out of any of the slots, and the cracks remained tightly interlocked. This can be seen in Figure
2.38, which shows transverse cracking in the grout and slab and a very small amount of separation of the
grout from the slab. Figure 2.39 shows another slot with some separation of grout and slab. Figure 2.40
shows transverse cracks in the grout.
Figure 2.34. Photograph of Joint 38, Section 556FD, in March 2002, showing
generally good condition. ( Blue line painted on surface shows HVS wheelpath).
Stage 5 Approved Version
UCPRC- RR- 2006- 02
39
Figure 2.35. Photograph of Joint 36, Section 557FD, in March 2002, showing
generally good condition.
Figure 2.36. Photograph of Crack 1, Section 559FD, in March 2002, showing
lack of fines, transverse cracking, and separation at edges.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
40
Figure 2.37. Photograph of Joint 33, Section 559FD, in March 2002,
showing a small amount of lack of fines and separation at the edges of
the dowel slot closest to the slab center.
Figure 2.38. Photograph of DBR slot, Section 557FD, in June 2003, showing
tight transverse cracks in grout and slab.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
41
Figure 2.39. Photograph of DBR slot, Section 559FD, Slab 39, in June 2002,
showing some separation of grout and slab.
Figure 2.40. Photograph of DBR slot, Section 558FD, Slab 42, in June 2002,
showing some transverse cracking in grout.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
42
3 OVERVIEW OF FWD AND HVS TESTING
This chapter presents details of the testing program on the Palmdale DBR sections and adjacent
sections using the Heavy Vehicle Simulator ( HVS) and the Falling Weight Deflectometer ( FWD). The
mobile deflection testing device used for this project is a Heavy Weight Deflectometer ( HWD) which is
the same as an FWD except that it is capable of applying heavier loads. Because it is a more commonly
known term, the HWD is referred to as an FWD in this report.
3.1 Measurement of Load Transfer Efficiency ( LTE)
As implied by the name “ load transfer efficiency” ( LTE), this property is measured through
application of load, and is a measure of what proportion of the load is transferred to the adjoining slab.
LTE is normally expressed as a percentage, with one hundred percent meaning that the two adjoining
slabs act as if they are one continuous slab, and zero percent meaning that the two slabs act entirely
independently.
However, for most testing it is not possible to calculate this value with the actual loads because
deflections, not the loads, are measured on the slabs. Thus this measure must be approximated using
calculations based on the deflections of the slabs.
Separate definitions of LTE were used for measurements taken using Joint Deflection
Measurement Devices ( JDMD) during HVS testing and for measurements taken using the FWD, because
the former uses a moving load and the latter uses a load dropped in one place.
3.1.1 Calculation Using JDMDs
JDMDs are used to measure the vertical movement of a single point on a concrete slab with a moving
wheel load ( normally the HVS wheel). JDMDs are attached to the side of the slab on both sides of the
joint, and a full deflection bowl is recorded as the wheel moves from one slab to another across the joint.
An example of the data is shown in Figure 3.1. Because a moving wheel is used, there is an “ approach
LTE,” which is the load transfer as the load approaches the slab, and a “ departure LTE,” which is the load
transfer as the load leaves the slab. Thus there are four possible LTE values for each joint ( two slabs, with
two LTE values each) on an HVS section because the wheel can traffic in both directions. On an in-service
pavement there can only be two, since the traffic is unidirectional for each slab.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
43
Wheel X Position ( mm)
Deflection ( mm)
y
y
y
y
peak
rel
peak
rel
1
2
2
1
Figure 3.1. Example of LTE calculation from JDMD measurements.
However, there are a number of different ways of defining LTE based on deflection. Normally, there
are four deflection values: the two peak values ( the superscripts in Figure 3.1 and in Equation 1 indicate
the instrument) and the two relative values, which are the deflection of each instrument when the other is
at its peak. If the instrument experiences its relative deflection before its peak, then it is an “ approach
LTE;” if it experiences its relative deflection after its peak deflection then it is a “ departure LTE.” There
are four possible calculations ( two per instrument) that can be performed from the four data values above:
2 1
1 2
2 1
2 1
approach rel depart rel
pair pair
peak peak
approach rel depart rel
single single
peak peak
LTE y LTE y
y y
LTE y LTE y
y y
= =
= =
( 1)
The “ pair” definitions shown in Equation 1 use the deflections of both instruments, when the
wheel is at the same location; the “ single” definitions use the deflection of only one instrument when the
wheel is at different locations. There are arguments for the use of both of these definitions: the single
definition in Equation 1 only uses the deflection of one slab, and so it cannot exceed a value of one, while
the pair definition has the load at a single location, and so compares the deflections of the two slabs under
Stage 5 Approved Version
UCPRC- RR- 2006- 02
44
the same loading condition. The pair definition as been implemented in the HVS database and used in
this report, because it provides more stable results, and therefore makes it easier to track changes in LTE
with damage.
Negative LTE is also possible when the approach slab is being pushed upwards by pumping.
With the pair definition, an LTE value greater than one is possible for the departure slab if it has a high
deflection ( caused by cracking, voids, or curling), and the approach slab has a low deflection.
3.1.2 Calculation Using the FWD
The FWD applies a single, dynamic load pulse to a slab close to the joint, and the peak deflection is
measured on either side of the joint. This loading is shown in Figure 3.2.
y1 y2
Figure 3.2. LTE testing using the FWD.
Load Transfer Efficiency ( LTE) is typically defined for FWD measurements as:
LTE = y2 / y1 ( 2)
where y2 and y1 are the peak deflections under the dynamic load.
The program Elmod 3 ( 26) was used to calculate LTE from FWD deflections for this report. To
calculate LTE, Elmod 3 makes use of Westergaard’s “ load transfer efficiency factor,” j, which is defined
by the equation:
j = 2 y2 / ( y1 + y2) ( 3)
Westergaard’s equation for stress at the bottom of a slab is given for a free edge. This stress can be
calculated with j at a joint with load transfer. The relationship between LTE and j is defined as:
LTE = j / ( 2- j) ( 4)
Stage 5 Approved Version
UCPRC- RR- 2006- 02
45
This calculation is close to the “ approach pair” LTE for the JDMDs, but it uses the peak deflections
of the slabs at both measuring points. Because of this it is less likely to provide values out of the zero to
one range.
3.2 Maximum Deflections
One other parameter was measured to evaluate the condition of load transfer at the transverse joints
and cracks: maximum deflection. Maximum deflection provides an indication of the energy applied by
the deflected slab to the underlying materials as well as to any water that is in the joint and under the
slabs. An approach for evaluating faulting based on the energy applied to the underlying layers by the
action of the joint was proposed by Hoerner ( 14), and a similar approach has been implemented in the
NCHRP 1- 37A software ( 11). This approach will be used to estimate faulting performance using the data
from these and other test sections in a later report.
Maximum deflection, referred to as “ peak deflection” in this report, is the maximum vertical
downward movement of the loaded slab. Under FWD loading this would be the maximum movement of
y1 in Figure 3.2. Under Heavy Vehicle Simulator ( HVS) loading this is the deflection of the approach slab
when the wheel is on the approach slab, which would be y1 in Figure 3.1, as the wheel passes from left to
right.
3.3 Schedule and Conditions of FWD and HVS Tests
3.3.1 Chronology of Testing
The chronology of deflection testing, HVS testing, and coring of the original HVS tests and DBR
sections at Palmdale is shown in Table 3.1.
3.3.2 FWD Test Conditions
The Dynatest Model 8082 Falling Weight Deflectometer ( FWD) Test System was used to generate
the nondestructive testing data analyzed for this report. The FWD generates a transient, impulse- type load
of 25– 30 millisecond duration at any desired load level between 27 kN ( 6,000 lbf) and 245 kN ( 55,000
lbf), thereby approximating the effect of a 50– 80 kph ( 30– 50 mph) moving wheel load. For this project,
test loads were normalized to 44 kN ( 10 kip) and 67 kN ( 15 kip). All the FWD tests were performed
separately in daytime and nighttime, and included testing along the center and the left edge of each slab.
The stationing for this project was carried out in units of feet. The starting point ( Station 0) is located at
the southern end of the test section, with stationing increasing northward.
Stage 5 Approved Version
UCPRC- RR- 2006- 02
46
Table 3.1. Timetable of Testing on Palmdale Sections
Event Time
Construction of Doweled and
Undoweled Pavements June 1998
FWD Test No. 1 June 19 1998
FWD Test No. 2 June 23 1998
FWD Test No. 3 August 1998
FWD Test No. 4 September 1998
FWD Test No. 5 January 1999
FWD Test No. 6 March 1999
HVS Test 532FD June – July 1999
HVS Test 533FD August – November 1999
HVS Test 534FD December 1999 – March 2000
HVS Test 535FD March – April 2000
HVS Test 536FD April – July 2000
HVS Test 537FD July – August 2000
HVS Test 539FD August – September 2000
HVS Test 540FD October – November 2000
HVS Test 538FD December 2000 – January 2001
FWD Test No. 7 February 2001
DBR Construction on Untested
Undoweled Joints and Cracks June 28- 30 2001
HVS Test 558FD August 2001 – March 2002
HVS Test 556FD March – August 2002
FWD Test No. 8 April 2002
HVS Test 557FD August – October 2002
FWD Test No. 9 October 2002
HVS Test 559FD October 2002 – March 2003
FWD Test No. 10 April 2003
FWD Test No. 11 June 2003
Coring of DBR Section June 2003
FWD Test No. 12 February 2004
During FWD tests, the 300- mm diameter loading plate is located at the slab corner. The
deflections under the loading plate and across the transverse joint ( or transverse crack) are measured by
Stage 5 Approved Version
UCPRC- RR- 2006- 02
47
geophones on the FWD. The spacing of geophones starting from the center of the loading plate is: 0, 200,
300, 800, 1200, 1600, and 2000 mm, with geophone numbering beginning with Geophone 1 at the 0 mm
distance, and Geophone 7 at the 2000 mm position. To measure deflections across a joint/ crack, the FWD
is positioned such that the joint/ crack of interest is between Geophone 2 ( 200 mm from loading plate) and
Geophone 3 ( 300 mm from loading plate). A summary of all FWD testing results is shown in
Appendix D.
3.3.3 HVS Test Conditions
The HVS is a 60 tonne mobile loading device that has one wheel ( dual or single) with variable load
and an 8- m long loading span. It can be run in a bi- or a unidirectional mode, channelized or with
programmed lateral wander.
The HVS loading was initiated under a 60- kN ( 13,500 lb) load ( on dual truck tires ( Goodyear
G159A radials) and then increased to 90 kN ( 20,250 lb) on dual truck tires with the tire pressure kept
constant at 689.5 kPa ( 100 psi). The final load used was 150 kN ( 33,750 lb) on an aircraft wheel, with tire
pressure maintained at 1450 kPa ( 210 psi). Channelized ( no wander), bidirectional loading was conducted
on the wheelpath for all tests over the center of the dowel bar group. The roof panels and some of the side
panels of the temperature control chamber were in place during the HVS tests to provide shading to the
test sections. The temperature control system was not used because it was previously shown to not
completely control temperatures and curling in the slabs ( 28); in addition some temperature variability
was desired during the testing.
3.3.4 Use of the HVS to Evaluate Joint Performance
Two important d