t
.l 6.
vs a.

 

.oti .2
.. f
I

 

. »

Allin.

. .. VII‘ u.
u 3‘ v A .z. H3 ‘1... .9...

(5

fl“

399 a

.
unfriu
can...»

‘14

.2:
I

 

S l’l
«and. 4
a
. ht ..
Lméii Y
. fix-50....
23......
10:31.. 2......
‘ 3...: u.»
Jill-I’li. ‘1.
1 5‘1... .
afirfiv
will 41;- .l
i . .
1.4... I}...
I: .I I
fimnqmgsfiafiwm
. all? J...
Wu... 1-1.! Rd
.1 ‘
4v.
1.... ii
. A .. nth»! ‘
“2... , 2012.-.!” nun-nay
«.113; 5
i... 6;. .i
y .54... . .
. 3 u...)
. i It“... a
‘ivuiu. cuh :a.}5.|3.!.(.h
. :25: . . .
:H: . 1......
3.3.
:5. L...

t;

LIBRARY
Michigan State
University

This is to certify that the
thesis entitled

EXPERIMENTAL CHARACTERIZATION OF THE PULLOUT AND
GAP OPENING BEHAVIOR OF MISALIGNED DOWEL BARS
UNDER THERMAL EXPANSION

presented by

Deepa Thandaveswara

has been accepted towards fulfillment
of the requirements for the

Master of Science degree in Civil EngineerinL

 

 

/

Mjor Professor’s Signature
09/06/05

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 c:/CIRC/DateDue.indd-p.15

EXPERIIVIENTAL CHARACTERIZATION OF THE PULLOUT AND GAP OPENING
BEHAVIOR OF MISALIGNED DOWEL BARS UNDER THERMAL EXPANSION

By

Deepa Thandaveswara

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Civil and Environmental Engineering

2005

ABSTRACT

EXPERIMENTAL CHARACTERIZATION OF THE PULLOUT AND GAP OPENING
BEHAVIOR OF MISALIGNED DOWEL BARS UNDER THERMAL EXPANSION

By
Deepa Thandaveswara

This thesis focuses on studying experimentally the fundamental pullout behavior of
misaligned dowel bars and gap-opening behavior of pavement joints with misaligned
dowel bars in plain concrete slabs, under thermal expansion. The variables considered in
the experimental plan include the number of dowel bars (1, 2, 3, and 5) and number
misaligned (all or alternate), misalignment type (horizontal, vertical, and combined),
orientation (uniform and non-uniform) and magnitude (0, Ms, 1/2 , % , l, and 2 in. over
half-length (9 in.) of the dowel bar). The effects of these parameters on the bond stress
between the dowel-concrete, load induced at different joint openings, and joint and slab
distresses (spalling, cracking, non-uniform joint opening, and uplift) were studied.

The initial slip/debonding stress found to be in the range of 10-50 psi irrespective
of the misalignment type and magnitude. Within a misalignment type, the load and
intensity of distresses increases with an increase in the misalignment magnitude and
number of dowels misaligned. Non-uniform orientation of misalignment of two dowel
bars requires more load per dowel bar at a given joint opening as compared to the
uniform orientation of dowels in the horizontal and combined misalignment types. For a
given misalignment magnitude, the overall trend of load versus type of misalignment
(horizontal, vertical and combined) is unclear. The load versus joint opening behaviors
obtained through this study can be used for development of analytical models to obtain

the stress states in dowel-concrete bond behavior subjected to pullout.

 

 

Copyright by
DEEPA THANDAVESWARA

2005

 

Dedicated to

My parents

iv

ACKNOWLEDGENIENTS

I would first like to express my sincere gratitude to Dr.Neeraj Buch for his constant
support, ideas, and guidance throughout this entire research study. I would also like to
thank my committee members Dr.Karim Chatti and Dr.Gilbert Baladi for their support. I
would like also to thank Dr.Amit Varma, co-principal investigator of the project for his
guidance during this research study.

I would like to thank Mr.Michael Backer, Mr.Doug Branch and other field
engineers at the Michigan Department of Transportation for sharing their field
experiences and facilitating field trips during different stages of the research. My
appreciation is due to the Michigan DOT for sponsoring this research work.

I would like to thank Mr.Siavosh Ravanbakhsh, lab manager and Mr.J.C Brenton,
lab technician for the help extended to me during the laboratory phase of my research. I
would like to thank my fellow graduate student Milind Prabhu for the enormous help and
assistance extended to me during the period of this research. I would also like to thank
Alan Hahn and Vinay Singhal for their assistance.

Finally, I would like to appreciate the emotional support provided to me by my
parents, Hyung Suk Lee, Golrokh Nossoni, Parama Sarkar, Swaroop Mannepalli, and

Praveen Desaraju during this time.

TABLE OF CONTENTS

CHAPTER 1 INTRODUCTION ........................................................................................ 1
1.1 Background ............................................................................................................... l
1.2 Research Objectives .................................................................................................. 7
- 1.3 Layout of the Thesis .................................................................................................. 7
CHAPTER 2 LITERATURE REVIEW ........................................................................... 10
2.1 Dowel bar Placement Methods ............................................................................... 10
2.1.1 Dowel Baskets ................................................................................................. 11
2.1.2 Dowel Bar Inserter ........................................................................................... 11
2.2 DOT Construction Practices ................................................................................... 15
2.3 Types of Misalignment ........................................................................................... 18
2.4 Construction Factors affecting Misalignment ......................................................... 19
Figure 2.3. Basic Types of Dowel Alignment Errors ............................................... 20
2.4.1 Cutting of Tie Wires ........................................................................................ 21
2.4.2 Wall of Concrete behind Paver ........................................................................ 22
2.5 Tolerance on Misalignment .................................................................................... 23
2.6 Effects of Misalignment .......................................................................................... 25
2.7 Experimental Studies on Dowel Misalignment ...................................................... 29
2.8 Field Measurement of Dowel bar Alignment ............... . .......................................... 34
2.8.1 Pachometer and Covermeter ............................................................................ 35
2.8.2 Ground Penetrating Radar ................................................................................ 36
2.8.3 MIT Scan-2 ...................................................................................... . ................ 3 8
2.9 Field Studies comparing Baskets vs DBI ................................................................ 42
CHAPTER 3 EXPERIMENTAL PLAN AND PROCEDURE ........................................ 49
3.1 Experimental Setup ................................................................................................. 49
3.2 Experimental Plan ................................................................................................... 53
3.3 Material Specifications ........................................................................................... 57
3.3.1 Steel Coupon Testing ....................................................................................... 58
3.3.2 Concrete Testing .............................................................................................. 60
3.4 Instrumentation ....................................................................................................... 62
3.5 Typical Casting-Testing Cycle ............................................................................... 66

3.6 Comparison of the Experimental Setup and Plan with previous studies ................ 71

CHAPTER 4 DISCUSSION OF RESULTS .................................................................... 73
4.1 Typical Bond Stress versus Joint Opening Behavior .............................................. 73
4.2 Overview of the Results .......................................................................................... 77

4.2.1 Horizontal Misalignment Tests ........................................................................ 77
4.2.2 Vertical Misalignment Tests ............................................................................ 83
4.2.3 Combined Misalignment Tests ........................................................................ 87

4.3 Comparisons of load vs joint behavior ................................................................... 91

4.3.1 One and Two Dowel tests ................................................................................ 92
4.3.1.1 Horizontal Misalignment .......................................................................... 93
4.3.1.2 Vertical Misalignment .............................................................................. 97

vi

4.3.1.3 Combined Misalignment ......................................................................... 102

4.3.2 Comparison across different number of dowels misaligned .......................... 107
4.3.2.1 Comparisons within same setup with different number of dowels
misaligned ........................................................................................................... 107
4.3.2.2 Comparison across different number of dowels setups .......................... 112

4.4 Load Zone Prediction .................................... 121
4.5 Effect on Misalignment on Pavement Performance ............................................. 123
CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS ................................... 130
5.1 Conclusions ........................................................................................................... 130
5.2 Recommendations for Future Research ................................................................ 136

vii

List of Tables

Table 2.1. DOT preferred dowel placement methods ....................................................... 15
Table 2.2. Sample Survey Form ....................................................................................... 16
Table 2.3. DOT Construction Specifications for Basket Assemblies ............................... 18
Table 2.4. Construction Factors affecting Misalignment .................................................. 19
Table 2.5. DOT Tolerance Specifications ......................................................................... 24
Table 2.6. Survey Results for DOT Preferences of Dowel Bar Tolerance Checks .......... 25
Table 2.7. Possible Effects of Dowel Misalignment on Pavement Performance ............. 29
Table 2.8. Summary of Field Study Results ..................................................................... 45
Table 3.1. Experimental Design Matrix ............................................................................ 54
Table 3.2. Mix Design of Concrete ................................................................................... 58
Table 3.3. Concrete Properties .......................................................................................... 60
Table 3.4. Statistical Parameters for Raw Strengths ......................................................... 62
Table 3.5. Instrumentation Location and Purpose ............................................................ 63
Table 4.1. Summary of Visual Observations in the Horizontal Misalignment tests ......... 78
Table 4.2. Summary of Visual Observations in the Vertical Misalignment tests ............. 84
Table 4. 3. Summary of Visual Observations in the Combined Misalignment tests ......... 87
Table 4. 4. Pair wise comparison of loads at 1A: 1n. joint opening (Horizontal
misalignment) ........................................................................................................... 93
Table 4. 5. Pair wise comparison of loads at % in. joint opening (Vertical Misalignment)
................................................................................................................................... 98
Table 4.6. Pairwise comparison of loads at 1A in. joint opening (Combined misalignment)
................................................................................................................................. 102
Table 4.7. Input Parameters for Sensitivity Analysis ......................... 125
Table A-1. Mix Design Specifications ............................................................................ 141
Table A-2. Description of Linear Variable Differential Transducer .............................. 142
Table A-3. Description of the Hydraulics (Actuators RC 156 and RC 256) .................. 142
Table B-1. Height Check Summary — One Dowel Tests ................................................ 152
Table B-2. Height Check Summary - Two Dowel Tests (North Dowel) ....................... 153
Table B—3. Height Check Summary - Two Dowel Tests (South Dowel) ....................... 154
Table 34. Height Check Summary - Three and Five Dowel Tests (Outer north dowel
through outer south dowel) ..................................................................................... 155
Table B-5. Horizontal Misalignment Summary — One Dowel Tests .............................. 156

Table B-6. Horizontal Misalignment Summary - Two Dowel Tests (North Dowel) 156
Table B-7. Horizontal Misalignment Summary - Two Dowel Tests (South Dowel) 157
Table B-8. Horizontal Misalignment Summary - Three and Five Dowel Tests (Outer

north dowel through outer south dowel) ................................................................. 158
Table B-9. Summary Table showing the various angles ................................................ 161
Table B—lO. Summary Table showing all angles for horizontal misalignment .............. 165

viii

List of Figures

Figure 2.1. Dowel Bar Installation sequence using basket assembly ............................... 12
Figure 2.2. Dowel Bar Insertion Sequence using Dowel Bar Inserter .............................. 13
Figure 2.3. Basic Types of Dowel Alignment Errors ....................................................... 20
Figure 2.4. Placement of Concrete while using Baskets ................................................... 23
Figure 2.5. Typical Pachometer ........................................................................................ 36
Figure 2.6. Typical Ground Penetrating Radar in use ...................................................... 37
Figure 2.7. Close up of MIT Scan—2 ................................................................................. 39
Figure 2.8. Typical Run of MIT Scan-2 in Field .............................................................. 39
Figure 2.9. Comparison of Baskets vs DBI ...................................................................... 44
Figure 3.1. Plan - Section view of the overall mold for 10in. slab ................................. 50
Figure 3.2. Experimental Setup and Components ............................................................ 51
Figure 3.3. Plan view of Uniform and Non-uniform Misalignment ................................. 56
Figure 3.4. Clockwise orientation of the dowel ................................................................ 57
Figure 3.5 . Standard V2 in. Round Tension Test specimen with 2 in. gage length (ASTM
E8-99) ....................................................................................................................... 59
Figure 3.6. Stages of Coupon Testing ............................................................................... 59
Figure 3.7. Typical Stress-Strain Curve for a coupon ...................................................... 60
Figure 3.8. Coupon Tests Strength Results ....................................................................... 61
Figure 3.9. Closeup of Instruments ................................................................................... 63
Figure 3.10. Instruments ................................................................................................... 64
Figure 3.11. Plan of Instrumentation Setup for 5- dowel test ........................................... 64
Figure 3.12. Instrumentation Setup for Typical 2- dowel setup with Instrumentation for
Vertical Response ....................... 65
Figure 3.13. 5-dowel Instrumentation Setup .............................. ; ...................................... 65
Figure 3.14. Typical Misalignment Check Setup (Surveying) ......................................... 67
Figure 3.15. Casting .......................................................................................................... 68
Figure 3.16. Fresh and Hardened Concrete Test Apparatus ............................................. 68
Figure 3.17. Concrete Hardened Properties Test Setup .................................................... 69
Figure 3.18. Cast Specimen .............................................................................................. 69
Figure 4.1. Typical Bond Stress versus Joint Opening curve ........................................... 73
Figure 4.2. Forces on a Misaligned dowel ........................................................................ 75
Figure 4.3. Typical Experimental Bond Stress versus Opening Behavior ....................... 76
Figure 4.4. Spalling in the 2H1U test ............................................................................... 79
Figure 4.5. Joint opening as a function of distance along the joint (2H 1U) ..................... 79
Figure 4.6. Cracking in 2H34NU ...................................................................................... 81
Figure 4.7. Load vs joint opening curve (2H34NU) ......................................................... 81
Figure 4.8. Crack pattern and Initial dimensions in 2H1NU ............................................ 82
Figure 4.9. Load vs joint opening (2H1NU) ..................................................................... 83
Figure 4.10. Vertical Uplift (2V1NU) .............................................................................. 85
Figure 4.11. Diagrammatic representation of Crack in 5V12NU ..................................... 86
Figure 4.12. Load vs joint opening curve (5V12N U) ....................................................... 86
Figure 4.13. Cracking in 2C34NU .................................................................................... 88

ix

Figure 4.14. Load vs joint opening curve (2C34NU) ....................................................... 89

Figure 4.15. Cracking in 2C1NU ...................................................................................... 90
Figure 4.16. Load vs joint opening curve (2C lNU) ......................................................... 90
Figure 4.17. Load vs joint opening curve (5C12N U) ....................................................... 91
Figure 4.18. Comparison of load vs joint opening curves for l-dowel Horizontal
Misalignment tests .................................................................................................... 94
Figure 4.19. Comparison of load vs joint opening curves for 2-dowel Non-uniform
Horizontal Misalignment tests .................................................................................. 95
Figure 4.20. Comparison of load vs joint opening curves for 2—dowel Uniform Horizontal
Misalignment tests .................................................................................................... 95
Figure 4.21. Comparison of load vs joint opening curves for 2-dowel Horizontal
Misalignment tests (One dowel misaligned) ............................................................. 96
Figure 4.22. Comparison of load vs joint opening curves for 2-dowel Horizontal
Misalignment tests (Non-uniform vs Uniform) ........................................................ 96
Figure 4.23. Comparison of loads at V: in. joint opening (Horizontal Misalignment) ...... 97
Figure 4.24. Comparison of loads at V2 in. joint opening (Horizontal Misalignment) ...... 97
Figure 4.25. Comparison of load vs joint opening curves for l-dowel Vertical
Misalignment tests .................................................................................................... 99
Figure 4.26. Comparison of load vs joint opening curves for 2-dowel Non-uniform
misalignment tests ..................................................................................................... 99
Figure 4.27. Comparison of load vs joint opening curves for 2—dowel Uniform Vertical
Misalignment tests .................................................................................................. 100
Figure 4.28. Comparison of load vs joint opening curves for 2-dowel Vertical ' '
Misalignment Tests (One dowel rrrisaligned) ......................................................... 100
Figure 4.29. Comparison of load vs joint opening curves for 2-dowel Vertical
Misalignment Tests (N on-uniform vs Uniform) .............................................. ' ....... 101
Figure 4.30. Comparison of loads at V: in. joint opening (Vertical Misalignment) ........ 101
Figure 4.31. Comparison of loads at V2 in. joint opening (Vertical Misalignment) ........ 102
Figure 4.32. Comparison of load vs joint opening curves for l-dowel Combined
Misalignment tests .................................................................................................. 103
Figure 4.33. Comparison of load vs joint opening curves for 2-dowel Non-uniform
Combined Misalignment tests ................................................................................. 104
Figure 4.34. Comparison of load vs joint opening curves for 2-dowe1 Uniform Combined
Misalignment tests .................................................................................................. 104
Figure 4.35. Comparison of load vs joint opening curves for 2—dowe1 Combined
Misalignment (One dowel misaligned) ................................................................... 105
Figure 4.36. Comparison of load vs joint opening curves for 2-dowel Combined
Misalignment (Non-uniform vs Uniform) .............................................................. 105

Figure 4.37. Comparison of loads at M1 in. joint opening (Combined Misalignment) 106
Figure 4.38. Comparison of loads at V2 in. joint opening (Combined Misalignment) 106
Figure 4.39. Comparison of load vs joint opening curves of 2-dowel V2 in. Horizontal

Misalignment (2H12NU, 2H12U, and 2H12AM) .................................................. 107
Figure 4.40. Comparison of load vs joint opening curves of 2-dowel V2 in. Vertical

Misalignment (2V12NU, 2V12U, and 2V12AM) .................................................. 108
Figure 4.41. Comparison of load vs joint opening curves of 2-dowel V2 in. Combined

Misalignment (2C12NU, 2C12U, and 2C12AM) ................................................... 108

Figure 4.42. Comparison of load vs joint Opening curves of 2-dowel 3% in. Horizontal

Misalignment (2H34N U and 2H34AM) ................................................................. 109
Figure 4.43. Comparison of load vs joint opening curves of 2-dowel —% in. Vertical

Misalignment (2V34NU and 2V34AM) ................................................................. 109
Figure 4.44. Comparison of load vs joint opening curves of 2-dowel 3%: in. Combined

Misalignment (2C34NU and 2C34AM) ................................................................. 110

Figure 4.45. Comparison of load vs joint opening curves of 5H12NU and 5H12AM 111
Figure 4.46. Comparison of load vs joint opening curves of 5V12NU and 5V12AM 111
Figure 4.47. Comparison of load vs joint opening curves of 5C12NU and 5C12AM 112
Figure 4.48. Comparison of load vs joint opening curves of 1H12 and 2-dowel V2 in.

Horizontal Misalignment (2H12NU, 2H12U) ........................................................ 113
Figure 4.49. Comparison of load vs joint opening curves of 1H1 and 2-dowel l in.
Horizontal Misalignment (2H1NU, 2H1U) ............................................................ 113

Figure 4.50. Comparison of load vs joint opening curves of 3H12NU and 5H12NU.... 114
Figure 4.51. Comparison of load vs joint opening curves of 1V12 and 2-dowel V2 in.

Vertical Misalignment (2V12NU, 2V12U) ............................................................ 115
Figure 4.52. Comparison of load vs joint opening curves of 1V1 and 2-dowel 1 in.
Vertical Misalignment (2V1NU, 2V1U) ................................................................ 115

Figure 4.53. Comparison of load vs joint opening curves of 3V12NU and 5V12NU.... 116
Figure 4.54. Comparison of load vs joint opening curves of 1C12 and 2-dowel V2 in.

Combined Misalignment (2C12NU, 2C12U) ......................................................... 116
Figure 4.55. Comparison of load vs joint opening curves of 1C1 and 2-dowel l in.
Combined Misalignment (2C1NU, 2C1U) ............................................................. 117
Figure 4.56. Comparison of load vs joint opening curves of 3C12NU and 5C12NU 117
Figure 4.57. Comparison of loads at V: in. across different tests of misalignment ......... 119
Figure 4.58. Comparison of loads at V2 in. across different tests of misalignment ......... 120
Figure 4.59. Comparison of load vs joint opening curves (3H12NU, 3V12NU, and
3C12NU) ................................................................................................................. 121
Figure 4.60. Comparison of load vs joint opening curves (5H12NU, 5V12NU, and
5C12NU) ................................................................................................................. 121
Figure 4.61. Regions for 1 and 2-dowel systems ............................................................ 122
Figure 4.62. Regions for 3 and 5-dowel systems ............................................................ 123
Figure 4.63. Dowel Looseness ........................................................................................ 125
Figure 4.64. Screenshot of the Traffic loading location in EverFE ................................ 126
Figure 4.65. Effect of dowel looseness on LTE near Wheel 1 of tandem axle .............. 128
Figure 4.66. Effect of dowel looseness on LTE near Wheel 2 of tandem axle .............. 129
Figure A—l. Cross section of the small mold for pullout test .......................................... 139
Figure A-2. Plan of the small mold with outside dimensions ......................................... 139
Figure A-3. Typical sectional view of the big mold ....................................................... 140
Figure A-4. Typical plan view of mold with 5 dowel bars ............................................. 140
Figure A-S. Calibration of 1.5 in. slider-1 ...................................................................... 144
Figure A-6. Calibration of 1.5 in. slider-2 ...................................................................... 144
Figure A-7. Calibration of 1.0 in. slider-l ...................................................................... 144
Figure A-8. Calibration of 1.0 in. slider-2 ...................................................................... 144
Figure A-9. Calibration of 1.0 in. slider-3 ...................................................................... 145

xi

Figure A-10. Calibration of 1.0 in. slider-4 .................................................................... 145

Figure A-l 1. Calibration of LVDT-l .............................................................................. 145
Figure A-12. Calibration of LVDT-2 .............................................................................. 145
Figure A-l3. Calibration of LVDT-3 .............................................................................. 146
Figure A-l4. Calibration of RC156 Actuator l .............................................................. 146
Figure A-15. Calibration of RC256 Actuator l .............................................................. 147
Figure A-16. Calibration of RC156 Actuator 2 .............................................................. 147
Figure A-17. Calibration of RC256 Actuator 2 .............................................................. 148
Figure A-18. Data Flow .................................................................................................. 149
Figure B-l. Horizontal plane showing angles measured from center of base plate to
misalignment ........................................................................................................... 160
Figure B-2. Vertical plane in the center of the plate ....................................................... 160
Figure B-3. Horizontal plane showing angles measured from center of base plate to
misalignment ........................................................................................................... 163
Figure B-4. Angles in the vertical plane, after forcing the misalignment of H .............. 164
Figure B-5. Showing the final computation of the horizontal misalignment ................. 164
Figure C-l. Bond Stress vs Joint Opening curve ............................................................ 168
Figure C-2. Bond stress vs joint opening curves for 1 d-owel Horizontal Misalignment
cases ........................................................................................................................ 169
Figure C- 3. Bond stress vs joint opening curves for 2- dowel Non-uniform Horizontal
Misalignment cases ................................................................................................. 169
Figure C-4. Bond stress vs joint opening curves for 2-dowel Uniform Horizontal
Misalignment cases ................................................................................................. 170
Figure 05. Bond stress vs joint opening curves for 2-dowel Horizontal Misalignment
cases (One dowel misaligned) ................................................................................ 170
Figure C-6. Bond stress vs joint opening curves for 3 and S-dowel Horizontal
Misalignment cases ................................................................................................. 171
Figure C-7. Bond stress vs joint opening curves for l-dowel Vertical Misalignment cases
................................................................................................................................. 171
Figure C-8. Bond stress vs joint opening curves for 2-dowel Non-uniform Vertical
Misalignment cases ................................................................................................. 172
Figure 09. Bond stress vs joint opening curves for 2-dowel Uniform Vertical
Misalignment cases ................................................................................................. 172
Figure C-lO. Bond stress vs joint opening curves for 2-dowel Vertical Misalignment
cases (One dowel misaligned) ................................................................................ 173
Figure C-1 1. Bond stress vs joint opening curves for 3 and 5-dowel Vertical
Misalignment cases ................................................................................................. 173
Figure C- 12. Bond stress vs joint opening curves for l-dowel Combined Misalignment
cases ........................................................................................................................ 174
Figure C-13. Bond stress vs joint opening curves for 2-dowel Non-uniform Combined
Misalignment cases ................................................................................................. 174
Figure C-l4. Bond stress vs joint opening curves for 2-dowel Uniform Combined
Misalignment cases ................................................................................................. 175

xii

Figure C-15. Bond stress vs joint opening curves for 2-dowel Combined Misalignment

cases (One dowel misaligned) ................................................................................ 175
Figure C-l6. Bond stress vs joint opening curves for 3 and 5-dowel Combined

Misalignment cases ................................................................................................. 176
Figure D-l. Spalling in the Horizontal Misalignment tests ............................................ 178
Figure D-2. Joint opening as a function of distance along the joint (5H12NU) ............. 179
Figure D-3. Spalling in the Vertical Misalignment tests ................................................ 180
Figure D-4. Joint opening as a function of distance along the joint (2V1NU) ............... 181
Figure D-5. Spalling in the Combined Misalignment tests ............................................. 182
Figure D—6. Joint opening as a function of distance along the joint (2C1U) .................. 183

xiii

CHAPTER 1 INTRODUCTION
1.1 BACKGROUND
Jointed plain concrete pavements (JPCP) are commonly constructed with contraction
joints to accommodate slab movements due to temperature and moisture variations.
Because such discontinuities constitute intrinsic planes of weakness, it is also desirable to
supply some means of transferring load across the discontinuity (Ioannides, 1990).

Aggregate interlock alone does not provide enough load transfer for good long-
terrn performance for most highway pavements due to heavy truck traffic. The American
Concrete Paving Association (ACPA) recommends that dowel bars should be used to
provide added mechanical load transfer where truck traffic exceeds 120 per day or
accumulated design traffic exceeds 4-5 million ESALs. Typically, this traffic will require
at least an 8 in. thick slab. For most highway applications, dowels are recommended, for
8 in. slabs or greater (ACPA 2004).

Dowel bars are smooth round bars placed across joints to transfer loads without
restricting horizontal joint movement. They also assist in maintaining the horizontal and
vertical alignment of slabs. Since dowels span the joint, daily and seasonal joint opening
does not affect load transfer across doweled joints as much as it does non-doweled joints.

Dowel bars have been used across transverse joints in PCC pavements at least
since 1917 (Teller and Cashell 1958). Justification for their use is provided by the fact
that edge-loading stresses and deflections are higher than the corresponding interior-
loading responses. Thus, if a portion of the applied edge load is transferred to an adjacent
slab, resulting stresses and deflections can be significantly reduced. Distresses commonly

exhibited by PCC pavements are attributed to high levels of stress or deflection (Snyder

1989). For this reason, distresses tend to appear most often in the vicinity of slab edges
and comers, the theoretically predicted locations of maximum system responses.
Deflection related distresses include pumping (Packard and Tayabji 1983) and faulting
(Darter et.al. 1985). High slab-bending stresses, on the other hand, may lead to premature
cracking. Excessive dowel-concrete bearing stresses may also result in Spalling and
faulting (Ioannides 1990).

A non-doweled joint provides load transfer through shear forces developed at the
rough joint interface due to aggregate interlock provided by the shape, size, and hardness
of the coarse aggregate. As a result, joint opening highly affects load transfer efficiency
(LTE) based on aggregate interlocking. That is why LTE of joints with aggregate
interlock is highly dependent on temperature: it is only high during warm weather when
the joint is tightly closed (small joint opening). In contrast, the load transfer mechanism
of dowel bars does not rely on warm weather or closed joints to transfer load across the
joint. As a result, dowels are effective throughout the year (Rufino et. a1. 2005).
Khazanovich and Gotlif (2003) performed a comprehensive analysis of LTE data from
the Long-Term Pavement Performance (LTPP) program. They showed that doweled
joints have much higher LTE than non-doweled joints.

Several field performance studies from the 1970’s through 1990’s, including
Highway Research Board (1962), Darter eta]. (1985), Smith et.al. (1990), Darter (1977),
FHWA (1989), and Khazanovich eta]. (1998), found that slabs without dowels at the
joints showed substantially more faulting than doweled ones at the same location. In
addition, some of these studies also concluded that dowels prevent comer breaks or

diagonal cracks. Darter (1977) considered faulting “the most serious distress” because it

had a significant effect on ride quality when it developed to a significant level requiring
maintenance. The importance of dowel bars was further underscored at the AASHO Road
test. An analysis was performed using NCI-IRP I-37A Mechanistic-empirical pavement
design guide (ARA 2004), assuming the cross-sections, traffic, climate, and materials
information from the AASHO Road Test. The analysis revealed that non-doweled joints
could have had up to seven times more faulting than doweled ones. The most prominent
advantage of using dowels is reduction in faulting, and consequently, maintaining
smoothness (Rufino eta]. 2005).

Design factors to be considered include the diameter, embedment length, and
spacing of dowels required to limit and control the magnitude of stresses developing in
each bar and in the surrounding concrete matrix. From a construction viewpoint, it is
important to install the dowels properly, i.e., in a horizontal plane, parallel to the
pavement centerline. If they deviate from the desired position, they are said to be
misaligned. Misalignment may result from either or both (a) misplacement (i.e. initially
placing the dowels in an incorrect position), or (b) displacement (i.e. movement during
the pavement operation). The amount of misalignment that can be tolerated is a matter of
some disagreement in the literature (Ross 1989).

Misalignment during dowel installation restrains the ability of the bars to slip
freely in the concrete, and may result in transverse slab cracking, comer breaks, and joint
Spalling around the dowel at the concrete face (Tayabji 1986). It is equally essential to
provide dowel corrosion protection for the duration of the design life of the pavement

(Schierer 1985).

Dowel bars lower edge and comer deflection and stress in the concrete slab and reduce
the potential for faulting, pumping, and comer breaks. Performance evaluation of in-
service concrete highway pavements show the use of dowels effectively reduces faulting.
Dowels also increase pavement service life by reducing deflections and stresses in the
slab by effectively transferring the load across the joint. Elongation of the dowel hole
reduces load transfer capabilities (ACPA 2004).

In the current DOT practices of dowel and joint design (AASHTO 1993)., the
design of dowels has been mostly based on experience, one rule of thumb being the
diameter of the dowel should be equal to ‘A; m of the slab thickness. FHW A (1990) also
recommends this guideline but limits the dowel diameter to a minimum of 1% in. for
highways. ACPA (1991) recommends, for highway pavements less than 10 in. thick, 1%
in. diameter dowels and for pavements 10 in. thick or greater, 1V2 in. dowels. A minimum
dowel diameter of 1% to 1V2 in. is needed to control faulting on highway pavements.
According to an ACPA survey (ACPA 2004a), the average dowel diameter used by state
highway agencies is 1% in. Loads on smaller dowels induce higher bearing stresses and
cause the concrete matrix around the dowel to deteriorate or elongate. Design methods
are available that check the allowable bearing stress between dowel and concrete (Huang
1993). Because concrete is weaker than steel, the size and spacing of dowels required are
governed by the bearing stress between the dowel and concrete, the allowable stress

being given by equation (1):
4—d .
fb :(ijc ...(l)

in which fb is the allowable bearing stress in psi, (1 is the dowel diameter in inches and fc’

is the ultimate compressive strength of concrete in psi. The maximum bearing stress is

determined theoretically by assuming the dowel to be a beam and the concrete to be a
Winkler foundation based on the original solution by Timoshenko and Friberg (1940)
(equation (2)).

=KB(2+flZI

3 ...(2)
4,6 Edld

 

0b = KY0

where maximum deformation of concrete under the dowel yo at the face of the joint is
expressed as given in equation (3):

P. (2 +flz)
0 = —-——3—— ...(3)
4 ,6 Ed! d
in which P. is the load on the dowel, z is the joint width, Ed is the Young’s modulus of the

dowel, Id is the moment of inertia of the dowel, and B is the relative stiffness of a dowel

embedded in concrete given by equation (4):

1d =_l_mz4 fl=4 Kd ...(4)
64 4Edld

in which K is the modulus of dowel support, which ranges from 300 to 1500 kci and d is

 

the diameter of the dowel. By limiting the bearing stress, the amount of faulting can be
reduced to the allowable limit.

According to the formulation presented by Tabatabaie et. al. (1979), the ILLI-
SLAB model considers the dowel bar as a thick-beam element whose shear stiffness (i.e.,
resistance to deformation in the vertical direction) is 12 C. This term expresses the shear
force in the dowel per unit vertical deformation of the dowel. The support provided by
the PCC matrix is modeled as a single spring that act as at the joint face and whose
stiffness is dowel-concrete interaction (DCI) parameter. This is the shear force transferred

by the dowel per unit deflection of the dowel with respect to the concrete matrix.

EdId

—— ...(5)
w3(1+¢)

C:

where (I) is the parameter accounting for through-the-thickness shear deformations arising

1215,11,,

2 ...(6)
GdAzw

in deep beams and is expressed as (1) =

where (1) = joint opening, A2 = cross-sectional area of dowel effective in shear = 0.9Ad

for solid round bars. By reference to Friberg’s analysis (1940), the expression for DCI is

as given in equation (7):

Pi _ 4.531321%

A..- (mm) ""7)

DCI =

 

where P, is the load transferred by any particular dowel, Ad, is the deflection of any given
dowel with respect to the concrete, and other parameters are as defined before.

The value of the critical dowel-concrete bearing stress, 0',” is calculated using equation

(8) (Ioannides 1990):

K

=—P 8
DCI c ()

”be

where Pc is the portion of the transferred load carried by the critical (design) dowel

As seen from the design methodology, the key parameter affecting the bearing
stress is the dowel diameter. In the NCHRP I-37A design methodology (mechanistic-
empirical design guide), bearing stresses have been linked to cumulative damage of the
joint. Dowel misalignment contributes to the magnitude of bearing stresses at the dowel-
concrete interface but to date, there is no structural model that accounts for the change in
bearing stress due to dowel misalignment, hence necessitating the need to study the effect

of dowel misalignment on pullout and joint opening behavior.

The research presented in this thesis is based on the laboratory phase of the MDOT
sponsored project titled ‘A laboratory evaluation of alignment tolerances for dowel bars
and their effect on joint opening behavior’.
1.2 RESEARCH OBJECTIVES
The objectives of this thesis are:
0 To investigate experimentally the fundamental pullout behavior of misaligned
dowel bars in plain concrete pavements under thermal expansion.
0 To investigate experimentally the gap-opening behavior of pavement joints with
misaligned dowel bars under thermal expansion.
Its hypothesized that the misalignment type and magnitude would affect the following:
0 Initial slip/debonding stress (bond stress at initial slip) (defined in chapter 4)
0 Type of joint and slab distresses observed
0 Post-slip behavior
1.3 LAYOUT OF THE THESIS
This thesis — ‘Experimental characterization of the pullout and gap opening behavior of
misaligned dowel bars under thermal expansion’ is outlined as follows.

Chapter 2 contains a synopsis of the technical papers and reports on several topics
related with this study. The topics of this chapter can be broadly classified into four
groups: construction practices focusing on dowel bar placement methods, definition,
factors and effects of misalignment, studies on misalignment - experimental as well as
field, modern technologies of detection of dowel bar position in the field. Dowel baskets
and dowel bar inserters are the two methods of dowel bar placement focused at during

various stages of the chapter.

Chapter 3 provides a detailed description of the experimental plan and procedure. An
experimental matrix, at the start of the chapter, addresses the different experimental
variables — number of dowel bars and number misaligned, misalignment type, orientation,
and magnitude, and sets out the plan for the testing. The various components of the
experimental setup are described. The concrete mix and dowel bars are as per the
Michigan DOT specifications. A detailed description of the material specifications and
quality control tests of steel and concrete is provided. Details of the various instruments
used in the tests, along with their calibration are presented thereafter. The instruments
described include linear variable differential transducers (LVDT), spring return linear
motion sensors (slider), pressure transducers, and hydraulic actuators. The chapter ends
with a description of the events in a typical casting-testing cycle.

Chapter 4 discusses the results obtained during the different tests carried out, as
per the experimental plan explained in chapter 3. A plot illustrating the typical bond
stress vs joint opening behavior explains the concepts of initial slip/debonding stress and
transverse contact. Distresses observed for different combinations of misalignment
include Spalling around the dowel-concrete interface, non-uniform joint opening,
cracking, and vertical uplift. This chapter explains all observations — distresses and load
vs joint opening in three subsections based on the misalignment types (horizontal,
vertical, and combined). Each subsection starts with an overview of the distresses and
load vs joint behavior, and then these details are substantiated with illustrations, provided
for clarity and conciseness of data representation. The nomenclature (specimen ID)

explained in chapter 3 is used while describing specifics of test results. Prediction of load

vs joint opening behavior, for a given number of bars and misalignment magnitude within
the gamut of the test matrix, is attempted.
Chapter 5 presents the conclusions of the research. This chapter also suggests

scope of future research.

CHAPTER 2 LITERATURE REVIEW
This chapter summarizes the findings of the literature review conducted as a part of the
research in five sections — dowel bar placement methods in field, construction practices
of some state highway agencies, types of misalignment, measurement of dowel bar
misalignment in the field, effects of misalignment, and field and experimental studies on
the effects of dowel bar misalignment. ‘
2.1 DOWEL BAR PLACEMENT METHODS
This section deals with practices in plain jointed concrete pavement construction, with a
focus on dowel bar placement methods and possible causes of misalignment.

Pavements are constructed using fixed form paving and slipform paving. Fixed
form paving is used generally for small jobs, complicated geometry pavements, or
variable width pavements, while slipform paving is used for larger jobs that require high
production rates. There are two main methods of dowel bar placement in the field: dowel
basket assembly and dowel bar inserter (DBI). Slipform paving Carr accommodate both
the methods of dowel bar placement while fixed form paving can accommodate dowel
baskets only.

Due to the increased cost of approximately $ 30,000/ interstate mile for basket-
assembly compared with that of the implanting method, investigations are underway to
check dowel alignment in existing pavements using both implanting schemes (Burati
eta]. 1983). DBIs have been used in Europe for a quarter century. In 1996, the Federal
Highway Administration (FHWA) officially encouraged the use of DBIs as an acceptable

alternate means of dowel bar placement in jointed concrete construction (Donahue 2003).

10

2.1.1 Dowel Baskets

Dowel baskets are simple truss structures used to hold dowel bars at the appropriate
height before PCC placement. Typically, dowel baskets span an entire lane width and are
fabricated from thick gauge wire. They are left in place after the PCC is placed but do
not contribute to the pavement structure.

When using dowel baskets, the dowels must be aligned and the dowel basket
firmly anchored to the base course. The FHWA recommends that the dowel baskets be
secured with steel stakes with a minimum diameter of 0.3 in. embedded at least 4 in. in
stabilized bases, 6 in. in treated permeable bases and 10 in. in untreated bases or
subgrade. Further, a minimum of 8 stakes per basket is recommended.

Figure 2.1 (a-d) illustrate a typical sequence of installation of dowel bars using
dowel baskets at a construction site in Michigan. The dowels come precoated with
lubricant from the manufacturer and are welded on alternate sides. Three nails each 12 in.
long are driven into either sides of a basket covering one lane (Figure 2.1 (a)). Then the
shipping/tie wires are cut (Figure 2.1 (b)). Figure 2.1 (c) illustrates the misalignment
check that is performed to ensure the basket is leveled. Figure 2.1 ((1) shows an installed
basket.

2.1.2 Dowel Bar Inserter

The dowel bar inserter is a device which mounts behind the conforming pan of the
slipform paver. The DBI can accommodate different dowel spacings, dowel depths,
skewed or square contraction joints, and both crowned and flat pavements. Figure 2.2 (a-
e) illustrate a typical dowel bar inserter and the sequence of operations in dowel bar

insertion.

ll

r «...Fdfisirq

 

(b) Cutting of tie wires

(a) Driving the nails into base course

 

((1) Installed basket

(c) Misalignment check

basket assembly

ing

on sequence [IS

Figure 2.1. Dowel Bar Installati

12

 

(d) Side and End Prongs (e) Insertion into Plastic Concrete

Figure 2.2. Dowel Bar Insertion Sequence using Dowel Bar Inserter‘
‘Source: Gomaco Inc. Website www.gomaco.com

The DBI is supported by rollers and is suspended above the inserter pan. The pan

confines the disturbance to the concrete surface where dowels are introduced into the

fresh concrete. As the slipform advances, the DBI and pan are towed by hydraulic
:ylinders. Vertical guides connect the DB1 and the pan without putting weight on the pan.
Thus, when the pan glides as it is towed along the concrete surface, the DBI moves with
lt.

When the desired joint location is reached, the automatic insertion cycle is
activated. A pan mounted dowel distributor then shifts allowing the dowels to drop from
:he magazine through the pan onto the fresh concrete. The vibration-isolated “inserting
fork assemblies” then vibrate the bars until they have reached their proper depth in the
:oncrete. Mounted on rubber-isolated beams, the vibrating forks reconsolidate the
:oncrete as the forks are pulled out of the slab. Once the forks have cleared the concrete,
:he vibration is stopped and the roller-supported DB1 and the pan are retracted. The
)scillating correcting beam, which is required when using the DBI, then refinishes the
:oncrete slab to a smooth concrete surface finish.

Because the dowels are being vibrated intensely and are inserted down to their
:‘inal resting place under the weight of only the dowel inserter assembly and light
rydraulic pressure, the concrete has enough time to flow around the bar rather than being

ii splaced.

l4

2.2 DOT CONSTRUCTION PRACTICES
An email survey was conducted amongst different state Departments of Transportation
(DOTS) in August 2004 to find out key features in pavement design and dowel bar
installation, with focus on:

o Dowel bar dimensions

0 Methods used for dowel bar placement and preferred method, if any

0 Tolerance in dowel bar placement

0 Quality control/misalignment check method adopted during construction

Initial bond stress specifications

The responses of the DOTS are summarized in Table 2.1 and a sample survey form is
shown in Table 2.2. As seen from Table 2.1, there is no consensus on the preferred
method of dowel bar installation. However, on an average, the states seem to prefer a

higher % of dowels installed using dowel baskets.

 

 

 

 

 

 

 

 

 

 

 

Table 2.1. DOT referred dowel placement methods"
DOT Basket DBI % DBI % Basket
Alabama Yes No 0 100
Colorado Yes I Yes 75 25
Florida Yes No 0 100
Nevada Yes Yes 25 75
North Carolina Yes Yes 25 75
Ohio Yes Yes 25 75
Pennsylvania Yes Yes 75 25
Virginia Yes Yes 25 75
Washington Yes Yes 25 75
Wisconsin Yes Yes 50 50

 

 

 

 

 

 

 

*Based on survey conducted in August 2004

15

Table 2.2. Sample Survey Form

 

Use this form to participate in a survey of Dowel Bar Installation
Practices currently used by State Highway Agencies in the United States.

 

Participant Details

 

 

Name

 

Title

 

 

Organization
Phone Number

 

Fax Number

 

Email Address

 

Pavement Design Practice

 

What is the typical range of transverse joint
spacing for each of the pavement type?

 

What is the typical range of PCC slab
thicknesses the network under DOT

jurisdiction?

 

What is the typical range of base
thicknesses in the network under DOT
jurisdiction?

 

What % of the network in your state is JPCP
and JRCP?

 

What are the typical diameters for dowel bars
used across transverse joints?

 

What is the typical length of the dowel bar?

 

What is the typical spacing between dowel
bars?

 

What are the typical paving mixture
requirements?

Target compressive strength:
Target flexural strength:

 

Dowel Bar Installation Practice

 

How are dowel bars placed along the
contraction joints?

DDowelBarlnserter [:laoth
DBasketAssemines

 

What are the allowable dowel alignment
tolerances for

Vertical skew:
Horizontal skew:
Horizontal translation:
Vertical translation:
Longitudinal translation:

 

How does your agency ensure specification
compliance with respect to dowel alignment
tolerances?

 

Have any studies on the effect of misalignment
on joint performance been carried out by your
DOT? If yes, please provide the reference

 

 

 

General Comments

 

16

 

As a part of the literature review, the detailed specifications of dowel bar placement of
five state DOTs including Michigan were reviewed and the findings are summarized in
Table 2.3.

As seen from Table 2.3, in general when using baskets, all DOTS (Michigan,
Ohio, Iowa, California, and Illinois) recommend use of stakes or pins, or bearing plates to
anchor the assembly. A minimum of 1 ft long nails are required to be driven into the
unstabilized bases/subbases. In case of stabilized bases/subbases, concrete anchors or
bearing plates are used.

The abovementioned DOTs require the following to be adopted while using
dowel bar inserters:

0 Concrete has to be placed and consolidated full-depth before insertion of the
dowel bars.

0 A light coating of oil or bond-breaking material has to be applied to the bars
before loading the bars in the dowel magazine.

0 The bars have to be inserted into the plastic concrete in front of the finishing
beam or screed and it has to be ensured that there are no voids around the dowel
bars.

0 Concrete has to be reworked and refinished to ensure that there is no evidence on
the surface of the completed pavement that there has been any insertion

performed.

17

Table 2.3. DOT Construction Specifications for Basket Assemblies

 

State

Cutting of

Tie Wires

Basket Installation Details

 

MI
(2004)

Yes

0 Stake shall engage bottom longitudinal spacer wire.

After staking, the bottom longitudinal spacer wire shall contact the base
material (unless non-penetrable) along its entire length.

For non-penetrable bases, permanent concrete anchors and stacking clips
shall be used to secure assembly to base.

 

OH
(2005)

Yes

At least eight Vz-in. diameter steel pins a minimum of 18 in. long are
driven at an angle to brace the assembly.

2 of these pins are driven opposite each other at each end of the
assembly, and the remaining pins are driven in staggered positions on
each side of the assembly.

 

IA
(2005)

Option of

Contractor

0 Securely staked or fastened to the base to line and grade.
0 Anchor pins should be 1 ft long.

 

CA
(2002)

Yes

A minimum of 8 alternating, equally spaced, concrete fasteners with
clips are used to anchor each 3.6 m assembly (4 per lower runner wire).
At least 10 concrete fasteners are used for assembly sections greater than
3.6 m and less than or equal to 4.9 m.

 

(2002)

 

 

Yes

 

Assembly is provided with two continuous bearing plates made of not
less than 2 in. width and not less than 0.04 in. thick sheet steel.

The bearing plates are attached by welding to the subgrade members or
by suitable clips and are punched to receive the protruding ends of the
upright supports and stakes.

The stakes are driven parallel to and next to the upright supports.
Bearing plates are not required on stabilized subbases.

At least ten nails shall be used for each assembly.

Bearing plates are punched to receive the nails.

For soil or granular subbase, metal stakes shall be used instead of nails,
and shall penetrate the subbase at least 12 in.

 

2.3 TYPES OF MISALIGNNIENT

The basic types of dowel alignment errors that can occur in concrete pavements are

horizontal translation, longitudinal translation, vertical translation, horizontal skew, and

vertical skew. A combination of the above could also occur. Figure 2.3 shows a typical

diagrammatic representation of the cross section and plan views of these errors. A shaded

bar denotes the original position of the bar and the dotted view is the position after

misalignment. Also shown in the table is a dowel bar with no misalignment.

18

 

2.4 CONSTRUCTION FACTORS AFFECTING MISALIGNMENT
With either method, care and attention to many details are required to achieve proper
dowel bar alignment (Yu 2005). For dowel baskets, the most critical factor appears to be
the manner in which the baskets are secured on the subbase or base prior to paving. If the
baskets are not adequately pinned down, the baskets may be shoved, rotated, or pulled
apart during paving, resulting in extreme dowel bar misalignments. The baskets may also
get bent during handling or during concrete placement by being walked on. For DBI
construction, the critical factors are the proper adjustment of the DB1 and PCC mix
design. When using a DBI, mix optimization is extremely important to ensure the dowel
bars do not become displaced after insertion. The PCC mix for DB1 construction must be
stable enough to hold the bars in place without displacing them during paving.
Construction factors that may affect misalignment as identified by Tayabji (1986)
have been summarized in Table 2.4. The following paragraphs discuss the possible effect
of some construction factors when using dowel baskets (leaving the tie wires intact and
the wall of concrete behind the basket) on dowel misalignment, if any.

Table 2.4. Construction Factors affecting Misalignment*

 

 

 

 

 

Dowel Baskets Dowel Bar Inserter
o Basket rigidity o Implanting machine operation
0 Quality control during basket fabrication 0 Strike-off after dowel placement
0 Care during basket transportation and o Consolidation (vibration) after dowel placement
placement 0 Location of saw-cut over implanted dowels
o Fastening of basket to subbase . Field inspection during construction
0 Location of saw-cut over basket
0 Paving operation
0 Field inspection during construction

 

*Ref: Tayabji (1986)

19

 

Initial Sawcut (Sealant 101m

L
“—fl
Concrete ih/‘im‘q to h/3
¢l .25" ¢1.25"E

W: h “—11.14: .__ “:1

 

 

 

 

 

 

 

 

(a) No Misalignment*

U ¢ h/2 [I

¢1.25't:--—-_ ............ ; ............. fl
Iijfi?*_—9n_>l h

(0) Vertical Translation*

 

 

 

 

 

 

 

 

 

 

 

 

 

A
\

 

((1) Vertical Skew*

Joint
\

 

 

 

 

(1) Horizontal Translation“

 

 

 

 

 

(g) Horizontal Skew

* Cross-sectional View
** Plan View

Figure 2.3. Basic Types of Dowel Alignment Errors

20

2.4.1 Cutting of Tie Wires

Most of the DOTS require the tie/spacer wires of the dowel basket assembly to be cut
before concrete placement (substantiated through Table 2.3). Theoretical investigation
(ACPA 2005) has indicated that leaving the tie wires intact has benefits and these have
been described in the following paragraphs.

Leaving tie wires intact will strengthen the dowel basket, making it more resistant
to movement and deflection while paving. This results in a smoother pavement, as well as
dowels that are better aligned. The strengthening offered by uncut wires is more critical
for taller dowel baskets used in thick pavements, such as heavy-use pavements. Dowel
baskets that have the tie wires out are more susceptible to spring-back problems, whereby
the basket is compressed when the paver passes above it, then springs back up after the
pressure is gone, resulting in a bump in the pavement.

Given the choice to cut the tie wires for purposes of location verification, or to
leave the tie wires to enhance pavement quality, ACPA recommends leaving the wires
intact. Since ACPA last conducted its review of state highway agencies practices in 1999,
at least 3 states (Iowa, Washington, and Wisconsin) removed the requirement for cutting
the tie wires (or spacer wires) in the dowel basket assemblies prior to paving. But around
20 state agencies still require concrete paving contractors to cut the tie wires prior to
placing concrete. The intent of this requirement is to eliminate or reduce the apparent
potential of the steel wires to lock the joint, or for the wires to cause micro-cracking in
the early ages of the concrete. The underlying belief is that the three to five small-

diameter wires, when crossing the joint, will restrict the shrinkage of the early-age

21

concrete. In addition, the wires could reinforce and prevent movement of the transverse
joint, and/or cause the concrete to crack.

These requirements are not well founded for a number of reasons. First, there are
always stresses that build up in the concrete pavement due to early-age concrete
shrinkage and temperature contraction. These are the same stresses that cause transverse
saw cuts in jointed pavement to become working joints. The stresses have to build up to
the point where they overcome the concrete strength, and then further build to overcome
other restraining forces for the joints to open up. The restraining forces include friction
provided by the subbase, and the amount of bonding between the concrete and the dowels
themselves. Once these friction and restraint forces have been overcome, the stress would
be transferred entirely to the tie wires.

For this mechanism to cause the concrete to crack, the tie wires must impart Stress
back to the concrete, and the total stress must be greater than the concrete strength at that
point in time to cause a crack. But an analysis of the mechanics shows that the tie wires
will fail one of two ways before they can cause damage to the concrete or lock the joint:

0 The wires themselves will yield, or

o The welds holding the wires to the basket will fail.
2.4.2 Wall of Concrete behind Paver
Whether the wall of concrete ahead of the paver would create an overturning moment
near the dowel baskets is a construction issue of key interest. Field visits to a mainline
pavement construction site in Michigan and interaction with site inspectors eliminated
this hypothesis. When using dowel baskets, the spreader places a significant amount of

concrete on the dowel basket keeping it in place as shown in and when the paver follows

22

the spreader it is ensured that consolidation is in the vertical direction and that the auger
does not push or cause a moment on the basket in the horizontal direction. Also, since six
pins are used on each basket, each driven at least a foot into the base, moment, if any, is
resisted because of the firmness of the basket. Since the slump of concrete is also very
low (about 2 in.), there is no flow induced when the spreader or the paver proceed

towards the baskets.

 

Figure 2.4. Placement of Concrete while using Baskets

2.5 TOLERANCE ON MISALIGNMENT

The review of existing construction specifications indicate that there is no consensus on
the practical limits placed on dowel misalignment tolerances (Table 2.5). The responses
on tolerance specifications and on the misalignment check during construction, obtained
through the previously mentioned survey, are summarized in Table 2.6. These tables
indicate that there is no consensus among the DOTS on: (a) the techniques used for
assessing dowel misalignment, and (b) the practical limits placed on dowel misalignment

in jointed concrete pavements.

23

Table 2.5. DOT Tolerance Specifications

 

 

 

 

 

 

 

 

 

 

 

Tolerance
State Baskets DBI
Misalignment: Dowel bars shall remain Misalignment: 1 V4 in. over the length
aligned (parallel) with each other and 1 1/8 of the bar in the horizontal and vertical
Michigan in. in both horizontal and vertical planes. planes.
(2004) Transverse Location and Depth: Dowels Longitudinal Location: 1 2 in. of
shall be placed middepth 11/2 in. Dowels planned longitudinal location.
shall be centered 1 ft 1 56 in.
Misalignment: Parallel to the pavement
surface and centerline 1 16 in. over 18
. . . 1n.
Wisconsin H0]: dowel bars m the ”1”“ posrtroncznd Transverse Location and Depth: 1 l in.
(2004) a rgnment usrng an engrneer-approv of the planned transverse location and
devrce during construction depth.
Longitudinal Location: 1 2 in. of
planned longitudinal location.
Misalignment: 1 V4111. per foot.
Ohio Transverse Location and Depth: Centerline of individual dowels shall be parallel to
(2005) each other, the surface and the centerline of the slab. Dowels shall be 1 92 in. on centers.
Dowels shall be placed mid-depth of the slab.
Misalignment: 1 1/8 in. over 18 in.
Transverse Location and Depth: Centerline
Iowa of individual dowels shall be parallel to the q
(2005) other dowels in the assembly 1 1/8 m. N/A
Spacing between dowels shall be 1 ft 1 V4 in.
Each assembly shall be placed so that the
bars are in a horizontal plane at T/2 1 1/2 in.
Misalignment: 1 0.354 in. over 18 in. in both horizontal and vertical directions.
Transverse Location and Depth: Parallel with the pavement lane centerline and surface
California of the pavement at mid-pavement depth.
Transverse location 1 1 in. from planned location.
Longitudinal Location: 1 2 in. of planned longitudinal location.
Illinois Misalignment: 1 1/8 in. over 1 ft. in the horizontal and vertical planes.
(2002) Transverse Location and Depth: Dowels, when used, shall be held in position parallel

 

to the surface and centerline of the slab by metal devices.

 

24

 

Table 2.6. Survey Results for DOT Preferences of Dowel Bar Tolerance Checks

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DOT Bond Stress Misalignment Tolerance Quabty Control
of Misalignment
DB1 Basket
Alabama No spec None V4 in. per 12 in. Field inspection
Colorado No spec No spec No requirement
. . Contractor's
Florida No spec None V1 1n. resmnsibility
Coring. Also
. . evaluating the
Nevada No spec V2 1n. over 18 m. the usefulness of
MIT Scan-2
North . V4 in., V4 in. 1 [IT
Carolina No spec V2 1n. opposing skew Scan-2
Ohio No spec None None Pachometer or coring
2,200-4,400 Alignment verified
lbs, prior to concrete
Pennsylvania depending on V4 in. over the dowel length placement.
pavement ~ Baskets assumed to
width be rigid.
No spec. Ensure that the dowels are
parallel to the surface and the slab
centerline.
Virginia No spec Parallel in the horizontal and vertical No requirement
plane.
16 in. tolerance from the midodepth of the
slab
Developing
Washington No spec % in. specifications for
MIT Scan-2
Wisconsin No spec V1 in. over 18 in. No spec

 

 

2.6 EFFECTS OF MISALIGNNIENT
Various studies experimental as well as field have been carried out since 1930’s to
evaluate the effects of misalignment on pavement performance and this section has a
compilation of these possible effects.

Segner and Cobb (1967) state that in general at least three types of failures,

excluding bearing failures around the periphery of the dowels, in concrete pavements can

25

be attributed to misaligned or locked dowel bars or other dowel bar defects. These
failures may be described as follows:

0 Cracks at or near the mid-span

0 Local Spalling at the contraction joint around the dowel bars

0 Flexural cracks between the mid-span and the contraction joint, frequently near

the ends of the dowel bars
Cracks near the mid-span are essentially tension failures caused by the lack of one or
more contraction joints to function (open) properly. Contraction joints not opening
properly causes two or more Slabs to be “locked” or “tied” together and function as a unit
rather than independently. This increased length under the action of initial shrinkage and
contraction due to temperature changes causes the tensile strength of the concrete to be
exceeded with cracks developing at more or less regular intervals. The failure of
contraction joints to open properly may be caused by bonding between the dowel bars
and the concrete due to deterioration of the dowel bar grease, misaligned dowels or a
combination of both. When this “locking” of the joint occurs, the contraction joint (which
is designed as a weakened plane in the Slab) actually becomes a reinforced plane with the
dowel bars functioning as reinforcing bars. Consequently, tension cracks develop in the
slab some distance from the joint. Since concrete is inherently weak in tension, the
critical period for this type of failure is the first few days after pouring, when the concrete
is very green.
Local Spalling around the dowel bars may be caused by dowel bar alignment

errors in any plane, but are most likely to occur when the dowel bars are misaligned in

the vertical plane. Flexural cracks between the mid-span and the contraction joint result

26

 

from bending action in the pavement slabs caused by deflection of the Slab ends. These
deflections in the Slab ends result from a combination of longitudinal slab movements
due to shrinkage and temperature changes, and dowel bar alignment errors in the vertical
or oblique planes. The deflections at the slab ends may actually force the end of each slab
section to act as a very short cantilever beam; hence, flexural or bending failures may
occur. It can be shown that deformations due to vertical alignment errors combined with
sufficient joint movement may be critical.

Donahue (2003) States that translation, both horizontal and vertical, does not have
as significant an impact as Skew. He states that longitudinal translation is a measure of
the bar’s effective length on the approach and leave slabs. It is not realistic to expect
every 18 in. bar to straddle a joint with 9 in. on either Side, but it is expected that a bar
have at least 6 in. on each side to ensure that it can adequately provide load transfer
across the slabs. A study conducted in the late 1950’s concluded that the dowel
embedment length required to provide full load transfer is five or more times the bar
diameter (T eller 1959).A recent study by Minnesota DOT (Bumham 1999) indicated that
an embedment length of only 2.5 in. is sufficient to keep faulting at an acceptable level of
V4 in. and provide LTE with less variability.

Donahue (2003) states that horizontal translation has little impact on load transfer
performance, unless grossly clustered enough to possibly create load transfer gaps and air
pockets. He also states that since the baskets are designed for 12 in. spacing and the DB1
slots are fixed for this spacing, there is virtually no chance of the abovementioned

occurring. Dowel depth is not as critical as skew, but still must be monitored to ascertain

27

that the bars are not being placed in higher stress zones where they could be deformed or
cause debonding or bearing failures in the PCC.

Bock and Okamoto (1989) state that the dowels with excessive vertical translation
from mid-depth can cause joints to spall. The capability of the dowel to transfer loads
across the joint can also be reduced when dowels have large levels of vertical translation.
Loss of load transfer can increase slab deflections under load causing pumping and loss
of subbase support. This can cause premature joint and corner cracking. Resistance to
movement is provided by subbase friction and locked joints. For slabs up to 20 ft,
resistance due to subbase friction is not a major problem. The magnitude of restraint
caused by locked joints depends on the degree of dowel misalignment, number of
misaligned dowel bars, and degree of dowel corrosion. Locked joints may result in
transverse cracking, comer breaks, and Spalling at the concrete face around the dowel.
Once a spall occurs around a dowel, load transfer effectiveness of the dowel may
decrease.

The effects of these basic types of dowel misalignment on pavement performance
have been summarized in Table 2.7 (after Tayabji 1987) in terms of the possible
distresses including Spalling, cracking, and loss of load transfer efficiency. As shown in
Table 2.7, skew dowel misalignments are more detrimental and they can cause all three
types of distresses at the pavement joints. In “in-service” pavements, the dowel
misalignments are probably combinations of these basic types of misalignments and
tentatively produce all three types of pavement distresses at the joint. The type and
magnitude of skew will impact the concrete-dowel bearing stress which leads to higher

cumulative damage at a joint.

28

 

The overall performance and distress of in-service concrete pavements are functions of:
(a) the pavement design, (b) environmental conditions including thermal gradients, (c)
joint spacing, ((1) applied loads and number of passes, and (e) the dowel misalignments.
The tolerance limits for dowel misalignment will depend on the other parameters (a - d
mentioned here) and the required number of passes (design life) before pavement failure

occurs in terms of the distress (Spalling, cracking) or performance (load transfer

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

efficiency < 70%).
Table 2.7. Possible Effects of Dowel Misalignment on Pavement Performance"
Type of Effect on Comment
Afigment Error Spalling Cracking Load Transfer
Horizontal _ _ Yes
translation
Longitudinal Y Depends on the magnitude of
. - - es .
translation translation
Vertical
translation Yes - Yes
Horizontal rotation of the slabs is
Horizontal skew Yes Yes Yes possible. If this rotation is restrained,
crackingis possible.
Vertical uplift of comers is possible,
Vertical skew Yes Yes Yes depending on the magnitude of the
skew.
Both rotation and uplift are possible,
Combined skew Yes Yes Yes depending on the magnitude of the
skew.
'After Tayabji (1987)

2.7 EXPERIMENTAL STUDIES ON DOWEL MISALIGNMENT

This section discusses the findings of the different experimental studies on the effect of
dowel misalignment.

Smith and Benham of Indiana (1938): They conducted tests on twenty specimens in four
groups of five test specimens. The test specimens were 4 ft X 4 ft 8 in., cast on the
ground, with a transverse contraction joint in the center (parallel to the 4 ft dimension),
each containing four dowel bars. Two slab thicknesses were studied (5 and 6 in.). The

dowel bars were 3%: in. in diameter, 24 in. long, and placed 12 in. on center. Dowel

29

misalignments included in each group were 0, Ms, 1/2, l, and 1.5 in. over 22 in. of the bar.
The outer dowels were horizontally misaligned (in opposing/non-uniform orientations)
while the inner two were vertically misaligned (in opposing/non—uniform orientations).

The contraction joints were filled with a piece of wood 5% in. thick and having a
vertical dimension equal to V!“ the slab thickness. A weakened plane was created by
installing a piece of No.24 galvanized iron, extending from the wood strip downward to
the subgrade. The dowel bars were passed through semi-circular openings, 3 in. diameter,
in the sheet metal. At the age of 28 days, the joints were opened to a distance of 3%: in.
using hydraulic jacks. After a few days, the process was reversed and the joints were
closed till their original position. Ten such cycles of movement were performed within
one year.

They found that for 6 in. thick slabs, misalignments in excess of l in. caused
Spalling when the joints were opened to % in. while for the 5 in. thick slabs, M: in.
misalignments resulted in minor distress. If the joint opening was limited to l/2 in.,
misalignments up to 1.5 in. did not result in any observable distress when the slabs were
pushed apart. Generally, the load required to open a contraction joint 1/2 in. did not exceed
3,000 lbs per dowel.

Segner and Cobb of the University of Alabama (1967): They conducted tests on 38 dog-
bone shaped specimens that consisted of two slabs each 2 ft 9 in. long and 3 ft. wide at
the joint. All slabs were 10 in. thick. Each specimen contained three dowels with the
center dowel misaligned. The dowel bars were 1.25 in. in diameter, 16 in. long, and

placed 12 in. on center. Dowel misalignments, defined over full length of the dowel bar,

30

 

were 0, V2, 3A, l, 1V2, and 3 in. in the horizontal, V4, V2, V4, and l in. in the vertical, and V2
and l in. in the oblique planes.

Weakened planes were formed in the specimens by means of V: in. thick redwood
strips held securely in place by slots cut in the side forms and a slotted 2 in. X 4 in.
extending transversely across the top of the specimen forms. The slots were cut 2.5 in.
deep and V4 in. wide to assure the uniformity of the needed joint dimensions.

Generally 6 specimens were poured at a time with three different errors in
alignment. After the concrete samples had reached their initial set, the top surface was
cured with a liquid curing compound. One sample of each alignment error was tested at 2
and 7 days of age. Two 20-ton capacity hydraulic rams connected to a common pump
were used to push the slabs apart. The specimens were loaded slowly (in approximately
ten load intervals) until each failed at its weakened plane. The load was then reapplied to
the specimens in approximately ten additional load intervals until the joint had opened 1
in.

They found that misalignments in the vertical plane are more critical than errors
of equal magnitude in the horizontal plane with vertical alignment errors being the most
severe. Loads required to produce a contraction joint opening of V2 and 3A in. vary as a
function of magnitude of misalignment. Misalignments in the horizontal plane can be as
high as % in. without causing any appreciable increase in load to produce joint openings
of V2 and 3A in. However, alignment errors of V4 in. in the vertical plane require a
significant increase in load to produce similar joint openings. Spalling was observed
when the joint opening reached approximately 0.9 in., which could have been caused by

the local crushing of the concrete around the periphery of the dowel bar. Load required in

31

..9 .
A's. - I Q t

opening a joint V2 in. for a 1 in. vertical misalignment was about 4000 lbs, and for a l in.
horizontal misalignment the load was about 2000 lbs.

Weaver and Clarke (1970): They conducted tests on specimens 10 in. thick slabs fitted
with 1 in. diameter dowel bars, 18 in. long, spaced at 12 in. on center, with misalignments
ranging from 4-16%. The variables considered were joint formation type (fractured and
plain), concrete age (7 and 28), and joint width (0.025, 0.05, 0.125, 0.25, and 0.375 in.).

In the joints with fractured faces, crack inducers of brass sheet 2.5 in. deep and
0.015 in. thick were used to reduce the section at the top and bottom of each specimen at
its midpoint; the joint was formed by cracking the specimen between the crack inducers.
The plain joints were formed by completely separating the halves of the specimen with a
% in. thick steel former; the former was fabricated in two parts, divided at the location of
the dowel bar, facilitating easy removal from the joint. 100 kN capacity hydraulic jacks
were used to load the specimens.

They observed that the rigidity of the dowels at the joint decreases with increasing
joint openings. The rigidity of a dowel bar (load/unit deflection) in a joint decreases
rapidly on initial application of load to a joint and then achieves a substantially constant
value at an equivalent misalignment (deflection/unit of joint width) of about 3%. The
measured rigidity of dowel bars increases up to an age of 7 days and little difference was
detected between specimens of age 7 and 28 days. They also recommended that
misalignments should not be greater than 4%.

Tayabji (1987): He conducted 16 tests on slabs that were 3 ft. wide by 3.5 ft. long
containing a single dowel and 33 tests on slabs 3 ft wide by 2 ft long fitted with two

dowel bars that were non-uniformly misaligned. Two slab thicknesses were studied (8

32

and 10 in.). The dowel bars were 18 in. long and placed 12 in. on center. Dowel
misalignments, defined over full length of the dowel bar, were 0, Vi, V2, 1, 2, and 4 in. in
the horizontal and the vertical planes. For each of the two dowel tests, both the dowels
had the same misalignment.

A 1A; in. thick steel plate was used to form the joint. Each specimen was cast on
two layers of polyethylene sheets. The initial tests were conducted on single-dowel
specimens and the pullout loads were low and hence the test procedure was modified and
the 2-dowel series was tested. Specimens were tested at 1, 3, 7, and 28 days and the joint
was opened to a maximum opening of V4 in. One slab section was held firmly while the
other slab section was pulled using a hydraulic jack. Pullout load was applied gradually
and uniformly to produce a joint opening of V: in. in about 1 minute. For each test, the
pullout test was performed three times. After each, the pulled slab was pushed back to
close the joint and the pullout test was repeated. For the second and third tests, the

maximum pullout load obtained was less than half that obtained for the first test.

From the single-dowel tests he found that a large portion of the pullout load was
required to open the joint 0.01 in. After the joint was opened 0.05 in., there was no
further increase in the pullout load. He did not find any significant difference in the
pullout load for the different levels of misalignment in the single-dowel cases whereas in
the 2-dowel cases although an increase in the pullout load was observed with increased
level of dowel misalignment, the absolute magnitudes were relatively low for
misalignment levels below 1 in. No Spalling was seen around dowel bars at the joint face
for specimens having misalignment levels of less than 1 in. at a maximum joint opening

of V: in.

33

The most stringent specification requires vertical and horizontal alignment to be within 'A;
in. per foot of dowel bar length; the least stringent requires only 3/3 in. per foot (ACPA
2004). The current ACPA recommendation suggests a value of 3/8 in. per foot, or 3 %.
There is, however, no consensus on the limit or tolerance of misalignment beyond which

it is detrimental to slabs.

2.8 FIELD MEASURENIENT OF DOWEL BAR ALIGNMENT

There are four methods used for the detection and measurement of dowel bar placement
in the field: coring, the pachometer and covermeter, the Ground Penetrating Radar
(GPR), and the MIT Scan-2. Amongst these methods, coring is the only destructive
method. By far, coring is the most accurate method but due to its destructive nature, its
use is limited.

While the importance of achieving dowel alignment is widely recognized, the
ability to monitor the placement accuracy of dowel bars effectively had been limited by
the lack of practical means of measuring the position and orientation of dowel bars
embedded in concrete (Yu 2005). The past difficulties in measuring dowel alignment had
several important consequences on concrete pavement construction, including the
following:

0 Limited validation testing for dowel alignment — Most agencies conducted only a
limited amount of coring to evaluate dowel alignment, leading to dowel alignment
evaluation based on extremely small samples.

0 Possibility of extremely strict dowel placement tolerance — Most agencies have
fairly strict tolerances on dowel placement accuracy, but those standards are based

on limited laboratory and field data. In some cases, the fabrication tolerances for

34

dowel baskets are adopted directly and used as the tolerance on dowel placement
accuracy, which leaves no room for any placement error during construction. The
actual dowel bar alignment needed to assure good pavement performance is
largely unknown at this time.

0 Limited usage of DBI - Because of the concern over the dowel alignment and the
lack of practical means of verifying dowel alignment in the past, DBIs are not
widely used in the US, and many highway agencies specifically prohibit the use
of the DB1, although the DB1 can offer significant advantage in construction cost
and speed.

In this section, more focus has been given to the GPR and MIT Scan-2 methods as they
are the most accurate methods, available to date. Details on the field studies that have
used the above methods and any observations during the studies are described in this
section.

2.8.1 Pachometer and Covermeter

Pachometer and coverrneter are battery-operated magnetic detection devices, which are
mainly intended to measure the depth of reinforcement in concrete, and to detect the
position of rebars. Its use has been extended to pavements to detect the location of dowel
bars. The device emits an electromagnetic field and detects disturbances in the field
caused by embedded metals. Figure 2.5 illustrates a pachometer in use.

Field Experiences: Fowler (1983) and Burati (1983) used the same electronic detector
model in Georgia and Alabama, respectively to measure misalignment. The detector was
considered accurate for horizontal measurements, but vertical measurements could not be

determined. Study by Fowler (1983) indicated that the accuracy obtained using the

35

 

electronic metal detector was V3 in. on horizontal rotation. Donahue (2003) used a
handheld pachometer or steel locator to measure actual dowel bar depths and used it as
ground truth data for calibration of GPR. The latest versions of the covermeters have an
accuracy of i 0.04 in. and are used based on procedures similar to BS 1881-2041988.
Study at Minnesota indicated that while accurate results could be obtained using a
covermeter, testing a large number of bars using this device is not practical (Yu and

Khazanovich 2005).

 

Figure 2.5. Typical Pachometer.
'Source: Donahue (2003)

2.8.2 Ground Penetrating Radar

The Ground Penetrating Radar (GPR) uses a radio wave source to transmit a pulse of
electromagnetic energy into a subsurface (in this case, concrete pavement). The
amplitude and arrival time of the reflected electromagnetic pulse (which originates from
the top of the dowel) is recorded for analysis (determination of spatial location of the
dowel). The GPR signal is characterized primarily by changes in reflection amplitude and

changes in the arrival time of specific reflections. The GPR record consists of a

36

 

continuous graphic display of reflected energy over a preset time interval. The depth to
the dowel can then be determined if the propagation velocity, and electromagnetic energy

through concrete are known or estimated. Figure 2.6 illustrates atypical GPR in use.

   

(a) Ground Penetrating Radar (GPR) (b) GPR in use

Figure 2.6. Typical Ground Penetrating Radar in use*
‘Source: Donahue (2003)

The resolution and depth penetration of the GPR tool is a function of the frequency of the
antenna employed and the conductance of material imaged. Higher frequency antenna
provide for better resolution, but less depth penetration. The GPR can penetrate resistive
materials, but cannot be transmitted through highly conductive materials (such as dowel
bars). When the GPR antenna crosses a dowel bar at right angles, the resulting GPR
image looks (visually) like an inverted U (hyperbola). The apex of the hyperbola
indicates the exact spatial location of the dowel bar.

Field Experiences-Advantages and Limitations: Study by Okamoto (1988) indicated that
radar is an effective tool for evaluating dowel bar misalignment, but needs improvement
with respect to horizontal misalignment and precision in this study was ~ V2 in. Study by
Donahue (2003) used a 1.5 GHz GPR and use of a single dielectric constant while using

GPR was attributed to lead to potential errors of :t Vs in. in dowel depth measurements (in

37

an absolute sense). However, he states that depth estimates at any single joint study site
would be accurate in a relative sense.

Interaction with pavement engineers at the Michigan Department of
Transportation and University of Missouri-Rona, who had used the GPR for similar
studies, revealed that the speed of data collection is walking speed and takes no more
than a few of minutes per joint. However, around 4 to 7 GPR profiles, perpendicular to
the dowel bar, have to be obtained, in order to obtain complete information of the dowel
bar location. The processing of data requires care and training in order to amass the
different runs. Also, post-processing cannot be carried out on the field. The starting point
of each GPR profile should be along the same horizontal line else it could lead to errors
or mismatches in the processed data.

According to a Concrete Pavement Technology Program (CPI‘P) project (Task
7F) (ACPA 2004) study, with GPR, variations in material properties, which can be
substantial along a project, can create random errors that significantly affect analysis
results. Water on the pavement surface or in the pavement structure can also affect
results.

2.8.3 MIT Scan-2
The MIT Scan-2 is a state-of-the-art device for measuring the position of metal bars
embedded in concrete. Figure 2.7 shows a close up of the MIT Scan-2 and Figure 2.8

shows the operation of the MIT Scan-2.

38

 

Figure 2.7. Close up Of MIT Scan-2 Figure 2.8. Typical Run of MIT Scan_2
in Field

The MIT Scan-2 utilizes an array of sensitive detectors and sophisticated data analysis
algorithms to produce very accurate results. The device emits a weak, pulsating magnetic
signal and detects the transient magnetic response signal induced in metal bars. The
methods of magnetic tomography are then used to determine the position of the metal
bars.

Unlike other devices that have been used in the past, which are general-purpose
instruments adapted to the dowel bar detection application, the MIT Scan-2 was
developed specifically for measuring dowel and tie bar alignments. As a result, the
device is simple to operate, efficient, and provides real-time results in the field.

Field Experiences-Advantages and Limitations:The MIT Scan-2 tests the entire joint at
once, providing results for all dowel bars placed in the joint in one shot. The testing
takes about 1 minute per joint and up to three lanes can be tested together. In an 8-hr
day, a 2-person crew can easily test 200 or more joints using the MIT Scan-2. It can be
run continuously for 8 hours on one charge of the battery. The MIT Scan-2 is shown to
provide very accurate results for dowel bars placed using a DBI. For DBI inserted bars,

the magnetic technology may be more reliable because the results are based on direct

39

measurements, rather than correlation to a calibration (Yu and Khazanovich 2005).
Because the MIT Scan-2 is essentially a metal detector, its operators must choose a
scanning location carefully, since magnetic fields and metal objects, close to the joint
being scanned (within about 3 ft of the bars) - such as power lines, tie bars over the
dowel, vehicles, marker nails in the concrete for saw cutting, even steel-toed boots, can
interfere with the measurements. The dowel baskets also interfere with the measurement
results; however, approximate results can be obtained if the following conditions are
met:

0 The dowel bars are epoxy coated

o The transport ties on the basket are either cut or removed
The accuracy of the MIT Scan-2 results depends on the position and the orientation of the
dowel bars (Yu and Kim 2005). It produces the most accurate results when the bars meet
the following typical placement tolerances:

0 Mean dowel depth 4 to 7.5 in.

0 Maximum vertical misalignment i 0.8 in.

0 Maximum horizontal misalignment i 0.8 in.

0 Maximum lateral position error (side shift) 5 2 in.
For bars meeting the limits of reliable results for MagnoNorm software (software used to
analyze MIT Scan-2 results) listed above, the estimated overall standard deviation of
measurement error is 0.12 in. in rotation, which means that the device can provide
measurement accuracy of i 0.2 in. with 95 % reliability. CPTP project (ACPA 2004)
results reported that MIT Scan-2 was found to be reliable, efficient, and accurate within i

0.08 in. when position errors are minimal. Accuracy depends on the degree of placement

4O

error. Within typical placement tolerances i 0.38 in. for vertical and horizontal
misalignment and 2 in. for side shift, the range of error is 10.16 in. With gross
misalignments, the error can be greater.

The California, Nevada, South Carolina, and Washington DOTs participated in
the CPTP field trials that evaluated the MIT Scan-2. Apart from being impressed with its
capabilities and practicality, they have provided the following suggestions/observations:

0 South Carolina DOT: Sample of 10 to 20 joints from each day’s production is
adequate to monitor performance. Testing every joint can slow production.
0 Washington DOT: 50 joints can be checked in 1 hr. Weston reported that minimal
training would be needed to interpret the positional data.
0 Nevada DOT: Sohila Bemanian commented that the trials were promising and is a
very powerful tool to inspect 100 % of the work on the first day of production.
Because the MIT Scan-2 operates on electromagnetic field, presence or absence of non-
conducting material does not affect the results. It is also not affected by changing
moisture conditions in concrete, so, the testing can be conducted at any concrete age,
including over fresh concrete. The test results are not affected by the presence of water on
the pavement surface.

The CPTP study compared the performance of MIT Scan-2 and ground-

penetrating radar (GPR) and found that both technologies can produce accurate results

and both have limitations as listed previously.

41

2.9 FIELD STUDIES COMPARING BASKETS VS DBI
Table 2.8 summarizes the results of the field studies comparing the two dowel bar
placement methods i.e. baskets and DBI in context of misalignment observed and
distresses as a result of this, if any.
Some of the key observations include:
o No method is significantly superior over the other but the DBI in many cases has
produced results comparable with the baskets and better with respect to average
depth and vertical misalignment.
0 Quality of dowel bar placement ultimately relies on quality of field inspection and
contractor consistency.
0 Occurrence of longitudinal translation is similar for both types of joints.
Longitudinal displacement is affected at least as much by location of the sawed
joint as it is by the actual dowel movement.
0 Individual dowel position has no effect on the alignment achieved.
0 Distribution of tilt is more symmetrical for basket joints than inserter joints
indicating that vertical misalignment may not be independent of paving direction.
Also it has been emphasized that it is advisable to do a test stretch on the pavement using
the DBI and a quick check on the placement in plastic concrete before proceeding with
the rest of the project. Also the DBI’s quality of placement depends on the manufacturer,
and a study has to be carried out before using a DBI from a different manufacturer.

The reference used for vertical height in case of DB1 is the concrete surface
whereas for baskets, the reference is the base grade. The misalignment is measured with

respect to the pavement surface and hence accuracy in depth of the dowel might be

42

impacted by even a slight change in the thickness of the concrete slab and the
comparisons might be misleading. The studies also recommend the use of a handheld
pachometer or a magnetic rebar locator to be available on all doweled PCC construction
projects, which would be useful in aligning sawn joints with the dowel bars and in
identifying missing dowels.

In the field study by Yu et.al. (2003) comparing the baskets vs DBI, a joint score
scheme was developed that takes into account both the number of misaligned bars in a
joint and the severity of misalignment. A joint score greater 10 indicates a greater risk of
joint problems, while joint score of 10 or less indicates relatively low risk of joint
problems. The basket section had nearly twice as many joints in the >10 category,
indicating a greater potential for joint problems. .

Yu et. al. (2005) compared 5 DBI projects and 7 basket projects located in 6 states
across the US, using the MIT Scan-2 and the results (Figure 2.9) indicate that both the
methods are comparable. The last two categories have been termed as high risk zones by
the authors.

In a dowel bar retrofit project in Washington, severe vertical rrrisalignments of 1—
2 in. were found in many dowel bars. However, from a visual standpoint, the pavement is

performing well (ACPA 2004).

43

% of bars

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100

90 — D

78.8 m 325:7
80 72.7
70 -
60 _
50 -
40 ..
30 ~

19.2
20 - 14.2 ~
10 ‘ I 3 6 5'7 17
0 l__’__|n'n'”'“ 1.6 ' 1.8 0.6
dso.4 0.4<d_<_0.6 0.6<d50.8 0.8<d$l d>1
Range of malignmem, in.

Figure 2.9. Comparison of Baskets vs DBI

 

 

00050.: :20.0:-:50_. 0:
80:00: :53 0:00.005. 80.00: ::0 .mn. 503:0: 00:85.5: ::00:.:.:m.0 oz 0

50500050.: .05:::.w:0. 0:0.00:< 0
:0:0:0: .0::0n.:0: 0:0.00:< o

 

000.00.: 3:22:55. 0: "0:805:20: 0805030.. 050:0?
:00000: 5.3 0:00.005. 00.00: ::0 an. 503:0: 00:80.5: 5005:50 02 0 8:00:09 .503. 0505005 ::0—=03. Anna—.9340 50.5:
5&0: .0309 o

.50:

:0: 0.5300000 03 :0.:0.>0: ::0::0:0 5:0: 0w0:0>0 ::0 5:0: 0m0:0>< o
005050.00

0005.0 0: :00000: 5.3 0:00.00: :05 ::0:0.0:00 0:05 003 .mn. 5:0: :0": o
050.. 0500 05 5 0:0:

.030: 503:0: 300.0 0:00:00 :5: ..0 0005:5000 000.:00 30: 03:: 50m 0

00:05:85.. 5 0w:0 55.0 .5. H ”:05000. 00:90:05. N50: :23. :0> :00: 80.00::

0 w5:.0: 0:00.00: 05 :53 300.0 .00.::0> :0 .0b:00 :00m b0> 03:: 50m 0 .5 w: :26 .5 0. H ”50:55:00.2 :0.:0=::0:00 ADE. 0: m3
.:0:0: 0.5. 0 3:58:00. 50.0 05 :0 05. 0:: 0: .0..0:00 ::0 8ch has. 6:50:2—

ED 0:: 5.3 00.80::0: 300.0 03:00.0: 0:00:08 0>0: 0: ::0: ::0: 0 0:80:00 0:: 0: 0:00: 0: 00:20.05 ::0: boa 5.80:2

550.052... .030n. o
3005 .030D o

00:80:53m

0305200.: 200000.. .350:

0:0: 3:050:00. ::005 0:05:03
.5 N H ::0.:0.0:0:: 35:25:05

0:00.00»: ..mQ

 

..030: 05 :0 ::0: 0:80:00
0:: 3.0.005 5 :.0 000: 5 .:0>030m ..030: 0:: :0 ::0 m::.:03 05 :0
w5:000 0 00 03:00:05 0. :50: :0: .030: 0: ... .w::.:03 0:03 0.030: .2. 0
0:00.85. 05 :0 :0::.0 5 :0:::000 00050.: ::0::0>00 :30.0: .030: 07. o
:38 :0 20:2:
.00.::0> 05 0: 0: :05000 w5::0.n.::. 5.3 .0:::00 0: 5:00.: 5:05.: :002
.::0E0>0:: .030: .0300 .3 0. :. 00 :55. :0300
05 ..0 :05000. .3 :02: 00:000. :0 :0:00t0 0. 50500030.: 35:25:04
50:55.0 .05:::.w:0_ :0 50.905. 0:: 5.0.02.8
::0 :0.:0:0: :0 05030:: ..0 8055.0 :0: 5: 0:00.00: :0 8:05.53 0
0:00.00:
35.5: :05 :0“ 00:05.0: 0:0 :0 0.00.5000 05 0: 5000.0 3:09:00
::0 00000 :005 5 0:0:0050000 05 :05 :0::0.n.::. :0 :00: :05 60.05. a: o

:0.:0:0: .0::0~.:0: ::0 .00.::0> o
::0:::w..0 35:35:01. 0

5:0O 0

"0:82:80:

:0:00:0Q

.902 00808.: .0550 "02:02
> .5 ::0 ::0 m

.5 Q... 003 ::0:::m..00.:: .0:0.:0:0m
:0.:0.0:0::

:0: .5 : H :0 > .m 000:230h
:... om w500% :59. .00 .5 m.

5:0. .5 w. .:0:0::0.: .5 mm: 0.030Q

0:00.00: :53 0:00.35.

03: ::0 350.95 0.030:
:53 0:05.05. 0:50:95 00:: ...
55:0: ::0:0 0 :050 :0>00

05 .:0 :00:0 3:05.00: 003
:0::0.n.::. 0: ... 00550 w5:.:
:00: ::0 .::0:w 5.3 :05:
0:503:00: .030: 55:05.:
05x0 5 w5:.:00: :0000:
:0: :0>00 0:: :0:.:0 0.030:

w500.0 :05 8:005:00
::0 ”5:05: 08-0.2
an.

”08: 5 0022: 08:0: 2-: .
000:0: 0.85:2: 080-2 .

.000:
0.0.06: E00: :0: 0:008:

 

 

80.00.050.053:

 

3:05:02
0:300:35— ::0 ::0—00.50002

:0 02:02 £58580

 

053$
«00,—. ::0 5:55 00:30:3—

 

 

0....8: 0:30 0:00: :0 0:05.50 .0.0 030:.

45

 

0:00.005.

0:50: :0 :0:0:.:05 0: :.:0:0 5030:: 0.5 :0: 503.00. ::0 500.0055
0: 0: 30::0 0:00..0:: .mn. 3.3.00 :0 0.030: 3.00.5 5.3 05030:: 0
0:00.05 50: :0 :0:.000: 0. :0.:0:..00:00 :0>0::5. o

00:00.0
505.55 5.3 0305:: .3. :0 :0>0.:00 0: :00 0.: :0 3.30.: 05”. o

0.030: 3.00.5 3.3.505 5 ::0 0:0: .030:

0.030: 3.00.2 o
0:.0> o

0500 00.: .
:0.:0:0: 3500.5... 0
:0.:50: .00.::0> 0
5:0: 0m0:0>< o

 

0000:0503:
05 5.3 0:50.. :300 350.3 5 5:00: 0: .0303 030.05 :0.:0:.00:00 3.:00 0:050:
00: :0.030: ..0 :0 0.:0..0>0 :0300. 3:0: 0.3035 0 3.3... o :0.:00:.: :05.0
.::0::0:5. 0. 0:50.. 3.300 :0: 3.0.35 0:0500< o 5 .5 m H 50.3.03: 3:55.33
.0:0:0:00 :00:: 05 :30:5 350:: .3 530. :3. 5 0...>00:0..
00:02, 00 0:00.: 0.0000 .0300 000 000.20 0000 :0 :0:: 20 :0 000:0 :0: :20 .0: x a "000000050033 :0 00-: 00 02:00 000
0. 305000.: .030: :0 5:0: 0: 30:00: 5.3 .mn. 0:: :0 :500 3.5:. 0 5:0: .mn. 3.0: :000 0305:: 00:: ...
.:0.:0:0: 3500.5: ::0 .00.::0> 5:0: 03:03 0: 30:00: -:.5 05 0>0:0 .5 x ”.030: :0 5:0n. .633 $.09
5.3 0:0: .030: :0 505000.: 005350.30 503.300 :0 23:00 0. .mQ 0 0003303: 50:80.3. 5:0: 0050..
0:00.09: :0::0.m.::. :. :0::0:
b30050: 0:03 0::05000.:0.: 3300.5: ::0 3:55.30. 0::.00:< o
.0:00..0:: 30.00: ::0 :0::0.:5.

50: :0: 0500 05 3.05:3 0:03 00:.0> :0.:50: 3500.:0: 0:500:< o
::00: 003 00050.: ::0 00:»: 50.59355 00033: :0.:0.0::00 07. o
.:::0:
003 00050.: :0 ::05:w..0 .030: :0 0:5» ::050>0: 50:: :00::0 02 o
.:0>0.:00 305:»..0 0:: :0 :00::0 0: :0: :0:.00: .030: .0::.>.::. o
.00:0:0::.: 05
0: :0:::.:::00 0>0: ::m.5 30:30.0:005 ::0 00:0:0::.: :0:00:::00 :5
0:00:05. :0:.00: 0:: 5 0.030: 05 0:03 :05 :0:».3 :83: 0:03 63:03
0:: :0 0:00.05 :0::0.:5. 05 :. 0.030: 05:05 :030.::. 0:300: ..0:0>O o
.00:0:0::.: 05
0: :0:::.:::00 0>0: ::w.5 :0:0:0.0:00:. ::0 00:0:0::.: 5:00:58 3..
.0:00..0:: :0:.00: 0:: 5 0.030: 05 0:03 :05 :0:w.3 :33: 0:03 .0m0:0>0

95503: 0503?.

 

 

05 :0 0:00.08: :0::0.:5. 05 5 0.030: 0:: :0:: :030.::. 0:300: ..0:0>O 0 50500055: 3300.:0: 0:500:00. 0 335.33 .35..
:0.:0:.002 :0.:00m
30.003000350. 0:805:00: ::0 50505000.). 0 :0 .0 0000.8 0
:0 :0:—:02 305005003 :0 ... : :Bm : 0.

 

 

 

 

35:00. 5." 0.0.0...

46

 

.3333 £038 mo 855::

05 9 585358 5 :25. .8530: .33 882.580 .«o 53:52

.8583 587.3 55. 5588 ED 05 5 58083 83 EoEouaEmE

555555. 205 35 9525a 55 .5858 on 8 3538 Pa 9533
“En meage— 53 Ho .3238 mm :0:» 55555; «c «3.58 850 .3 o

55% some .53 50553355 .85? mo 5:555; 98 50553.85
35:? we 3&3 Eon Ho 8:2 5 58.93 555 .033 5 85558505 :5 o
.3052: 53 he 5585: on 9 Roman gnu—u €35. owfio>< o

.5532

~52 58.35 05 30:8 8: moot So 32 05 p: 8 $533. 5.5 38058

$553533
50535q5fl .3655m53 o
50553352 35555 o
52.55352 .8555 o
Sac—u owfla>< o

 

35.85 8: 5 .5582 .55. 05 .5 305555 on :3 “55355.55 o ..ES~EES~
656055 960— 05 5 5.33538 3557. war—cu maxoh £575 50 Snow .9.-—
855555 B 9 822 8588 5:35 as 5 503% 05 .555 . .52. 5552.3 5:55 ..Es..§ Ea: saufio
355523

.355. mo 39: Son 5.5 3:56 5 cog—mash E55595. .6 352.580 0
5582:. w5>aa .50 535355 on 8 «.508 50552355 .Scontom c
55625 w5>aa
.8 5095355 on 8: .35 5053285. 3555 85 55.355 .955.
8:35 55 $52 .9743 he .aofiofiém 208 5 :2 no 50555.85 0
£8.93
53 an 5953 358:6 585.55 0: 5 205 .38: 3:350:
53 .5: 3559 4553536 35535.8— .fiaov 335V :0 33m 0

EoEoanmE E55555; 0
50555352 358525 o
50553352 B535 0

Egon owEo>< o

..Eu352um

.5 _ H ”cons—mafia. 35555

.5 3 hon .5 X H ”50553—352

.5 5 u 535-55 ”552. 538
“8520—5.

558% 5% 5852 Cum .5 2

db .5 S © 95— .5 3 ..5 3.7;»
”58% .59. .5 2 .08 .5 2

db .5 9 © w=o_ .5 mm ..5 3.7“?
.5 fl Ba .5 X H ”8:228.

558% 555. 505:2 Gun .5

GmQ gov—mama 5285? do;
3mm .88.ng $on .31
Gmfl «850.8 0:5: .34

 

 

.3335 9 35558 =03 3.55% an . 2 .20 .5 2 © 9.2 .5 8 ..5 22-5 6%: scaflo .58:
69.8535 .333
93659593565“ Bougam 53 «5593802 mo .8 a 3:93 o
5 35»: 55.8583 . ... a cam . a

 

 

 

 

Sang ad «Eng.

47

 

.5209. an 2: 55
EB REE—3E: :0 $8582— :onwE a :2 :288 “87.3 05 .3: wdA
::a ”.5 m6 9 0.8 motowofio 5253338 238 20.: 05 he Son 5 o
.owSa .5 ed 9 mmd 05 E was REE—«£8 mo 093582.
SawE 05 9 ~20on 26 33 8:09. :5 05 E was 8%..«030 05 .«o
uwfifiéom 58on 05 3:82. .88: on ha: :288 ED 05 £8039:
:52. wEmzmo “:0:—:3“ 330: 6:89.: 2: mo am: 05 ho was: 5 a
G: :3 m> Q» Gm: 5:3
3.58:2. 5:08 .833 05 .3:w=am_E wan mo 52:2: “0 $58 5 o
6238

E08523? 32:9 :5 35:82
..ESususum

N-:8m .52 .435»:

.E 3 “SE8: 830: .823

a _ Ba .5 34...

3:085:86: 32:? ::a 13:09::

6:35:05 :3:
5:85:08: m:
350.. 8: 3.29

SEot—au
80.8.5
:o .53
.3833

::a 50 $7.: some :0308 0:0

 

 

 

 

“8.3: 05 E 85 52: 3.53 o: g 5:08 an 2: :_ E253? .038 o ..3:E~3n 383 532.320— 5%
gag—UEOUWMEfi-gh gouggflfiua—flwuulaeafloz :cfluom
an :a a 3:93 o
no .8502 gaseofiuunm . a. a cam .« m

 

 

$3.33 QN 033—.

48

CHAPTER 3 EXPERIMENTAL PLAN AND PROCEDURE

This chapter describes the experimental setup and plan, instrumentation and material
testing details, and typical preparation steps in a casting-testing cycle.
3.1 EXPERIMENTAL SETUP
The experimental investigations were conducted on laboratory-scale pavement specimens
with doweled contraction joints. Each pavement specimen consisted of two 48 x 24 x 10
in. concrete slabs connected at the joint using steel dowel bars. Steel dowel bars that are
1.25 in. in diameter and 18 in. long are placed at the mid-depth at the joint, with equal
lengths (9 in.) embedded in each concrete slab, and are placed 12 in. on center. The
number of dowel bars at the joint and their misalignment type, magnitude, and uniformity
varied from specimen to specimen, according to the test matrix presented in the next
section. Figure 3.1 shows the overall dimensions (plan view) of the concrete after
demolding (for a l-dowel test). Figure 3.2 (a-c) illustrates the different parts of the
experimental setup and details of the components. Each concrete slab is supported on
smoothened rollers placed on hardened steel plates. There is an ’/s in. gap between the
concrete slabs that serves as a full—depth contraction joint. The pavement specimen is
tested by pushing apart the two concrete slabs using hydraulic actuators simulating
thermal expansion.

Essential parts of the molds are made from steel sections. The concrete slabs are
cast in-place directly in the molds and are ready to test after the concrete sets. The joint
between the concrete slabs was formed using a 'A; in. thick aluminum plate with circular

holes of appropriate diameter to pass the dowel bars through it. This aluminum plate is

49

left in place after casting the concrete, and it does not hinder the experiment or the
separation of the slabs in any way.

In all, three molds were fabricated for specimens accommodating one and two
dowels and one mold for three and five dowel specimens. The mold consists of the
following, starting from the ground up: base rails, solid rollers, base plate, and channels.
The channel sections are held in place with the help of bolts located at maximum spacing
as per the AISC Steel Manual. The key mold details (cross sectional and plan view) are

illustrated in figures A-l through A4 in Appendix A.

 

 

 

 

I8 sap
1750 a 5 4.750

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.1. Plan - Section view of the overall mold for 10 in. slab

(All dimensions are in in. unless specified otherwise)

50

#80): Cutout

C 10 xlS.3 stool channel

 

(a) Experimental Setup
:15 . v4 “" -,
' Dunc] hanging assembly-N8
in. (liameEcr Lhreadcd " ' ‘
,houk‘

_.
.

l

.

=5

l . .
‘ Loud dist“!

union:~ ,-

:
plate ‘

 

cradle for 7
biting the actuator

(c) Hydraulics

Figure 3.2. Experimental Setup and Components

51

The dimensions mentioned in the following paragraphs are in context of the small mold
but the function is the same in both the small and the big molds. The base rails (54 x 6 x
1/2 in. thick — 2 in number and 54 x 10 x 1/2 in. thick) have a polished surface to support
the solid rollers of 2 in. diameter and 54 in. long. The base rails help in providing a flat
surface. The base plates (53 in. x 27 in. x 1/2 in. thick -— 2 in number), with the bottom
surface smoothened are placed on the solid polished smooth rollers. This smoothness is
required to ensure that when the pullout load is applied using the actuators, the base rails,
solid rollers, and base plate move with minimal frictional resistance relative to each
other.

The cast and west channels are structural steel sections C10x15.3 attached with a
removable steel box cutout (13 x 6 x V2 in. thick) with the corners having fillets. The
north and south sides are similar to the east and west channels but do not have the box
cutout. The north and south channels including box cutouts are removed during
demolding. The steel box cutout joins the two base plates together thereby providing
height and rigidity to the mold. The purpose of the box cutout is to provide space for
placing the actuators and the steel spreader plates during the pullout of the slabs.

The permanent section of C 10x 15.3 (48 in. long) stays bolted to the base plate
throughout the duration of the experiment. The structural steel section used for hanging
the dowel bar assembly is illustrated in Figure 3.2 (a-b). A slot of 5% x 3/3 in. on the
channel section provides the % in. finely threaded mild steel bar enough room to move in
a horizontal direction. The dowel hanging assembly shown consists of threaded U-hooks
that are bolted to supporting channels. This specially designed assembly allowed fine

adjustments to the dowel position in the vertical and horizontal directions.

52

A new test setup was added, with the size of the specimen, nearly double of the previous
test setup for the one and two dowel bars, to accommodate three to five dowel bars.
These specimens consisted of 2 concrete slabs, each 96 x 36 x 10 in.

The specimens were tested by pushing apart the concrete slabs, using hydraulic
actuators, shown in Figure 3.2 (c). The hydraulics and the testing procedure are described
later in the chapter.

3.2 EXPERIMENTAL PLAN
The following variables were taken into account while designing the experimental
matrix:

0 Number of dowels and number misaligned

0 Misalignment type

0 Misalignment magnitude

0 Orientation
The experimental design matrix is shown in Table 3.1 and is explained in the following
paragraphs.
Number of dowels: Tests were carried out on two slab sizes — the smaller slabs (2’ X 4’)
accommodating either one or two dowel bars and big slabs (3’ X 8’) accommodating
three or five dowels. On the 2-dowel systems, tests with either one or both misaligned
were performed. On the 5-dowel systems, tests with all dowels misaligned, and alternate
dowels misaligned (outer and center dowels misaligned) were performed. A total of 51
unique misalignment combinations have been tested (15 one-dowel, 25 two-dowel, 3

three-dowel, and 8 five-dowel, respectively).

53

Table 3.1. Experimental Design Matrix

 

Number of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Slab Drmensrons Dowels ID Mrsalrgnment Magnitude, m.
1A Aligfl 0
11114 M:
11112 1/2
1H34 Horizontal %
1H1 l
1H2 2
lVl4 IA
1 1v12 l/2
1v34 Vertical 1%
1V1 l
1V2 2
1C14 %
r
1C12 Combined /2
1C34 V4
1C1 1
2 2A Aligned 0
r
2H12U Horizontal + A, + V2
2H1U + l, + l
‘ r
.2 WIN Vertical + 95, + /2
2 slabs each (Uniform) 2V1U + 1, + 1
r r
(48 x 24 x 10 in.) 2C12U Combined + b, + /2
2C1U + l, + l
2H14NU + ‘14, - Mr
1 _ r
2H12NU Horizontal + /2, /2
2H34NU + 3’4, - 3%:
2H1NU +1, -1
2V14NU + l/4, - 1/4
r _ 1
2. 2V12NU Vertical + /2, /2
(Non-uniform) 2V34NU + -%, - -%
2V1NU +1. -1
2C14NU + ‘A, - %
r _ r
2C12NU Combined + /2, /2
2C34NU + %, - 3%:
2C1NU +1, -1
2H12AM . + I/2, 0
Horizontal
2H34AM + %, 0
(O 2 b ZVIZAM Vertical + V2, O
ne ar
3
misaligned) 2V34AM + A, 0
ZCIZAM . + V2, 0
Combined
2C34AM + %, 0

 

54

 

Table 3.1 (contd.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Slab Dimensions N‘st‘fi'l's“ n) Misalignment Magnitude, in.
3 3H12NU Horizontal
. 3V12NU Vertical + V2, - V2, + V2
(Non‘un'fo’m) 3C12NU Combined
5 5H12NU Horizontal
- r _ r l _ r r
2 slabs each (Non-uniform) 5V12NU Vertical + /2, /2, + /2, /2, + /2
. 5C12NU Combined
(96 X 36 X 10 1n.) ,
5H14AM Horizontal 1 r
5 . +A,O,-A,O,+%
5V14AM Vertical
(Outer and Center .
dowel misaligned) 5H12AM Horizontal
5V12AM Vertical + V2, 0, - V2, 0, + V2
5C12AM Combined

 

Misalignment Type: Three types of misalignment have been focused on — horizontal,

vertical, and combined misalignment. Combined misalignment refers to equal magnitude

of misalignment in the vertical and horizontal directions. A total of 16 cases of vertical

and horizontal misalignment each, and 14 cases of combined misalignment have been

tested.

Misalignment Orientation: In cases of multiple bars, the effect of two orientations was

studied — uniform and non-uniform misalignment. Uniform misalignment has been

termed for cases where the consecutively misaligned dowel bars are parallel to each other

in the direction of misalignment. In non-uniform misalignment, the misalignment of

consecutive dowels is opposite to each other. Uniform and non-uniform misalignments

are illustrated in Figure 3.3. A total of 6 uniform and 20 non-uniform misalignment cases

were tested.

55

 

 

     
   

 

'12”): ~ ‘ _ ; ”1'25 " . - A .. ,-
" , 1299. , ' , , I . < 1299. ’_ Non-32:53:“),
1' . ‘ : A. . .. ‘t < 1299 . l Q . > 1299 , ’. 90%le
. 12,; L'>Umfm1nlyMisahgned p, , .
> . ‘. ,: 3 \ DowelBars ,, I‘
~ ' .12»: ' - 2 " - ’

 

 

 

Figure 3.3. Plan view of Uniform and Non-uniform Misalignment

Misalignment Magnitude: Tests with misalignment magnitudes of 0, 'A, 1/2, 3A, and l in.
over half-length of the bar (9 in.) have been carried out. The choice of these magnitudes
were based on three factors — maximum accuracy, maximum probable limit of
misalignment in the field, and limit at which distresses develop. The aligned one and two
dowel specimens are treated as the control specimens.

Nomenclature of test ID: Each test is identified with an ID and the nomenclature is based
on the various variables discussed above. The first number is the number(s) of dowel bars
in the specimen. The second letter is the misalignment type, A for aligned (or straight), H
for horizontal, V for vertical, and C for combined. The number(s) following the
misalignment type is the misalignment magnitude in inches per half length. An '/4 in.
misalignment is denoted as 14, l/2 in. as 12, % in. as 34, and l as l in. If there is more than
one dowel bar in the test specimen, then the alphabets following the misalignment
magnitude is the relative orientation of the dowels. Non-uniform misalignment is denoted
as NU, uniform as U, and alternate bars misaligned as AM. Clockwise (CW) is

considered positive and is used to define the orientation of the misalignment. A plan view

56

 

and section view of a dowel bar with clockwise combined misalignment is shown in
Figure 3.4. The transverse joint runs in the North-South direction. Clockwise direction is
defined with reference to the East side for horizontal and the North side for vertical
misalignment. The north most dowel has a clockwise orientation in all cases and the

orientation of the following dowels depends on the orientation type.

 

 

I Joint I I /lfi |

Slab Slab

w J. ....... . w e w ..... E

 

   

 

 

 

(a) Cross-sectional View (b) Plan View

Figure 3.4. Clockwise orientation of the dowel

3.3 MATERIAL SPECIFICATIONS

The MDOT specifications require the dowel bars to be made of billet steel grade 40 as
per AASHTO specification M31, and to have minimum yield strength of 40,000 psi and a
minimum ultimate strength of 70,000 psi. The dowel bars are epoxy coated as per
AASHTO M254. Most of the steel dowel bars were obtained from the same heat and
batch as far as possible. The concrete mixture is the MDOT pavement mix grade P1 the
mix design of which is summarized in Table 3.2. The detailed mix design specifications

are shown in Table A-1 of Appendix A.

57

Table 3.2. Mix Design of Concrete

 

Source of Concrete: Plant l4-East Lansing

 

SSD

 

 

 

 

 

 

 

 

 

 

 

 

 

Material Class: Source Wei gilt (lb /y d3) Yield, ft’
Cement ASTM 0150 Type I: Essroc 564 2.87
Fine Aggregate 2NS: Builders Aggrejates (#34-86) 1275 7.65
Coarse Agggegate 6AA: MLO LS(#71-3) 1720 10.81
Water 256 4.1
Air Content 6.5 % 1.77

Total 27.2

Admixtures added:
ASTM C-494A Water Reducer Type A MB 200 N 3.0 oz/c
ASTM C-260 Air Entrainer MB Microair 1.2 ozjc
Desired Plastic Concrete Properties

Slump 3 in.

Concrete Unit Weight 142 pcf

Air Content 6.5 %

3.3.1 Steel Coupon Testing

Nine test coupons were obtained from 3 dowels bars. The specimen size and testing was
in accordance with ASTM E8-99 as shown in Figure 3.5. Figure 3.6 (a-c) illustrate the
different stages in coupon testing.

A typical stress-strain curve is shown in Figure 3.7. The properties derived from
each of the curves include (i) yield strength, (ii) ultimate strength, and (iii) the elastic
modulus of steel. Figure 3.8 (a-c) summarize these properties for all the coupons tested.
The horizontal line in each plot indicates the minimum required value of the
corresponding parameter as specified by ASTM A615. The averages of the yield strength,

ultimate strength, and elastic modulus of steel obtained are 69 ksi, 95 ksi, and 30,270 ksi,

respectively.

58

 

 

 

Dowel Bar

 

 

D= nominal diameter = 0.500 in.

A: length of reduced section = 2% in.
G = gage length = 2.000 :1: 0.005 in.
R = radius of fillet = 3/n in.

1.: left, M=middle, R=right

Figure 3.5 . Standard V2 in. Round Tension Test specimen with 2 in. gage length
(ASTM E8-99)

\!

    

       

     

(byCoupon at the Intiation of ' (c)>Coupon at Failure
Necking

(a) Coupon before testing

Figure 3.6. Stages of Coupon Testing

59

 

100
90 a
80 -

f

Y ..
%

Slope = Es

50 4

40-
30-
20—
10~

Stress, ksi

 

 

 

0 i r
0.00 0.05 0.10

l l I

0.15 0.20 0.25
Strain, inJin.

l l

0.30 0.35

Figure 3.7. Typical Stress-Strain Curve for a coupon

3.3.2 Concrete Testing

The fresh and hardened concrete properties measured and their corresponding ASTM
standards are listed in Table 3.3. Concrete cylinders (4 X 8 in.) were cast at the same time
as the pavement specimens. The hardened concrete properties were measured at three

specimen ages -3-day, 7-day, and 28-day. 3-day and 7-day correspond to the days of

demolding and slab testing, respectively.

Table 3.3. Concrete Properties

 

 

 

 

 

 

 

 

 

 

 

Type of Test Property Measured ASTM Standard
Slump (in.) ASTM C143
Fresh Concrete Unit Weight (pct) ASTM C138
Air Content (%) ASTM C138
Temperature (°F) ASTM C1064
Compressive Strength (psi) ASTM C39
Hardened Concrete Split Tensile Strength (psi) ASTM C496
Flexural Strength (psi) ASTM C78

 

 

60

0.40

mm ‘lnfiua-us melanin

 

 

 

 

 

 

 

In In W V) In V) W
V o v Ox 3 o v
VD W W V M M
1 1 l l l 1

I 1 T 11 l V 1
1111111131111111-1l|].I1iI11'-1|1111111111‘l1l1ll11l111 11‘ 1111:11‘111n111l'11|L11|'1||11l1llrl1|i11
at. .‘1 i 1.9 “a? m!‘
'15». ‘ . a , ,1 11),: 12.1 m . .‘ 1.. m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l
T
I 1 '3': I. "111ll1i n11”: 26111: :I1iltl
l I 1‘ . ‘1‘ l11+11‘111l1.11i11 111I1‘111
. . w I‘M M W11
.12
11 1,1
WI 1'11: ,Iii'lili"'i‘. 121 1“ .1|-.1i'1 .! =1i'l‘1.l.‘,1 1‘511113'4311
n f l r 1 1

 

E

O
m

8

8

m Wants PPM

M
5
00
t:
1.3
U)
a:
am§
2
«xi-gag
an
mag
63
M

ULmMER
(a) Yield Strength

Sample [D

8(IN ‘snlnpow 395813

5
5

r 2.0E+0

. 0.0E+00

 

_- 3.0E+05
.. 2.5E+05

 

i 1.SE+0
.- 1.0E+05
.- 5.0E+04

1 'ti i!:,:illl'-Ie!i.i1 .il 1

2.. 1: 2: .1 if

111‘

 

 

H-‘I 6‘ L21
i

 

Mn r 3'1 II n 'ui't H w .: mm! 1
. J
‘ 1

 

 

 

 

 

 

 

 

 

r'HWlMl I1 .I

  

1|
1.311I2..‘ili 13:1i1‘rli

.H 21.1mmiml:.lnlfllluivmnn .1111
1.;

“i i:

 

 

 

 

 

 

50000

61

 

§

 

ls’l ‘snlnpow 993915!

Sample ID
aLmMIR
(c) Elastic Modulus
Figure 3.8. Coupon Tests Strength Results

Fresh properties were tested on only one sample per test and the averages over all the
batches tested are:

0 Temperature: 70.8 °F

0 Slump: 3.6 in.

0 Unit Weight: 144.5 pcf

- Air Content: 6.1 %
The number of samples used for the hardened concrete tests consisted of 3 cylinders for
compression and 2 cylinders for spilt tension on each of the three days of testing. 1, 2,
and 3 beams, respectively, were tested on the 3“, 7m, and 28'” day. The averages of all the
strengths at 28 days and the sample sizes associated with their calculation are presented

in Table 3.4.

Table 3.4. Statistical Parameters for Raw Strengths

 

 

 

 

 

Property Average, Hi 1
Compressive Strength (45)fi 4991
Split Tensile Strength (32? 392
Flexural Strength (39). 609

 

 

 

'The number in braces is the number of samples tested.

The variation across batches was very high but the variation within a batch was not as
high, which is as expected. Also, the compressive strength exhibited less variance than
the flexural and split tensile strengths.

3.4 INSTRUMENTATION

This section describes the instrumentation used in the measurement of the slab responses
and the calibration of the instruments. Table 3.5 describes the location of the various
instruments and the response measured. After vertical uplift was observed in the two
dowel non-uniform vertical misalignment cases, sliders were used at the slab comers

where vertical uplift was expected, i.e. in vertical and combined misalignment cases.

62

Figure 3.9 (a-d) show stand alone pictures of the instruments while Figure 3.10 (a-b)

show a close up of the same in use.

Figure 3.11 shows a diagrammatic representation of LVDT and slider locations in

a 5-dowe1 test. Figures 3.12 and 3.13 illustrate the instrumentation setup in a two dowel

test with expected vertical uplift and a five dowel test.

Table 3.5. Instrumentation Location and Purpose

 

 

 

Transducer (LVDT)

Instrument Response Measured Location
Placed perpendicular to the joint at
Spring Return Linear Motion Measure Joint Openin dowel bar locations and at 4 or 8 in.
Sensor (Slider) g from the longitudinal edges depending
on the slab size.
Linear Variable Differential Measure Joint Opening Placed perpendicular to the Jomt at

dowel bar locations.

 

Pressure Transducer

 

 

Measure pressure applied
to the actuators

 

Placed in the box cutouts.

 

 

 

[3610-15
(a) Linear Variable Differential
Transducer

    

 

(b) 9610 Linear Motion Position Sensor

(1.0 in. stroke)

 

(c) PX303-OISGSV Pressure (d) 9615 Linear Motion Position Sensor

Transducer

(1.5 in. stroke)

Figure 3.9. Closeup of Instruments"

*Source: www.0mega.com (a,c). www.beiduncan.com (b.d)

63

 

(b) Pressure Transducer and Hydraulic Jack

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(a) LVDTs and Sliders
Figure 3.10. Instruments
North
‘ o Slider Actuator
for Vertical Uphfi North Transverse
Slider
Slider
12 in. c/c
LVDT and Slider
96 in. LVDT and Slider
LVDT and Slider
Slider
South Transverse
Slider 1 1
1 o Actuator
36 in. H 36 in.

Figure 3.11 . Plan of Instrumentation Setup for 5-dowel test

 

. a
. E

.

...”,
My

a.
A
.”

a

w
,

f

z

1» .
. as”?

. . —
.
.37.

s

'

 

Figure 3.12. Instrumentation Setup for Typical 2-dowel setup with Instrumentation

for Vertical Response

.7

r

A .
...”..e

....

V,

W

, ...-...-
“......

.5.

....

M...

r;

_ ”an
5

s

 

-dowel Instrumentation Setup

Figure 3.13. 5

65

The pullout load for the experiments was provided by the hydraulic actuators and hand-
pumps. Each hand pump can be connected to either a RC 156 or a RC 256 actuator. The
specifications for the LVDT and the pressure transducer are summarized in Tables A-2
and A-3 of Appendix A. The calibration of the instruments is described in Appendix A.
3.5 TYPICAL CASTING-TESTING CYCLE

The steps in a typical casting-testing cycle are mold assembly, rough misalignment,
misalignment check, casting and material testing, demolding and instrumentation setup,
and testing. The length of a typical casting-testing cycle is 14 days (2 weeks).

Rough Misalignment: After the mold has been assembled, the dowels are placed in the U-
hook hanging assembly and are misaligned as per the desired combination. As per
MDOT construction specifications R-40-E, the. spacing of the dowels is 12 in. c/c for no
misalignment and the dowels are suspended at mid-depth of the slab i.e. 5 in. for a 10 in.
slab. After rough misalignment using a tape measure, each dowel is welded on one side
and greased on the other half as per R-40-E. Alternate dowel bars are welded on the same
side. The greasing is, however, done just before casting.

Surveying: The dowel misalignments are adjusted and measured accurately (before
placing the concrete), using two total electronic stations (theodolites), a reference point
(base plate), trigonometric principles, and surveying techniques. The accuracy of the
measurement system (0.005 rad.), was adequate for determining dowel misalignment
angles with confidence. The misalignment checks in the dowel with respect to the base
plate are described in Appendix B. The theodolites were used to measure the various
angles between the dowel bar, base plate, and the horizontal and vertical planes. For each

of the misalignments in the test setups shown in Table 3.1, the calculation of

66

 

misalignment using the theodolite was compared to misalignment obtained using the tape
measure. The maximum error tolerated is z % in.

A typical surveying setup is shown in Figure 3.14. Details on the calculations for
typical cases of surveying for horizontal and vertical misalignment are given in Appendix
B. Tables B-l through B-8 provide a summary of the surveying of the various cases.
Tables B-l through B-4 summarize the comparison of height checks and tables B-5
through B-8 summarize the horizontal misalignment checks.

The average error was found to be 0.093 in. and 0.028 in. in vertical and

horizontal measurement, respectively.

 

Figure 3.14. Typical Misalignment Check Setup (Surveying)

Casting: The concrete slabs are cast in-place directly in the molds placed and are ready to
test 7-days after the concrete sets. The aluminum separator plate is left in place after
casting the concrete, and it does not hinder the experiment or the separation of the slabs
in any way. The test specimens along with beams and cylinders for quality control are

cast using the MDOT approved paving mix delivered by ready mix concrete suppliers.

67

Figure 3.15 shows casting of a 5-dowel test setup. Figure 3.16 shows the fresh concrete
properties apparatus and molds for hardened concrete properties specimens. Before
casting the main specimens, quality control tests are performed on a sample of concrete
to ensure it meets the specifications for slump and air content, else the concrete is

rejected. Figure 3.17 (a—c) show typical hardened concrete properties test setups.

 

Figure 3.16. Fresh and Hardened Concrete Test Apparatus

Demolding and Instrumentation Setup: A typical specimen containing one dowel is

shown in Figure 3.18. The test specimens are demolded typically three days after casting,

68

after it had been ensured that the concrete had achieved at least 50 % of the target 28—day

compressive strength (3500 psi). The instrumentation is setup as described in section 3.4.

 

(3) Compression (b) Flexure (c) Split Tension
Figure 3.17. Concrete Hardened Properties Test Setup

10 in.

 

Figure 3.18. Cast Specimen

Testing: Pullout tests on the slabs under thermal expansion are conducted 7 days after
casting. The protocol maintained during the entire testing regime was to continuously
measure the induced load in the dowel bar due to joint opening of up to l in. The total
load measured as the joint opened was divided by the number of dowels in the test

specimen.

69

The specimens are tested by pushing apart the concrete slabs using hydraulic actuators,
which are of lS-kip capacity for the small slabs and of 25-kip capacity for the big slabs,
each having a stroke of 6 in. The actuators were used to apply controlled monotonic
longitudinal loading. The hydraulic pressures in the actuators were synchronized using
split flow and needle valves. The concrete slabs were pushed apart very slowly at the
approximate load rate of 20 lbs/min. followed by the opening displacement rate of
approximately 0.02 in./min.

Pressure transducers were used to measure the hydraulic pressures in the
actuators. The loads applied by the actuators were estimated using the measured
hydraulic pressures and the calibrated actuator piston areas. This procedure for
determining the applied loads using the measured hydraulic pressures was calibrated and
validated prior to testing. Several instruments including spring-return linear motion
sensors (sliders) and linear variable displacement transducers (LVDTs) were used to
measure the transverse opening of the joint. The vertical and horizontal displacements of
the slabs were also measured using sliders and LVDTs. The sliders had measurement
ranges of either 0-1 in. or 0-1.5 in. with accuracies of 0.02 in. and 0.03 in., respectively
and the LVDT had a measurement range of 0-1.5 in. with accuracy of 4.5 x 10'3 in. All
the instruments were calibrated prior to conducting the experimental investigations,
details of which are provided in Appendix A.

The data from the different instrumentation was collected at the rate of 6 scans per
second, using a data acquisition system, the details of which is given in Appendix A. The
data acquisition system is capable of handling all the different instruments — sliders,

LVDTs, and pressure transducers. The software interface used to collect and process data

70

 

is Little General Version 6.1. The calibration files, obtained as described in earlier
sections are input into the software and, the data is acquired in the desired units of
measurement. As mentioned earlier, quality control tests on concrete were conducted in
parallel on the 7m day also.

3.6 COMPARISON OF THE EXPERIMENTAL SETUP AND PLAN WITH
PREVIOUS STUDIES

In this section, the different features of the experimental setup and plan, described in the
previous sections of this chapter, are compared with that of the previous experimental
studies. Different features of the experimental setup were based on lessons learnt from
previous experimental studies described in the literature review (chapter 2).

Experimental Setup: Specimen configuration is similar to the dog-bone shaped specimens
used by Segner and Cobb (1967) to facilitate the placement of the hydraulic jacks during
the pullout test. The plain transverse joint between the concrete slabs was formed using a
‘/s in. thick aluminum plate with circular holes of appropriate diameter to pass the dowel
bars. This concept was derived from Tayabji (1987). In the present research, each
concrete slab is supported on smoothened rollers placed on hardened steel plates.
Smoothness was to ensure that when the pullout load is applied using the actuators, the
base rails, solid rollers, and base plate move with minimal frictional resistance relative to
each other and the pullout load acts perpendicular to the joint face. The method for
misalignment check is unclear in the literature available. In the present research,
surveying principles were applied using total electronic stations that had an accuracy of
0.005 radians and the maximum error tolerated in this study was i 'A; in. Dowel hanging

assembly consisting of % in. finely threaded mild steel U-hooks that are bolted to

71

supporting channels were used to support the dowel bar, on the lines of the assembly used
by Segner and Cobb (1967). This specially designed assembly allowed fine adjustments
to the dowel position in the vertical and horizontal directions. Corner reinforcements near
the box cutouts were used to prevent excessive build up of stresses at the box cutouts
comers. Segner and Cobb (1967) had used shoulder reinforcement near the axes of
application of the load.

Experimental Plan: One of the main limitations of previous pullout studies includes the
low rate of data collection accompanied by a high rate of loading. The limitations of the
previous laboratory studies have been addressed in the following manner:

0 Bond stress behavior has been studied.

0 In this research study, the data was Collected at the rate of 6 data points per
second and the displacement rate of approximately 0.02 in./min, which is
approximately 1600 data points collected for each 0.1 in. joint opening increment.
This rate of data collection helps in the understanding of the dowel-concrete bond
behavior, especially at smaller joint openings where debonding takes place and
the slope of the load vs joint opening curve is very high.

0 Study of the relative effect of the misalignment orientation on load vs joint
opening behavior has been conducted.

0 Tests on specimens accommodating up to five dowels have been carried out

which is close to simulating half a lane.

72

 

CHAPTER 4 DISCUSSION OF RESULTS

This chapter deals with the discussion of the results obtained from the various pullout
experiments conducted. The discussion presented includes typical bond stress versus joint
opening behavior (hypothesized and observed), overview of the results that includes
distress observations, and the load vs joint behavior results for the different tests and load
zone prediction given misalignment magnitude and number of dowels.

4.1 TYPICAL BOND STRESS VERSUS JOINT OPENING BEHAVIOR

As mentioned earlier in chapter 3, the load induced per dowel bar is calculated by
multiplying the pressure by the corresponding calibrated piston area and dividing by the
total number of dowel bars. The bond stress T, is then calculated by dividing the load by
the circumferential area of the greased side of the dowel bar (7t*1.25*9). It is
hypothesized that the bond stress versus joint opening curve has two distinct regions: (a)

fully bonded region (OA) and (b) post-slip/debonded region (BC) as shown in Figure 4.1.

 

A
l Transverse contact or mechanical
friction / rmax
'5
a C
O
J:
m
1:
i:
c
m
A /Initial Slip/debonding stress, Tb
O >

 

 

 

umax

Joint Opening, u, in.

Figure 4.1. Typical Bond Stress versus Joint Opening curve

73

The first region is the high initial slope region where there is bond between the dowel bar
and concrete. The applied pullout force which can be resolved in the axial and normal
directions of the dowel bar is opposed only by the dowel-concrete bond force in this
region as illustrated in Figure 4.2 (a). Slip occurs after the bond stress per dowel
increases beyond a certain threshold value. The magnitude of the bond stress at the point
of debonding or initial slip is denoted as the initial slip/debonding stress (13b), and is
calculated using equation 4.1:

Pb
7” = 7d)L

 

..(4.1)

where Fb is the force at initial slip/debonding in lbs, D is the dowel bar diameter in inches
and L is the embedment length of the dowel bar (9 in.). Section 914.07 of MDOT (2003)
requires this stress not to exceed 60 psi. PENNDOT requires the force at initial slip to be
in the range of 2,200-4,400 lbs, depending on the pavement width (based on survey
conducted in August 2004).

As further thermal expansion is induced, the forces induced in the dowel bar
increase and depending on the misalignment, the induced force could form a plateau or
increase causing failure of the test specimen. It should be noted that only the forces
induced due to joint opening were measured in the lab and not the stress states at the
interface of the dowel and concrete. Each of the misalignment type, magnitude, and
orientation would cause a specific certain stress state zone at the interface along the
length of the bar and at the joint, which cannot be studied directly through lab
observations/tests. Due to the presence of misalignment and joint opening, normal

bearing stress in concrete is introduced and due to the frictional effect across the

74

interface, shear stresses are developed as shown in Figure 4.2 (b).The effects of these

parameters on the post-slip joint opening behavior are presented later in the chapter.

 

 

 

 

 

 

Concrete 12
b
V 0
4/ K .
”fumed Force (P)
"'7 Tb PsinG
(a) Before Initial Slip
Concrete

0

Psin0
Shear Forces due to
Mechanical Friction

 

 

(b) Post-slip

Figure 4.2. Forces on a Misaligned dowel

PcosB (Axial Force)

Applied Force (P)

W P0030 (Axial Force)
#71”

After the experimental plan was completed it was found that for any specimen tested, the

bond stress vs joint opening behavior always showed an initial bond behavior between

the dowel and concrete at less than 0.01 in. joint opening. The experimental curves

closely followed the regions or shape of the curve proposed in Figure 4.1. A typical

experimental curve obtained is shown in Figure 4.3 .

75

 

160
140 4
120 i
100 i

Bond Stress, psi
00
o

 

 

 

60 a
40 4
20 _
0 r r r r 1 r r r i
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure 4.3. Typical Experimental Bond Stress versus Opening Behavior

It was found that ‘tb is independent of the misalignment magnitude, orientation, and the
number of bars misaligned and was always in the range of 10-50 psi. Figure C-l
illustrates the Tb for all the gamut of tests conducted. The bond stress vs joint opening
graphs substantiating these results are presented in Appendix C.

The slip strain can be estimated as the joint opening divided by the embedded
length (9 in.) of the greased side of the dowel. The bond shear stress-slip strain responses
for the l-bar straight specimens (with zero misalignment) provide information regarding
the overall longitudinal bond interaction between straight (aligned) dowels and the
surrounding concrete and are critical in the analytical modeling of the dowel-concrete
behavior. The bond shear stress-slip strain responses for the remaining l-bar specimens
provide information regarding the longitudinal bond between the misaligned dowel and
the surrounding concrete, while including the effects of friction and bearing stresses

(from dowel misalignment).

76

I. I'L'nf‘ -

 

4.2 OVERVIEW OF THE RESULTS
The discussion focuses on two main aspects:
0 Overview of the type of joint and slab distresses observed; and
0 Effect of misalignment type and magnitude on load vs joint opening behavior

(post-slip behavior)

The results are grouped as per the misalignment types— horizontal, vertical, and combined
misalignment. Overall, the following distresses were observed during the tests:

0 Bearing failure/Spalling of concrete near the dowel-concrete interface

0 Cracking

0 Non-uniform joint opening

0 Vertical uplift
The first three distresses are destructive even without traffic and/or restraints from the
adjacent slabs and base while non—uniform joint opening and vertical uplift in the current
setup of just thermal expansion are not destructive. The significance of the different
distresses is described in section 4.5.

During the pullout test, the slab that opens or pulls out is the slab that contains the
greased side of the dowel. Because of the restraint due to welding on the other side it is
the greased side that slips out after debonding during the pullout test. Hence in all cases
where spalling was observed, spalling was found to occur at the dowel-concrete interface
at the joint face near the greased end of the dowel.

4.2.1 Horizontal Misalignment Tests
This section discusses the key visual observations in the horizontal misalignment tests. In

the single dowel tests, spalling at the dowel-concrete interface was observed in the 2 in.

77

test only. In the 2, 3, and 5-dowel tests, different combinations resulted in different
distresses: spalling at the dowel-concrete interface, non-uniform joint opening, and
cracking. Spalling was the predorrrinant distress in these tests. The spalls ranged from 2
in. by 2 in. to more than half-depth of the slab. Table 4.1 summarizes the visual
observations for the different horizontal misalignment tests.

Typical illustrations of spalling and non-uniform joint opening observed in the
2H1U test are shown in figures 4.4 and 4.5, respectively. Illustrations of other horizontal
misalignment tests where spalling and non-uniform joint opening occurred are shown in
figures D-l (a-d) and figure D-2 of Appendix D.

Table 4.1. Summary of Visual Observations in the Horizontal Misalignment tests*

 

 

 

 

 

Specimen ID Visual Observations
1A, 1H14, lHl2, 1H34, 1H1 None
1H2 Spalling on the West joint face (3 in.
X 31/4 in.)
2A, 2H14U, 2H12U, 2H34U None
0 Spalling on the East joint face (3.5
in. 7 in.)
2H1U 0 Non-uniform joint opening (At the

end of test, south edge at about V4
in. higher opening than north edge)

 

 

2H14NU Spalli$
2H12NU Spalling on the East jornt face (2 in.
X 2 in.)

 

o Spalling on the west joint face

 

 

 

 

 

 

 

 

 

2H34NU 0 Cracking at 0.72 in. joint opening_
0 Spalling on the west joint face
2H1NU 0 Cracking at 0.67 in. Em openinL
2H12AM, 2H34AM None
3H12NU Spalling
0 Spalling near the outer and center
dowels
Non-uniform 'oint o nin (At
H12NU ' 1 PC 8
5 the end of test, north edge at
about 1/4 in. higher opening than
south edge)
Spalling near the outer and center
5H12AM dowels
5H14AM None

 

*Spalling dimensions should be viewed in conjunction with the corresponding photograph

78

 

 

Figure 4.4. Spalling in the 2H1U test
. z : Z ' A

{ South_Edge

4in.

S_LVDT

6 in.
A: Slab_Center
6 in.

N_LVDT
E.

4 in.

I North_Edge

 

 

l

 

 

l 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
Joint Opening, in.

Figure 4.5. Joint opening as a function of distance along the joint (2H1U)

79

Cracking of slabs in later stages of the tests (at joint openings higher than V2 in.) were
observed in the 2H34NU and 2H1NU tests are described below.

Cracking in the 2H34NU test: At a joint opening of 0.716 in., cracking occurred in the
west slab, which resulted in a sudden drop of load from 4198 to 515 lbs, resulting in a
joint movement of 0.065 in. in about ‘/2 a second. One full depth crack appeared
instantaneously and it split the west slab into two halves at the position of the dowel bar
on the north side. The dimensions and position of the crack are shown in Figure 4.6 (a).
The crack at its point of initiation and at the end of the test is shown in Figure 4.6 (b) and
(c), respectively. The sudden drop in pressure is captured in the load versus joint opening
curve shown in Figure 4.7. As further load was applied, the crack continued to open,
pushing the two halves of the slab apart rather than opening the joint. Thus the test was

stopped at a joint opening of 0.87 in.

 

 

 

 

 

 

 

 

 

s
E W
Z 1A in.
I
\fix
2' o-
1/8 in, 1.5 in.
N

 

 

 

 

 

(a) Crack Pattern and Initial dimensions

80

 

(b) Crack initiation (c) Crack at end of test

Figure 4.6. Cracking in 2H34NU

 

4000-

   

Crack Initiation

Load, lbs
b)
§

1000-

 

 

 

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.7. Load vs joint opening curve (2H34NU)

Cracking in the 2H1NU test: At a joint opening of 0.668 in., cracking occurred in the
west slab, which resulted in a sudden drop in load from 4891 to 1476 lbs. One full depth
crack appeared instantaneously and it split the west slab into two parts at the position of

the dowel bar on the north side. The dimensions and position of the crack are shown in

81

Figure 4.8. Since the pattern is to the similar to the 2H34NU test, the photographs have

 

 

 

 

 

 

 

 

not been provided here.
S
E W
:7 5/8 in.
x

2' m
1/2 in. 3%: in.

N

 

 

 

 

 

 

Figure 4.8. Crack pattern and Initial dimensions in 2H1NU

The sudden drop in pressure is captured in the load versus joint opening curve as seen in
Figure 4.9. As further load was applied, the crack continued to open, pushing the concrete
pieces apart rather than opening the joint. Thus the test was stopped at a joint opening of

0.857 in. and the crack width at the joint at the end of the test was % in.

82

 

   

Crack Initiation

20001

 

1000

 

 

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.9. Load vs joint opening (2H1NU)

4.2.2 Vertical Misalignment Tests

This section discusses the key visual observations in the vertical misalignment tests. In
the l-dowel tests, there were no distresses observed. In the 2, 3, and 5-dowel tests,
different combinations resulted in different distresses: spalling at the dowel-concrete
interface, non-uniform joint opening, vertical uplift, and cracking. Table 4.2 summarizes
the visual observations for the different vertical misalignment tests.

Illustrations of spalling and non-uniform joint opening in the vertical
misalignment tests are shown in figures D-3 (a-d) and figure D-4 of Appendix D. In the
2V1NU test, in addition to surface spalling and non-uniform joint opening, the northwest
and southeast ends of the slabs lifted up (the welded sides of the two dowel bars). Figure
4.10 (a) gives a diagrammatic representation of the uplift and Figure 4.10 (b) and (c)

show the vertical uplift at the NW and SE faces, respectively. The vertical uplift of the

83

slabs in the 2V12NU and 2V34NU tests is similar to that observed in the 2V1NU test and
have not been shown. The magnitude of the uplift was not recorded.

Table 4.2. Summary of Visual Observations in the Vertical Misalignment tosts"

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Specimen ID Visual Observations
lA,lV14,1V12,1V34,1Vl,1V2 None
2A, 2V14U, 2V12U, 2V34U None
2V 1U 0 Spalling on the West joint face (4.5 in. X 23/4 in.)
0 Spalling on the East joint face (2 in. X 1% in.)
2V14NU None
2V12NU, 2V34NU Spalling and vertical uplift
0 Spalling on the West joint face (5 in. X 3% in.)
0 Spalling on the East joint face (3% in. X 2% in.)
2V1NU 0 Non-uniform joint opening (At the end of test,
south edge was at about 0.13 in. higher opening
than north edge)
0 Vertical uplift
2V 12AM, 2V34AM None
3V12NU Spalling
Spalling near the outer and center dowels
Cracking at a joint opening of 0.862 in.
5V12NU o Non-uniform joint opening (At the end of test,
north edge at about V4 in. higher opening than
south edge)
5V14AM None
5V12AM Spalling near the outer and center dowels

 

 

*Spalling dimensions should be viewed in conjunction with the corresponding photograph

Cracking of the slab in later stages of the test (at joint openings higher than V2 in.) was
observed in the 5V12NU test. At a joint opening of 0.862 in., the pressure started
dropping drastically Figure 4.11 and after about 2.5 seconds, a hairline crack was
observed in the east slab, near the center dowel bar, diagrammatically shown in Figure
4.11. This crack resulted in a drop of load from 3902 to 2832 lbs, in one second. The
sudden drop in pressure is captured in the load versus joint opening curve shown in
Figure 4.12. As further load was applied, the crack continued to open, pushing the two
halves of the slab apart rather than opening the joint. Thus the test was stopped at a joint

opening of 0.93 in.

84

   

 

E

(3) Di agrammatic representation

   

(13) NW side (c) SE side

Figure 4.10. Vertical Uplift (2V1NU)

85

96 in

 

 

36 in.

North 36 in.

Figure 4.11. Diagrammatic representation of Crack in 5V12NU

 

 

 

 

6000

5000 ~
g 4000 r Cracking
I:
a
.n 3000
3.. z
8.
1:
3 2000 +
u-l

1000 —

0 r T I l I I I T I
0 0.1 0.2 0. 3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure 4.12. Load vs joint opening curve (5V12NU)

86

4.2.3 Combined Misalignment Tests

This section discusses the key visual observations in the combined misalignment tests. In
the l-dowel tests, the only distress observed was spalling at the dowel-concrete interface
in the l in. test. In the 2, 3, and 5-dowel tests, different combinations resulted in different
distresses: spalling at the dowel-concrete interface, non-uniform joint opening, and
cracking. Spalling was the predominant distress in these tests. The spalls ranged from 2
in. by 2 in. to more than half-depth of the slab. Table 4.3 summarizes the visual
observations for the different combined misalignment tests. Illustrations of spalling and
non-uniform joint Opening in the combined misalignment tests are shown in figures D-5
(a—f) and figure D6 of Appendix D.

Table 4.3. Summary of Visual Observations in the Combined Misalignment tests*

 

 

 

 

Specimen ID Visual Observations
1A, lCl4, 1C12, 1C34 None
1C1 Spalling on the West joint face (2% in. X 1% in.)
2A, 2C14U, 2C12U, 2C34U None

 

o Spalling on the East joint face (3 in. X 2V4 in.)
o Spalling on the West joint face (2% in. X 37/3 in.)

 

 

 

 

 

 

 

 

 

 

 

 

ZCIU 0 Non-uniform joint opening (At the end of test, south
edge was at about V4 in. higher opening than north
edge)

2C14NU Spalling
2C12NU Spalling on the East joint face (5 in. X 4 in.)
2C3 4NU o Spalling on the west joint face.
0 Cracking at 0.95 in. jornt opening
2C lNU 0 Spalling on the east joint face (3.67 in. X 2V4 in.)
0 Cracking at 0.45 in. joint opening
2C12AM, 2C34AM None
3C12NU Spalling
0 Spalling near the outer and center dowels at 0.58
5C12NU in.
o Cracking at 0.824 in. joint opening
5C12AM Spalling near the outer and center dowels
5C14AM None

 

*Spalling dimensions should be viewed in conjunction with the corresponding photograph

Cracking of slabs observed in the 2C34NU and 2C1NU tests are described below.

87

Cracking in the 2C34NU test: At a joint opening of 0.949 in., cracking occurred in the
west slab, which resulted in a sudden drop of load from 4060 to 1124 lbs, resulting in a
joint movement of 0.032 in. in about V3"l a second. Two cracks formed on the concrete
above the north dowel bar, the positions of which are diagrammatically shown in Figure
4.13 (a). One of the cracks was a full-depth crack appeared instantaneously and split the
west slab into two halves at the position of the dowel bar on the north side. The cracks at
their point of initiation are shown in Figure 4.13 (b). The sudden drop in pressure is

captured in the load versus joint opening curve shown in Figure 4.14.

 

 

 

 

__i—NL (b) Crack initiation on top of the north

dowel
(a) Crack Pattern and Initial dimensions

Figure 4.13. Cracking in 2C34NU

88

E

 

i

g

Spallin g

i

Cracking \

Load per bar, lbs

i

i

 

 

 

o
‘h

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure 4.14. Load vs joint opening curve (2C34NU)

O

Cracking in the 2C1NU test: At a joint opening of 0.447 in., two closely spaced cracks
formed in the west slab near the south box cutout, which resulted in a sudden drop of load
from 5550 to 4691 lbs. On further separation of the slabs, it was seen that the crack
observed at the surface was a result of the crack that was initiated near the south dowel
bar. At the surface, the cracks did not open significantly, but at the west joint face, the
crack propagated from one dowel bar to the other. No further cracks were formed. The
position of the cracks is shown in Figure 4.15 (a). The slabs were pushed further apart
after the test and the southwest box cutout comer separated eventually (Figure 4.15 (c)).
The sudden drop in pressure is captured in the load versus joint opening curve shown in

Figure 4.16.

89

 

__l~

00

 

 

 

(a) Crack pattern

 

(c) Stages of separation of the slabs

Figure 4.15. Cracking in 2C1NU

 

50001

Load per bar, lbs

Crackin
/ g

 

 

 

eggs:

0

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure 4.16. Load vs joint opening curve (2C1NU)

90

Cracking in the 5C12NU test: At a joint opening of 0.58 in., spalling occurred near the
outer dowels which resulted in a drop in pressure. Cracking of the slab in later stages of
the test (at joint openings higher than V2 in.) was also observed. At a joint opening of
0.824 in., a hairline crack occurred in the east slab, near the center dowel bar, the location
similar to the location in the 5V12NU test. This crack resulted in a drop in load from
4108 to 1568 lbs, resulting in a joint movement of 0.046 in. in about V3“! 3 second. The
sudden drop in pressure is captured in the load versus joint opening curve shown in

Figure 4.17.

 

6000 -1 r

5000-

Spallin g
3000 r ' Cracking
/

Load per bar, lbs

 

 

 

0 I I I I I I I r I

o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.17. Load vs joint opening curve (5C12NU)

4.3 COMPARISONS OF LOAD VS JOINT BEHAVIOR
The key conclusions that are also substantiated by the plots in this section can be
summarized as below:

0 Within a misalignment type, the load increases with an increase in the

misalignment magnitude.

91

0 Non-uniform orientation of misalignment of two dowel bars requires more load
per dowel bar at a given joint opening as compared to the uniform orientation of
dowels in the horizontal and combined misalignment types and it is vice versa for
vertical misalignment.

0 Degree of locking is higher in the non-uniform orientation of dowel bars. Non-
uniform orientation of misalignment of dowels is more destructive.

- For a given misalignment magnitude, the load required per bar increases as the
number of dowels misaligned increases.

0 For a given misalignment magnitude, the trend of load versus misalignment type
(horizontal, vertical and combined) is inconclusive. However, some trends were
observed within specific tests and are listed later.

The discussion is divided into two main subsections (x = H, V, or C):

0 Comparisons within same number of dowels tests (within 1x, 2xU, and 2xNU
series)

0 Comparisons across different number of dowels tests (2xAM vs 2xNU/U and
SxAM vs 5xNU;lx vs 2xNU/U and 3xNU vs 5xNU)

4.3.1 One and Two Dowel tests

The plots in this section (figures 4.18 through 4.38) collectively illustrate the
comparisons of load versus joint opening behavior for different misalignment magnitudes
within a given misalignment type (horizontal, vertical, or combined). In general, the load
per dowel increases with an increase in the misalignment magnitude. This trend is more
evident in the 2-dowel tests (non-uniform and uniform orientations). However, there are

some exceptions due to experimental variability. Since there are overlaps in the curves

92

corresponding to different misalignment magnitudes at smaller joint openings, bar graphs
of loads at V: and V2 in. joint openings have been plotted for better representation in the
area of interest (joint opening up to V2 in.).

4.3.1.1 Horizontal Misalignment

Figures 4.18, 4.19, 4.20, and 4.21 illustrate the comparisons of load vs joint behavior in
the l-dowel, 2-dowel (non-uniform and uniform orientations), and 2-dowel setup with
one dowel misaligned tests, respectively.

Table 4.4 summarizes the comparisons of loads at a 1A in. joint opening derived
from the bar graphs (Figures 4.23 and 4.24). Similar trends hold good for the loads at a V2
in. joint opening. The last column indicates either a percent increase or decrease of load
at a V: in. joint opening as we move from test 1 to test 2. ‘+’ indicates an increase in the
load and ‘—’ indicates a decrease in the load. For example, when the horizontal
misalignment is increased from V2 in. to 1 in. in a l-dowel setup, this is accompanied by a
43 % increase in the load at a V4 in. joint opening (first row of Table 4.4).

Table 4.4. Pair wise comparison of loads at V4 in. joint opening (Horizontal

 

 

 

 

 

misalignment)
Test 1 - load Test 2 - load % change from Test 1 to Test 2
1H12—21001bs 1H1 —3000 lbs +43 %
2H12NU — 2600 lbs 2H1NU — 3500 lbs +35 %
2H12U — 1400le 2H1U — 2900 lbs +107 %
2H12U — 1400 lbs 2H12NU - 2600 lbs +86 %

 

 

 

 

 

For a given misalignment type and magnitude, the non-uniform orientation of
misalignment in the 2—dowel tests has higher loads at a given joint opening as compared

with the uniform orientation tests (figures 4.22 through 4.24).

93

Cracking, as in the 2H34NU and 2H1NU tests, is captured on the load versus joint
opening curve by an instantaneous drop in pressure at the time when the crack occurs,
which was approximately at 0.67 and 0.72 in. joint openings, respectively (Figure 4.19).
Spalling also leads to a drop of pressure, but in most of the tests where spalling occurred,
the drop has been observed to have occurred over a significant time and cannot be
captured by mere inspection of the load versus joint opening curve. Non-uniform joint
opening, as in the 2H1U and the 5H12NU tests, is captured by observing the variation of

the joint opening magnitudes along the joint through plots similar to Figure 4.5.

 

Load per bar, lbs

 

 

 

 

0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.
Figure 4.18. Comparison of load vs joint opening curves for l-dowel Horizontal
Misalignment tests

94

ll

 

 

:2"?

 

 

 

lin.
U}
'3 o 75 in ‘
is 0.25 in. f‘
‘9 - ' Cracking
a; 3000 ~ 0.5 m.
e /
a x
O l I I I I I I I I I
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure 4.19. Comparison of load vs joint opening curves for 2-dowel Non-uniform

Horizontal Misalignment tests

 

6000

5000*

Load per bar, lbs

 

 

 

 

0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.20. Comparison of load vs joint opening curves for 2-dowel Uniform

Horizontal Misalignment tests

95

I"

 

 

 

 

 

 

 

5000 ~
U)
a 4000 -
l-
n
.D
g. 30m 7 M
.3 2°00 “ 0.5 in.
1000 ,
O I I f I I I fi I I I
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure 4.21. Comparison of load vs joint opening curves for 2-dowel Horizontal
Misalignment tests (One dowel misaligned)

 

6000

l in. Non-Unifofm

   
  

   
 
 
 

0.5 in. Non-Uniform

' l in. Uniform

 

Oin.‘“w-0.5in.Uniform.—A awn. .1: . _ 314mg

 

O I I I I I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure 4.22. Comparison of load vs joint opening curves for 2-dowel Horizontal
Misalignment tests (Non-uniform vs Uniform)

96

 

as

4000 *

Load at 0.25 in., lbs

1111

 

 

 

 

 

 

 

   

 

 

 

 

 

 

o lozsl 0.5 [0.75, 1

Single

[2

 

 

 

 

 

s ‘ s E
s S s g
\ \ \ =
s § \ § =
§ § S s 2-
s .\ \ E
0 l0.25[0.5I0.75| 1 0.5
Non-Uniform All
Misaligned
Two Three
Mkalignment, in.

025' 0.5 0.5

DC
Misal

All

 

Five

Figure 4.23. Comparison of loads at V: in. joint opening (Horizontal Misalignment)
6000

 

Load at 0.5 in., lbs

I]

, ,LA, 7 A

ll

 

 

 

 

 

 

s

///////////1
/////////////////.

 

1
fl

 

 

 

 

 

 

Single

0 Iozslosloasl 1 l 2

 

0 I025. 0.5 Ions] 1

Non-Uniform

Two

  

0.5
All

 

 

Three

Misalignment, 'm.

0.25 0.5 0.5

O-C
Misal

All

 

Five

Figure 4.24. Comparison of loads at V2 in. joint opening (Horizontal Misalignment)

4.3.1.2 Vertical Misalignment

Figures 4.25, 4.26, 4.27, and 4.28 illustrate the comparisons of load vs joint behavior in

the l-dowel, 2—dowel (non-uniform and uniform orientations), and 2-dowel setup with

one dowel misaligned tests, respectively. As seen from the plots, in general, the load per

97

dowel increases with an increase in the misalignment magnitude, but the magnitude of
the increase is lesser than in the horizontal tests.

Table 4.5 summarizes the comparisons of loads at a V: in. joint opening derived
from the bar graphs (Figures 4.30 and 4.31). Similar trends hold good for the loads at a V2
in. joint opening. The format of Table 4.5 is similar to Table 4.4.

Table 4.5. Pair wise comparison of loads at V1 in. joint opening (Vertical

 

 

 

 

 

Misalignment)
Test 1 — load Test 2 - load % change from Test 1 to Test 2
1V12—2700 lbs 1V1 -3100le +15 %
2V12NU — 2300 lbs 2V1NU — 2900 lbs +26 %
2V12U — 2500 lbs 2V1U — 3980 lbs +59 %
2V12U — 2500 lbs 2V12NU — 2300 lbs -8 %

 

 

 

 

 

For a given misalignment type and magnitude, the uniform orientation of misalignment in
the 2-dowel tests exhibit higher loads than the non-uniform counterparts, with some
experimental variability. At a V4 in. joint opening, the trend is valid for both the V2 and 1
in. misalignment series, while at a V2 in. joint opening, the trend reversed in the V2 in.
misalignment test and remained the same in the 1 in. misalignment test (Figures 4.29
through 4.31). The reason why uniform orientation tests exhibit a higher behavior is due
to the physical restraints preventing uplift in the direction of pullout of the slabs,
indirectly leading to a higher locking effect.

Non-uniform joint opening, as in the 2V1NU test, is captured by observing the
variation of the joint opening magnitudes along the joint through plots similar to Figure
4.5. Vertical uplift, however, was not captured through measurement and was visually

observed during the tests (2V12NU, 2V34NU, and 2V1NU tests).

98

 

  
    

 

 

 

6000

5000 r
E. 4000‘ 1- MW’WMWM
g 2in. W
a 3000 7 “n' 0.75 in. ' ¢
'3 ‘W‘m05 in. . M” ‘ ‘ ~. ‘
3 2000 — 0.25m. 0m _ 4. ., .. Inn..- _

1000

0 1 r r r r r r i I
0 0. 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure 4.25. Comparison of load vs joint opening curves for l-dowel Vertical

 

 

 

 

 

Misalignment tests
6000
5000 2
U)
a 4000 —
:37 .
h 3000 2 \ 0.75 in.
8- “. . I _ g. | - . . I 41' , W
'5 M 0'5 m'
3 2000 — ,1
A W 0.25 in. W
{/l - -V v - - ”::0 in Wv
1000 j" r '
O I I I I I 7 I I I I
0 0.1 0 2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Oneninz. in.
Figure 4.26. Comparison of load vs joint opening curves for 2-dowel Non-uniform
misalignment tests

99

Load per bar, lbs

 

 

 

 

 

 

 

Load per bar, lbs

 

 

6000
5000 r
4000 A
1 in. W
3000 ~
2”) — O 5 In. W
1000 ,. ~ - - -~+orn.—w fl _ w : “Wm-w
I
O I I I I I I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure 4.27. Comparison of load vs joint opening curves for 2-dowel Uniform
Vertical Misalignment tests
6000
5000 -
4000 r
3000 " 0.75 in.
2000 2 /
0.5 in.
1000 ~
0 I I f I T I I I I
0 0. 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Joint Opening, in.

Figure 4.28. Comparison of load vs joint opening curves for 2-dowel Vertical
Misalignment Tests (One dowel misaligned)

100

 

 

 

 

 

 

 

5000 —
.8 4000 .
:1 _ lin. Uniform .. J
.8 1 1n. Non-Uniforgl’mw
g 3000 -
.3 j ...a‘ . 0.5 in. Non-UnromMW
3 2000 _ W 0.5 in.Uniform
W
10001 '“v* w ”“Oin. -~-‘ “~::~m=~
0 . fi fi , , , , , ,
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening,in

Figure 4.29. Comparison of load vs joint opening curves for 2-dowel Vertical

Misalignment Tests (Non-uniform vs Uniform)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

    

 

 

 

 

6000
5000 ‘
Si
5'“ 4000 - =
3 3000 . E
3 N ‘ \‘ : g
a mo— i s g g
.3 $§ \ E E t 2
1000 « §§ N g E 2 2
\ § E E x .
\ = = \ \
0 , ,
0 I0.25I0.5I0.75I 1 I 2 §I0:5I 0.5I0.75I1 05 075 0.5 0.25 0.5 0.5
Non-Uniform Uniform One All O-C All
isaligned Misal
Single Two 'Ihree Five
Misalignment, in.

Figure 4.30. Comparison of loads at V4 in. joint opening (Vertical Misalignment)

101

 

Load at 0.5 in., lbs

6000

5000 ‘

2000

a?

as

 

J

 

 

 

ill

 

 

 

////////1
/////////
///////////.

E
s

   

 

o firzsIosImsI 1 I

 

Single

2

0 I0.25I0.5I0.75I 1

Non-Uniform

 

Two

 

Misalignment, in.

 

 

 

 

0.5
All

 

 

Three

0.25 0.5 0.5

O-C
Misal

All

Five

Figure 4.31. Comparison of loads at V2 in. joint opening (Vertical Misalignment)

Table 4.6 summarizes the comparisons of loads at a VI in. joint opening derived

4.3.1.3 Combined Misalignment

hybrid of the behavior exhibited in the horizontal and vertical tests.

in. joint opening. The format of Table 4.6 is similar to Table 4.4.

Figures 4.32, 4.33, 4.34, and 4.35 illustrate the comparisons of load vs joint behavior in
the 1-dowel, 2-dowel (non-uniform and uniform orientations), and 2-dowel setup with
one dowel rrrisaligned tests, respectively. Since combined misalignment constitutes

horizontal as well as vertical misalignment, the load versus joint opening behavior is a

from the bar graphs (figures 4.37 and 4.38). Similar trends hold good for the loads at a V2

Table 4.6. Pairwise comparison of loads at V: in. joint opening (Combined

 

 

 

 

 

 

 

misalignment)
Test 1 - load Test 2 - load % change from Test 1 to Test 2
1C12 — 1900le 1C1 — 3300 lbs +74 %
2C12NU — 2900 lbs 2C1NU - 4600 lbs +59 %
2C12U— 1800 lbs 2C1U-3700 lbs +105 %
2C12U - 1800 lbs 2C12NU - 2900 lbs +61 %

 

 

 

 

102

For a given misalignment type and magnitude, the non-uniform orientation of
misalignment in the 2-dowel tests has higher loads at a given joint opening as compared
with the uniform orientation tests (figures 4.36 through 4.38).

Cracking, as in the tests of 2C34NU and 2C1NU is captured on the load versus
joint opening curve by an instantaneous drop in pressure at the time when the crack
occurs, which was approximately at 0.45, 0.95, and 0.82 in. joint openings, respectively
(Figure 4.33). Spalling also leads to a drop of pressure, but in most of the tests where
spalling occurred, the drop has been observed to have occurred over a significant time
and cannot be captured by mere inspection of the load versus joint opening curve.
However, in the 2C34NU test, spalling induced the occurrence of cracking, and this is
captured by the drop in the pressure at about 0.22 in. joint opening. Non-uniform joint
opening, as in the 2C1U test, is captured by observing the variation of the joint opening

magnitudes along the joint through plots similar to Figure 4.5.

 

 

 

 

 

6000
5000 ~
m
e 4000 .
I:
.8 1m.
3 3000 -
D: .
.3 0.25m. x ...4...
3 2000 4 0.75m .. .. “Mg, g
1000— ' _ W
O I I I I I I I I F fl
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.32. Comparison of load vs joint opening curves for l-dowel Combined
Misalignment tests

103

 

5000‘

.3000,

2000-

Load pe bar, lbs

1000'

0#

 

I

0.1

 

lin. Cracking
\ . 1‘
, I~h*~ *“.~q“

Spalling 0.5 in., p I
0.75 in.
0.25 in.

Cracking\_

Oin. - « :‘fr‘w ” ”WW“ 2....

 

I I T I I

0.2 0.3 0.4 0.5 0.6
Joint Opening, in.

Figure 4.33. Comparison of load vs joint opening curves for 2-dowel Non-uniform

Combined Misalignment tests

 

6000

:2“

Load per bar, lbs

N

 

 

 

 

 

 

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.34. Comparison of load vs joint opening curves for 2-dowel Uniform

Combined Misalignment tests

104

 

5000‘

4000i

Load per bar, lbs
9

LII

.5-

 

 

 

1000 -
O T T I T I f T I f
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.35. Comparison of load vs joint opening curves for 2-dowel Combined
Misalignment (One dowel misaligned)

 

   
  

 

 

 

 

 

6000
Cracking
5000 ~
1 in. Non-Uniform \
E *. 4‘ .
2. “:3 I lin.Uniform J. #50
.8 ' '
g 3000 2 0.5 in. Non-Uniform
'u
3 2000 —
"I ‘0.5 in. Uniform
. . -- M—AWH’WW
1000 01n km ._- __
O I I I I I I I I I T
0 0. l 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.36. Comparison of load vs joint opening curves for 2-dowel Combined
Misalignment (Non-uniform vs Uniform)

105

é

 

Load at 0.25 in., lbs
N I.» -h Ur
§ § § §

2

 

 

 

 

III

C

 

 

 

 

 

 

W//////////////////////

 

0 I0.25I 0.5 I0.75I

 

Single

1

 

 

1

S

s

s

ts ,

0 I0.25I 0.5 I0.75I l 0.5
Non—Uniform Uniform

Two

Misalignment, in.

 

One
Misaligned

 

 

 

Three]

 

 

 

 

 

0.5

O-C
Misal

 

0.5
All

Five

Figubggb 4.37. Comparison of loads at V: in. joint opening (Combined Misalignment)

 

5000*
4000‘
3000-

2000 . ~~ ...

””2“” = II

Load at 0.5 in., lbs

 

 

 

 

 

 

 

 

   

’//////////////////A

S
\
‘ \
i s
s
s s
s
S S
t\ b

 

7/////////////////////////

 

0 I0.25I 0.5 I0.75I

 

Single

1

 

0 I0.25I 0.5 I0.75I

Non-Uniform

b

Two

Misalignment, in.
Figure 4.38. Comparison of loads at V2 in. joint opening (Combined Misalignment)

106

 

 

 

All

 

 

Three

 

 

 

 

 

0.5
O-C

 

Misal

0.5
All

Five

 

 

4.3.2 Comparison across different number of dowels misaligned
The comparisons covered in this subsection can be further categorized as (x = H, V, or
C):

0 Number of dowels misaligned is different but the total number of dowels remains

the same (2xAM vs 2xNU/U and 5xAM vs 5xNU)

0 Number of dowels in the setup are different (1x vs 2xNU/U and 3xNU vs 5xNU)
4.3.2.1 Comparisons within same setup with different number of dowels misaligned
The first comparison, illustrated in figures 4.39 through 4.44, compares the 2-dowel tests
with one dowel misaligned (2xAM) against both misaligned (2xNU/U). In the V2 in.
misalignment tests (horizontal, vertical, and combined), the 2AM tests lie below the 2NU
tests, but above the 2U tests, with intersections in the lower joint openings due to
experimental variability. In the 3A in. misalignment tests, only the 2NU series tests were

carried out and the 2AM tests lie below the 2NU tests (figures 4.42 through 4.44).

 

 

 

 

 

 

 

6000

5000 ‘
3 4000 ~
5
i 3000 - 2H12NU "fvfl
8.
.2 WM
3 2°00 ‘ 2H12AM

2H12U wag .4; .. _ #3le
1000 r 4m
0 I I I I I? T I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.39. Comparison of load vs joint opening curves of 2-dowel V2 in. Horizontal
Misalignment (2H12NU, 2H12U, and 2H12AM)

107

 

 

 

 

 
 

 

 

 

5000 ,
§ 4000 .
I:
a
.D
8. 3000 I 2V12NU
,, dkkw
a 2000 . 2V12AM
.3 2V12U w
1000 ~
0 I I I I I I I I I I
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.40. Comparison of load vs joint opening curves of 2-dowel V2 in. Vertical

Misalignment (2V12NU, 2V12U, and 2V12AM)
6000

 

 

 

 

 

 

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.41. Comparison of load vs joint opening curves of 2-dowel V2 in. Combined
Misalignment (2C12NU, 2C12U, and 2C12AM)

108

 

 

 

 

6000

5000 -
§ 4000 ~
g“ 2H34NU
.n
'- 3000 r
8.
'8

1000 -

0 I I I 1 I I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.42. Comparison of load vs joint opening curves of 2-dowel 3%: in. Horizontal
Misalignment (2H34NU and 2H34AM)

 

 

 

 

6000
5000 —
E 4000 ~
I:
a
.D
'- 3000 2 - . WW
8. 2V34NU
'3 2V34AMFYWWW
3 2000 -
1000 -
O T I I I I I I I I
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.43. Comparison of load vs joint opening curves of 2-dowel % in. Vertical
Misalignment (2V34NU and 2V34AM)

109

 

E
1

 

5. a?

Load per bar, lbs

2%:

2C34NU
2C34AM

0 g I I I I I I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

 

 

 

Joint Opening, in.

Figure 4.44. Comparison of load vs joint opening curves of 2-dowel % in. Combined
Misalignment (2C34NU and 2C34AM)

There is just one misalignment magnitude (V2 in.) available for comparison of trends
across different number of dowels misaligned in the 5-dowel tests (5x12NU vs 5x12AM)
(figures 4.45 through 4.47). In the horizontal misalignment tests, the outer and center
dowels misaligned (AM) test has higher loads at a given joint opening than the all dowels
misaligned tests (Figure 4.45), but in the vertical and combined misalignment tests it is
vice versa (figures 4.46 and 4.47). Also the SNU tests were more destructive than the
SAM tests.

In general, it can be concluded that for a given misalignment magnitude, the load

induced per dowel increases as the number of dowels misaligned increases.

110

6000

 

5000 7

5H12AM
3000 ‘

2000 ‘ 5H12NU

1000—/

0 I I I I I r n I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Load per bar, lbs

 

 

 

Figure 4.45. Comparison of load vs joint opening curves of 5H12NU and 5H12AM
6000

 

5000“

i

5V12NU
Cracking /

5V12AM WW-

Load per bar, lbs
U.)

i

1000 ‘

 

 

 

O I I I I I I l I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.46. Comparison of load vs joint opening curves of 5V12NU and 5V12AM

111

6000

 

 
   
 

5000 t
. 5C12NU
3 mm
“a f Spalling /
a
f M 5C12AM
3.
g Crackin/
.3

 

oéééé

 

 

I I I I I I I I I

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.
Figure 4.47. Comparison of load vs joint opening curves of 5C12NU and 5C12AM

O

4.3.2.2 Comparison across different number of dowels setups

This section presents a pair wise comparison of 1. vs 2-dowel tests and 3 vs 5-dowel tests.
A comparison across the two pairs will not be valid because the specimen molds, and
hence the volume of concrete in the two setups, is different. The 2-dowel (5-dowel)
specimens underwent more significant distress than the l-dowel (3-dowel) tests because
the total applied load (load per dowel x number of dowels) is higher.

Horizontal Misalignment: In the V2 and 1 in. misalignment tests, the 1H lies between the
2HNU and 2HU tests, with the 2HNU lying above the 1H test (figures 4.48 and 4.49). As
illustrated in Figure 4.50, the load per bar at a given joint opening increases with an

increase in the number of bars from three to five.

112

6000

 

 

 

 

5000 ‘
E 4000 -
I:
a
Q
h 3000 ‘ 2H12NU
3.
'U
a
3 2000 ‘ 1H12 WWW
WW4
1000 g 2H12U
O I r r I I I F r I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure 4.48. Comparison of load vs joint opening curves of 1H12 and 2-dowel V2 in.
Horizontal Misalignment (2H12NU, 2H12U)

 

Load per bar, lbs

 

 

O I I I I I

 

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure 4.49. Comparison of load vs joint opening curves of 1H1 and 2-dowel 1 in.
Horizontal Misalignment (2H1NU, 2H1U)

113

 

 

§§ s 3 8

 

 

E

I:

a

.1:

i-

3. SHIZWMWW
“g 3H12NU

.3

1000 4
O I I j I I I I I I

 

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure 4.50. Comparison of load vs joint opening curves of 3H12NU and 5H12NU

Vertical Misalignment: In the V2 in. misalignment tests, the 1V12 test lies between the
2V12NU and 2V12U tests, with the 2V12NU test lying above the 1V12 test (Figure
4.51). However, in the 1 in. misalignment tests, the 1V1 test lies below both the
2V1Uand 2V1NU tests, with the 2V1U test lying above the 2V1NU test (Figure 4.52).
As illustrated in Figure 4.53, the load per bar at a given joint opening increases with an

increase in the number of bars from three to five.

114

 

 

 

 

 

6000
5000 ~
E 4000 ~
5
a
I- 3000 ~
3- ZVIZWWHW—Iw
‘ I ‘
E 2000 — WIZMWWWWI
2V12U mm...
1000 ~
0 I I T I I I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.51. Comparison of load vs joint opening curves of 1V12 and 2-dowel V2 in.
Vertical Misalignment (2V12NU, 2V12U)

6000

 

Load per bar, lbs

 

 

 

 

0 I I I l I T I I I F
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.52. Comparison of load vs joint opening curves of 1V1 and 2-d0wel 1 in.
Vertical Misalignment (2V1NU, 2V1U)

115

 

é

3,”

3?

5V12NU
3V12NU

i.

Cracking

Load per bar, lbs

2000+

1000*

 

 

 

0 j I I I I I I I I

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.53. Comparison of load vs joint opening curves of 3V12NU and 5V12NU
Combined Misalignment: In the V2 in. misalignment tests, the 1C12 test lies between the
2C12NU and 2C12U tests, with the 2C12NU test lying above the 1C12 test (Figure 4.54).
However, in the l in. misalignment tests, the 1C1 test lies below both the 2C1U and

2C1NU tests, with the 2C1NU test lying above the 2ClU test (Figure 4.55).

 

 

 

 

 

 

 

6000

5000 I W
3 4000 - 2C12NU
I:
a
.n 3000
h ..a
8.
.3 WW

1000 ‘ 2C12U *‘ ~ A.—..

0 I I I I I I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.54. Comparison of load vs joint opening curves of 1C12 and 2-dowel V2 in.
Combined Misalignment (2C12NU, 2C12U)

116

 

 

 

 

6000
Cracking
5000 7
2C1NU

w W
e 4000 ~
g M“ ZCIUW
‘- 3000 7 "
8. f 1C1
,3 W
3 2000 7

1000 ‘

0 I I I I I T I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure 4.55. Comparison of load vs joint opening curves of 1C1 and 2-dowel l in.
Combined Misalignment (2C1NU, 2C1U)

 

        
 

 

 

 

6000

5000 7
3 W 5C12NU
'2 3C12NU
g 3000 d Spalling
8. /
'3 Cracking
,3 2000 7

1000 7

O I I I I I I I I 7
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure 4.56. Comparison of load vs joint opening curves of 3C12NU and 5C12NU
A comparison of the loads at V4 and V2 in. joint openings for different misalignment
scenario has been presented in figures 4.57 and 4.58. In general, these graphs substantiate

the fact that for a given misalignment magnitude, the trend of load versus misalignment

117

type (horizontal, vertical and combined) is inconclusive. However, the following trends
of loads at V4 and V2 in. joint openings have been observed:

0 2-dowel non-uniform orientation: V < H < C

0 2-dowel uniform orientation: H < C < V

o 3 and 5-dowel non-uniform orientation (all misaligned): H << V< C
In the l-dowel misalignment and the alternate dowels misalignment series, the trend is
inconclusive. The trend of load vs misalignment type, for the 3 and 5-dowel non—uniform
misalignment series presented before is further supported by figures 4.59 and 4.60.

It has to be to noted that two tests may have the same load at a V4 in. joint opening
but might exhibit totally different behavior at a later joint opening because the loads are
equal but not the stress states at the dowel-concrete interface. For example, the 1H34 and
2H34NU curve have similar loads at V: in. joint opening but the load vs joint opening
curve for the 1H34 test flattened and formed a plateau at later stages and no distress
whatsoever was observed while in 2H34NU the stresses due to lockup caused the curve
to follow an increasing trend and lead to a sudden drop at a joint opening of 0.72 in. as a
result of cracking. Thus, these bar graphs are useful to get an overview of the trend of the
load vs misalignment features but concluding purely based on these graphs is not
warranted. Before making any conclusions, the trends have to be analyzed in conjunction

of the corresponding load vs joint opening curves.

118

Eofiuwm—ammfi we 88“ 3.9.0.55 mucous .5 S 3 meme. he comings—co Sm... Pqu

.3 453332

9E BE 039 swam

an: 3922
3 0-0 .2 Bo gnome: 533152

We md mad md mud no ~ md fl mhd m6 mad o N H mud m6 mmd o

U>EU>I>IU>EO>EU>EU>=U>EU> U>=U>EU>E >EU>EU>=U>EU>E

cog

ooom

95m

ooov

ooow

 

coco

5‘“ “u; SZ'O 19 PBO'I

119

“5:53—33. he 33:58.51 39.3 .5 «\a “a meme. he acme—«Econ. .wm.v 0.53%

.3 2533.32

95 85 can swam

an: Benin:
as. 0.0 =< Bo Eon: 83:52

m6 md mad We mud md _ m6 d mud md mmd o N fl mud md mmd o

U>EU>=>EU>EU>EU>=U>EU>EU> U>EU> U>m >EU>EU>EU>EU>E

o

82
m.

ooom B
p
B
1?

can 0
ms.
m.

ooov ,.»I
q
S

ooow

 

oooe

120

6000

 

5000 7

4000 I M
3C12NU
W

'- 3000 7 3V12NU

3H12NUWW

Load pe bar, lbs

2000 7

 

 

 

0 I T I I I I I I I

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Joint Opening, in.

Figure 4.59. Comparison of load vs joint opening curves (3H12NU, 3V12NU, and

 

 

 

 

3C12NU)
6000
5000 7
5C12NU W
§; 4000 - 44-" 5V12NUMWM
I- .. -~
.8 W
I. 3000 ~ ,
8. 5H12NU
lg W
3 2000 w
1000 7
0 I I I I I I H I I
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Joint Opening, in.
Figure 4.60. Comparison of load vs joint opening curves (5H12NU, 5V12NU, and
5C12NU)
4.4 LOAD ZONE PREDICTION

Based on the load vs joint opening behavior data from the l-dowel and 2-dowel tests,
regions were developed, where the load vs joint opening curve would be expected to lie,

given the number of bars and misalignment orientation (figures 4.61 and 4.62). The

121

misalignment magnitudes covered under the regions vary from O or aligned to 1 in.
misalignment, over half length of the bar i.e. 9 in. Since the load per bar curves for
different misalignment types lie close to each other the region has been generalized to all
misalignment types. The range or extent of the load regions in the 1 and 2—dowel tests
were used to develop similar regions for the 3 and S-dowel tests. It has to be noted that
Figure 4.61 has been developed for small slab specimens (2’ X 4’) and Figure 4.62 has
been developed for big slab specimens (3’ X 8’). However, it has to be noted that these

regions are applicable to similar volumes of concrete as in the lab setup only.

 

     
   

       
   

 
 
    
 

 
 
   
    
       

     

      
  

     
 

        

 

   

 
 

 

  

 

 

6000
5000 7
4000- git," I} ' »
é “UV/j 2 bar
= :' ,1 '_. - , .
..a 3000 7
g.
g I.
E200“ _
. ' -°"‘.3.‘ if“: fifti‘iix \{éi'xi‘gizé’fis‘o‘fi’z‘é‘gx‘iiyi‘ :%S§;:§“‘§§\§'\\Q$‘t‘~ ’ ‘ ‘
1000 ”\N‘Wmmmss%\&
s“ lbar “‘ “f" "
o a . . . .
0 0.1 0.2 0.3 0.4 0.5
Joint Opening, in.

Figure 4.61. Regions for 1 and 2-dowel systems

122

 

 

 

6000-—~ _,_ -*- 7— — —‘ —- —— —— —— w
5000—
34000-
=’
.9
33000.
n.
'u
a
32000.
1000.
0 I I I I I I I I I I
0 0.1 0.2 0.3 0.4 0.5
Joint Opening, in.

Figure 4.62. Regions for 3 and S-dowel systems

4.5 EFFECT ON MISALIGNMENT ON PAVEMENT PERFORMANCE

The next step is to study the potential effect of the different distresses observed in the lab
due to misalignment on pavement performance. Vertical uplift in itself as in the current
experimental setup did not cause any destruction. But it could potentially lead to faulting
and or cracking when traffic load acts on the slab. Non-uniform joint opening in the
current setup did not lead to any destruction because the longitudinal edges of the slab
had no restraint. In the lab, non-uniform joint opening was observed in the uniform
(2H1U and 2C1U) as well as non-uniform orientation (2V1NU, 5H12NU, and 5V12NU)
tests. In case of non-uniform orientation it was accompanied by other distresses also.
However, in field conditions adjacent lanes and shoulder could cause increase in the
concrete stresses and cause cracking. Poblete et.al. (1988) conducted a field study on
thermal deformations of undoweled slabs and they found that in the presence of positive

gradients higher than the “built-in curl”, the slabs are normally expanded and since

123

rotation at their transverse edges is restricted, a cylindrical shape around the longitudinal
axis is imposed on the slab with its longitudinal edges supported. Thus, misalignment
conditions that could lead to non-uniform joint opening could possibly cause cylindrical
shape of the slab due to gradients and restraints and hence lead to lack of or partial
support in the central zone of the slab and possibly cause cracking.

These possible effects cannot be verified as there is no structural model that
accounts for the change in bearing stress due to dowel misalignment. Relating spalling at
the dowel-concrete interface to pavement performance was attempted as described in the
following paragraphs.

Spalling at the dowel-concrete interface could possibly lead to dowel looseness
which is defined below. Dowel looseness is said to occur when voids develop underneath
the dowel bar when the concrete near the dowel has got crushed and these crushed
concrete particles are removed. Dowel looseness is also possible if the dowel-concrete
bearing stress is very high (FHWA 2005). Dowel looseness can arise from poor
construction techniques as well as damage to the surrounding concrete under cyclic
loading (Davids 1998).

EverFE2.24 developed by Davids (1998) was used to study the theoretical effect
of dowel looseness on different pavement responses. Dowel looseness as defined in
EverFE2.24 is illustrated in Figure 4.63. The assumed gap 7 tapers parabolically from a

maximum value to zero over length L of the embedded dowel.

124

Slab
Y Y
F I I
_ Dowel
N0 gap L No gap Boundary
of
Concrete
|

Figure 4.63. Dowel Looseness

 

A sensitivity analysis was performed on a 2-slab system with 15 ft joint spacing for
different levels of looseness, the inputs of which are summarized in Table 4.7. L was
maintained at 4 in. in all the cases. The loading configuration is shown in . The thermal

gradients were based on the data for Detroit obtained using the Enhanced Integrated

 

 

 

Climatic Model (EICM).
Table 4.7. Input Parameters for Sensitivity Analysis
Parameter Values considered
Slab thickness and dowel diameter 10 in., 1.25 in. spaced 12 in. on center
System 0 Slab on 6 in. granular base and 10

in. sand subbase (elastic foundation)
0 Slab on subgrade with effective k
(dense liquid foundation)

 

 

 

Y 0, 0.0025, 0.005, 0.01. 0.02 in.
Thermal gradient (over 10 in.) 22.3 °F, 0.1 °F, -6 °F, 93 °F
Analysis parameters 0 Vertical deflection

 

 

 

0 Load transfer efficiency

 

125

 

l5fi

Figure 4.64. Screenshot of the Traffic loading location in EverFE

The relative vertical displacements of the two slabs grew with an increase in y. The
deflections on the loaded slab increased with an accompanying decrease in the unloaded
slab deflections as the gap V was increased. The load transfer efficiency (LTE) at the two
wheel locations of the axle closer to the joint was studied and the effect of y on loss of
load transfer at these two locations is illustrated in figures 4.64 and 4.65. Figure 4.66
corresponds to the wheel location closer to the slab edge. In general, load transfer
efficiency decreases with increasing dowel looseness up to a threshold value of the gap
after which it tends to flatten out. The loss in LTE is more evident in the negative
gradient cases. For a negative gradient of —9.3 °F, a dowel looseness of 0.0025 in. leads to
a loss of LTE of nearly 30% when the slab is placed on an elastic foundation while the
same gap leads to a loss of LTE of nearly 50% when the slab is placed directly on a dense
liquid foundation. At points closer to the edge, the loss in LTE was higher and in some
cases, there was no load transfer. For slabs on elastic foundation with positive gradient,

there was just about 5% loss in load transfer. A smaller gap is required in dense liquid

126

foundation systems to produce the maximum decrease in LTE. Beyond a gap of 0.005 in.,
the magnitude of dowel looseness does not cause any further impact on the deflections or
load transfer.

The difference in behavior between the elastic and dense liquid foundation is due
to the different subgrade models used in the software. In the elastic layer model, as the
loaded slab contacts the subgrade and displaces it downward, there is a loss of support
under the unloaded slab. With a dense liquid directly below the slab, the subgrade under
the unloaded slab provides continuous support that is independent of the displacement of

the loaded slab, tending to increase vertical displacements between the two slabs.

127

~08 80053 me ~ .355 .32. PHI— : $2.88. .953. «a Seam .34. 953.0

x 930:0 as, Deanna .5 22m I m5 5 a3 A3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.80583 .9399
8.0 20.0 00.0 n00.0 0
_ . _ 0

I 0_

I on

"r man u I on

"no 0.0 II/ I 0v

me o. I on

ma m a- I 00

I on

I 0.”

I 00

000

8893 8.3:” I EA E 33 30
A: 98:88. .225
00.0 300.0 80.0 300.0 0

. . W _ llilllllllli o
"wommm - I 2

I? I. III. 0 a

a. ma- I on

I 0v

I 0m

I 00

I 0“.

I 0w

I 00
000

% 51.1.1 “I ssoq

%‘GI.L'I“!SSVI

x 238% as, Deanna so 9% I E: Q

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.86582 .35:
8.0 20.0 _0.0 m00.0 0

p _ p O
I 0—

I cm

I on

"no ma- 1 0.?

"mo 0.. \IIIIIIIIO/ I on

ma 0.0 TIIIIIIIIII / I 00

mo QNN u I 0h

I cm

I 00
000

Sam? Loam—Im I mi...— 03
.5 .8658:— .225
8.0 200.0 30.0 300.0 0

_ k _ o
I 0_

I on

I 0m

I 0v

I on

2- - 8

no I 0b

are. I 0 m

0.0

mam. I 00
mo 81. I II!!! 8_

% ‘EIL'I

% m1

128

2.8 82.5. a: a .855 .8: 0.5 .5 8288. .26.. a .85 .8... 2.5..—

e. 280:0 5? 833m 8 92m I m5 5 9.3 .3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.5 .8588. .269
8.0 0.0.0 .0.0 30.0 0

p h p o

I 0.

I 0N

.m I

n... m N \\ I on

I 0..

mo ..0 IIIIIIIIIIIIIIIII I on

"re. I I 00

n... Q. I E

I ow

I 00
I 00.

5293 .86.-.“ I E9 5 $09 30
.5 £92.88. .369
.00 900.0 30.0 300.0 0
p b . O
QNN n IIIIIIIIIIIIIII

me I 0.

mo ..0 I 0N

"H09 W on

I n... 2- I 9.

I 0m

I 00

I 0..

I 00

I 00
00.

 

 

 

 

% ‘EIL'I‘IISSO'I

%‘EI.L’I“I$°'I

x 268% as, 029:... 8 new I m5 .3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.5 .8588. .269
8.0 0.0.0 .0.0 m00.0 0
p p _ C
no m.0-
n... 0- 1
... s m
%
n... m.m~
539$ .66.-.“ I PHI. .3
5 £85.86. .369
.00 0000.0 n00.0 300.0 0
F _ _ O
I 0.
I 0N
I on
I 0.» 1
I a m
"mo man I 8 %
n... 0. I 0..
II... ..0 I 00
I 00
m.m~ I I
H... II/ 85

 

 

 

 

129

CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS

5.1 CONCLUSIONS

To gain an insight into the effects of dowel misalignment on joint opening behavior, the
objectives of this thesis were proposed as: to investigate experimentally the fundamental
pullout behavior of misaligned dowel bars and gap-opening behavior of pavement joints
with misaligned dowel bars in plain concrete pavements, under thermal expansion. It was
initially hypothesized that the misalignment type and magnitude would affect the initial
slip/debonding stress, type of joint and slab distresses observed, and post-slip behavior.

To systematically accomplish these objectives and test the hypotheses, an
experimental plan was developed that focused on the following variables/misalignment
features: number of dowel bars (1, 2, 3, or 5) and number misaligned (all or alternate),
misalignment type (horizontal, vertical, or combined), orientation (uniform or non-
uniform), and magnitude (0, 1A , l/2 , 3%: , 1, or 2 in.). Pullout tests simulating thermal
expansion were carried out on all the tests listed in the experimental plan.

The experimental investigations were conducted on laboratory-scale pavement
specimens with doweled contraction joints. Each pavement specimen consisted of (a) two
48 x 24 x 10 in. concrete slabs connected at the joint using 1 or 2 steel dowel bars or (b)
two 96 x 36 x 10 in. concrete slabs connected at the joint using 3 or 5 steel dowel bars.
Steel dowel bars that are 1.25 in. in diameter and 18 in. long were placed at the mid-depth
at the joint, with equal lengths (9 in.) embedded in each concrete slab, and placed 12 in.
on center. Each concrete slab was supported on smoothened rollers placed on hardened
steel plates. The joint between the concrete slabs was formed using an ‘AI in. thick

aluminum plate with circular holes of appropriate diameter to pass the dowel bars

130

through it. After 7 days, the pavement specimen was tested by pushing apart the two
concrete slabs using hydraulic actuators simulating thermal expansion at the approximate
load rate of 201bs/min. followed by the opening displacement rate of approximately 0.02
in./min.

The concrete slabs were cast in-place directly in the molds. The concrete mixture
was MDOT pavement mix grade P1 and the dowel bars were made of billet steel grade
40. After rough misalignment using a tape measure, each dowel was welded on one side
and greased on the other half as per MDOT specification R-40-E. The dowel
misalignments were adjusted and measured accurately (before placing the concrete),
using total electronic stations (theodolites) and surveying techniques. The accuracy of the
system was 0.005 rad.

Sliders and linear variable displacement transducers (LVDTs) were used to
measure the transverse Opening of the joint. The data from the different instrumentation
was collected at the rate of 6 scans per second, using a data acquisition system.

The experimental setup/test procedure developed for the pullout of the slabs
overcame most of the limitations of the previous dowel misalignment experimental
studies. This thesis focused on the following additional details, in addition to, the basic
pullout behavior studied in previous studies: bond stress behavior of dowel-concrete
during pullout, relative effect of the misalignment orientation on load vs joint opening
behavior, and tests on slabs accommodating five dowel bars. The high rate of data
collection enabled better understanding of the pullout behavior, especially at smaller joint
openings where debonding takes place between the dowel bar and concrete and the slope

of the load vs joint opening curve is very high.

131

 

As a word of caution, all conclusions presented herewith have to be interpreted keeping
in mind experimental variability and are valid for conditions of thermal expansion only.
Bond versus joint opening behavior. The bond stress versus joint opening curve has two
distinct regions: (a) fully bonded region and (b) post-slip/debonded region. The first
region is the high initial slope region where there is bond between the dowel bar and
concrete. The applied pullout force which can be resolved in the axial and normal
directions of the dowel bar is opposed only by the dowel-concrete bond force. Slip
occurred after the bond stress per dowel increased beyond a certain threshold value which
was found to be in the range 10-50 psi in the gamut of tests conducted. As further thermal
expansion is induced, the forces induced in the dowel bar increase and depending on the
misalignment, the induced force could form a plateau or increase causing failure of the
test specimen due to cracking. Due to the presence of misalignment and joint opening,
normal hearing stress in concrete is introduced and due to the frictional effect across the
interface, shear stresses are developed.

Visual observations: The different joint and slab distresses observed during the tests
include spalling of concrete near the dowel-concrete interface, cracking, non-uniform
joint opening, and vertical uplift. The first three distresses are destructive even without
traffic and/or restraints from the adjacent slabs and base while non-uniform joint opening
and vertical uplift in the current setup of just thermal expansion are not destructive.
Spalling was always observed only at the joint face of the slab containing the greased

side of the dowel as this side slips out during the pullout test.

132

'.

 

 

Load versus joint opening behavior: The key conclusions obtained in context of the load

vs joint opening behavior are described below:

Within a misalignment type, the load the load required per bar and the intensity of
distresses increased with an increase in the misalignment magnitude.

Degree of locking is higher in the non-uniform orientation of dowel bars and hence
non-uniform misalignment tests were more destructive.

For a given misalignment magnitude, the load required per bar and the intensity of
distresses increased as the number of dowels misaligned increased.

For a given misalignment magnitude, the overall trend of load versus type of
misalignment (horizontal, vertical and combined) is unclear. However, the loads at M:
and V2 in. joint openings followed typical trends in the 2, 3, and 5-dowel all
misaligned tests. In the 2-dowel non-uniform and uniform orientation tests, the trend
was V < H < C and H < C < V, respectively. In the 3 and 5-dowel non-uniform
orientation series an H << V< C trend was observed. In the l-dowel misalignment

and the alternate dowels misalignment series, the trend is inconclusive.

The conclusions, visual observations as well as load versus joint opening behavior, are

explained in context of the each misalignment type in the following paragraphs.

Horizontal Misalignment

In the single dowel tests, spalling at the dowel-concrete interface was observed in the
2 in. test only. In the 2, 3, and 5-dowel tests, different combinations resulted in
different distresses: spalling at the dowel-concrete interface, non-uniform joint
opening, and cracking. Spalling was the predominant distress in these tests with spalls

ranging from 2 in. by 2 in. to more than half-depth of the slab.

133

 

Cracking occurred in tests with higher misalignment magnitudes and of the non-
uniform orientation (2H34NU and 2H1NU) at high joint openings (equal to or greater
than 0.5 in.).

Non-uniform joint opening was observed in uniform (2H1U) as well as non-uniform
misalignment (5H12NU) tests.

In the 5-dowel all as well as alternate misalignment tests, spalling was found near the
outer and center dowels.

Non-uniform orientation of misalignment of two dowel bars required more load per
dowel bar at a given joint opening as compared to the uniform orientation of dowels

due to higher degree of locking.

Vertical Misalignment

In the single dowel tests, no distresses were observed. In the 2, 3, and 5-dowel tests,
different combinations resulted in different distresses: spalling at the dowel-concrete
interface, non-uniform joint opening, vertical uplift, and cracking.

Vertical uplift occurred in the tests with non-uniform orientation of dowels (2V12NU,
2V34NU, and 2V1NU).

Non-uniform joint opening was observed in tests with non-uniform orientation
(2V12NU and 5H12NU).

In the 5-dowel all as well as alternate misalignment tests, spalling was found near the
outer and center dowels.

Uniform orientation of misalignment of two dowel bars required more load per dowel
bar at a given joint opening as compared to the non-uniform orientation of dowels. In

the test setup, there is more restraint against vertical uplift than horizontal rotation

134

 

and hence, in vertical misalignment tests the uniform orientation exhibited higher
load vs joint opening behavior.

Combined Misalignment

0 In the single dowel tests, spalling at the dowel-concrete interface was observed in the
1 in. test only. In the 2, 3, and 5-dowel tests, different combinations resulted in
different distresses: spalling at the dowel-concrete interface, non—uniform joint
opening, and cracking. Spalling was the predominant distress in these tests with spalls
ranging from 2 in. by 2 in. to more than half-depth of the slab.

o Cracking occurred in tests with higher misalignment magnitudes and of the non-

 

uniform orientation (2C34NU, 2C1NU, and 5C12NU) at high joint openings (equal to
or greater than 0.5 in.).

o Non-uniform joint opening was observed in only one test (2C1U).

o In the 5-dowel all as well as alternate misalignment tests, spalling was found near the
outer and center dowels.

o Non-uniform orientation of misalignment of two dowel bars required more load per
dowel bar at a given joint opening as compared to the uniform orientation of dowels
due to higher degree of locking.

Based on the load vs joint opening behavior data from the l-dowel and 2-dowel tests,

regions were developed, where the load vs joint opening curve would be expected to lie,

given the number of bars and misalignment magnitude, under the lab conditions.
Spalling which was the main form of structural distress observed in the lab could
possibly cause dowel looseness. A sensitivity analysis using EverFE2.24 revealed that

looseness leads to higher relative vertical displacements between the loaded and the

135

unloaded slabs and thus losses in load transfer efficiency. When subjected to a negative
gradient of -9.3 °F, looseness of 0.0025 in. causes a loss of LTE of nearly 30%.

One of the significant contributions of this study apart from the conclusions stated
above includes the load vs joint opening behavior from the single dowel bar tests. The
bond shear stress-slip strain responses for the straight specimens (with zero
misalignment) provide information regarding the overall longitudinal bond interaction
between straight (aligned) dowels and the surrounding concrete and are critical in the
analytical modeling of the dowel-concrete behavior. The bond shear stress-slip strain
responses for the remaining specimens provide information regarding the longitudinal
bond between the misaligned dowel and the surrounding concrete, while including the

effects of friction and bearing stresses (from dowel misalignment).

5.2 RECOMMENDATIONS FOR FUTURE RESEARCH

To arrive at a definitive conclusion, tests with traffic as well as thermal loading have to
be carried out and these were beyond the scope of this thesis. However, recommendations
regarding the effects of different misalignment features on joint opening behavior are
arrived at.

Each of the misalignment type, magnitude, and orientation would cause a specific
certain stress state zone at the interface along the length of the bar and at the joint, which
cannot be studied directly through lab observations/tests. The experimental results do not
provide comprehensive knowledge of the mechanics of dowel-concrete interaction
including: (a) the particular shear and normal stresses induced in the surrounding
concrete by the dowel bar, (b) the effects of dowel misalignment on the stress-states

induced in the concrete, and (c) the structural distresses produced by excessive stresses

136

 

and stress concentrations. This knowledge is virtually impossible to obtain
experimentally. Hence, analytical investigations based on the experimental results
presented in this thesis, are needed to develop better understanding of the basic dowel-
concrete interaction mechanics. Analytical studies on these lines have been initiated in

the later phases of the MDOT project study.

137

.._..z

 

...-n- 4lfi"

 

APPENDIX A ‘
DETAILS OF EXPERINIENTAL SETUP AND COMPONENTS

138

3:8 sup (mam plate)

 

to
J.

‘W

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

C10x153
10 o
" (ia
A325 Bolts
In PL 112' lhk
0.00 00
Solid Rollers
mn-
Figure A-l. Cross section of the small mold for pullout test E
118
I*—-——i’l50 6.5—Lg? 1750
4: 4* 4* ¢ 4} ¢ + if 4*
fl— +: '+ If.
(I I
+ +
41> +
13
4b +
48.00 4} 4r
4} +
13
I} <9
{I
l I
I
.I'_ . .. filmf—iv
+ II + 4% 41> 4* +
53.34

 

 

 

 

Figure A-2. Plan of the small mold with outside dimensions

139

 

 

 

bolts

 

. . ClOX 15.3
'/21n.dra

 

 

l I—_—Q_——. Q U Q:— A
Plate V’ m' “““k Solid Rollers 2 III. dia

Figure A-3. Typical sectional view of the big mold

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

77.3350
0 9 O 9 0 0 ‘0 O O O 9 0 O 0
u n
I
° L ‘3 °
I: F a
u o
a n
a a
o o
3 0
101.2000 - -
9 O
O 0
0 0
0 0
0 0
o 0
°IL———36. 3&0000———-I*°
'1‘ 1/8
0 0
o o n o a o I a 0 II n o n o n

 

Figure A-4. Typical plan view of mold with 5 dowel bars

140

Table A-1. Mix Design Specifications

Specification: 2003 Standard Specifications
Grade of Concrete: P1
Intended Use of Concrete: Pavement Form

 

 

 

 

 

 

 

 

 

 

Material Class Specific Gravity Absorption %
Cement I/IA 3.13
Fine Aggregate 2N8 2.61 1.47
‘ Coarse Aggegate 6AA 2.66 1.50
Cement content : 564 lb/yd3 B/Bo
Air Content (design): 6.5 % Specification Tolerance
(specified): 6.5 %
R.W.C : 1.15 Theoretical Yield

Fly Ash Content, lb/yd3 : o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Weight of Coarse Aggregate and Water Proportions
Aggregate Quantities, lblyd3 of concrete

(Dirk/as 86) Fine Aggregate Coarse Aggregate Total
(Oven Dry) (Oven Dry) Water

88 1330 1711 233

39 1313 1730 287

90 1297 1750 286

91 1280 1769 286

92 1264 1788 285

93 1247 1808 284

94 1231 1827 283

95 1214 1847 282

96 l 198 1866 281

97 1181 1886 280

93 l 164 1905 279

 

 

 

Typical Unit Weight (dry, loose) of coarse aggregate as described above is 93 lb/ft3

141

:O.72
: 1.5 %

: 100.00%

 

Table A-2. Description of Linear Variable Differential Transducer

 

 

 

 

 

 

Model LD610-15

Excitation Voltage 2 15 V@ 18 mA maximum
Output i 10 V DC

Stroke 2 15 mm

Total Stroke 30 mm (1.18 in.)

 

 

 

Table A-3. Description of the Hydraulics (Actuators RC 156 and RC 256)

 

 

 

 

 

 

 

 

 

 

Name Capacity Stroke Cylinder Effective Area Weight
(ton) (in.) (in?) (lbs)
Actuator RC 156 15 6 3.14 15
Actuator RC 256 25 6 5.16 22
Calibration Procedure

A. Sliders and Linear Variable Differential Transducers
The calibration procedure of the LVDTs and sliders is described below:

- Precalibrated steel pieces of thicknesses 0.1 in. through 1.5 in. and 0.1 in. through
1.0 in. were used, for the 1.0 in. stroke and 1.5 in. stroke sliders, respectively.

0 The sliders/LVDTS were excited at their respective excitation voltages (+5 V DC
for sliders and t 15 V DC for LVDTs) and the output voltage was recorded at
every 0.1 in. increment using a multimeter.

0 Graphs were plotted for the transducer stroke displacements (measured in inches)
versus the output voltages.

o The calibration graphs for the sliders and LVDTs are presented in figures A-5
through A-13, respectively. The graphs were found to be linear in all cases.

B. Calibration of Hydraulics and Pressure Transducers
The calibration procedure of the actuators and pressure transducers are described below:
0 For calibration of the actuators and pressure transducers, a MTS pre-calibrated

machine in the lab was used to measure the load from the actuators.

142

The actuators were placed one at a time between the two compression plates of
the machine. A 10 V excitation DC voltage was applied to the pressure transducer
to obtain the output between 0.5 V to 5 V.

On application of pressure from the hand pump, the MTS machine readout gave
the amount of load that the actuator was exerting on the plates.

This readout at every 1000 psi load step was noted along with the voltage from
the transducer.

Graphs showing the load from the MTS vs. voltage were plotted and are shown in
figures A-l4 through A-17.

A trendline superimposed on this graph showed that the relation of load vs. output
voltage was linear and this calibrated curve was input in the data acquisition

software.

143

“E". "mwmfl

3558 .5 3 s 5585.8 .53 2:5...

 

 

 

 

 

 

 

 

 

> .853»
v m N _ o
\\
38.0 + >83 u e \\
255... .5 m4 5 5585.8 .e-< 2:5...
> 33.5»
a. m N 5 o

_ _ p —

 

 

 

 

\\

 

\

 

286 + 2.83 u._s\.\

 

.\

 

\\

 

 

 

o x: a! o
O O O
“u! ‘iuaumndaa

°9
o

N.—

Eoze .5 3 5 55.9.5.5 .3, 2:5..—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>Ju§§
a. m N 5 o
_ _ p p o
\ Nd
z m
885 + >525 u e\\. m
3
\ w
. an...
\ mo .m.
l _
2
5-85.8 .5 m5 5 8585.5 .m.< 8:5
>85;
5. m N 5 o
— p b p o
\\ g
v.0 G
\ ed .m.
\ M
. . wd m
380+>§8ue .\ _ an...
\0\ NA .m.
\« #4
i- 2

 

144

955 55 5585.6 .25 2:5..—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 2 n 5 m- 2- 2-
+5 .\
as \. n
WW) 5
5855 + >855 u 5 \\ w. . m
\ c — am.
a:
.185.“ .5 5.5 55 55525.5 53. 955.5
> 65.5.5
m e N N _
p h h F o
\ N5

\\ a

5.5 .m.

5255 + >825 u 5 \\ m.

55 m

5 5 ..m.

\\

 

 

‘

 

 

N.—

_.E>._ .5 55585.6 .=.< 855.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

> 55.5.5

2 2 n 5 m- e- 2 -
.. \\
K n
325 + >885 n 5 \. ...
nee—5:5 .5 A: me 5:95:50 .a.< unsur—
> 55.5.5
n v m N 5
b p _ o
55
N55 + >525 u 5 \\

\ ed
\ wd
< _

_

 

 

1!! ‘iuawamdfla

1!! ‘inamaovldfla

145

Displacement, in.

MTS Load, lbs

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_____ __ __ __} 6_55 __ _57 . -. . A 5
5“ 7 ’ ”y“ £00595 l 0.9164
5 ___—_ ,5 R2: I
-15 -10 -5 0 5 10 15
Voltage, V
Figure A-13. Calibration of LVDT-3
35000
30000 - —~ ___ — __ — ——
25000 4» —— a —
20000 ~
15000 ~ ——
10000 +—
5000 ,.
0
0

 

Voltage, V

Figure A-14. Calibration of RC156 Actuator 1

146

35000

 

25000 5

i

MTS Load, lbs
['3
§

 

 

 

 

 

 

  

y = 10208x - 5088.5

 

 

 

 

 

 

 

 

 

 

 

2.5 3 3.5 4
Voltage, V

Figure A-lS. Calibration of RC256 Actuator l

 

 

 

MTS load, lbs

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

y = 6213.6x - 3093.8

 
 

 

Voltage, V

Figure A-16. Calibration of RC156 Actuator 2

147

 

 

45000

35000

30000 T

1

MTSLoad,lbs
8 .. 8
8 g 8

 

 

 

 

 

 

 

 

 

10000 4 —
5000 — - —
0 l l r f l
O 2.5 3 3.5 4 4.5
Voltage,V

Figure A-17. Calibration of RC256 Actuator 2

148

 

Data Acquisition System

The data from the different instrumentation was collected at the rate of 6 scans per
second, using a data acquisition system. The data acquisition system is capable of
handling all the different instruments — sliders, LVDTs, and pressure transducers. The

data flow is illustrated in Figure A-18.

Chassis

PCI 60524:

  

Computer

 

Figure A-18. Data Flow

The data acquisition system is designed to hold the chassis and provide access and power

to each channel on the chassis. The chassis is the main unit which consists of all the
channels of the data acquisition system. The chassis model is a SCXI-1001. The chassis
consists of a series of modules connected together to form a unit. A module consists of a

series of channels which are used for certain instrument types. For instance, module 1

149

consists of a series of modules connected together to form a unit. A module consists of a
series of channels which are used for certain instrument types. For instance, module 1
consists of the first 32 channels and modules 2-7 all consist of eight strain gage channels.
The modules are cards (SCXI-1102B or SCXI-1520) which can be individually removed
from the chassis. Module 1 is a NI DAQ card termed SCXI-1102B and is the only
module used in the pullout tests. Out of the 32 channels, channels 0-15 are three-wire
connections, 16-25 are thermocouple connections, and channels 26-31 are four-wire
connections. The three wire connections accommodate the sliders and the pressure
transducers (Channels 0-15) while the four wire connections (Channels 26-31)
accommodate the LVDTs. External power supplies are linked to the data acquisition
system to provide the necessary excitation voltages for the different instruments.

The calibration files corresponding to the different instrumentation are fed into
the computer. The instruments are connected to the different channels on the data
acquisition system using connectors and wires. As shown in Figure A-18, the connection
between the computer and the data acquisition system is established using a cable and the
Analog-Digital card PCI 6052-E present in the computer. Measurement and automation
explorer (MAX) is a DAQ program set up by National Instruments and is responsible for
the configuring the computer to read the DAQ. This data is then converted to the
corresponding measurement units by the software interface (Little General Version 6.1)
using the calibration files mentioned earlier. Little General run off the configuration of

MAX.

150

APPENDIX B
SURVEYING SUMMARY AND CALCULATIONS

151

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5:_v 552v 050.0 550.0 00.5 00.5 30.5 550.5 _
552v 52v 5:5 55A 200 550.0 5N5 55.5 5mN.5 N55 55.0
552v 55v 550.0 58.0 5N5 55.5 NON.5 555.5 56:55.0 U
52v 52v 5:5 55A 0500 500.0 05.5 05.5 555.5 555.5 5.0
52V 52v 5:5 552A 050.0 550.0 55.5 5N5 055.5 555.5 5N.0
53v 5_:v 5N0.0 No.0 00.5 00.5 550.5 050.5 N
52v 5:5 5:5A w:v 5:5 552A NN_.0 550.0 00.5 00.5 NN_.5 5No.5 _
52v 5:5 53A 52v 5:5 55A N500 005.0 00.5 00.5 N505 005.5 55.0 m
52v 5:5 55A 52v 5:5 55A 500.0 050.0 00.5 00.5 500.5 050.5 5.0
52v 5:5 5:_A 55v 5050 55.0 00.5 00.5 505.5 550.5 5N.0
53v 53v N500 550.0 00.5 00.5 N505 550.5 N
w:v 5:5 5SA 52v 5:5 5_:A 000.0 550.0 00.5 00.5 500.5 NNO.5 —
55v 55v 5000 N500 00.5 00.5 500.5 N505 Ecl— >
55v 52v 5:5 55A N500 550.0 5N5 55.5 35.5 55.5 55.0
55v 52v 550.0 0N0.0 05.5 05.5 555.5 0N5.5 5.0
52V 52v 5:5 55A 550.0 N500 55.5 5N5 505.5 555.5 5N.0
5_:v 55v NN0.0 5N0.0 00.5 00.5 NNO.5 wN0.5
52v 5:5 525A 525v 550.0 N500 00.5 00.5 550.5 N505 0 23555
55v 52v 5:5 5:5A wN0.0 550.0 00.5 00.5 58.5 550.5
555 53m 555 58>» 58m .83 “mam «85> 555m .83 .5 .55:—
.5 ..35m .5 .5 .535: 3:53 .5 5.32.5 . 2.5.5
. .855 85855 55:85 555: 555552

 

 

 

 

 

 

 

58:. .26: 25 1 535555 58.5 555: 5.5 8555

152

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5Sv 5Sv 555.5 555.5 555.5 555.5 555.5 55.5.5 55.5.5 0
55v 25 555.5 555.5 555.5 555.5 555.5 555.5 5.5.5

52v 55 53A 5Sv 555.5 555.5 555.5 555.5 555.5 555.5 55.5.5 55
5Sv 52v 5:5 55A 555.5 555.5 5 5 555.5 555.5 5.5.5

55v 55 5:7 52v 5.5 5:? 555.5 555.5 555.5 555.5 555.5 555.5 55.5.5 >
53v 52v 55 5:? 555.5 55 5.5 555.5 555.5 555.5 35 5.5.5

52v 55 5:2 52v 5:5 5:? 555.5 555.5 5.5 5.5 555.5. 555.5 55.5

52v 5:5 53A 52v 55 53A 555.5 555.5 555.5 555.5 555.5 555.5 :55
55v 525v :55 555.5 5.5 5.5 555.5 555.5 5 u
5Sv 552v :55 555.5 555.5 555.5 555.5 555.5 55.5
5Sv 55v 5:5 55A 555.5 555.5 555.5 555.5 555.5 555.5 5.5

55v 55 52? 52v 55 5:? 555.5 555.5 555.5 555.5 555.5 555.5 55.5
5Sv 5Sv 555.5 :55 5.5 5.5 555.5 555.5 2.5
55v 25 555.5 555.5 555.5 555.5 555.5 555.5 3.55
55v 5Sv 555.5 555.5 5 5 555.5 555.5 5 :
5Sv 55v 555.5 555.5 5 5 555.5 555.5 55.5

55v 5:5 55A 52v 55 5:? 555.5 555.5 555.5 555.5 555.5 555.5 5.5
525v 52v 555 5:? 555.5 555.5 5 5 555.... 555.5 55.5

55v 55 5:2 52v 55 53A 5555 555.5 5... 5.5 555.5 555.5 2.5
5Sv 55v 55 55A 555.5 555.5 555.... 555.5 555.5 555.5 :55
53v 53v 555.5 5.5.5.5 5 5 555.5. 555.5 5 >
5Sv 55v 555.5 555.5 55.5. 555.5 35.5 555.5 55.5

52v 5:... 5:7 52v 55 5:? 555.5 555.5 555.5 555.5 555.5 555.5 5.5

55v 55 525A 52v 55 5:? 555.5 555.5 555.5 555.5 555.5 555.5 55.5

52v 55 53A 5Sv 555.5 555.5 555.5 555.5 555.5 555.5 5 25555
.555 .83 .555 .83 .555 .83 .5555 as: .5 .558

.5 59:5 .5 .355: .533. .5 ”5:555:55 555.552 25.

 

.585.“ 2.555.:

 

 

55:85 5555:

 

 

 

 

25.55: 5.870 355.5 .955: 95,—. I 5:55:55 58:0 5563 .N.m 5.55.5

153

 

:23: 5.59 was .025 a: u 585% V.85 Ema: .mé 2.5

154

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

03v 05v 000.0 30.0 04. 0 0 mom.”V 30.0 0.0
002v 02V 05w 0_:A 08.0 000.0 0 0 000.0 0004q 0 .5500
05v 05v 000.0 03.0 0.0 04V 030 $06 0.0 ”0 650 >0
0Sv 05v 500.0 000.0 0 0 000.»‘ 000.0 0
02v 0.8 020A 002v 000.0 000.0 0 v 0.0 000;.V 53.0 0.0
0:v 05 03A 05v 000.0 30.0 000.0 000.0 0000 30.0 0.0
02v 0:“ 020A 0:_v 000.0 000.0 000.0 000.0 02.0 000.0 0.0
02v 05 002A 02v 020 000.0 000.0 000.0 02.0 0006 0.0 =< E0
002v 0_:v 90.0 000.0 000.0 000.0 03.0 000.0 0.0
02V 02v 0:“ 0—2A 000.0 0000 000.0 000.0 000.0 0000 0.0
- 22v - 30.0 - 0 - 000.»V 0.0
- 02V 05“ 22A - 000.0 - 0 - 00—0 0 08:00
- 02v 0.3 002A - 3.0.0 - 0 - 0N0.»V 0.0 $0
. . a. .25
- 002v - 03 0 - 0 - 03 0 0
- 02v 0.3 22A - 000.0 - 0 - 000.0 0.0
02v 05 020A 02V 05 002A 000.0 0000 0d 0.0 0006 000.0 0.0
02v 0_:v 000.0 v8.0 0 0 000.0 0004V 0 .5500
02v 05 002A 02V 05“ 002A 30.0 00.0 0.0 00 3.0.0 03.4V 0.0 M0 .630 00
02v 02v 05 22A 0.0.0.0 N000 0 0 0N0.0 0004V 0
02v can 00:A 02v 05 0_:A 000.0 000.0 04V 0.0 000.”V 30.0 0.0
02v 05 BEA 02V 000.0 000.0 00 0.0 000d 000.0
02V 0>V 05“ 002A 05.0 000.0 0.0 04V 0010 0006 =< >0
02v 0:“ 002A 02v 98 002A 20.0 000.0 0% 0.0 2.0.»v 000.0
02v 0:5 0_:A 0_:V N000 500.0 0 0 N000 000%
02v 0.8 002A 0_:v 000.0 000.0 0 0 000.0 000.0 0.0 =< 30
02v 05 020A 020v 05.0 20.0 0 0 050.0 0004v
02v 05 002A 0:_v 000.0 000.0 04» 0.0 000.»‘ 000.0
0_:v 02V 05» 002A :00 000.0 0.0 0.». 00v.0 000;.V =< 00
02v 05” 002A 00:V 050.0 000.0 06 0.0 050.0‘ ~0v.0
«mam— «83 “mam “no; «mam 08>? unam— «00>?
. . .5 "05.323 .5 .285 02.00.52
.5 ...ctm 5 «Sta 5 .503: 50.8.5. £00.52 £9.69 2:0.
828: .552 3%:

 

 

0953. 5:8 53:: 50:85 3.53. 5.5: .3360 38,—. .325 3E 05“ 095,—. I mama—5:0 58:0 306: .VA— 0.0—ah.

155

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

05\5v 052v 53.0 050.0 0005 000.5 53.5 0505 315
052v 025v 5N00 N000 0.0 0.0 0550 N000 3:00
02v 05 05\5A 02v 5555 052A N000 500.0 0005 0005 N005 5005 5
05\5v 025v 050.0 30.0 005.0 005.0 005.0 505.0 05.0 0
025v 052v 0N0.0 500.0 0.0 0.0 0N0.0 500.0 0.0
052v 05\5v 0N0.0 0N0.0 00N.0 00N.0 05N.0 05N.0 0N.0
05\5v 05v 05 0555A 5000 000.0 000.5 0005 5005 000.5 315
05\5v 052v 5.500 050.0 0.0 0.0 30.0 050.0 DI00
05\5v 0555v v8.0 N000 000.5 0005 050.0 000.0 5
£5 oSv Sod 086 RS 83 853 £50 Po :
052v 052v 000.0 0N0.0 000.0 000.0 000.0 0N0.0 0.0
02v 5555 025A 05\ 5v 000.0 500.0 0N.0 0N.0 050.0 50N0 0N.0
«mam 58>? 53m .83 E 583 «mam 583 .5 5:055
.5 £8.50 .5 08.5.5 85855.5 .5 695359 55394.. .5 ”0155353 .nw. am. 25.5.5.
50.8.5. 8:359 .5 .52
295.505 55.870 88.5. 595.5 95.5. I FEE—::0 5:0:—50553555 55:55.85.— .0-m 255,—.
05\5v 0555v 5500 03.0 0005 000.5 550.5 000.0 5
05\5v 052v 0N0.0 N000 005.0 005.0 055.0 N050 05.0
055v 02v 5:8 0: 5A 5500 000.0 005.0 005.0 005.0 000.0 20I050 U
052v 025v 0N0.0 N000 000.0 000.0 05v0 03.0 0.0
05v 5:8 052A 052v 000.0 N000 00N.0 00N.0 050.0 NON.0 0N.0
0:5v 05v 05 055A N000 000.0 000.N 000.N N00.N 000.N N
052v 0\5V0§ 052A 5000 005.0 0005 000.5 5005 0055 5
05\5v 055v 000.0 0N0.0 005.0 005.0 005.0 055.0 05.0 E
05\5v 0555v 5500 5N00 000.0 000.0 550.0 5N00 0.0
05\5v 05\5v 000.0 0N0.0 00N.0 00N.0 3N0 05N.0 0N.0
«mam 58>? 55 58>» 53 58>? «mam 583
. . . .5g5hotam .5 .555 25.5.5.
.5 59:”.— .5 3.5.5.5 358.: 5 8:535 55394.. 50:85. 3:350— 50558555

 

 

 

 

 

 

38.5. 595.5 020 ... 5.555550 550550553552 58.85535 .04— 255.5.

 

156

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0555v 052v 050.0 000.0 005.0 005.0 505.0 005.0 05.0
052v 025v 0N0.0 5N0.0 0.0 0.0 5550 0510 0.0 U
05\5v 052v N500 500.0 005.0 005.0 N050 000.0 05.0
05\5v 025v N30 0500 0.0 0.0 0050 050.0 0.0 m5
05\5V 052v N500 30.0 000.5 000.5 N505 30.5 DI5
0:5v 052v 30.0 050.0 0.0 0.0 30.0 050.0 DI00
05\5V 025v 0N0.0 N30 000.5 000.5 000.0 000.0 5 U
05\5v 0555v 30.0 N500 005.0 005.0 0050 N050 05.0
02v 05 05\5A 05\5v 000.0 550.0 0.0 0.0 505.0 550.0 0.0
05\5v 05\5v 5N0.0 000.0 00N.0 00N.0 55N.0 00N.0 0N.0
05\5v 05\5v 5N0.0 5N0.0 000.5 000.5 5N0.5 5N0.5 DI5
05\5v 052v 53.0 30.0 0.0 0.0 550.0 30.0 DI00
0555v 05\5v 000.0 550.0 000.5 000.5 500.0 000.0 5
05\5v 05\5v 050.0 000.0 005.0 005.0 005.0 005.0 05.0 5.5
05\5v 0555v 000.0 30.0 000.0 000.0 000.0 00520 0.0
02v 5:8 05\5A 02v 5:8 05A N000 V000 0N.0 0N.0 N00 300 0N.0
«mam 585$ unam— umo>> «mam— umo>> «mam— 583 .5 .552:
5095.50 .5 £8.55 858.2 .5 695859 55394 .5 £55053 éw=am55>5 25.5.5.

 

 

 

50:85. 8:859

 

 

 

 

A5955 5:500 385. 59.25 95.5. I 0.55550 55550555555 55555.8: .545 255,—.

157

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

052v 052v 500.0 0.0 500.0 0.0

- - - - - - - - 0.0 BEoU
05\5v 052v 0 550.0 0.0 005.0 0.0 50 830 I0

- - - - - - - . 0.0
052v 055v 5N0.0 0.0 05520 0.0
052v 052v 30.0 050.0 0 0 0.0 30.0 050.0 0.0
0555v 05\5v 0N0.0 03,0 0 0 0.0 5550 5050 0.0
05\5v 025v 050.0 500.0 0.0 0.0 050.0 0050 0.0 =< 550
05\5v 05\5v 050.0 000.0 0 0 0.0 N050 000.0 0.0
05\5v 052v 050.0 050.0 0.0 0.0 N010 050.0 0.0
05\5v 052v 03.0 000.0 0.0 0.0 050.0 000.0 0.0

- - - - - - - - 0.0

. . . . . . 85:00
05W5v 05W5v 000 0 000 0 0-0 0-0 00.0 0 00wV 0 MW 50 52:0 U0
052v 02v 5E5N 05\5A 500.0 505.0 0 0 0.0 500.0 000.0 0.0
05\5v 0555v 000.0 N000 0 0 0.0 000.0 N000 0.0
052v 052v 050.0 500.0 0 0 0.0 050.0 500.0 0.0 =< 550
0555v 052v 03.0 000.0 0 0 0.0 050.0 000.0 0.0
05v can 015A 052v 500.0 050.0 0.0 0.0 03.0 050.0 0.0
052v 055v N500 050.0 0.0 0.0 N500 050.0 0.0 :< 00
05\5v 0555v 0N0.0 550.0 0.0 0.0 0N0.0 550.0 0.0
Ban 5855 53m .83 «mam 583 «mam 58>?
.5 .5 .5 5% .5 .555 0950555535 255.5.
.5 59...”.— . . 503.555 50553535 29.25
. 53.505 858.; .3535: _a=8< .
8:559

 

 

A5955. 55:8 .5555 5055.55 5955. 55.8: 58305 38.5. 59.25 05.5 05 3.5.5. I 5.35550 55550553555 5355.5: .055 0555.5.

158

Surveying Calculations
The algorithm used to compute the horizontal and vertical misalignments is explained
here, along with examples to illustrate the procedure.
Measurement of Vertical Misalignment using Electronic Theodolite

Notations:

L = Distance between Station A and Station B
Horizontal Angles

Ga: from Station A to the base plate (center of the base plate)
6b = from Station B to base plate (center of the base plate)

Vertical Angles

0t.I = from Station A to the base plate (at the level of the base plate)
ab = from Station B to base plate (at the level of the base plate)
Ba = from Station A to the dowel bar (at the level of the center of dowel bar)

[3b = from Station B to the dowel bar (at the level of the center of dowel bar)

A = Station A B = Station B
C = Base Plate (center) D = dowel bar (center)

E = height at eye level (center)

159

'E

180 - (9,. 91,)
A ‘ 9a

9b
V
B

Figure B-l. Horizontal plane showing angles measured from center of base plate to
misalignment

Referring to Figure B-l, in the horizontal plane,

For A ABE using sine rule: AE = Lcosec(Ba' + 9b )sin(6b)

A B. LE

Iv

C

 

 

Figure B-2. Vertical plane in the center of the plate
Now with respect to the dowel at the center of the plate (Figure B-2):
The vertical height from the eye level to the base plate is EC = AE tan (aa ).
The vertical height from the eye level to the center of the dowel bar is ED = AE tan(,Ba ).

The difference in the height taken from the eye level to the base plate and the center line
of the dowel bar gives the vertical height, V = EC — ED

Therefore, vertical misalignment = V — 5 in. (mid height of the dowel)

160

Example Calculation for Vertical Misalignment of V2 in:
The sample calculations shown are for a vertical misalignment of I/2 in on the east side of

the test setup. Similar set of calculations are carried out on the west side.

Table B-9. Summary Table showing the various angles

 

EAST SIDE (L = 9.2')

 

 

 

 

Slab Designation Location Theodolite Theoretical
Degrees Minutes Seconds Degrees
9a 25 12 40 25.211
9b 74 5 l 0 74.850

 

1V12 (16 in per 012 333 44 20 333.739

half'cng‘h) ab 331 47 20 331.789
Ba 335 39 40 335.661
[36 314 32 10 314.536

 

 

 

 

 

 

 

 

 

 

 

Referring to Figure B-l, for A ABE using sine rule,
AE = Lcosec(6a + 0b )sin(0b) = 108.227 in
The vertical height from the eye level to the base plate is (Figure B-2):

EC = AEtan(aa) = 53.397 in

The vertical height from the eye level to the bottom edge of the dowel bar is:

ED = AEtan(fla) = 48.9549 in
The difference in the height taken from the eye level to the base plate and dowel bar
gives the distance from the base plate to the bottom edge of the dowel bar,

V = EC— ED = 4.4429 in
Perfectly aligned bar will be at (base plate to center line of dowel bar) = 5 in.
Misalignment = 5 — 4.4429 = 0.557 in.

Error from desired misalignment = 0.557 — 0.5 = 0.057 in.

161

Measurement of Horizontal Misalignment using Electronic Theodolite
Notations:
L = Distance between Station A and Station B

Horizontal Angles

0a = from Station A to the base plate (center of the base plate)
0b = from Station B to base plate (center of the base plate)
ya: from Station A to the base plate (horizontal misalignment on the base plate)

Yb = from Station B to base plate (horizontal misalignment on the base plate)
Vertical Angles

0ta = from Station A to the base plate (at the level of the base plate)

ab = from Station B to base plate (at the level of the base plate)

53 = from Station A to the misaligned position on the base plate

81, = from Station B to the misaligned position on the base plate

ma: from Station A to the dowel bar (at the level of the center of dowel bar)

tub = from Station B to the dowel bar (at the level of the center of dowel bar)

A = Station A B = Station B

C = Base Plate (center) D = dowel bar (center)

E = height at eye level (center) E = height at eye level (misaligned)
C’ = Base Plate (misaligned) D’ = dowel bar (misaligned)

In the following surveying description the following are assumed:
1. The height of both the stations is equal

2. The two stations are parallel to the horizontal misalignment

162

 

 

 

 

l*—'—’l
E' E
lb 23
96
f
A Ga Ya Yb E

Figure B-3. Horizontal plane showing angles measured from center of base plate to
misalignment

Referring to Figure B-3, in the horizontal plane,

For AABE, using sine rule:
AB = L cosec (6a + 9b )sin (4%) BE '= L cosec(ya + 7b )sin(}'a )

To make sure that the two stations are parallel, the perpendicular distances (Z0 and Z) are

compared:
20 = AEsin(Ba) 2b = BE'sin(yb)

If, Z a E Z b , then the two stations are parallel to the horizontal misalignment.

163

Wa

 

 

Cl

Figure B-4. Angles in the vertical plane, after forcing the misalignment of H

The vertical height from the eye level to the base plate is (Figure B-4):
E’C’ = AE' tan(5a)
The vertical height from the eye level to the center of the dowel bar is
E’D’ = AE' tan(t//a)
The difference in the height taken from the eye level to the base plate and the center line

of the dowel bar gives the vertical height: V’ = E ’C ’ — E ’D’

Therefore, vertical misalignment = V’ — 5 in. (mid height of the dowel)

 

 

 

 

vii—>1
E' E
Zb Za
96
f
A 93 Ya Yb B

Figure B-5. Showing the final computation of the horizontal misalignment

164

For A ABE’, in horizontal plane, AE' = Z cosec(ya)

Using cosine rule for AAEE ' (Figure B-S):

 

H = JAEZ + (AE')2 -— 2(AE)(AE')cos(9a - 7a)

The H obtained is then compared with tape measure as a final check.

Example Calculation for Horizontal Misalignment of V2 in:

Table B-10. Summary Table showing all angles for horizontal misalignment

NORTH WEST SIDE (L = SOUTH EAST SIDE (L =

Slab Location 5-7') 5.2')
Designation Theodolite Theoretical Theodolite Theoretical

Deg Min Sec Deg Deg Min Sec Deg
9a 56 58 40 56.978 57 30 40 57.511
0b 49 0 20 49.006 53 19 0 53.317
ya 57 30 0 57.500 57 6 40 57.111
Yb 48 41 40 48.694 53 47 30 53.792
01.1 317 58 30 317.975 319 53 20 319.889
ab 321 0 30 321.008 321 30 0 321.500
Ba 321 6 40 321.111 323 13 40 323.228
56 324 5 10 324.086 324 52 0 324.867
8,, 317 52 40 317.878 320 3 50 320.064
51, 321 11 0 321.183 321 22 20 321.372
W 321 0 30 321.008 323 24 0 323.400
Wt, 324 17 50 324.297 324 42 0 324.700

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1H12 ($6 in per half length)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Referring to Figure B-3, in the horizontal plane,
For A ABE, using sine rule,

AE= L cosec (19a + 9b )sin (0b) = 53.54 in.
BE'= Lcosec(ya + 7b )sin(ya) = 56.09 in.
To make sure that the two stations are parallel, the perpendicular distances, Z0 and Zb are

compared: Za 2 AE sin(19a) = 45.16 in. and Z}, = BE'sin(yb) = 45.27 in.

165

If, Z a 5 Z b , then the two stations are parallel to the horizontal misalignment.
The vertical height from the eye level to the base plate is (from Figure B-4):
E’C’ = AE' tan(6a) = 45.12 in.
The vertical height from the eye level to the center of the dowel bar is
E’D’ = AE' tan(1//a) = 40.02 in.
The difference in the height taken from the eye level to the base plate and the center line
of the dowel bar gives the vertical height:
V’ = E’C’ — E’D’ = 5.09 in.
Therefore, vertical misalignment after forcing the misalignment in the horizontal plane
should be = V’ — 5 in. (mid height of the dowel) = 0.09 in
For A ABE’, in horizontal plane,
AE. = Zcosec(ya) = 53.89 in

Using cosine rule for AAEE ' (from Figure B-S):

 

H = JAEZ + (AE)2 - 2(AE)(AE')cos(19a — ya) = 0.5168 in

The H obtained is then compared with tape measure as a final check.

166

APPENDIX C

BOND STRESS CURVES

167

9:3 «5:25 .53. m> 325 6:3— ._.0 2:3..—

.5 figs:

2m 66:: 63

we: eofiamg

:< v-0 :< 0:0 ESE—D ESmEDéoZ

md no mud md who We m We . fl who Do
U mu 2 ID ID ED EU ID $0 $0 EU>EU I

mad 0 N ~ who no mud o
EU>EU 30 EU I

 

1s Sutpuoqad/dus 12111111

lsd ‘sseu

168

Horizontal Misalignment

 

 

   

 

.‘ Mikmmfl 1 «vi-9"“ * .IIW'Mafl'i‘W

 

 

 

 

0 F j I T T I r I

0 0. l 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 l

 

Joint Opening. in.
Figure C-2. Bond stress vs joint opening curves for l-dowel Horizontal
Misalignment cases
160
140 -
I is»

120 . I
0.75 in. “ NW
10° ‘ 0.25m. ‘ .
V 0.5 in. ‘ Cracking
80 a

Bond Stress, psi

 

 

 

0 I 1 1 r 1 f I 1 r f

O 0.] 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure C-3. Bond stress vs joint opening curves for 2-dowel Non-uniform Horizontal
Misalignment cases

169

 

 

 

16o ___._Ai,, -—~———— ——-—

140*

120 *

§

Bond Stress, psi
00
c

 

 

 

 

 

 

0.5 in
40 .4
0 in. WM
20 -
O r T r I I I v 1 I
o 0.1 0.2 0.3 0.4 05 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure C-4. Bond stress vs joint opening. curves for 2-dowel Uniform Horizontal
Misalignment cases
160
140 ~
120 -

§

Bond Stress, psi
00
o

oQ7s in.
60 ~ '
... m“ 0.5111.
40 ~

20‘

 

 

 

o I I T r I fl I I

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure C-S. Bond stress vs joint opening curves for 2-dowel Horizontal
Misalignment cases (One dowel misaligned)

170

 

 

 

 

 

 

 

 

 

 

160 if __ A h.
140 .
120 4
.. 100
in
e l VON-t!" “" “fl *1» "
g 80 4 5 bars 5 bars _, '
1: Outer-Center All Misaligned 3 bars
M ed
g 60 ‘ mugn ,2:
411/
,/
20 -
0 I I I I I I I T i
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure C-6. Bond stress vs joint opening curves for 3 and S-dowel Horizontal
Misalignment cases
Vertical Misalignment
160
140 -
120 WW
2.... WM)
... 100 -
a ’ rue
3 1 in.
5’, 30— .. ‘ (”Simw- ‘J
'u .r ' ’,v" v flWM05 in.
g """.,-'r’d W " “ ‘
6° ‘ 1””. 025 in. g 1
...-W o ”-..... 1...-.. ~
. ' ,,<
40 I" v
20
o I I I T I I I j r I
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Joint Opening, in.

Figure C-7. Bond stress vs joint opening curves for l-dowel Vertical Misalignment
cases

171

 

 

140*

120 7

Bond Sm, psi
8 §

6’

E

 

 

 

 

 

 

 

 

 

 

/ - “W - A F 4 - Momw
/.
20 /
o I ‘Y 7 l l 1 I r I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure 08. Bond stress vs joint openingcurves for 2-dowel Non-uniform Vertical
Misalignment cases
160
140 -
120 -
. 1 in.
E 100 ‘ .
g 80
'8
8 60 -
0.5 in.
W
40 .1
WV n, A W 43—0 in. M
20 1'
0 l 1 T 1 1 1 fi fi r 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure C-9. Bond stress vs joint opening curves for 2-dowel Uniform Vertical
Misalignment cases

172

160 4~

 

 

 

 

 

 

 

 

 

,
140 4
120 4
- 100 ‘
8.
(g 30 ‘ 0.75 in.
'6
fl 5‘! i
e g.. m I. I. ~ (~—
m 60“ .._4v -"~“..._
"' 0.5 in.
4.0 4
20 4
o I l T f I I I I I
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure 010. Bond stress vs joint opening curves for 2-dowel Vertical Misalignment
cases (One dowel misaligned)
160
140 -
120 4
5 bars
All Misaligned Cracking
100 4
'2. 3 bars
(‘3 80 . 5 bars
'3 Outer-Center WW
3 60 _ Misaligned
40 “~ 1“
20 4
0 I T I I I I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure C-ll. Bond stress vs joint opening curves for 3 and S-dowel Vertical
Misalignment cases

173

Bond Stress, psi

Bond Sm, psi

Combined Misalignment

 

 

 

 

 

160 4-— — ~44 '44-————— — -———4 ——~
140 4
120 4
100 4 1 in.
80 4
025 in
60 ‘ . x 0.75 in. 95 ‘n
I“ ‘ “WI-
1 ‘ 0 ‘ . [5:31. i
40 11, . ° "“ ‘6...
I, ‘h‘ x “D -
20 4
0 I r I I I I I I I
o 0.1 0.2 0.3 0.4 05 0.6 0.7 0.8 0.9
Joint Opening, in.
Figure 0 12. Bond stress vs joint opening curves for l-dowel Combined
Misalignment cases
160
Cracking

 

 

 

 

0.5 in i
l
Cracking
\
0.5 0.6 0.7 0.8 0.9
Joint Opening, in.

Figure C-13. Bond stress vs joint opening curves for 2-dowel Non-uniform
Combined Misalignment cases

174

 

 

16o 44— 4444—4—44 _44 -_-¥__,~___- -V, 2 a

140 3

Bond Stress, psi

 

 

 

 

   

 

 

 

 

0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.
Figure C-l4. Bond stress vs joint opening curves for 2-dowel Uniform Combined
Misalignment cases
160
140 4
120 4
100 4
’1.
3, 80 ~ 0.75 in.
1, .
d
O
a: 60 a
0.5 in.
40 —4
20 - g
l
i
0 I T I I I I I r I A
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure C-15. Bond stress vs joint opening curves for 2-dowel Combined
Misalignment cases (One dowel misaligned)

175

160 444 444—4 4 44 4- --

 

 

Bond Stress, psi

 

 

 

140 .
5 bars MVHWMN '
“MW“ MIA-MM»
._ Outer-Center 1‘
1m - Spalling‘ Misaligned
3 bars
80 7
Cracking
60 .4
4° 7
20 a
0 I I I I I I I I I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Joint Opening, in.

Figure C-16. Bond stress vs joint opening curves for 3 and S-dowel Combined
Misalignment cases

176

 

APPENDIX D
VISUAL OBSERVATIONS

177

1"1 n‘zr. MW

 

 

3 1/‘9,

(a) West joint face (1H2) (b) East joint face (2H12NU)

      
   

4.1 411“. 1. 111".

(c) North outer dowel (5H12NU) 444 44: 4 4 4 44 5%! 44 k- 44-4
((1) South outer dowel (5H12NU)

Figure D-l. Spalling in the Horizontal Misalignment tests

178

 

 

 

 

 

 

 

 

 

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2
Joint Opening, in.

0.1

0.0

S_Edge

12 in.

12 in.

N_Edge

Figure D-2. Joint opening as a function of distance along the joint (5H12NU)

179

 

(a) East joint face (2V1NU) (b) Westjoint face (2V1NU)

 

(c) East joint face (2V1U) (d) West joint face (2V1U)

Figure D-3. Spalling in the Vertical Misalignment tests

180

 

 

V

 

A

\ \\ V ‘ \\i

 

‘[ South_Edge

3 in.

Slider_South_LVDT

 

 

 

 

 

 

l

 

 

 

 

 

 

‘
' A

 

 

 

 

Slider_Non.h_LVDT

3 in.

{ Nonh_Edge

 

 

 

 

 

 

 

l

 

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

Joint Opening, in.

Figure D-4. Joint opening as a function of distance along the joint (2V1NU)

181

1%

        

2 1/4”
(a) East joint face (1C1)

(b) East joint face (2C1U) (c) West joint face
(2C1U)

   
     
   
    

 

(d) East joint face (2C1NU) 7 (e) Eastjoint face (2C12NU)

‘._.-_>

thdwel bar (2C34NU)
Figure D-S. Spalling in the Combined Misalignment tests

(f) West joint face near the Sou

182

 

 

 

LIA

J I ; ~ > x South_Edge
\ \ 4in.

 

‘ v
“ North_Edge

 

 

J,

 

 

 

 

 

 

 

 

 

 

 

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

Joint Opening, in.

Figure D-6. Joint opening as a function of distance along the joint (2C1U)

183

10.

11.

REFERENCES

American Association of State Highway and Transportation Officials (AASHTO),
AASHTO Guide for Design of Pavement Structures, American Association of
State Highway and Transportation Officials, Washington, DC, 1993.

American Concrete Paving Association, Database of State Department of
Transportation Concrete Pavement Practices, downloaded from
www.pavement.com, 2004a.

American Concrete Paving Association, Design and Construction of Joints for
Concrete Highways, Technical Brief No. TBOlQ-OlP, Concrete Paving
Technology, America Concrete Paving Association, 1991.

American Concrete Paving Association, Concrete Pavement Technology Update,
ACPA, FHWA-IF-O4-025, www.fhwa.dot.gov/pavement/conhome.htm, Aug
2004.

American Concrete Paving Association, Concrete Pavement Progress-Technical
tips and techniques, ACPA, Volume 40, Number 4, CPP-090604,
http://www.pavement.com/CPP/2004/CPP-090604gdf, Sep 2004.

American Concrete Paving Association, Concrete Pavement and Research
Technology Update, Number 6.01, www.pavement.com/techserv/RT6.01.pdf, Jan
2005.

Applied Research Associates, Guide for Mechanistic-Empirical Design of new
and rehabilitated pavement structures, Final Report of Project No. I-37A,
National Cooperative Highway Research Program.

Bock B.T., Okamoto P.A., Evaluation of Dowel Placement using a Dowel Bar
Inserter, Proceedings of the 4th International Conference on Concrete Pavement
Design and Rehabilitation, Purdue University, West Lafayette, Indiana, April
1989, p. 641—649.

Burati, J.L., Beeson, M.G., Hill, H.S. Jr., Comparative analysis of dowel
placement in Portland-Cement Concrete Pavements, Transportation Research
Record 924, 1983, p 19-27.

Bumham, T, A Field study of PCC Joint Misalignment near Fergus Fall,
Minnesota, Final Report No. MN/RC-l999-29, Minnesota Department of
Transportation, 1999.

California Department of Transportation, Standard Specifications for construction
of local streets and roads,
http://www.dot.cagov/hq/esc/oe/proiect_plans/HTM/stdplns-dual~new02.htm,
2002.

184

12. Cashell, Harry D., Performance of Doweled Joints under repetitive loading,
Public Roads, Volume 30, Number 1, April 1958. -

l3. Darter,M.I., Design of zero-maintenance plain jointed concrete pavement, Report
No. FHWA-RD-7-l l 1, Federal Highway Administration, 1977.

14. Darter et.al., Concrete Pavement Evaluation Systems, NCHRP Report 277,
Washington, DC, 1985.

15. Davids, W.G., Modeling of Rigid Pavements: Joint Shear Transfer Mechanisms
and Finite Element Solution Strategies, PhD thesis, University of Washington,
1998.

16. Donahue, J .P, Investigation of Dowel Bar Placement Accuracy with a Dowel Bar
Inserter, Research Investigation RIOl-049, Missouri Department of
Transportation, April 2003.

17. Federal Highway Administration (FHWA), The Benefits of using dowel bars,
Technical Paper No.89-03, Washington, 1989.

18. Federal Highway Administration (FHWA), Computer-Based Guidelines For
Concrete Pavements Volume Ila-Design and Construction Guidelines and
HIPERPAV 11 User's Manual, FHWA-HRT-04-122, Feb 2005.

19. Fowler, G. and Gulden,W., Investigation of Location of Dowel Bars Placed by
Mechanical Implantation, FHWA/RD-82/153, Georgia Department of
Transportation, Office of Materials and Research, May 1983.

20. Friberg, B.F, Design of dowels in transverse joints of concrete pavements, ASCE
Transactions 105, 1940, p. 1076-1095.

21. Highway Research Board, The AASHO road test — Report 5, Special Report 61E
about Pavement Research, Publication No. 954, Division of Engineering and
Industrial Research, Washington, DC, 1962.

22. Huang, Yang H., Pavement Analysis and Design, Prentice-Hall, 1993, p.192-200.

23. Illinois Department of Transportation, Standard Specifications for Road and
Bridge Construction, http://www.dot.state.il.us/desenv/gifspecZOOZ/sec4000df.
2002.

24. Ioannides, A. M., Korovesis, G. T., Analysis and design of doweled slab-on-grade
pavement systems, Journal of Transportation Engineering, v 118, n 6, Nov-Dec,
1992, p 745-768.

25. Ioannides, Anastasios M et.al., Control of faulting through joint load transfer
design, Transportation Research Record No. 1286, 1990, p. 49-56.

185

26. Iowa Department of Transportation, Standard Specifications for Road and Bridge
Construction, hith/wwwerl.dot.state.ia.us/Oct 2004/GS/frames.htm, 2004.

27. Khazanovich, L and Gotlif, A, Evaluation of joint and crack load transfer - Final
Report, Report No. FHWA-RD-02-088, Federal Highway Administration, 2003.

28. Michigan Department of Transportation, Standard Specifications for
Construction, http://mdotwas l .mdot.state.mi.us/public/specbookl, 2003.

29. Michigan Department of Transportation, Office Memorandum on Surface

Condition Survey and Dowel Bar Restraint — Recycled Concrete Pavement
Research Project 788-99, Dec 1989.

30. Nichols, TH. and Hoffman, G.L., Machine Insertion of Plastic-Coated Dowel
Bars in PCCP, Research Project 78-8, Materials and Testing Division,
Pennsylvania Department of Transportation, Harrisburg, May 1980.

31. Ohio Department of Transportation, Construction and Material Specifications,
http://www.dot.state.oh.us/construction/OCA/Specs/Rewrite2005/20058pecBooH
2005 Whole 3-17-2005%20to%20Printer.pdf, 2005.

32. Okamoto, P.A., Field Evaluation of Dowel Placement along a Section of [-45 in

Texas, Transportation Research Record No. 1186, Transportation Research Board,
1988, p 16-34.

33. Okamoto, P.A., Field Evaluation of Dowel Placement along a Section of [-90
near Janesville, Wisconsin, Report to Wisconsin Department of Transportation,
District 1, Madison, Wisconsin, August, 1988.

34. Packard and Tayabji, Mechanistic design of concrete pavements to control joint
faulting and subbase erosion, International Seminar on Drainage and Erodibility
at the Concrete Slab-Subbase-Shoulder Interfaces, Mar 1983, Paris.

35. Parry, J, Dowel Bar Placement: Mechanical Insertion versus Basket Assemblies,
Final Report, Study No. 88-10, Wisconsin Department of Transportation, 1989.

36. Poblete et.al., Field Evaluation of Thermal deformations in Undoweled PCC
Pavement slabs, Transportation Research Record No. 1207, Transportation
Research Board, 1988, p 217-228.

37. Ross, Rutkowski, A Designed system of Load transfer dowels for Joints in PC C
pavement, Final Report WI 89-05, Jan. 1989.

38. Rufino, Mallela, and Darter, Dowel bar and slab size recommendations based on
the Mechanistic-Empirical pavement design guide, 8th International Conference
on Concrete Pavements, Aug 2005.

186

39.

40.

41.

42.

43.

45.

46.

47.

48.

49.

50.

51.

Schierer, D.L, Paper on dowel bars, prepared for FHWA, Washington, DC.
(unpublished), 1985.

Segner, E.P., Jr. and Cobb, J.R., A Study of Misaligned Dowels in Concrete
Pavements, HPR Report 32, Alabama Highway Department, Aug 1967.

Smith, AR. and Benham, S.W., Effect of Dowel-Bar Misalignment across
Concrete Pavement Joints, ASCE Transactions, Paper No.2002, 1938, p.1133-
1162.

Smith, K.D. et.al., Performance of Jointed Pavements, Vol. I, Evaluation of

Concrete Pavement Performance and Design Features, Federal Highway
Administration, FHWA, 1990.

Snyder, M.B., Dowel load transfer systems for fidl-depth repairs of jointed
Portland cement concrete pavements, PhD thesis, University of Illinois, Urbana,
1989.

. Soriano, A., Dowel Bar Placement and Displacement, Publication No.SK9l-O9-

X, South Dakota Department of Transportation, Mar 1995.

Tabatabaie, A.M. et.al., Longitudinal joint systems in slip-formed rigid
pavements: Vol. II — Analysis of load transfer systems for concrete pavement,
Report No. DOT/FAA/RD-79/4, US. Department of Transportation, Federal
Aviation Administration, Washington, DC, 1979.

Tayabji, S.D., Dowel Placement Tolerances for Concrete Pavements,
Transportation Research Record 1062, Transportation Research Board, 1986,
p.47-54.

Tayabji, S.D., Okamoto, A.P., Field Evaluation of Dowel Placement in Concrete
Pavements, Transportation Research Record 1110, Transportation Research
Board, 1987, p.101-109.

Teller, L.W., and Cashell, H.D, Performance of dowels under repetitive loading,
Public roads, 30 (l), 1958, p. 1-24.

Wisconsin Department of Transportation, Standard Specifications,
http://www.dot.wisconsin.gov/business/engrserv/construction-librarv.htm, 2004.

Darter,M.I., Design of zero-maintenance plain jointed concrete pavement, Report
No. FHWA-RD-7-111, Federal Highway Administration, 1977.

Yu, HT and Khazanovich, L, Applied Research Associates, Final Report of Use

of Magnetic Tomography Technology to evaluate Dowel Placement, Federal
Highway Administration Contract No. DTFH61-03-C-00120, Feb 2005.

187

 

52. Yu, H.T, Kim, H.B, Use of MIT Scan-2 in Dowel Alignment Evaluation, 5m
International Conference on Pavement Technologies Proceedings Vol. II, May

2005.

188

i

i.“-
... 4 Minoan—h
R | t ...

 

   

IIIIIIIIIIIIIIIIIIIIIIIIIIIII

1111111111 111111111111111111111111111