Final_Report

FHWA-NJ-2002-19
Cause and Control of Transverse Cracking in Concrete

Bridge Decks
Final Report
December 2002
Submitted by
Dr. M Ala Saadeghvaziri Rambod Hadidi

Professor Research Assistant
Department of Civil and Environmental Engineering

New Jersey Institute of Technology
Newark, NJ 07102
NJDOT Research Project Manager

Mr. Raj Chawla
In cooperation with
New Jersey
Department of Transportation
Division of Research and Technology
and
U.S. Department of Transportation
Federal Highway Administration
DISCLAIMER STATEMENT
“The content of this report reflects the views of the author(s)

who is(are) responsible for the facts and accuracy of data
presented herein. The contents do not necessarily reflect the
official views or policies of the New Jersey Department of
Transportation of the Federal Highway Administration. This
report does not constitute a standard, specification, or
regulation.”
TECHNICAL REPORT STANDARD TITLE PAGE
1. Report No. 2. G overnment Acc essi on N o. 3. Recipient’s Catalog No.
FHWA-NJ-2002-019
4. T i tl e and S ubti t l e 5. Report Date
Ca u se a nd Co nt r o l o f T r a n s ve r s e Cr a c k i n g i n Co ncr et e B r id g e De c e mb e r 2 0 0 2
Dec k s 6 . P erfo rm i ng Org ani zati on C od e
7. A ut hor (s) 8. Performing Organi zati on Report No.
M . Ala Sa a d e g h va zir i a n d Ra mb o d H ad id i
9. Perf orming Organi zation Nam e and A ddress 10. W ork Unit No.
Dep a r t me n t o f C i vi l a nd E n v ir o n me n ta l E n g i nee r i n g
Ne w J ers e y I n st it ute o f T echno lo gy 11. C ontract or G rant N o.
Ne wa r k, NJ 0 7 1 0 2 -1 9 8 2
13. T ype of R eport and P eriod C overed
12. Sponsoring Agenc y Name and Address
M a r c h 2 0 0 0 to Dec e mb e r 2 0 0 2
New Jersey Department of Transportation F e d e r a l H i g h wa y Ad m i n i s t r a t i o n
PO 600 U.S. Department of Transportation
Trenton, NJ 08625 Washingt on, D.C. 14. Sponsoring Agenc y Code
15. Supplementary N otes
1 6 . A bstract
Many concrete bridge decks develop transverse cracking and most of these cracks develop at early ages, some right after
construction, and some after the bridge has been opened to traffic for a period of time. Transverse cracks usuall y occur
when concrete is set and widen with time. These cracks have been observed in most geographical locations, and on many
superstructure types. It i s estimated that m ore than 100,000 bri dges i n the U ni ted S tates develop earl y transverse cracks.
These cracks are typi call y full depth l ocated 1-3 m (4-12 ft) apart al ong the length of the span and are usual l y observed
over transverse reinforcement. It has been reported that predominant form of deck cracking is transverse cracking. These
cracks reduce the service life of the structure and increase maintenance costs, which is of paramount importance in
highway maintenance activities. Transverse cracks accelerate reinforcement corrosion, especially in regions where deicing
chemi cal are appli ed. C orrosi on damage has been observed even on epoxy coated rei nforcing bars. Freeze-thaw cycl es of
water in cracks and leakage of water to supporting structures may also reduce service life of structure.
Cracks in concrete occur when a restraint mass of concrete tends to change volume. Volume change in concrete depends
on the properties of its constituents and their proportions as well as environmental conditions such as ambient temperature
changes and humi di ty. R estraint, whi ch i s basical l y due to com posite action of deck and gi rder, depends on desi gn
characteristics of the bridge (i.e., structural design factors). Construction techniques also contribute to volume change
and/or to degree of restraint of concrete mass.
Factors associated with mix design/material and construction procedures have been the subject of a significant number of
research studies over the past several decades. Structural design factors, however, have not been the subject of much
research i n the past and they were the m ai n thrust of thi s research stud y. U si ng 2-D and 3-D l inear and nonli near finite
element models many design factors such as girder stiffness, deck thickness, girder spacing, relative stiffness of deck to
girder, amount of reinforcements, etc. were studied. The research study also included a comprehensive review of the
existi ng l iterature as wel l as surve y of 24 bri dges i n the state of N ew Jerse y. R esul ts of each research task are presented
and discussed i n details. Furtherm ore, based on anal ytical resul ts and li terature revi ew effect of various factors are
quantified and specific recommendations for possible consideration in design are made. These are classified under the
three m aj or categori es, nam el y: 1) materi al and mi x desi gn, 2) construction practi ce and ambi ent conditi on factors, and 3)
structural design. A sim ple window application program to more accurately estimate deck stresses during design is also
devel oped under this stud y. Future research needs are al so i denti fi ed.
1 7 . Key W ords 1 8 . Dist ribution Stat ement
Co ncr ete , D ec k, Cr ac ki ng, Cr ac ks, T r a n s ver se,

B r id ge s, Ca u se s o f Cr ac ki n g , Co ntr o l o f
Cr a c ki n g
1 9 . Sec urit y Class (of t his report ) 2 0 . Sec urit y Classi f. (of this page) 21. N o of
21. No. ofPages
Pages 2 2 . Pric e
U ncl as s i fi ed U ncl as s i fi ed 2 0203

4
Form DOT F 1700.7 (8-69)
ACKNOWLEDGEMENT
This research study was supported by the New Jersey Department of
Transportation/FHWA and the National Center for Transportation and Industrial
Productivity (NCTIP) at NJIT. The assistance of NJDOT staff, especially Fred Lovett
from the Bureau of Materials, Jose Lopez and Gerry Sellner from the Bureau of
Structures & Design, and Raj Chawla the project monitor in conducting the field surveys
and/or providing material and design data on bridges is greatly appreciated. The results
and conclusions are those of the authors and do not necessarily reflect the views of the
sponsors.
iv
TABLE OF CONTENTS
DISCLAIMER STATEMENT ...........................................................................................II
FORM DOT 1700.7 .......................................................................................................III
ACKNOWLEDGEMENT ............................................................................................... IV
TABLE OF CONTENTS ................................................................................................. V
LIST OF FIGURES...................................................................................................... VIII
LIST OF TABLES ........................................................................................................ XII
LIST OF SYMBOLS .................................................................................................... XIV
SUMMARY ......................................................................................................................1
Structural Design Factors.............................................................................................1
Mix Design (Material) ...................................................................................................1
Construction Practices .................................................................................................2
Training and Implementation........................................................................................2
Future Research Needs ...............................................................................................2
INTRODUCTION .............................................................................................................3
Problem Statement ......................................................................................................3
Research Objectives....................................................................................................5
Research Approach .....................................................................................................5
Report Organization.....................................................................................................6
LITERATURE REVIEW...................................................................................................7
Background..................................................................................................................7
Causes of Transverse Deck Cracking..........................................................................7
Material and Mix Design Factors..................................................................................9
Construction practice and Ambient Condition Factors ...............................................14
Structural Design Factors...........................................................................................17
Research Needs ........................................................................................................20
FIELD STUDY ...............................................................................................................21
Objectives and Scope ................................................................................................21
Data Sources .............................................................................................................23
Database....................................................................................................................26
Evaluation of Survey Results .....................................................................................27
Structural Design Factors...........................................................................................27
Material Properties and Mix Design Factors ..............................................................36
Construction Techniques and Ambient Condition Factors .........................................38
Remarks.....................................................................................................................39
v
CAUSES OF VOLUME CHANGE IN CONCRETE AND RESTRAINED SHRINKAGE
TESTS...........................................................................................................................43
Volume Change and Cracking in Concrete................................................................43
Bridge Deck Cracking ................................................................................................48
Restrained Shrinkage Tests.......................................................................................49
FINITE ELEMENT ANALYSIS ......................................................................................55
Two Dimensional Model of Deck and Girder..............................................................55
Results of Two Dimensional Analyses .......................................................................60
Three Dimensional Model of Deck and Girder ...........................................................81
Results of Three Dimensional Analyses ....................................................................87
A SIMPLE METHOD FOR ESTIMATING STRESSES CAUSING TRANSVERSE
CRACKING IN CONCRETE BRIDGE DECKS ...........................................................101
Compatibility Equations ...........................................................................................101
Adequacy of Assumptions........................................................................................105
Results and Comparison to Finite Element Analysis................................................108
Other Boundary Conditions......................................................................................108
Loading definition for simplified procedure...............................................................113
Windows Based Application.....................................................................................116
CONCLUSIONS AND RECOMMENDATIONS ...........................................................117
Structural Design Factors: Recommendations.........................................................117
Mix Design: Recommendations ...............................................................................119
Construction Practices: Recommendations .............................................................119
Future Research Needs ...........................................................................................120
REFRENCES ..............................................................................................................123
APPENDIX A: BRIDGE INFORMATION AND SURVEY RESULTS ..........................126

Organization.............................................................................................................126
Bridge Information and Survey Results....................................................................127
Structure Number 0206-165.....................................................................................128
vi
vii
LIST OF FIGURES
Figure 1. Corrosion due to transverse deck cracks in a bridge in New Jersey. ...............3
Figure 2. Transverse deck cracks on a bridge deck in New Jersey.................................4
Figure 3. Mechanism of transverse cracking. (a) Concrete is poured. (b) Concrete
shrinks, (c) Due to restraint from girder, concrete shrinkage produces downward
deflection. (d) Tensile stress is developed in deck, which causes transverse cracks.
.................................................................................................................................8
Figure 4. Effect of source and type of concrete on curing temperature (Babaei and
Purvis, 1994). .........................................................................................................11
Figure 5. Cracking as a function of bar size, slump and cover (Dakhil et al., 1975). .....13
Figure 6. Geographical distribution of bridges surveyed................................................22
Figure 7. Visual bridge evaluation form. ........................................................................24
Figure 8. Typical NJDOT Inspection/Testing datasheet. ...............................................25
Figure 9. Percent of cracked bridge decks with different end condition.........................29
Figure 10. Boundary condition: girders are bounded.....................................................29
Figure 11. Structural detail of the end condition shown in Figure 10. ............................30
Figure 12. Percent of cracked bridge decks with different skewness. ...........................31
Figure 13. One dimensional finite element model..........................................................33
Figure 14. Modeling of traffic loads for 1D model. .........................................................34
Figure 15. A picture of microfilm of design sheets. ........................................................35
Figure 16. Rt. 133 bridge deck crack.............................................................................40
Figure 17. Watson Creek Bridge (West Bound): Cracked continuity joints (bridge
No.1130-152). ........................................................................................................41
Figure 18. Watson Creek Bridge (East Bound): Un-cracked continuity joints (bridge
No.1130-153). ........................................................................................................42
Figure 19. Drying shrinkage of concrete........................................................................44
Figure 20. Crack in plastic concrete due to uneven settlement. ....................................45
Figure 21. Effect of w/c and silica fume on tensile stress produced by autogenous
shrinkage of concrete (Paillere et al., 1989). ..........................................................46
Figure 22. Effect of w/c, cement fineness, silica fume and admixtures on autogenous
shrinkage (Tazawa et al., 1995). ............................................................................47
Figure 23. Temperature distribution across the section for hot weather (Emanuel and
Hulsey, 1978). ........................................................................................................48
Figure 24. Ring test apparatus used by different researchers (Burrows, 1998).............50
Figure 25. Ring test apparatus used by Krauss and Rogalla (1996). ............................51
Figure 26. Cement rating criteria based on Blaine’s work (Burrows, 1998). ..................51
Figure 27. Bar test for studying cracking tendency of concrete (from Burrows, 1998)...53
Figure 28. Cracking frame developed in Germany (from Burrows, 1998)......................54
Figure 29. Typical results of cracking frame test (from Burrows, 1998).........................54
Figure 30. Cross-section of typical slab-on-girder bridge superstructure.......................56
Figure 31. Finite element model of deck, girders and diaphragms. ...............................56
Figure 32. Lateral nodal forces along the deck assuming uniform displacements.........57
Figure 33. Two-dimensional (2D) finite element model. ................................................58
Figure 34. Deck bottom and top stresses caused by 55-microstrain uniform deck
shrinkage for different boundary conditions............................................................60
viii
Figure 35. Deck bottom stresses caused by 55-microstrain uniform deck shrinkage for
different span lengths and boundary conditions. ....................................................62
Figure 36. Deck top stresses caused by 55-microstrain uniform deck shrinkage for
different span lengths and boundary conditions. ....................................................62
different deck thickness and boundary conditions. .................................................63
different deck thickness and boundary conditions. .................................................64
different girder spacing and boundary conditions. ..................................................65
different girder spacing and boundary conditions. ..................................................66
different ratio of girder/deck moment of inertia and boundary conditions. ..............67
different ratio of girder/deck moment of inertia and boundary conditions. ..............68
different composite moment of inertia and boundary conditions.............................69
different composite moment of inertia and boundary conditions.............................70
different area of longitudinal reinforcement and boundary conditions. ...................72
different distribution of longitudinal reinforcement and boundary conditions. .........73
different distribution of longitudinal reinforcement and boundary conditions. .........74
Figure 49. Suggested pouring sequence and lengths for different boundary conditions.
...............................................................................................................................76
Figure 50. Shrinkage stress profile. ...............................................................................79
Figure 51. Deck top and bottom stresses for different boundary conditions and (a)
shrinkage profile 1, (b) shrinkage profile 2, (c) shrinkage profile 3. ........................80
Figure 52. 3D finite element model................................................................................83
Figure 53. Bond-slip relationship (Houde, 1973). ..........................................................84
Figure 54. Shear Connector Detail. ...............................................................................84
Figure 55. Shear Connector load slip behavior (Yam an Chapman, 1968). ..................85
Figure 56. Concrete material model in tension. .............................................................86
Figure 57. Measured concrete deck strain and temperature of a simply supported
bridge. ....................................................................................................................87
Figure 58. Deformed shape of the girder and cracked deck at the end of analysis (pin-
roller boundary condition). ......................................................................................88
Figure 59. Deck cracks at end of span (bottom) and mid span (top) of the bridge at the
end of analysis(pin-roller case)...............................................................................88
Figure 60. Mid span girder deflection vs. shrinkage strain (pin-roller case)...................89
ix
Figure 61. Deck top and bottom stress over girder at mid span for pin-roller boundary
condition. ................................................................................................................89
Figure 62. Deck top stress over the girder at mid span and quarter span for pin-roller
boundary condition. ................................................................................................90
Figure 63. Deck top and bottom stress across the slab at mid span for pin-roller
Figure 64. Reinforcement stress over girder at mid span and quarter span for pin-roller
Figure 65. Deck cracks at end of span (bottom) and mid span (top) of pin-pin bridge at
the end of analysis. Circles indicate cracks for each element.(pin-pin case)..........93
Figure 66. Mid span girder deflection vs. shrinkage strain for pin-pin boundary condition.
...............................................................................................................................93
Figure 67. Deck top and bottom stress over girder at mid span for pin-pin boundary
condition. ................................................................................................................94
Figure 68. Deck top stress over the girder at mid span and quarter span for pin-pin
Figure 69. Deck top and bottom stress across the slab at mid span for pin-pin boundary
condition. ................................................................................................................95
Figure 71. Deformed shape of the girder and cracked deck at the end of analysis (fixed-
roller boundary condition). ......................................................................................96
Figure 72. Deck cracks at fixed end of span (bottom left), roller end of span (right end)
and mid span (top) of fixed-roller bridge at the end of analysis. .............................97
...............................................................................................................................97
Figure 74. Deck top stress at 1/3, half and 2/3 of span from fixed end for fixed-roller
Figure 75. Deck bottom stress at 1/3, half and 2/3 of span from fixed end for fixed-roller
Figure 76. Top reinforcement stress at 1/3, half and 2/3 of span from fixed end for fixed-
roller boundary condition. .......................................................................................99
Figure 77. Stress of deck elements for fixed-fixed boundary condition........................100
Figure 78. Temperature profile along the section. .......................................................102
Figure 79. Compatibility forces and moments and temperature distribution. ...............102
Figure 80. Finite element model of deck and girder (ANSYS V5.5, 1998)...................106
Figure 81. Distribution of nodal forces in longitudinal direction for pin-roller deck girder
system between deck and girder..........................................................................107
Figure 82. Distribution of nodal forces in direction normal to deck for pin-roller deck
girder between deck and girder. ...........................................................................107
Figure 83. Superposition principle applied for finding deck Stresses in different
boundary conditions. ............................................................................................109
Figure 84. Comparison of deck bottom stresses for different boundary conditions. ....112
Figure 85. Comparison of deck bottom stresses for different boundary conditions. ....113
Figure 86. AASHTO temperature gradient profile for steel and concrete girders.........115
Figure 87. Typical input and output window of application. .........................................116
x
Figure 88. Details of controlled composite action connector. ......................................122
Figure 89. Stress-strain curve for typical hyperealstic materials..................................122
Figure 90. Crack on the deck (Bridge No 0713-151). ..................................................132
Figure 91. Construction vehicle traffic on the bridge (Bridge No. 0713-151). ..............133
Figure 92. Crack on the deck (Bridge No. 0713-151). .................................................133
Figure 93. Crack extending into the parapet (Bridge No. 0713-151). ..........................134
Figure 94. Crack on deck and parapet (Bridge No. 0173-151). ...................................135
Figure 95. A view from underneath the bridge (Bridge No. 1013-151). .......................137
Figure 96. Girder end condition at the abutment (Bridge No. 1013-151). ....................138
Figure 97. Cracked continuity joint (Bridge No. 1130-152). .........................................143
Figure 98. Another cracked continuity joint (Bridge No.1130-152). .............................143
Figure 99. End condition in simply supported spans (Bridge No 1130-152). ...............144
Figure 100. Transverse cracks on the deck (Bridge No.1130-153). ............................147
Figure 101. Un-cracked continuity joint (Bridge No.1130-153). ...................................147
Figure 102. Deck cracks at the bridge end (Bridge No. 1143-466)..............................157
Figure 103. Wide view of deck cracks at the bridge end (Bridge No. 1143-466). ........158
Figure 104. Close view of a crack (Bridge No. 1143-166). ..........................................158
Figure 105. Girders embedded in end diaphragm (Bridge No.1143-168)....................162
Figure 106. Continuity joint (Bridge No. 1143-168)......................................................163
Figure 107. Transverse cracks on bridge deck (Bridge No. 1143-169). ......................165
Figure 108. Repaired transverse crack (Bridge No. 1143-170). ..................................167
Figure 109. Close up view of repair patch (Bridge No. 1143-170). ..............................168
Figure 110. Cracks on the deck (Bridge No.1143-171). ..............................................170
Figure 111. Another marked crack on the deck (Bridge No.1143-171)........................171
Figure 112. Close up view of deck cracks at the ends (Bridge No.1143-172). ............173
Figure 113. End diaphragm (Bridge No.1143-176)......................................................181
Figure 114. A view from underneath the bridge (Bridge No.1143-176). ......................182
Figure 115. Transverse deck cracks on Bridge No.1143-176 (NJ Turnpike in the
background). ........................................................................................................182
Figure 116. Transverse (and longitudinal) cracks on deck (Bridge No.1143-177). ......184
Figure 117. A view of the bridge (Bridge No.1149-168)...............................................186
xi
LIST OF TABLES
Table 1. Construction data for bridges surveyed. ..........................................................22
Table 2. Results of 1D FEA ...........................................................................................35
Table 3. Shrinkage correction factors for initial moist curing (AASHTO). ......................45
Table 4. Geometric and design information for the bridge modeled (Hackensack Ave
over NJ Route 4). ...................................................................................................55
Table 5. Deck top and bottom tensile stresses caused by 55-microstrain uniform deck
shrinkage for different boundary conditions............................................................60
Table 6. Deck top and bottom stresses caused by 55-microstrain uniform deck
shrinkage for different span lengths and boundary conditions................................61
shrinkage for different deck thickness and boundary conditions. ...........................64
shrinkage for different girder spacing and boundary conditions. ............................65
shrinkage for different ratio of girder/deck moment of inertia and boundary
conditions. ..............................................................................................................67
Table 10. Different ratio of girder/deck moment of inertia used in calculation and
corresponding deck and girder dimensions. ...........................................................68
Table 11. Different ratio of composite moment of inertia used in calculation and
corresponding deck and girder dimensions. ...........................................................68
shrinkage for different composite moment of inertia and boundary conditions. ......69
Table 13. Deck top stresses caused by 55-microstrain uniform deck shrinkage for
Table 14. Deck top stresses caused by 55-microstrain uniform deck shrinkage for zero
area of longitudinal reinforcement and boundary conditions. .................................71
shrinkage for different distribution of longitudinal reinforcement and boundary
conditions. ..............................................................................................................74
Table 16. Deck top and bottom stresses caused deck dead load..................................75
Table 17. Residual top and bottom deck stresses caused by pouring sequence (a) for
pin-roller boundary condition. .................................................................................76
Table 18. Residual top and bottom deck stresses caused by pouring sequence (b) for
pin-pin boundary condition. ....................................................................................76
Table 19. Residual top and bottom deck stresses caused by pouring sequence (c) for
fixed-roller boundary condition. ..............................................................................77
Table 20. Residual top and bottom deck stresses caused by pouring sequence (d) for
fixed-pin boundary condition...................................................................................77
Table 21. Residual top and bottom deck stresses caused by pouring sequence (e) for
fixed-fixed boundary condition................................................................................77
Table 22. Residual bottom deck stresses caused by pouring sequence (f) for two span
continuous girder....................................................................................................77
Table 23. Residual top deck stresses caused by pouring sequence (f) for two span
xii
Table 24. Residual bottom deck stresses caused by pouring sequence (g) for three
span continuous girder. ..........................................................................................78
Table 25. Residual top deck stresses caused by pouring sequence (g) for three span
Table 26. Deck top and bottom stresses caused by different shrinkage profile in deck for
different boundary conditions. ................................................................................79
Table 27. Design characteristic of modeled bridge......................................................106
Table 28. Comparison of FEA and compatibility equations results for the pin-roller
supported deck-girder system. .............................................................................108
Table 29. Comparison of compatibility equations for plane stress and plane strain
assumption. ..........................................................................................................108
Table 30. Comparison of FEA and compatibility equations results for the different
boundary conditions. ............................................................................................109
Table 31. Basis for temperature gradients (AASHTO LRFD 1998). ............................115
xiii
LIST OF SYMBOLS
Agirder , AComposite = Cross sectional area of girder and composite section
AElement = Area of Element
Ari = Total cross sectional area of reinforcement layer i
Ai = Cross sectional area of section i
Ctop , Cbottom = Top and bottom reinforcement cover
d ti = Distance from centroid to top fiber of section i
d bi = Distance from centroid to bottom fiber of section i
d bg , d tg = Distance of top and bottom fiber of girder from centroid
d bc , d tc = Distance of top and bottom fiber of composite section from
centroid
d rebar = Reinforcement diameter
Ec = Modulus of elasticity of concrete
Ei = Modulus of elasticity of section i
E Element = Modulus of elasticity of element
E girder , Ecomposite = Modulus of elasticity of girder and composite section
Fi = Force resultant of stresses at interface section i and i+1
Fri = Force in reinforcement layer i (positive denotes tensile force)
fr = Modulus of rupture of concrete
f c′ = Compressive strength of concrete
hi = Height of section I
I Girder , I Composite = Girder and transformed composite section moment of Inertia
Ii = Moment of inertia of section i
K vs , K h = Size and humidity correction factors
L, L1 , L2 , L3 = Span length
l element = Element length
M , M1, M 2 = Redundant Moment
P = Redundant force
Qi = Moment resultant of stresses at interface section i and i+1
Ri = Curvature of section i
S b i , S ti = Bottom and top section modulus of section i
Sr i = Section modulus at level of reinforcement
t = Time in days after measured after curing
t = Deck thickness
ti = Time
T0 , T1 , T2 ,..., T8 = Girder and deck temperature changes
xiv
Tri = Temperature at reinforcement layer i
u = Reinforcement bond stress
v/s = Volume to surface ration
w = Reinforcement slip
αi = Coefficient of thermal expansion of section i
α ri = Coefficient of thermal expansion of rebar layer i
ε Bottomi = Strain at the bottom of section i (Tensile strain is positive)
ε Topi = Strain at the top of the section i (Tensile strain is positive)
ε sh ,t = Shrinkage strain
ε sh,u = Ultimate Shrinkage strain
σ DeckBottom , σ DeckTop = Deck bottom and top stresses (Tensile stress is positive)
∆σ DeckBottom , ∆σ DeckTop = Additional deck bottom and top stresses (Tensile stress is
positive)
xv
SUMMARY
A significant number of concrete bridge decks develop transverse deck cracking, often
at early ages. Cracks in concrete occur when a restraint mass of concrete tends to
change volume. Amount of volume change in concrete depends on mix design and
construction procedures. Factors associated with mix design/material and construction
procedures have been the subject of a significant number of researches over the past
several decades. Restraining effect arises from the composite action between the deck
and girder, which is mostly controlled by structural design factors although partly
dependent on construction practices too. Structural design factors have not been the
subject of much research in the past and they were the main thrust of this research
study. Using 2-D and 3-D linear and nonlinear finite element models many design
factors such as girder stiffness, deck thickness, girder spacing, relative stiffness of deck
to girder, amount of reinforcements, etc. were studied. The research study also included
a comprehensive review of the existing literature as well as survey of 24 bridges in the
state of New Jersey. Based on the results of this research study the following
recommendations are made:
Structural Design Factors
• Specify an upper limit on actual concrete strength vs. the design value.
• Minimize the ratio of girder to deck stiffness.
• Boundary restraints should be consistent with design.
• Time-dependent loadings must be considered in design of bridges with integral
abutments.
• Employ more flexible superstructures.
• Use uniform reinforcement meshes.
• Design should consider AASHTO Article 3.12 (A simple design tool has been
developed under this study to facilitate this).
Mix Design (Material)
• Reduce cement content to 650-660 lb/yd3, and consider using fly ash.
• Limit water cement (w/c) ratio to 0.4-0.45.
• Use AASHTO specification Type II cement.
• Adopt a restraint shrinkage test.
• Consider using type K shrinkage compensating concrete when available.
• Use aggregate size and shape as discussed in the report.
1
Construction Practices
• As specified in NJDOT Specs, ensure that curing starts immediately after

finishing and wet cure for at least 7 consecutive days.
• If “early-open” is not an issue consider 14-day wet curing.
• Make use of evaporation rate chart proposed by ACI and cast the deck in mild
temperatures.
• Record wind speed and humidity during construction for future reference.
• Give consideration to pouring sequence as proposed under this report.
Training and Implementation

Many of the recommendations are straightforward and can be implemented
immediately. However, couple of these recommendations may have implications and/or
may require some training. NJIT will work with NJDOT, specifically Bureaus of
Structures & Design and Materials, to coordinate implementation of these
recommendations.
Future Research Needs

Several areas of future research have been identified that are all very important. For
example the proposed connector called Controlled Composite Action (CCA), as
discussed in last chapter, has great potential and should be seriously considered for
future investigation. CCA has the high potential to eliminate, as oppose to minimizing,
the problem of deck cracking.
2
INTRODUCTION
This chapter introduces and explains transverse deck cracking in concrete bridge decks
and its effects on the bridge performance. It also discusses the needs for further
research, such as this study, by presenting and discussing the results of prior research
activities on this subject. The objectives and plan for this research study are also
discussed in some details. Finally, the report organization is presented.
Problem Statement
New Jersey Department of Transportation’s (NJDOT) Bureau of Construction and
Materials have noticed an increase in the number of cracking in concrete bridge decks.
Literature indicates that indeed many concrete bridge decks develop transverse
cracking and most of these cracks develop at early ages, many, right after construction.
These cracks are typically full-depth and spaced 3 to 10 ft apart along the length of the
bridge. Transverse cracks can accelerate corrosion of reinforcing steel, deteriorate deck
concrete, possibly cause damage to underneath components of the bridge, and damage
bridge esthetic. As a result of these adverse effects of transverse cracking, the
maintenance costs will increase and ultimately the service life of the bridge system will
be shortened.
A picture of a full width transverse crack in a bridge deck is shown in Figure 1 Although
not quite clear, this is shown to emphasize the extent of damage that a crack can cause
once it is initiated. The crack has propagated all the way into the monolithic parapet.
The arrow shows the location of the crack. As indicated by the dark water trace on the
underneath girder the reinforcements are also corroded. Maintenance and repair costs
associated with damage like this to our infrastructure put a significant burden on
highway agencies’ resources. Figure 2 shows picture of another bridge deck crack that
is quite wide and extends the full width of the deck.
Figure 1. Corrosion due to transverse deck cracks in a bridge in New Jersey.
3
Figure 2. Transverse deck cracks on a bridge deck in New Jersey.
There have been significant numbers of studies on the cause of transverse deck
cracking, as it will be discussed in the next chapter. However, the causes are not yet
fully understood and the problem still exists. Previous studies were mostly focused on
concrete mix design and improvement through changes to construction practices to
alleviate shrinkage problem. In many instances there are major disagreements on the
factors affecting transverse cracking indicating the need for further research. For
examples, Dakhil et al. (1975) report a direct relationship between an increase in
cracking and an increase in concrete slump while Cheng and Johnston (1985) have
observed a decrease in transverse cracking in concrete bridge decks with increasing
slump. Contradiction on the effect of girder type is another important example on the
lack of full understanding of the causes of this phenomenon. Meyers (1982) indicates
that structures supported on wide flange and composite steel-plate girders exhibited
much more cracking than those constructed on other systems. However, the results of
our survey showed that the percentage of the cracked decks supported on prestressed
concrete girder are higher. Furthermore, the review of the construction and mix design
documents for the bridges surveyed and NJDOT 1996 Standard Specifications for Road
and Bridge Construction and 1998 Supplemental Specification (hereafter NJDOT
Specs) show that most of the important recommendations of previous studies are
already included in mix design and construction of bridge decks in New Jersey.
However, transverse cracks are still observed on some of the newly constructed bridge
4
decks indicating the contribution of other factors and the need for further study to
identify these factors and propose remedies.
Research Objectives
The overall objective of this research is to investigate the cause(s) of transverse
concrete deck cracking and to develop solutions for possible implementation in design
and construction of new concrete bridge decks. Thus, the tasks to achieve these
objectives are as follow:
• Literature review and evaluation of previous research recommendations along

with review of current design and construction practices in New Jersey,
• Development of a database and statistical analysis based upon survey of bridge
decks, and evaluation of results in light of literature review,
• Development of 3-D finite element models and parametric study,
• Development of simplified analytical procedures for design purposes,
• In-field evaluation and response measurements bridge,
• Development and formulation of design recommendations.
This report discusses the results of the study. In the following section the research
approach for tasks conducted during this study is briefly discussed.
Research Approach
There have been significant numbers of studies on the cause of transverse deck
cracking. These studies were reviewed to identify their most important
recommendations and findings. These recommendations were compared to NJDOT
Specs (1998) to identify any discrepancy and to make recommendation for possible
adoption. The literature findings are also critically reviewed to adopt appropriate
research approach and to identify areas for further work.
24 bridges were surveyed in New Jersey. Their conditions were not known a priori. The
criteria for selecting these bridges were that i) they are built after 1994 (5 year old age
at the time of this research), and ii) they span more than 75-ft, single or multiple spans.
It turned out that most bridges selected were cracked, although there were a few
uncracked bridges among those surveyed. During the surveys crack information such
as spacing, width and location are qualitatively recorded. The design and construction
documents for the bridges surveyed were also reviewed. Based on the information
collected a database was developed and it was used in the subsequent statistical
analysis. The objective of the statistical analysis was to identify major factors causing
transverse deck cracking in New Jersey. Using these results and the knowledge gained
through literature survey a narrower list of factors were selected for a more focused
study using finite element analysis.
5
Several Finite Element (FE) models are developed. A comprehensive linear and
nonlinear Finite Element Analysis (FEA) were conducted to study crack patterns on
bridge deck and quantify the effect of design factors affecting deck transverse cracking.
These factors were selected based on the literature review and field surveys conducted
within initial phases of this study. Based on the results of the FEA further
recommendations are made to reduce crack tendency in bridge decks.
To study the effect of ambient and hydration temperature as well as shrinkage on deck
strains and stresses, four bridge decks were instrumented. Temperature and strains
were monitored and recorded during each test. These data are presented and analyzed
to identify the effect of hydration, daily temperature, and shrinkage on transverse deck
cracking.
Finally, a simplified method to evaluate stresses in concrete bridge deck due to

shrinkage, hydration effects and ambient temperature changes was developed. It is
proposed that this method be used during design stage to evaluate different
superstructure design parameters with respect to transverse deck cracking. To further
facilitate use of this method, a user-friendly windows based application is developed,
which implements this method.
Report Organization
Results of this study are organized in seven chapters and one appendix. Following this
introductory chapter on problem statement and research objective, the literature reviews
are presented and discussed in chapter two. Chapter three discusses the details of field
study, data gathering, database development, and discussion of field study results.
Chapter four examines the causes of volume change in concrete and relevant test
methods. 2D and 3D finite element study results are presented in chapter five. Details of
the simple method developed to estimate deck stresses are discussed in chapter six.
Finally, chapter seven presents recommendations of this study. There are also an
appendix, which contains details of the information collected for each bridge during field
surveys.
6
LITERATURE REVIEW
This chapter presents the results of a comprehensive literature review on the cause of
transverse deck cracking. It includes experimental and analytical research results as
well as survey studies on the effects of different factors on concrete deck cracking.
Consistent with past work on the subject, causes are classified under three categories,
namely: 1) material and mix design, 2) construction practice and ambient condition
factors, and 3) structural design. Finally areas for further research are identified
Background
Many concrete bridge decks develop transverse cracking and most of these cracks
develop at early ages, some right after construction, and some after the bridge has
been opened to traffic for a period of time. Transverse cracks usually occur when
concrete is set (Iowa DOT, 1986; Kosel et al., 1985; PCA, 1970) and widen with time
(Cady et al., 1971; Horn et al., 1972; Ramey et al., 1997). These cracks have been
observed in most geographical locations, and on many superstructure types (Krauss
and Rogalla, 1996). It is estimated that more than 100,000 bridges in the United States
develop early transverse cracks (Krauss and Rogalla, 1996). These cracks are typically
full depth (Krauss and Rogalla, 1996; La Fraugh, 1989; Mc Keel, 1985) located 1-3 m
(4-12 ft) apart along the length of the span (Cheng and Johnson, 1985; Kosel et al.,
1985; PCA, 1970) and are usually observed over transverse reinforcement (Krauss and
Rogalla, 1996; Iwoa DOT, 1986; Mc Keel, 1985; PCA, 1970; Ramey et al., 1997). It has
been reported that predominant form of deck cracking is transverse cracking (Cady et
al., 1971; Mc Keel, 1985; PCA, 1970; Ramey et al., 1975). These cracks reduce the
service life of the structure and increase maintenance costs, which is of paramount
importance in highway maintenance activities. Transverse cracks accelerate
reinforcement corrosion, especially in regions where deicing chemical are applied (Mc
Keel, 1985; Perfetti et al., 1985). Corrosion damage has been observed even on epoxy
coated reinforcing bars (Perfetti et al., 1985). Freeze-thaw cycles of water in cracks and
leakage of water to supporting structures may also reduce service life of structure.
Although transverse cracking in bridge decks has been one of the main concerns of
designers and researchers for decades, effect of contributing factors on transverse
cracking is not fully understood yet. This chapter examines the state of knowledge on
transverse deck cracking, discusses areas for further research, and presents a set of
recommendations for reducing possibility of transverse deck cracking.
Causes of Transverse Deck Cracking

Cracks in concrete occur when a restraint mass of concrete tends to change volume.
Volume change in concrete depends on the properties of its constituents and their
proportions as well as environmental conditions such as ambient temperature changes
and humidity. Restraint, which is basically due to composite action of deck and girder,
7
depends on design characteristics of the bridge. Construction techniques also
contribute to volume change and/or to degree of restraint of concrete mass. Mechanism
of transverse cracking is shown in Figure 3.
Figure 3. Mechanism of transverse cracking. (a) Concrete is poured. (b) Concrete

shrinks, (c) Due to restraint from girder, concrete shrinkage produces downward
deflection. (d) Tensile stress is developed in deck, which causes transverse cracks.
Volume change in concrete bridge deck, as influenced by the properties of concrete, is

caused by drying-shrinkage, autogenous shrinkage, plastic shrinkage, thermal
shrinkage, and creep. Drying shrinkage is change in concrete volume due the change in
water content during the time after exposure to atmosphere. Autogenous shrinkage is
the change in concrete volume without change in its water content and usually occurs in
very low w/c ratios. Plastic shrinkage is referred to shrinkage caused by excessive
evaporation of surface water. Cooling of concrete after initial hydration is the cause of
thermal shrinkage. Creep strains, on the other hand, tend to counteract the effect of
shrinkage. Although relative importance of these factors is not completely quantified yet,
drying shrinkage and thermal shrinkage are considered by many studies to be the major
cause of concrete deck cracking cracking (Babaei et al., 1997, 1987; French et al.,
1999; Krauss and Rogalla, 1996; La Fraugh et al., 1985; PCA, 1970). Earlier studies
reported correlation between deck cracking, drying shrinkage (Babaei and Hawkins,
1987; La Fraugh, 1985; PCA, 1970) and higher placement temperatures (PCA, 1970).
Babaei and Purvis (1994) measured thermal and drying shrinkage of different mixes
and showed that mixes with higher thermal and drying shrinkage values tend to crack
more. Krauss and Rogalla (1996) through analytical study and further instrumentations
of bridge decks showed that drying shrinkage and temperature changes through the
8
section are responsible for deck cracking. In another study, Ducret et al. (1997) have
measured that mixes with lower peak hydration temperature produces less residual
stress in concrete. In a recent study, Frosch et al. (2002) through field instrumentation
and test of constructed deck in laboratory have conclusively shown that drying
shrinkage is the most important cause of transverse cracking.
Restraint of deck by girder against its volume change provides the condition for
cracking. Ducret et al. (1997) have measured that reducing the ratio of cross sectional
area of girder to deck reduces risk of cracking. However, relative effect of different
factors on the restraint of the composite system is not fully understood yet. Very little, if
any, effort has been done to reduce the restraint of deck girder system by changes in
design.
Construction practice, such as curing procedures, pouring sequence, and form type can
also affect deck cracking. Cady et al. (1971), in their study of 249 bridges in
Pennsylvania, have shown that the bridge decks constructed by certain contractors
have more transverse deck cracking than other decks in the study and concluded that
construction practice plays a major role in cracking of concrete bridge decks. Several
researchers (e.g. Horn et al., 1975; Kochanski et al., 1990; PCA, 1970) have
emphasized effect of curing and weather. Although construction methods may increase
or decrease risk of cracking but cracking have been observed on decks built with
different construction techniques. Consequently, transverse deck cracking cannot be
solely attributed to a certain type of construction technique.
In the following sections, the state-of-the-art on the causes and control of transverse
deck cracking are presented under three main categories of material and mix design,
construction practice and ambient conditions, and structural design factors.
Material and Mix Design Factors

Aggregate
Type, size, volume and properties of aggregate have pronounced effects on concrete
properties. Suggestion on aggregate from prior studies include using largest possible
size of aggregate (Babaei and Purvis, 1994; Kosel et al., 1985; Krauss and Rogalla,
1996; PCA, 1970), maximizing aggregate volume (French et al., 1999; Kochanski, 1990;
Kosel et al., 1985), and using low shrinkage aggregate (Horn et al., 1972; Krauss and
Rogalla, 1996; PCA, 1970) to reduce cracking. To identify low shrinkage aggregate
Babaei and Purvis (1994) further recommended that coarse aggregate absorption
should not be more that 0.5 percent and the fine aggregate absorption should not be
more that 1.5 percent and that aggregates should have high specific gravity. In general,
concrete mixes with good quality, clean, low shrinkage aggregate with high aggregate to
paste ratio have been observed to perform better.
9
Water Content
While Horn et al. (1975) noticed little correlation between cracking and water content,
Schmitt and Darwin (1999) found increased cracking with increased water content and
recommended reducing water content. Other researchers (Babaei and Hawkins, 1985;
Babaei and Purvis, 1994; Issa, 1999) have also made similar recommendation. Schmitt
and Darwin (1999) further suggested that volume of water and cement should not
exceed 27% of total volume of concrete, and Babaei and Purvis (1994) recommended
the maximum water content of 192 kg/m3 (323 lb/yd3).
Cement Type
Several studies have found that cement type has a significant effect on cracking. It is
accepted among researchers that use of type II cement reduces cracking and several
studies recommended the use of type II cement in bridge deck construction (Horn et al.,
1975; Krauss and Rogalla, 1996; Babaei and Purvis, 1994; Kosel et al., 1985; La
Fraugh, 1989). The better performance of type II cement is usually attributed to reduced
early thermal gradient and shrinkage of type II cement Figure 4 shows the effect of
cement type and source on curing temperature as reported by Babaei and Purvis
(1994).
Cement Content
Many studies have observed more cracking with higher amount of cement in the
concrete mix (French et al., 1999; Krauss and Rogalla, 1996; La Fraugh, 1989; Iwoa
DOT, 1986; Kochanski et al., 1990; Kosel et al., 1985; La Fraugh, 1989; Schmitt and
Darwin, 1999). The adverse effect of higher cement content is usually related to higher
drying shrinkage, higher temperature rise during hydration and higher early modulus of
elasticity of concrete. Different amounts of cement have been recommended as the
maximum acceptable cement content in concrete mixes:
• 360 kg/m3 (611 lb/m3) (Iwoa DOT, 1986)

• 370 kg/ m3 (620 lb/yd3) (La Fraugh, 1989)
• 446 kg/ m3 (725 lb/yd3) (Babaei and Purvis, 1994)
• 385-390 kg/ m3 (650-660 lb/yd3) (French et al., 1999)
10
Figure 4. Effect of source and type of concrete on curing temperature (Babaei and
Purvis, 1994).
Water/Cement Ratio
Schmitt and Darwin (1999) noticed reduced cracking with reduction in water cement
ratio. Indeed other studies also recommend reduction of water cement ratio in concrete
mix to reduce cracking (French et al., 1999; Iowa DOT, 1986; Kochanski et al., 1999;
PCA, 1970). Reducing water cement ratio of concrete is believed to reduce shrinkage of
concrete. However Stewart and Gunderson (1969) found no relationship between high
w/c ratio and cracking. The following maximum water cement ratios have been
recommended:
• 0.48 (PCA, 1970)

• 0.41 (Iowa DOT, 1986)
• 0.40 (Kochanski et al., 1990)
• 0.40-0.45 (Ramey et al., 1997)
It is also recommended to reduce water cement ratio using water reducers and
pozzolans (La Fraugh, 1989).
Concrete Strength
There has been significant increase in concrete strength during the past decades.
Increased strength, which is usually accompanied by increase in cement content,
increase in paste volume and higher hydration temperatures, is blamed to cause more
cracking in concrete decks. In fact, Krauss and Rogalla (1996) related the increase in
deck cracking since 1970s to AASHTO’s 1973 increase of minimum strength from 3000
11
psi to 4500 psi and lowering w/c from 0.53 to 0.445. It may seem that high early
strength of concrete may reduce cracking but since the strength gain of concrete is
usually accompanied by a gain in modulus of elasticity, it can’t be easily said that higher
strength reduces cracking. There are no general agreements among studies that
considered this factor. Schmitt and Darwin (1999) noticed increased cracking with
increased compressive strength. However, Ramey et al. (1997) recommended
increasing compressive strength. On the other hand, Krauss and Rogalla (1996)
recommended the use of concrete with low early strength (i.e. 60 –90 day strength
should be specified).
Slump
There are many contradictions in results of the studies performed so far on the effect of
slump on deck cracking. Dakhil et al. (1975) in their experimental study reported
increased cracking with increasing slump (Figure 5). Some studies have recommended
reducing the slump (PCA, 1970; Babaei and Hawkins, 1987; Isaa, 1999; Kosel, 1985;
Schmitt and Darwin, 1999) and proposed values for maximum slump:
• 50 + 12 mm (2 + ½ in.) (PCA, 1970)

• 60+12 mm (2 ½ + ½ in.) (Iowa DOT, 1986)
However, Stewart and Gunderson (1969) found no relationship between high slump and
cracking, and Krauss and Rogalla (1996) mentioned that there is no relation between
slump and cracking tendency. On the other hand, Cheng and Johnson (1985) even
noticed a decrease in transverse cracking with an increase in slump.
Air Content
Air content is usually used to increase freeze thaw durability of concrete. But it may be
advantageous to use high values of air content in moderate and warm climates. Cheng
and Johnson (1985) observed that increase in air content reduces cracking. Schmitt and
Darwin (1999) even noticed significant decreases in cracking with air content more than
6% and recommend at least 6% air content. French et al. (1999) recommend air content
of 5.5-6%. In contrary, Stewart and Gunderson (1969) found no relationship between air
content and cracking.
12
Figure 5. Cracking as a function of bar size, slump and cover (Dakhil et al., 1975).
Admixtures
Effect of different types of admixtures on cracking is not completely understood yet.

Cady et al. (1971) reported that the use of retarder is not an important factor. Other
researchers (Horn et al., 1975) have reported the same observation. However, some
studies (Krauss and Rogalla, 1996; La Fraugh, 1989) encourage the use of retarders,
as they believe reduced rate of early temperature rise and early gain of modulus of
elasticity would reduce deck cracking. Studies (Krauss and Rogalla, 1996; Schmitt and
Darwin, 1999) have shown that use of silica fume may significantly increase cracking if
precautions are not taken to prevent plastic cracking. Krauss and Rogalla (1996) also
discourage use of calcium chloride and triethanolamine admixtures.
13
Construction practice and Ambient Condition Factors
Weather Condition and Concrete Temperature
Weather condition during placement of concrete and relative concrete temperature can
greatly affect deck cracking. While hot and cold weather conditions may result in poor
quality concrete, it is also believed that restraint to thermal variations contributes to
cracking (PCA, 1970). Thermal stresses developed at early age in concrete deck
depend greatly on concrete temperature and weather conditions. Concrete temperature
rises as a result of hydration while girder temperatures remains relatively unchanged.
This temperature change will cause thermal stresses in the section. Weather condition
and solar radiation may also increase these stresses resulting in more cracking. Studies
(Cheng and Johnson, 1985; Mayers, 1982; Schmitt and Darwin, 1999) have shown that
hot and cold weather increases cracking and several values for allowable ambient
temperatures, and concrete temperature during placement are proposed such as:
• Maximum concrete placement temperature of 27oC (80oF) (PCA, 1970)

• Minimum ambient temperature of 7.2oC (45oF) (Cheng and Johnson, 1985)
• Minimum and maximum ambient temperature of 4 and 32oC (40 and 90oF) and
reducing temperature difference between deck and girder. (French et al., 1999)
• Maximum concrete placement temperatures 27oC (80oF) (Krauss and Rogalla,
1996)
• Concrete temperature of at least 5-10oC (10-20oF) cooler that ambient
temperature (Krauss and Rogalla, 1996)
• Girder temperature of 12-24oC (55-75oF) should be maintained in clod weather
(Babaei and Purvis, 1994)
Some other studies specified the allowable differential temperature of deck and girder,
for example:
• Temperature difference of at least 12oC (22oF) for at least 24 hours is

recommended by Babaei and Purvis (1994)
Low levels of humidity and high wind speed can also increase cracking. Plastic
shrinkage cracks are often attributed to higher evaporation rates than concrete
bleeding, where evaporation rates increase with high temperatures, low humidity, and
high wind speed. Evaporation rates of concrete under different conditions can be found
using evaporation chart. Krauss and Rogalla (1996) recommended that special
consideration should be taken when evaporation rates are more that 1.0 kg/m2/hr (0.2
lb/ft2/hr) for normal concrete and 0.5 kg/m2/hr (0.1 lb/ft2/hr) for low w/c ratio concrete.
PCA (1970) recommends testing mixes for bleeding. Kochanski et al. (1990)
recommend estimating evaporation rate and reducing it to a maximum of 1.25 kg/m2/hr
(0.25 lb/ft2/hr).
14
Curing
Curing has a pronounced effect on the properties of hardened concrete such as

durability, and strength. Adequate and timely curing is a key factor in reducing cracking.
Importance of cracking is emphasized by almost all studies. Initial fogging (Horn et al.,
1975; Stewart et al., 1969), early curing (Horn et al., 1975; PCA, 1970; Babaei and
Hawkins, 1987), sprinkling water on concrete surface (PCA, 1970), applying wet
burlaps, and applying curing compounds (Stewart et al., 1969) are among the
recommendations proposed in literature. Krauss and Rogalla (1996) recommended
following procedure for curing:
• Use of fog nozzle water spray in hot weather to cool concrete and to cool the
steel and forms immediately ahead of placement – ponding of water on the forms
or plastic concrete should not be allowed.
• Use of wind breaks and enclosures when the evaporation rates exceed 1
kg/m2/hr (0.2 lb/ft2/hr) for normal concrete or 0.5 kg/m2/hr (0.1 lb/ft2/hr) for low
water cement ratio concretes susceptible to plastic cracking
• Application of water mist or monomolecular film immediately after strike-off or
early finishing.
• Application of white-pigmented curing compound as soon as bleed water
diminishes.
• Application of prewetted burlap as soon as concrete resist indentation – the
burlap must be kept continuously wet by continuous sprinkling or by covering the
burlap with plastic sheeting and periodic sprinkling.
• Continuation of wet curing for a minimum of 7 days, preferably 14 days – curing
should be extended in cold weather until the concrete has gained adequate
strength.
Extended curing time is also suggested by La Fraugh (1989). Kosel and Michols (1985)
and Frosh et al. (2002) recommended minimum curing of 7 days for type I and 14 days
for type II cement. To reduce temperature, Kochanski et al. (1990) further
recommended that decks be covered with permeable membranes rather than
impermeable ones.
Pour length and Sequence
Although earlier studies (Cheng and Johnson, 1985; Perfetti et al., 1985) reported that
puor length and sequence do not seem to influence cracking, however later studies
suggested that pour length, sequence, and rate may have some effects on deck
cracking and recommended specifying pouring sequence (Iwoa DOT, 1996) and
avoiding pouring irregularities (Horn et al., 1975). Kochanski et al. (1990) recommend
pouring concrete at a rate faster than 0.6 span lengths per hour. In his analytical study,
Issa (1999) attributes cracking to sequence of pour and recommends placing concrete
first in positive moment regions. Ramey et al. (1997) recommend a detailed pouring
procedure as follows:
15
• Place complete deck at one time when possible.
• Place simple span bridges one span per placement or if span is long place divide
deck longitudinally and place each stripe at one time. If this cannot be done too,
then place the center of span first and then place other portions.
• If multiple placements should be made on continuous beams, place middle spans
first and observe 72 h delay between placements. Use bonding agent to enhance
bond at joint.
Time of Casting
There is an indication that evening and nighttime casting can reduce the extent of
cracking. PCA (1970) recommended nighttime casting. Krauss and Rogalla (1996) also
recommended early or mid evening placing.
Finishing
It has been reported that early finishing reduces cracking (Horn et al., 1975; Krauss and
Rogalla, 1996). Horn et al. (1975) noticed that hand finishing increases cracking.
However, mechanical grooving is recommended by Krauss and Rogalla (1996). Stewart
and Gunderson (1969) also found that applying water or grout to concrete surface
during finishing operation has adverse effects on cracking.
Revolutions in Concrete Truck
Horn et al. (1975) noticed that excess revolution in truck do not affect cracking.
Vibration of Fresh Concrete
Sufficient vibration of concrete is essential to good concrete. Issa (1999) considers

insufficient vibration of fresh a contributing factor in concrete cracking. Horn et al. (1975)
noticed that under-vibrated areas tend to develop more cracks, however, over vibration
of concrete don’t cause any noticeable effect.
Construction Loads
Effect of traffic and construction loads on deck cracking is also not completely known.
Furr and Fuad (1982) found that no deterioration can be attributed to traffic in adjacent
lanes during construction, and Manning (1981) showed that good quality plain and
reinforced concrete is not adversely affected by jarring and vibrations of low frequency
and amplitude during the period of setting and early strength development. However,
Issa (1999) attributes cracking to weight and vibration of machinery
Dead load of structure also affects concrete deck cracking (Krauss and Rogalla, 1996).
It has been suggested that shoring girders may reduce deck cracking (Babaei and
Hawkins, 1987)
16
Form Type
Inconsistent results have been reported on the effect of form type on deck cracking.
Issa (1999) attributes cracking to weight of the forms and their deflection. Based on
survey results, Cady et al. (1971) reported that Stay-In-Place (SIP) forms perform better
than removable forms. Cheng and Johnson (1985) reported that use of SIP or
conventional forms have little effect on transverse deck cracking. While Krauss and
Rogalla (1996) have found that SIP forms sometimes increase cracking. Through an
experimental study, Frosh et al. (2002) have concluded that additional restraint from
stay-in-place forms contribute to cracking and recommended that some other type of
forms be used.

It should be noted that research on design factors in general, is very limited. However
there are a few studies, which have considered these factors qualitatively. These are
discussed in following sections.
Girder Type, Boundary condition, and Spacing
Several studies (Krauss and Rogalla, 1996; PCA, 1970; Cheng and Johnson, 1985;
Mayers, 1982; Frosh et al., 2002) have found that decks on steel girders tend to crack
more when compared to deck on concrete girders. It is believed that since concrete
girders conduct heat slower than steel girders (i.e. lower temperature gradients),
thermal stresses in concrete girder bridges are lower than steel girder bridges,
consequently, less cracking tendency is expected. Krauss and Rogalla (1996) found
that cast in place concrete girders and young prestressed girders have the best
performance while deep steel beams have performed worse. They also discouraged
design of prestressed composite bridges.
Girder end condition also has pronounced effect on deck cracking (French et al., 1999).
Cracking is more prevalent on continuous spans when compared to simple spans
(Krauss and Rogalla, 1996; Mayers, 1982; Cady et al., 1971; Cheng and Johnson,
1985), but cracks are observed on both types of spans. There seems to be no
significant difference among transverse cracks patterns on different type of structures.
Portland Cement Association study (PCA, 1970) indicated that regardless of type of
span the same pattern of uniformly spaced cracks are observed on decks supported on
steel girder.
Stud Configuration and Properties
Composite action of the deck and girder is basically due to response of shear studs.
However, there is no significant effort to reduce deck restraint through changes in stud
configuration. Although Krauss and Rogalla (1996) have found that girder restraint and
studs cause significant cracking, they don’t provide any suggestion on how to reduce
17
girder restraint through change in stud configuration and properties. The only
recommendation comes from French et al. (1999) where they have recommended fewer
studs with smaller rows and lengths but they don’t specify any practical guidelines.
Concrete Cover
Based on their experimental study, Dakhil et al. (1975) reported that concrete cover
over reinforcement is the most important factor affecting crack formation (See Figure 5).
Increased cover depth reduces risk of cracking, however, excessive increase in cover
depth increases probability of settlement cracks over reinforcement. Different values are
proposed as the optimum value of the cover depth over top reinforcing bars:
• Minimum of 38 mm. (1.5 in.) (PCA, 1970)

• 88 mm (3.5 in.) (Babaei and Hawkins, 1987)
• 50 mm (2 in.), where deicing chemicals are used use 64 mm (2 ½ in.) and
maintain 76 mm (3 in.) limit (Ramey et al., 1997)
• 38 - 76 mm (1.5 - 3 in.)
Contrary to these studies, Meyers (1982) found that decks with cover of 76 mm (3 in)
and more seems to be more susceptible to cracking.
Deck Thickness
Increase in deck thickness reduces deck cracking (French et al., 1999; Krauss and
Rogalla, 1996; Kochanski et al., 1990; Ramey et al., 1997; Horn et al., 1972; Mayers,
1982). Meyers (1982) observed that bridge decks with deck thickness of 25 cm (10 in.)
and more are less susceptible to cracking. Kochanski et al. (1990) recommends deck
thickness of 8 ½ to 9 in. French et al. (1999) recommended decks with thickness
greater than 16 cm (6 ¼ in.).
Reinforcement Type, Spacing, Size and Distribution
Reinforcing detail is of paramount importance in controlling cracks in concrete

structures. Bar size, type, spacing, and distribution affects cracking tendency of
concrete decks greatly. Dakhil et al. (1975) in their experimental study reported that
cracking increases with an increase in bar size (See Figure 5). Some other studies
(Babaei and Hawkins, 1987; Schmitt and Darwin, 1999) also observed the same
behavior and recommended limiting the bar size of deck. Since longitudinal bars control
deck stresses and reduce cracking tendency, an increase in the amount of longitudinal
reinforcement without increasing bar size is another recommendation proposed in some
studies (Krauss and Rogalla, 1996; PCA, 1970; Kochanski et al., 1990; Horn et al.,
1972; Frosh et al., 2002). Kochanski et al. (1990) as well as Ramey et al. (1997)
recommended the maximum top transverse bar size of No.5 and an increase in
longitudinal reinforcement. Krauss and Rogalla (1996) recommend use of No. 4 bars
with maximum spacing of 15 cm (6 in.). It is also suspected that due to subsidence of
fresh concrete over the reinforcing bars and formation of weak plane, deck tends to
18
crack over transverse reinforcing bars. So, French et al. (1999) also recommended
limiting transverse bar size and/or maximize transverse bar spacing.
Epoxy coated bars are used to control corrosion of reinforcing bars. Meyers (1982)
found that decks with epoxy bars tend to show more cracking. The same finding is
reported by Krass and Rogalla (1996) However contrary to their findings a study by
Iowa DOT (1986) recommended use of epoxy coated rebars to control cracking.
Details of construction are also important. Horn et al. (1975) noticed that tightly tied
reinforcements develop more small cracks initially than loosely tied reinforcements but
ultimately cracking was the same. Issa (1999) attributed some cracking to insufficient
reinforcing detail at joints between new and old decks
Ramey et al. (1997) suggest following recommendation for reducing deck cracking:
• Limiting the size of deck reinforcement to No. 5

• Reversing lying transverse and longitudinal rebars in the top mat and staggering
top and bottom rebars so as not to create significant plane of weakness and
using higher percentage of longitudinal steel
• Using ρ =0.002 for top mat longitudinal steel and using the same for bottom mat
and trying to use No. 4 bars
• Reducing splices
• Extending deck transverse steel full width
Section Stiffness
Results of the research studies on the effect of section stiffness on deck cracking are a
bit confusing. While Babaei and Hawkins (1987) suggested minimizing the flexibility of
structure, Ducret et al. (1997) showed in their study that with an increase in the ratio of
girder to deck area (which can be related to reducing flexibility) cracking tendency
increases. This finding is in agreement with the findings of French et al. (1999) and
Krauss and Rogalla (1996) who also reported that reducing deck stiffness reduces
cracking.
Since restraint volume change of deck is the principal cause of deck cracking, reducing
section stiffness seems to decrease deck cracking, however, this statement needs to be
verified with further research.
Vibration and Impact Characteristics
Perfetti et al. (1985) found no relationship between frequency of vibration of

superstructure, speed and impact parameters and transverse cracking. However,
Babaei and Hawkins (1987) suggested reducing the amplitude and frequency of
structure vibration under live load.
19
Traffic
Although some studies (Krauss and Rogalla, 1996; Stewart et al., 1969; Cady et al.,
1971) reported no relationship between daily traffic of bridge and tendency for deck
cracking, others (Mc Keel, 1985) observed that bridges that carry fewer trucks at lower
speeds exhibit less cracking than those that carry large number of truck at higher
speeds.
Research Needs
Despite of the large number of studies on concrete deck cracking, transverse deck
cracking is still a problem faced by many transportation agencies worldwide. There are
still areas on cause and control of concrete deck cracking that need to be investigated:
Many studies have considered material factors. However, there is no general

consensus on the range of different mix design factors (i.e. w/c ratio, cement content,
water content, etc.). More research in needed to quantify the acceptable range of
different factors.
Although research have shown that restraint of deck due to composite action of deck
and girder is the main cause of cracking, there is no significant effort to quantify and/or
reduce this restraint through design recommendations. Particularly, effect of shear studs
characteristics and section stiffness on deck cracking is almost unknown.
There seems to be more agreement among different researchers on the effect of

construction factors, however, due to constant change in construction techniques and
introduction of new concrete products like HPC, research is needed to identify
appropriate construction procedure for these products.
Standard repair procedures need to be developed and employed by transportation
agencies to repair already cracked decks. These procedures can be incorporated into
contract specification (Krauss and Rogalla, 1996).
A promising approach to reduce deck cracking is the use of shrinkage compensating

cement and/or fiber reinforcement as well as prestressed deck systems (Krauss and
Rogalla, 1996). However, more research is required before successful applications of
them.
Design tools for evaluation of deck stresses for various geometries, boundary
conditions, concrete mix and ambient condition are required (as specified by AASHRO
LRFD (1998) article 3.12.4). Such tools may be based on the earlier work of Krauss and
Rogalla (1996). One such a tool has been developed in this study and presented in
chapter 6.
20
FIELD STUDY
This chapter presents scope and objectives of the field study and provides the details of
data collection, data base development, statistical analysis, and the features of the
database. In the following sections, field study is discussed and the methods and
sources for different types of data in the database are explained. This chapter also
presents the results of field study and the statistical analysis of based on the data,
which were collected during the field study. Results and their limitations are presented
under three major categories: structural design, material properties and mix design, and
construction. In each part, results are compared to similar studies performed by other
researchers. These results are followed by discussions of research findings and some
interesting observations that were made during the field study
Objectives and Scope

A field study was initiated with the survey of 24 bridges in the state of New Jersey and
by collecting design and construction data on these bridges for subsequent statistical
analysis (Appendix A contains the data collected for all bridges). In the following section
the objectives and scope of the field study are discussed.
Objective
The main objective of these surveys was to identify factors that affect transverse deck
cracking in bridges in the state of New Jersey. Another objective of these surveys was
the evaluation of current deck mix design and construction practices in the state of New
Jersey based on the survey results and the results reported in the literature. Based on
this evaluation, recommendations with regard to material and mix design as well as
construction practices were made to improve bridge deck performance. Another
important objective of this field study was to help to focus the efforts of the second part
of the research study by narrowing down the list of important factors that need to be
investigated in more details. These are factors that either have not received proper
attention during past research, as discussed, or because these factors are related to
particular design and/or construction practice unique to the state of New Jersey. As it
will be discussed, the surveys and subsequent evaluation of the data do indeed support
the initial thrust of the research endeavor with regard to more emphasis on design
factors. These factors were investigated in details under this study using analytical and
finite element analyses.
Scope
The field surveys included 24 bridges from central and northern New Jersey. 20 of
those bridges surveyed were located in the Mercer County while Bergen, Essex,
Hunterdon and Monmouth Counties each had one bridge among the bridges surveyed.
Figure 6 shows the geographical distribution of the surveyed bridges.
21
The bridge condition with regard to transverse deck cracking was not know prior to their
selection. The only criteria used for selecting these bridges were span length (longer
than 85-ft) and age (built after 1994 and considered new). Both prestressed and steel
girder bridges were considered. Results showed that the inventory included both
cracked and uncracked bridges. However, majority of the bridges, 18 bridges or 75%,
were cracked. Table 1 shows the construction year of the bridges included in the
survey.
Figure 6. Geographical distribution of bridges surveyed.
Table 1. Construction data for bridges surveyed.

Year Built 1995 1996 1997 1998 1999 2000
No of Bridges 6 2 0 13 2 1
Figure 3.
All bridges considered have composite decks. 8 of them have steel girders while the
rest were supported on prestressed concrete girders. The bridges surveyed bridges
have various span lengths with the maximum span length of 175 ft. Most of them
consisted of 1, 2 or 3 spans (83 percent), but it also included bridges with up to 12
spans.
There were different support conditions among the bridges surveyed. There were 8
spans with simply supported steel girders and 17 continuous multiple spans with steel
22
girders. Considering prestressed concrete girders, 11 spans were simply supported and
the rest of the spans (i.e., 43 spans) were continuous multiple spans. Note that
continuous spans are continuous in interior spans only and the exterior supports are
simply supported.
Except for one bridge, all of the bridges were open to traffic at the time of field survey.
Field survey included visual walk-by evaluation mostly of the top of the deck. During the
field surveys, bridge decks were visually evaluated and the crack information (crack
type, spacing, size and approximate location in the deck) were recorded qualitatively for
each bridge. The survey logs also included information regarding bridge location, type,
span number, span type, wearing surface and type of bearing. Furthermore, structural
plans and construction and mix design information for the bridges were collected from
NJDOT and important aspects of design and construction were determined for each
bridge. The survey observations and collected data are reported individually for each
bridge in appendix A.
Data Sources
There were three major source of data employed in development of the data base that
was consequently used for statistical analysis. These are:
• Field survey forms

• Structural plans
• NJDOT Inspection/Testing datasheets
Field Survey Form
Figure 7 shows the visual bridge evaluation form, which was specifically developed for
the field survey. Additional information, such as photos of cracks and structural
components, were also collected for some of the bridges in conjunction with completing
this form, which are all included in Appendix A. However, the crack data and general
information about the bridge were two main parts of information collected during the
field surveys.
Structural Plans
Structural plans were another source of information. In coordination with NJDOT staff at
the Bureau of Bridge Design, structural plans for all bridges were obtained and
important design information such as bridge dimensions, deck details, and girder details
were extracted.
23
Date of Evaluation: Structure Location:
Region:
Date Constructed: Type of Cracks (use additional forms for > 3
Number of Spans: span):
Single
Multiple Span Span 1 Span 2 Span 3
Simply Supported
Continuous Transverse
Degree of Skewness (quantify if possible): Longitudinal

None (straight) Others
Mild
Severe
Location of Cracks within the Deck:
Type of Beam:
Prestressed Span 1 Span 2 Span 3
Steel
Others Center
Deck Surface: End(s)

Concrete
Others
Latex
Unknown
Spacing of Cracks:
Deck Surface Texture:
Sawcut Span 1 Span 2 Span 3
Turf Drug
None – New Construction 0 – 2’
3’ – 5’
Type of Bearings:
Steel Bearings > 5’
Elastomeric Pads Irregular
Others:
Pictures/Video Available: Size of Cracks (quantify if possible):

Yes
Span 1 Span 2 Span 3
No
Additional Note: Hair Line
Typical
Wide
Team Members:
1. 2.
3. 4.
Figure 7. Visual bridge evaluation form.
24
NJDOT Inspection/Testing Datasheets
These datasheets, which are completed during the construction, contain different data
regarding construction and the mix design for each bridge. The results of the strength
and slump tests are also reported on these forms. These forms
are part of the documents that NJDOT holds for each bridge. Figure 8 shows a typical
datasheet. One part of these forms is completed on the day of casting and the other
parts are completed after conducting strength tests. Mix design, strength test results
and temperature measurements are three important parts of these datasheets that were
used in this study.
Figure 8. Typical NJDOT Inspection/Testing datasheet.
25
Database
The information in the database (see Appendix A) is divided into five major categories
for each bridge. These are: general information, structural design information, material
properties and mix design information, construction information, and crack information.
General Information
This part of the database contains information regarding bridge location and
construction year. This information is derived from the survey forms.
Structural Design Information
Using the structural plans, detailed information about the bridge design is extracted.
These data include:
• Bridge dimensions
• Number of spans
• Traffic direction
• Girder type
• Span type (e.g., if continuous at interior spans or simply supported)
• Span length and width
• Framing information (e.g., spacing of girders in each span)
• Deck design information (i.e. rebar details, thickness, cover depth and wearing
surface)
• Girder properties (e.g., area, depth, and moment of inertial)
• Shear studs spacing
Material Properties and Mix Design Information
This information is derived from NJDOT Inspection/Testing datasheets. Important mix

design information like water content, cement content, cement type, admixtures and
slump are reflected in the database. Note that since during the construction slump in
measured several times, the average value of these tests are taken as the
representative value of the slump for the deck concrete during casting. Similarly,
average compressive strength of all concrete deck specimens are reported as the
compressive strength of the deck.
26
Construction Information
In this part, based on the NJDOT Inspection/Testing forms the average air temperature
during the casting period and the average value of concrete temperature at the time of
casting for all the measurement is reported as the representative values.
Crack Information
Crack information, which was recorded during the field surveys, is reported in this part.
This information includes: crack type, approximate crack location on the deck, crack
spacing and its size.
The database program Microsoft Access (MS Access, 2000) was used to store and
organize the data. Microsoft Excel (MS Excel, 2000) program was mostly used for
presentation and analysis of the data. These two software packages are also employed
for statistical analyses. Charts are extensively used in presenting the result of analyses.
Evaluation of Survey Results

Factors identified relevant to deck cracking are evaluated for all bridges surveyed and
are compared with the results of other important researches (for an extensive literature
review please refer to chapter two). Results of literature review were used in the
selection of these factors. Of course, availability of data was another criterion in
determining the factors considered. In the following sections these results are discussed
and summarized using tables and graphs. However, the raw data for all bridges are
reported in Appendix A.

Design factors included in the survey are detailed in the Appendix. These factors are as
follow:
• Girder type
• End Condition
• Skewness
• Type of bearing
• Surface texture
• Wearing surface
• Deck thickness
• Bar size and spacing
27
Girder Type
Girder type is considered an important factor by various researchers. Some of the

pervious works such as NCHRP Report 380 (Krauss and Rogalla,1996) and Minnesota
DOT’s research (French et al., 1999), show that decks supported on steel girders are
more likely to crack than those supported on concrete girders. However, results of this
study show that 94 percent of decks supported by prestressed concrete girders were
cracked whereas only 38 percent of steel girder bridge decks cracked. The bias in this
case is more likely due to the fact that a great number of bridges with prestressed
concrete girders had more restraint (fixity) at their supports. This is a very important
factor, as it will be discussed throughout this report. Nevertheless, the contradiction
highlights the fact that there are other factors that play a significant role that complicate
the problem. As it will be discussed, it appears that it is the relative stiffness of the deck
with respect to the girder stiffness that is more important than the girder type.
End Condition
This study shows a good correlation between the end fixity and the cracking tendency of
the bridge decks. As it is shown in Figure 9, by increasing the fixity of the end supports
the percentage of cracked-decks increases. Note that with reference to this figure the
continuous end condition refers to situations where the girder is continuous over internal
supports. Fixed condition is when the abutment end is fixed (i.e., the end of the girder is
built into the abutment wall or integral abutment). The term fixed end should be used
with some cautions here. Within the context of these data, this end condition refers to a
situation as shown in Figure 10. This type of construction, where the end diaphragm is
cast around the girders, is quite rigid. This type of construction is typical construction for
bridges built on Route 133. This type of construction is considered simply supported in
design stage. As long as the diaphragm is uncracked, i.e., under low level of forces, the
connection will act like fixed support. Probably for higher level of forces, such as
ultimate, the diaphragm will crack and the girders will behave like simply supported
beam as intended. It is difficult to determine the actual rotational stiffness provided by
this type of diaphragm. However, under shrinkage stresses, which are much smaller
than ultimate stresses, the connection quite likely acts like a fixed support. The effect of
end condition will be discussed in more detail later in the report. Nevertheless,
widespread cracking of bridges with similar end conditions, which actually prompted this
research investigation, supports the pronounced effect of end conditions and rotational
rigidity on transverse deck cracking.
28
100
90
80
Percent cracked
70
60
50
40
30
20
10
0
Simply Mixed Continuous Fixed
Supported
End condition
Figure 9. Percent of cracked bridge decks with different end condition.
Figure 10. Boundary condition: girders are bounded.
29
Figure 11. Structural detail of the end condition shown in Figure 10.
NCHRP report 380 (Krauss and Rogalla, 1996), K-Tran (Schmitt and Darwin, 1999) and
Minnesota DOT report (French et al., 1999) are among research studies that report
increased cracking for fixed girders compared to those with pin-ended girders. Since the
number of bridges surveyed is not enough, the effect of the end condition can’t be
evaluated for different girder types but comparing the number of the continuous steel
and prestressed concrete girder bridges shows that 100 percent of prestressed
concrete girder bridges cracked whereas only 25 percent of those continuous bridges
with steel girder cracked. However, this result is not reliable due to small number of
bridges in the sample.
Skewness
There seems to be no direct relationship between the degree of skewness and the
potential for transverse deck cracking. Figure 12 shows that the percentage of cracked
bridge decks do not follow a consistent trend with respect to skewness. Considering the
overall percentage of cracked bridge decks in this study which is equal to 75 percent, it
seems that the data on the graph is just some variation with respect to this number
which exists in any statistical analysis. Similar result is reported in NCHRP Report 380
(Krauss and Rogalla, 1996).
30
90
80
70
Percent cracked
60
50
40
30
20
10
0
None Mild Severe
Skewness
Figure 12. Percent of cracked bridge decks with different skewness.
Type of Bearing
There were two types of bearings in bridges surveyed. Steel bearing and elastomeric
pads. Elastomeric pads where only used with girders on route 133 bridges, where as
discussed, the end diaphragm where cast around the girders (see Figure 10, 11). The
survey shows that all bridges with elastomeric pads are cracked but for steel bearings
the percent of cracked bridge decks is 56. This could be due to the fact that probably
steel bearings allow rotation more freely. Thus, it seems that the type of bearing may
have an affect on transverse deck cracking. However, the result should be viewed with
cautions because the elastomeric pads were only used in bridges that also had a
different end diaphragm as discussed under end condition.
Surface Texture
Since the dominant texture for bridges surveyed is saw-cut texture and there is only one
bridge decks with other type of surface texture, comparison between different textures is
not possible.
Wearing Surface
80 percent of the surveyed bridges had concrete wearing surface whereas latex
concrete was the wearing surface for the remaining bridge decks. 84 percent of the
bridge decks with the concrete wearing surface were cracked while only 20 percent of
the bridge decks with the latex concrete surface developed cracks. Thus, there is an
indication that latex concrete can reduce the cracking but due to the small number of
31
bridges in the latex sample (5 bridges), this result should be treated carefully and one
cannot draw a general conclusion.
Deck Thickness
The average thickness of the cracked decks was about 8.75 in while this average for
the un-cracked bridge decks is around 9 in. This shows that an increase in the deck
thickness reduces cracking. Similar results are also reported by other researchers (e.g.,
Krauss and Rogalla, 1996).
Bar Size and Spacing
All of the bridges surveyed use a mesh of #5 or #6 bars spaced 5 to 7 in as the top
mesh (except in the negative moment areas). This study can’t identify any significant
relationship between the bar size and bar spacing and transverse deck cracking in the
range of available data. It is generally accepted that smaller bar size and closer spacing
can reduce cracking.
One Dimensional Model of Deck and Girder
Based on the survey results it appears that the total stiffness of bridge and relative
stiffness of deck and girder are important factors in transverse deck cracking. To further
investigate the effect of bridge stiffness on transverse cracking, a one dimensional finite
element model is developed. Structural plan and design sheets for surveyed bridges
were also reviewed. The results show that increasing the stiffness of bridge increases
the possibility of deck cracking. The details of Finite element study and its results are
presented below.
Objective: One-dimensional model is used to evaluate the response of the bridges

under service condition. The most important response parameter of interest in this
model is the maximum deflection under service condition. A one-dimensional linear
model of one girder and effective portion of deck (AASHTO, 1998) is developed with
ANSYS (V5.5, 1998) to investigate service response of bridges. Based on the structural
plans of the bridge, this model is dimensioned and analyzed.
Mesh and Elements: Beam elements are used to model the bridge. The span length is
divided into 100 portions and each portion is modeled with beam element. Based on
structural plans of bridge, composite moment of inertia of girder and effective portion of
deck (AASHTO, 1998) are evaluated and specified through out the span length. Figure
13 shows the model.
32
Figure 13. One dimensional finite element model.
Boundary Condition: Based on the bridge being modeled, pin or roller boundary
condition is applied on different nodes (Figure 13).
Material Properties: The only material property used in the one-dimensional model is
elastic modulus. Elastic modulus of steel and concrete are considered to be 29E6 and
3.83E6 psi respectively.
Loading: Based on AASHTO (LRFD, 1998) provisions, two types of loading should be
considered for calculating maximum service load deflection; moving truckload and 25%
of moving truckload plus lane load (Figure 14). Based on AASHTO code, distribution
factors equal to number of lanes divided by number of girders are multiplied to each
loading case.
33
Figure 14. Modeling of traffic loads for 1D model.
The values of truck and lane loads used in analysis are based on AASHTO Articles
3.6.1.2.2 and 3.6.1.2.4. The truckload is moved throughout the span, using birth and
death option of ANSYS to simulate moving truck. No dynamic effects are included
directly.
One Dimensional Model Results
The result of the one-dimensional analysis is compared to the AASHTO (LRFD, 1998)
optional deflection limit of L/800 to study the effect of bridge stiffness on deck cracking.
Thirteen bridges among surveyed bridges are modeled and analyzed in this part of
study. Where available, the FEA calculated deflections are also compared to the
deflections calculated on design sheets at design time of the bridge. Unfortunately the
deflection check for all the bridges could not be retrieved, however there is a good
agreement between FEA results and calculated values where available. Figure 15
shows a microfilm picture of design sheets reviewed in this study to obtain design
deflection values.
34
Figure 15. A picture of microfilm of design sheets.
Table 2 shows the summary of the analysis. Based on this table, the average ratio of
actual to limit deflections for cracked bridges is 0.09 while this ratio for bridges without is
0.20. This result suggests that as the stiffness of bridge increases, the possibility of
cracking also increases. The deflection L/800 limit can be set as benchmark to compare
the stiffness of bridges, however since the study is based on a limited number of
bridges no particular ratio can be proposed for comparison.
Table 2. Results of 1D FEA
Row Bridge Name Description Span Length No. of No. of DF

No. (mm) Lanes Girders
1 ST. # 1130-156 Ramp Span 1, Simple, G2 16404.19948 1 5 0.2
2 ST. # 1130-156 Ramp Span 2,3,4, Continuous, G7 46505.91 1 5 0.2
3 ST. #1103-158 Span 1, Simple, G2 31233.60 2 6 0.33
4 ST.#1013-151 Span 1, Simple, S3 40682.4147 2 7 0.29
5 ST.#1136-154 Span 2, Simple, 21S 42322.83465 2 5 0.4
6 ST.#1143-167 &168 Span 1, simple,G2 27279 2 5 0.4
7 ST.#1143-169 &170 Span 1,2,3 Continuous, G2 36576 2 5 0.4
8 ST.#1143-171 &172 Span 1,2 Continuous, G2 32000 3 6 0.5
9 ST#1143-173 Span 1,2 Continuous, G2 24840 2 6 0.33
10 ST#1143-174 &175 Span 1, Simple, G2 27430 2 5 0.4
11 ST#1143-176 & 177 Span 1,2,3, Continuous, G2 40538 2 5 0.4
12 ST#1130-154 Span 1,2 , Continuous, G102 52493.44 3 8 0.375
13 ST#0206-165 Span1, Simple, G112 27482.8 3 7 0.43
35
Results of 1D FEA (Continued)
Truck 25% Truck + Lane Final
Row Ix (Composite) ∆ ∆ ∆ L/800 Ratio Bridge Status
No. 4
(cm ) (mm) (mm) (mm) (mm) (actual/Limit)
1 1029986.27 3.28 1.670 3.28 20.51 0.16 Not Cracked
2 Variable 8.027 5.290 8.03 58.13 0.14 Not Cracked
4 60760442.97 10.64 8.660 10.64 50.85 0.21 Not Cracked
6 3186616.46 2.56 1.480 2.56 34.10 0.08 Cracked
7 6896776.47 5.49 2.910 5.49 45.72 0.12 Cracked
8 5973539.01 4.28 2.260 4.28 40.00 0.11 Cracked
9 3136103.71 2.52 1.170 2.52 31.05 0.08 Cracked
10 4919769.65 1.68 0.980 1.68 34.29 0.05 Cracked
11 7088042.67 4.56 3.880 4.56 50.67 0.09 Cracked
13 Variable 13.66 8.79 13.66 34.35 0.40 Cracked
Material Properties and Mix Design Factors

Based on the NJDOT Inspection/Testing datasheets that contain the information about
mix design and construction practice several material properties are extracted and used
in the development of the database. The material properties that are recorded in the
NJDOT Inspection/Testing datasheets and used in this study are:
• Cement Content
• Water Content
• Water cement ratio
• Air content
• Cement Type
• Slump
• Compressive strength
• Admixture
Note that the range of data for these factors is quite narrow and this fact should be
considered in interpreting the results. However, this may be a good thing and supportive
of the research thrust on design factors. That is, the data are within a narrow band and
most of them well within the recommended range made by other research. Still majority
of the bridges have cracked supporting the fact that design factors play a significant role
in causing and/or controlling transverse deck cracking.
Cement Content
The cement content in the deck concrete for bridges surveyed is in the range of 611 to
735 lb/yd3 with most of the bridge decks built with cement content of 700 lb/yd3 (19
36
bridges). This study shows that the cracking occurs on both decks with high cement
content and low cement content, albeit within that narrow range. 80 percent of decks
with cement content of 700 lb/yd3 and more cracked, whereas 50 percent of decks with
lower cement content cracked. As for those 19 bridges with exactly 700 lb/yd3 cement
content 73 percent are cracked. Thus, there is an indication that lower level of cement
content reduces cracking but considering the distribution of data (i.e., narrow range and
the fact that majority had one cement content) this result cannot be emphasized. Other
researchers also reported increased cracking with an increase in cement content. (such
as Krauss and Rogalla, 1996; French et al., 1999; Babaei and Purvis, 1996; Shmitt and
Darwin, 1999)
Water Content
The water content for the deck concretes is between 31.5 lb/yd3 and 35 lb/yd3, where 19
bridge
decks have water content of 31.8 lb/yd3 and water content for 23 decks is in the range
of 31.5 lb/yd3 to 32.6 lb/yd3. The ratio of cracked bridges is 77 percent. Previous studies
reported that cracking increase with an increase in water content (Krauss and Rogalla,
1996; Shmitt and Darwin, 1999; French et al., 1999; Babaei and Purvis ,1996), but due
to the narrow range of data in this part no general conclusion can be made.
Water Cement Ratio
The w/c ratio for the bridges surveyed is between 0.44 and 0.36, where 19 bridges have
w/c of 0.38. Again, due to the narrow range of data and their distribution this part is also
inconclusive. But it should be noted that the range of the w/c ration in bridges surveyed
more or less is in the range recommended by other researchers.
Air Content
The average air content for the bridge decks is in the range of 5.1 to 6.7. There is no
indication of the effect of higher or lower air content on deck cracking in the bridges
surveyed based on the results.
Cement Type
Type II cement is used in all decks for bridges surveyed. The cement manufacturer is
also one company for 92 percent of bridges. Literature also recommends the use of
type II cement to minimize transverse deck cracking.
Slump
The average slump of the concrete used in these bridge decks is in the range of 3 to 4
inches. Cracking is observed in decks with both high slump (4 inches) and low slump (3
37
inches). It seems that for the range of slump observed in these bridges, it has no effect
on the transverse deck cracking.
Compressive Strength
The compressive strength of the bridge decks is in the range of 4500 to 6623 psi, which
is a broad range. The average compressive strength of un-cracked bridge decks is 5640
psi, whereas this average for cracked bridges is 5730 psi. This shows that cracked
bridges have slightly higher compressive strength, which is in agreement with previous
studies (e.g., Krauss and Rogalla, 1996). In fact comparing these strengths with the
4500 psi, which is the required strength for design, it is observed that the average
compressive strength is about 1200 psi more than that specified in design. Reducing
this margin, which partly means reducing cement content, may reduce the potential for
deck cracking.
Admixture
Water reducer and air entraining agent is used in all bridges surveyed. Also, 79 percent
of the deck bridges have retarder agent in their mix. There is no indication of increase or
decrease in cracking because of the use of these admixtures.
Construction Techniques and Ambient Condition Factors

All the data that are available and used in the development of the database are those
provided on the Inspection/Testing sheets. Unfortunately, no documented data is
available about curing method and time of curing, which are among the most important
factors with regard to transverse deck cracking. Therefore, it may be a good practice to
include more data on construction methods in the inspection and quality control forms.
Among data that may be included are: wind velocity, humidity, curing method, curing
period, and placement length. Note that measurement of humidity is required based on
the NJDOT Specs (1998).
Thus, based on the available data, in this study the effect of the following factors are
considered:
• Air and concrete temperature

• Month of placement
Air and Concrete Temperature
The study shows that the cracked bridge decks were cast in slightly higher
temperatures and with slightly higher concrete temperatures. The average air
temperature in the time of casting for cracked bridges is 64oF while this number is 60oF
for un-cracked bridges. Also, the average concrete temperature at the time of casting
for un-cracked bridges is 76oF, while this number is 73oF for un-cracked bridges.
38
Literature indicates (e.g., Babaei and Purvis, 1996) that placement of bridge decks in
very high and very cold weather increases the possibility of cracking. However, the data
show that none of these 24 bridges surveyed were cast in very hot or very cold weather.
Month of Placement
As it was just mentioned, literature indicates that deck placement in very hot and very
cold weather increases transverse cracking. The month of placement can be a good
indicator of this situation. The decks for none of the bridges surveyed were cast in the
winter. Construction season for the bridges surveyed are as follow: 37.5 percent in the
spring, 33.3 percent in the summer and 29.2 percent in the fall. 43 percent of bridge
decks cast in the fall developed cracking, while this number is 75 and 89 for the bridges
cast in the months of summer and spring, respectively. This can indicates that casting
the decks in mild weathers can reduce the potential for deck cracking.
Remarks
Review of the data collected and comparison with the literature and with the 1998
NJDOT Specs (1998) show that many material and mix design recommendations are
already satisfied. Due to the limitation of the data such a conclusion cannot be made
with regard to construction factors. For example, it could not be determined if 7 day wet
curing is employed. However, consistent with the literature recommendation, NJDOT
Specs (1998) does require 7 day wet curing. This is a very important factor in controlling
transverse deck cracking and every effort must be made to adhere to the Specifications.
In summary, NJDOT Specs (1998) contains many of the recommendations made as
results of prior research. There are also some factors are consistent with research
recommendation. For example, none of the bridges surveyed were cast in the winter
and all surveyed bridge decks had relatively low w/c ratio.
With regard to design and material property factors, higher compressive strength of the
tested specimens compared to design value shows that the structural design
requirement on the compressive strength could have been satisfied by the use of lower
cement in the mix. Reducing the cement content in turn can reduce drying shrinkage,
thermal shrinkage and temperature differentials during casting, which are believed to be
the dominant causes of transverse cracking. Therefore, it is strongly recommended to
reduce the cement content. Considering the design compressive strength of 4500 psi
for bridge decks, this recommendation can still be satisfied quite easily. If the current
guideline provides any incentive for higher than design compressive strength it should
be revoked. It is also noticeable that except for high amount of cement content, other
parameters included in the study satisfies recommendations made in the literature. Use
of type II cement, adequate air content (>5%), satisfactory w/c ratio (<4.5) and use of
the water reducer and retarder agents are indications of good mix design as required by
literature to reduce transverse cracking.
39
Despite significant research work and enhanced knowledge on the effect of mix design
and construction practice on concrete deck cracking, the current knowledge on the
effects of structural design factors on deck cracking is limited. The design
recommendations proposed in the literature are mostly based on engineering judgment
and have not been quantified in details through analytical and/or experimental studies.
For example, it is accepted that reducing deck restraint or increasing deck thickness
can reduce the possibility of deck cracking, however, no specific guidelines and/or
values have been recommended. The observations of bridges on route 133 (see Figure
10, and 11), as mentioned in section 3.7, suggest that the design factors can be
important in deck cracking. Note that the end diaphragms on all of these bridges are
cast around the girders. As shown in Figure 3.16, the cracks in these bridges are
parallel to the bridge longitudinal axis and become normal to the axis as they propagate
towards the center of the span. Considering the not so common design of the end
diaphragms and its high rigidity compared to a pinned case, the question arises about
the possible contribution of design factors to significantly increase transverse deck
cracking. On the other hand, the same factors can be employed in a balanced design to
provide remedies.
Figure 16. Rt. 133 bridge deck crack.
The observations made from the field survey of two parallel bridges crossing the
Watson Creek (EB and WB) further support the need for more knowledge on the effects
of structural design factors. These two bridges are almost identical and built on EB and
WB side of route I-195. Structural designs as well as the mix designs for these two
bridges are very similar to each other (see appendix A). During the surveys transverse
cracks were observed on the eastbound side but the westbound side had no transverse
40
deck cracking. Examination of the joint continuity for these two bridges showed that the
joints on the westbound side were cracked, while the continuity joints on eastbound side
were intact (see Figure 17 and 18). That is, the bridge with cracked joint (less
rigidity/constraint) did not develop any cracking in the deck, while the one with
uncracked joint (more rigid/restraint) did develop transverse deck cracking. These
observations point to the role of structural design factors in deck cracking and the need
for further research to enhance or knowledge on these factors.
Figure 17. Watson Creek Bridge (West Bound): Cracked continuity joints (bridge
No.1130-152).
41
Figure 18. Watson Creek Bridge (East Bound): Un-cracked continuity joints (bridge
No.1130-153).
42
CAUSES OF VOLUME CHANGE IN CONCRETE AND
RESTRAINED SHRINKAGE TESTS
This chapter presents an overview of the causes of volume change in concrete relevant
to bridge deck cracking. Effects of four types of shrinkage, ambient temperature
changes, and creep are examined. Laboratory tests currently utilized to measure the
cracking tendency of restrained concrete are introduced. The chapter concludes with a
summary.
Volume Change and Cracking in Concrete

Concrete cracks as a result of restraint volume change of concrete and this section
describes the basic mechanisms of the volume change in concrete. There are seven
basic causes of volume change in bridge deck concrete relevant to deck cracking;
drying shrinkage, autogenous shrinkage, plastic shrinkage, thermal contraction, creep,
ambient temperature changes, and solar radiation.
Restraint of concrete during volume change causes cracking. Effect of restraint on

bridge decks will be discussed comprehensively in the following chapters. As mentioned
previously, this project focuses on alleviating the deck-cracking problem by reducing
restraint of concrete deck through changes in design. It should be emphasized again
that although the authors recognize the key role and importance of reducing volume
change in concrete (as described herein) but they also believe that, as a safeguard,
current design of bridge decks can be improved to accommodate volume change of
concrete with less cracking.
Drying Shrinkage
Drying shrinkage is the result of water loss of hardened concrete. Volume of concrete
reduces as water withdraws from concrete. However, only part of (40-70 percent) the
shrinkage is recoverable with future wetting cycles.
Drying shrinkage depends on concrete properties. It is strongly influenced by w/c ratio,

volume to surface ratio and cement content. Both ACI (Committee 209, 1978) and
AASHTO (LRFD, 1998) codes provide equations for estimating drying shrinkage values
as a function of time. These equations do not consider the effect of cement type and its
composition. Where, as experimental results show, shrinkage strains are significantly
influenced by cement properties (Burrows, 1998).
Drying shrinkage in a restraint concrete can produce significant tensile stresses, which
may result in cracking. Figure 19 shows a graph of drying shrinkage strains vs. time for
different volume to surface ratios.
43
Figure 19. Drying shrinkage of concrete.
ACI (Committee 209, 1978) recommends the following equation for predicting drying
shrinkage of concrete.
ti
ε sh ,t = ε sh ,u (1)
ti + α
Where: α = 35 for moist cured concrete and 55 for steam cured concrete
ti = Time in days after measured after curing
ε sh ,u = Ultimate shrinkage=780E-6 in/in in standard condition
The value of the ε sh ,u should be multiplied by following factors when other than standard
condition exist:
For other than standard humidity a correction factor should be applies to equation.
For 40<H<80 Percent K H = 1.40 − 0.010 H (2)
For 80<H<100 K H = 3.00 − 0.30 H (3)
For curing period other than 7 days for moist cured concrete and 3 days for steam cured
concrete multipliers in Table 3 should be used. For v/s ratios (Volume to surface) other
than 1.5” correction factor K vs should be employed.
v
K vs = 1.20 exp(−0.12 ) <0.2 (4)
s
44
Table 3. Shrinkage correction factors for initial moist curing (AASHTO).
Moist Curing Duration (Days) Factor

1 1.2
3 1.1
7 1.0
14 0.93
28 0.86
90 0.75
ACI (Committee 209, 1978) also suggests a second method for computing K vs where:
K vs = 1.23 − 0.038t t>6” (5)
Where t is deck thickness. This equation is valid only during the first year of curing,
which is the concern of this study.
There are other factors for Slump, Cement Content, Air Content and Temperature. But
ACI (Committee 209, 1978) states that for slump less than 5”, fine aggregate between
40 to 60 percent, cement content between 470 and 750 lb/yd3, and air content less than
8 percent these factors are approximately equal to unity.
Plastic Shrinkage and Plastic Settlement
Plastic shrinkage and plastic settlement occur in plastic concrete. Plastic shrinkage is
the result of excessive evaporation of water from concrete before hardening. The rapid
evaporation of water from surface layer of concrete and the subsequent contraction of
this layer causes tensile stress due to restraint from wet concrete layers below. This
may result in plastic shrinkage cracks. Plastic settlement is caused by uneven
settlement of fresh concrete over the obstructions like reinforcement (Figure 20).
Figure 20. Crack in plastic concrete due to uneven settlement.
Prevention of plastic shrinkage and plastic settlement cracks is well understood. Plastic
shrinkage is controlled by controlling the rate of evaporation of water from concrete. The
45
rate of water evaporation basically depends on wind velocity, air temperature, relative
humidity, concrete temperature, cement content, and aggregate size. ACI monograph
(Committee 308, 1986) can be used to estimate the rate of evaporation (Fig 24). Special
curing procedure should be applied when evaporation rate exceeds certain limits.
Plastic settlement cracks can be eliminated by use of smaller reinforcements, and

adequate and timely vibration of concrete. It is believed that the plastic settlement in
decks produces weak planes over reinforcement, which may later crack by other
effects.
Autogenous Shrinkage
Autogenous shrinkage is the concrete shrinkage without loss of water. This kind of
shrinkage occurs at low w/c ratios and significantly increases with use of silica fume,
HRWRAs (High Range Water Reducing Admixtures) and finer cement. In the past, this
type of shrinkage was insignificant. However, with the downward trend of w/c ratio in
concrete mixes, use of silica fume, finer cements, and widespread use of HRWRAs this
type of shrinkage came into attention. Stresses produced by this type of shrinkage can
add up to locked in stresses due to thermal contraction, drying shrinkage and ambient
effects and may cause cracking. Paillere et al. (1989) measured the tensile stress
produced in a test specimen as a result of autogenous shrinkage. They observed
increasing stresses in concrete mixes with the same amount of cement (716 lb) but
different w/c ratios and silica fume. Figure 21 shows their result.
Figure 21. Effect of w/c and silica fume on tensile stress produced by autogenous
shrinkage of concrete (Paillere et al., 1989).
Tazawa et al. (1994 and 1995) have clearly shown that autogenous shrinkage increases
with lower w/c ratios, higher cement fineness, silica fume and use of HRWRA (Figure
22).
46
Figure 22. Effect of w/c, cement fineness, silica fume and admixtures on autogenous
shrinkage (Tazawa et al., 1995).
Thermal Contraction
Concrete temperature rises after placement due to hydration. If the concrete is restraint,
subsequent cooling and increase in modulus of elasticity produces tensile stresses in
concrete and may cause cracking a few days after placement. It is also possible that
these locked in stresses add up to stresses produced by other causes.
It has been found (Sprinegnschmid et al., 1994) that selecting low heat cement may not
completely solve the problem. Because it ignores the effect of rate of increase in
modulus of elasticity, and tensile strength. Therefore, a better approach is required for
selecting appropriate cement. The test frame developed for thermal contraction tests
(Sprinegnschmid et al., 1994) are described in the next section as a possible
alternative.
Cement and its chemical composition have strong effect on magnitude of thermal
contraction. Based on the work of Sprinegnschmid and Breitenbucher (1990), for a low
thermal cracking tendency, cement should have low alkali content, be course grounded,
and have high sulfate content.
Lower placement temperatures, higher air content, (Breitenbucher and Mangold, 1994)
and slower cooling rates (Chui and Dilger, 1993) also reduces the thermal cracking
tendency
Ambient Temperature Changes and Solar Radiation
47
Daily temperature variations and solar radiation produces temperature differential
throughout the section. Stresses produced by this temperature differential add up to the
locked in stresses from different types of shrinkage. Similarly, seasonal variation of
temperature can produce internal stresses.
Figure 23 shows the temperature distribution across the depth of a composite beam
obtained using a finite element analysis for hot weather (Emanuel and Hulsey, 1978).
Similar results are also available for cold weather. AASHTO (LRFD, 1998) has
recommendation for estimating temperature differentials in the deck-girder section for
design purposes.
Figure 23. Temperature distribution across the section for hot weather (Emanuel and
Hulsey, 1978).
Creep
Creep is the property of concrete to deform with time under sustained stresses. Creep
reduces stresses from sustained stresses, thus, reduces deck cracking. So, concrete
with high creep tendency, especially during first few months after casting is desirable.
High w/c ratio, low strength, and soft aggregates produce concrete with high creep. ACI
(Committee 209, 1978) and AASHTO (LRFD, 1998) have equations for estimating
creep strains as a function of time.
Bridge Deck Cracking

Magnitude and time variation of different factors mentioned above determines the
performance of a bridge deck vis-à-vis of cracking. Burrow (1998) describes the causes
48
of 52 cracked bridge decks out of 100 surveyed by Krauss and Rogalla (1996) as
follows:
• “22 cracked during first week: The cause must be thermal contraction plus (if w/c
is below 0.5) autogenous shrinkage
• 6 cracked in next six weeks: It is too late for thermal contraction and too early for
drying shrinkage. Therefore it must be autogenous shrinkage adding to the
locked in stresses from thermal contraction.
• 8 cracked with in a year: Probably stresses from drying shrinkage adding to the
locked in stresses from thermal contraction and autogenous shrinkage that have
not been relieved by creep.
• 16 cracked after one year: Drying shrinkage is taking its toll.”
Restrained Shrinkage Tests

Restraint shrinkage tests evaluate the cracking resistance of cement or concrete. Any
cement or concrete can be evaluated by comparing the test values against an
established evaluation criterion. These tests have been used in many countries such as
France, Canada, Japan, and Germany to control concrete cracking, but very few have
been used in the U.S. Several testing devices with different dimensions, configuration
and objectives have been developed so far. However, all of these devices pursue the
common goal of classifying cements/concretes based on their cracking tendency. These
devices fall into two main categories:
Ring Tests
Ring test was invented by Roy Carlson (1942) of MIT to measure cracking resistance to
shrinkage. In this test cement mortar or concrete is cast around a steel ring and the time
to occurrence of first crack is measured. The elapsed time untill occurrence of first crack
is used as a measure of cracking tendency of cement. Based on the purpose of the test,
curing and temperature changes may be introduced. Figure 24 shows different sizes of
this device used by different researches. Figure 25 show a picture of the ring used by
Krauss and Rogalla (1996)
Different methods are used to record cracking time. Visual inspection is the most basic
method. Blaine (1953) applied a conducting paint to the ring and recorded the crack
time by measuring the voltage change applied to the paint. Krauss and Rogalla (1996)
installed strain gages on the inside of the steel tube and captured cracking time as
sudden change in strain readings. Nowadays, different technologies are available for
capturing cracking time.
Figure 26 shows the recommended criteria based on Blaine’s work for crack resistant
cement (Burrows, 1998).
49
Ring tests provide valuable information on cracking tendency. However, there are two
drawbacks. Ring tests do not fully restrain the cement mortar or concrete due to friction
at the interface of cement and steel. Furthermore, there is no direct method of
measuring stresses in concrete.
Figure 24. Ring test apparatus used by different researchers (Burrows, 1998).
50
Figure 25. Ring test apparatus used by Krauss and Rogalla (1996).
Figure 26. Cement rating criteria based on Blaine’s work (Burrows, 1998).
51
Bar Tests
In this test concrete or cement mortar is cast in the shape of a bar and is hold by two
grips at both ends to create restrained condition. The cracking tendency is expressed in
terms of level of tensile stress in the specimens as a result of shrinkage. Figure 27
shows the apparatus used in these tests. In a new version of this test, which was
developed in Germany two invar bars are used to connect the grips together. In this
frame like apparatus, stress is measured indirectly by measuring stress in the invar bars
(Figure 28). Using this frame, a test method has been developed in the University of
Munich (Sprinegnschmid et al., 1994) to investigate thermal contraction. In this method,
without the application of any external heat or cooling, concrete is cooled down to
ambient temperature during 4 days. If the concrete has not cracked during 4 days, it is
then cooled down artificially at a rate of 1.8 F per hour. The specific temperature at
which cracking occurs is defined as cracking temperature. Figure 29 shows typical
results of this test.
Bar Tests can be used to measure thermal contraction, autogenous shrinkage and
drying shrinkage in a more realistic manner than ring tests. Through the use of
controlled environment and specimen size, it is also possible to separate different types
of shrinkage more easily and measure them independently.
52
Figure 27. Bar test for studying cracking tendency of concrete (from Burrows, 1998).
53
Figure 28. Cracking frame developed in Germany (from Burrows, 1998).
Figure 29. Typical results of cracking frame test (from Burrows, 1998).
54
FINITE ELEMENT ANALYSIS
This chapter describes development of several finite element (FE) models to study
different aspects of transverse deck cracking and presents results of the analyses.
Objectives and details of implementation for each model are presented. Effect of
different design factors on stresses causing transverse cracking is studied using a 2D
FE model. A nonlinear 3D FE model is used to model cracking and study crack patterns
on concrete bridge deck with different boundary conditions.
Two Dimensional Model of Deck and Girder

Two-dimensional finite element model is used to study the effect of several design
factors on transverse bridge deck cracking caused mainly by drying shrinkage. A linear
plane stress model of a single girder and the tributary deck with its reinforcements is
developed using ANSYS (V5.5, 1998). Geometric and design information used in
development of the finite element model (Table 4) corresponds to one of the bridges
surveyed (i.e., Hackensack Ave over NJ Route 4). This bridge is built in 1998 in Bergen
County, New Jersey. It is a two-span simply supported bridge on steel girders. The two
spans are similar in design. This bridge has shown cracking all over both spans.
Table 4. Geometric and design information for the bridge modeled (Hackensack
Ave over NJ Route 4).
Span Length 1082”

Span Width 564”
Girder Spacing 7 Girders@7’-1”
Deck thickness 8.5”
Longitudinal Bars #5@6”(B), #5@15”(T)
Transverse Bars #6@7”(T&B)
Girder Bottom Flange Thickness 2”
Girder top Flange Thickness 1.25”
Girder Bottom Flange Width 24”
Girder Bottom Top Width 20”
Girder Web Thickness 1”
Girder Web Height 35.25”
Shear Stud Spacing 12” (Midspan@71’) 9” (Support@18’)
The choice of plane stress model is made after detailed 3D FE analysis of the same
bridge structure. Since it is computationally expensive to model and analyze a complete
bridge structure in detail, usually a single girder and the corresponding portion of the
deck is modeled and analyzed. As a single girder and its tributary deck are considered
separate from the rest of the structure, it is required to somehow take into account the
effect of the adjacent portion of the structure. Figure 30 shows typical girder and deck
bridge cross section and highlights the single girder and its deck, which is modeled. To
determine the correct type of finite element model that can be used to model the single
girder and its deck a detailed 3D FE model was developed and analyzed. Results of this
55
analysis showed that 2D plane stress and 3D stand alone models of single girder and
its tributary deck are good approximations.
Portion to be modeled in detail
Figure 30. Cross-section of typical slab-on-girder bridge superstructure.
The 3D model is shown in Figure 31, which is the right portion of a bridge with a cross
section shown in Figure 30. The girders are modeled using elastic shell elements and
the deck is modeled with elastic solid elements. Since the overall nature of
superstructure response is being investigated the reinforcements are not modeled and
full composite action is assumed (i.e, deck and girder nodes are connected).
Diaphragms are also modeled using truss elements. Their location and dimensions are
chosen based on structural plans. Elastic modulus of steel and concrete are assumed to
be equal to: 29E6 psi and 3.83E6 psi, respectively.
Figure 31. Finite element model of deck, girders and diaphragms.
The span modeled is simply supported. So the nodes on the bottom flange of the
girders at one end is restrained in three directions and at the other end is only
restrained in transverse and vertical directions. It is assumed that due to symmetry of
56
the system, the middle girder will not have displacement in the X direction (transverse
direction). Consequently, one end of the diaphragms, which are connected to middle
girder are restrained.
Figure 32 shows the graph of required lateral nodal forces to move the edge of the deck
in Figure. 31, 2.5-cm (1-inch) toward the center of the bridge. These values are
obtained by specifying uniform unit displacement on the edge nodes of the deck and
obtaining the restraining forces at each node. It can be said that these forces would be
applied on the single girder and its tributary deck model provided that they have the
same 2.5 cm (1 in) displacement. If the single girder and its deck were completely
restrained by the adjacent structure (Figure 31) the required forced at each node of
Figure 31 should be:
E Element AElement = 26 E 6 × 0.22 × 0.56 = 3200 KN (6)
Where E Element (26E6-KN) is element modulus of elasticity, AElement is element area. 0.22-
m (8.5-inch) is deck thickness and 0.56-m (22-inch) is element length. Considering the
relative magnitude of actual forces in Figure 32 compared to Equation 1 result (3200-KN
vs. 400KN), it can be concluded that the restraining effect of girders on each other in the
lateral direction is negligible. This means that plane stress elements can be used in 2D
models and that stand-alone 3D models are good approximations.
14000
12000
10000
Force(KN)
8000
6000
4000
2000
0
0 20 40 60 80 100
Length Ratio From One End(Percent)
Figure 32. Lateral nodal forces along the deck assuming uniform displacements.
57
Since it is shown that it is justifiable to use plane stress 2D model to study the behavior
of a single girder and its corresponding deck, the details of 2D FE model will be
presented.
2D Model Details
Deck and Girder Elements: One girder and tributary deck are modeled using plane
stress elements. The span is divided into 50 sections along its length. The deck and
girder are modeled with 3 and 5 elements through the depth, respectively. The
thickness of plane stress element is varied according to actual thickness of deck and
girder flange/web in the transverse direction. Figure 33 shows the FE model a close up
view of it.
Reinforcement Elements: The two layers of longitudinal reinforcement are modeled with
truss elements. These elements are placed along the boundary of the deck plane stress
elements at two different depths within the deck and are connected to the same nodes
as plane stress deck elements. Based on the design detail for the bridge being
modeled, total area of truss elements is chosen equal to total area of longitudinal
reinforcements at the corresponding depth.
Figure 33. Two-dimensional (2D) finite element model.
58
Boundary Conditions: Based on the case being considered, different boundary
conditions are applied on different nodes or set of nodes. A roller support is modeled by
restraining the bottom node of the girder in vertical direction. A pin support is modeled
by restraining the bottom node of the girder in the vertical and horizontal directions. To
model a fixed end, all nodes at the cross section are restrained from moving in all
directions. Figure 33 shows a girder with pin and roller supports.
Modulus of Elasticity: Elastic modulus of steel and concrete are assumed to be equal to:
29E6 psi and 3.83E6 psi, respectively.
Steel and Concrete Coefficient of Thermal Expansion: Winter and Nilson (1986) state
that the concrete coefficient of thermal expansion should be in the range of 4E-6 to 7E-6
in/in per degree of Fahrenheit. Khan, Cook, and Mitchell (1998) measure the values of
5.3e-6 to 5.5e-6 in/in per oF for maturing normal weight concrete. In this study the
values of 5.5E-6 and 6.25E-6 in/in per oF are assumed for concrete and steel,
respectively.
Loading: Based on the case under consideration several types of loading are
considered for each analysis, which are described in a case-by-case basis.
59
Results of Two Dimensional Analyses
Effects of several design factors on stresses causing transverse cracking are studied in
this section. The design parameters of the base model (shown in Table 4) have been
changed to study their effect on the stresses causing transverse cracking (Deck
longitudinal stress, i.e. Sxx component of stress in Figure 31).
Effect of Boundary Condition
Increasing boundary condition restraint of deck and girder system increases tensile
stress produced in deck, which may ultimately cause cracking. Figure 34 and Table 5
show the deck top and bottom stresses caused by 10oF temperature decrease in the
deck, while rebar temperature is held constant. This temperature decrease models the
effect of uniform shrinkage strain equal to 5.5E-5 or 55 microstrain.
Figure 34. Deck bottom and top stresses caused by 55-microstrain uniform deck
shrinkage for different boundary conditions.
Table 5. Deck top and bottom tensile stresses caused by 55-microstrain uniform
deck shrinkage for different boundary conditions.
Deck Bottom Stress Deck Top Stresses

Boundary Condition (Section)
(psi) (psi)
Pin – Roller (Mid Section) 72.59 36.17
Pin – Pin (Mid Section) 71.86 40.94
Three Span Continuous (Over Support) 121.21 131.25
Two Span Continuous (Over Support) 130.84 155.35
Fixed – Roller (8%L From End Section) 131.40 145.81
Fixed – Pin (8%L From End Section) 182.74 208.80
Fixed – Fixed (8%L From End Section) 210.30 210.30
60
Effect of Span Length
Span length does not affect stresses in the deck considerably. This is an interesting
result as it shows that the magnitude of stresses developed in the deck do not depend
on span length. Effect of 50% change (increase and decrease) in span length on deck
stresses caused by 55-microstrain uniform deck shrinkage is shown in Table 6. Neither
deck bottom stresses, nor deck top stresses changes considerably (less than 10%) with
span length for all boundary conditions considered. Figure 35 and 36 show the graph of
top and bottom deck stresses for different span lengths considered.
shrinkage for different span lengths and boundary conditions.

(psi) (psi)
Percent of Original Length 50%L 100%L 150%L 50%L 100%L 150%L
Pin – Roller (Mid Section) 72.59 72.59 72.59 36.17 36.17 36.17
Pin – Pin (Mid Section) 71.92 71.86 71.84 40.55 40.94 41.06
Three Span Continuous (Over 116.21 122.05 129.40
121.21 131.25 130.84
Support)
Two Span Continuous (Over 123.98 132.64 152.58
130.84 155.35 154.84
Support)
Fixed – Roller (8%L From End 128.32 131.71 145.74
131.40 145.81 146.28
Section)
Fixed – Pin (8%L From End Section) 173.66 182.74 184.18 204.29 208.80 210.71
Fixed – Fixed (8%L From End 210.30 210.30 210.30
210.30 210.30 210.30
Section)
61
different span lengths and boundary conditions.
different span lengths and boundary conditions.
62
Effect of Deck Thickness
Increasing deck thickness reduces deck stresses for all except fixed-fixed boundary
conditions. Thus, thicker decks are preferred over thinner decks with respect to
transverse cracking. However, practical limits of deck thickness should be considered.
Table 7 shows effect of thickness on deck stresses due to shrinkage. Concrete cover is
held constant while deck thickness is changed. Figures 37 and 38 show graphically the
effect of deck thickness on bottom and top stresses for different boundary conditions.
These stresses are produced by 55-microstrain uniform concrete deck shrinkage.
different deck thickness and boundary conditions.
63
different deck thickness and boundary conditions.
shrinkage for different deck thickness and boundary conditions.

(psi) (psi)
Deck Thickness 7” 8.5” 10” 7” 8.5” 10”
Three Span Continuous (Over 132.38
121.21 111.41 139.96 131.25 123.51
Support)
Two Span Continuous (Over 143.60
130.84 119.32 163.06 155.35 148.61
Support)
Fixed – Roller (8%L From End 143.17
131.40 120.81 154.41 145.81 138.40
Section)
Fixed – Fixed (8%L From End 210.30
210.30 210.30 210.30 210.30 210.30
Section)
Effect of Girder Spacing
Increasing the girder spacing will reduce deck stresses produced by volume change of
concrete in all but fixed-fixed deck girder system.
64
Table 8 shows results of the finite element analysis for the example bridge with 50,
100, and 150 percent of original spacing and different boundary conditions. These
stresses are for 55-microstrain uniform concrete shrinkage. Figures 39 and 40 show
deck top and bottom stresses for different spacing and boundary conditions.
shrinkage for different girder spacing and boundary conditions.

(psi) (psi)
50%W 100% 50%W
Percent of Original width 150%W 100%W 150%W
W
121.21 97.98 163.98 131.25 108.95
Support)
130.84 104.83 187.53 155.35 132.10
Support)
131.40 69.96 179.02 145.81 122.49
Section)
210.30 210.30 210.30 210.30 210.30
Section)
different girder spacing and boundary conditions.
65
different girder spacing and boundary conditions.
Effect of Relative Flexibility of Girder to Deck
Increasing the ratio of girder to deck moment of inertia increases deck stresses for all
but fixed-fixed boundary condition. Table 9 shows deck stresses for different boundary
conditions and three different ratios of girder to deck moment of inertia, while composite
section moment of inertia is held constant. In other words all the sections have the
same composite section stiffness, however, contribution of deck and girder is different.
Deck thickness and girder height are changed to obtain different relative flexibility but
equal composite section inertia, all other aspects of design are held constant. Table 10
shows contribution of deck and girder in each case and the corresponding deck
thickness and girder height. Concrete cover is held constant while deck thickness is
changed. Figures 41 and 42 shows the graph of deck top and bottom stresses for
different boundary conditions and various ratios of girder to deck moment of inertia.
These results show that for all but fixed-fixed boundary condition, the potential for deck
cracking reduces as the ratio of girder to deck moment of inertia decreases. Since
composite section moment of inertia is held constant, change in relative stiffness
(flexibility) will not affect bridge serviceability requirements while reducing deck-cracking
potential. There is no change in stresses for fixed-fixed boundary condition.
66
shrinkage for different ratio of girder/deck moment of inertia and boundary
conditions.

(psi) (psi)
Ratio of Girder to Deck Inertia 4.13 6.18 9.66 4.13 6.18 9.66
121.21 128.25 125.28 131.25 137.15
Support)
130.84 139.91 150.64 155.35 160.52
Support)
131.40 139.80 140.49 145.81 151.70
Section)
210.30 210.30 210.30 210.30 210.30
Section)
different ratio of girder/deck moment of inertia and boundary conditions.
67
different ratio of girder/deck moment of inertia and boundary conditions.
Table 10. Different ratio of girder/deck moment of inertia used in calculation and
corresponding deck and girder dimensions.
Deck Deck Girder

Section Web Height Composite Girder/Deck
Thickness Inertia Inertia 4
Number
(in)
(in) 4
(in ) 4
(in )
Inertial (in ) Ratio
1 7.5 36.474 2988 28855 64853.1 9.66
2 8.5 35.25 4350 26876 64852.5 6.18
3 9.5 34.1 6073 25089 64852.8 4.13
Effect of Composite Section Moment of Inertia
Increasing composite section moment of inertia increases deck stresses slightly for all
except fixed-fixed and fixed-pin boundary conditions. Composite section moment of
inertia is varied by changes in web height as shown in Table 11. The deck top and
bottom stresses are shown in Table 12 and Figures 43 and 44.
Table 11. Different ratio of composite moment of inertia used in calculation and
corresponding deck and girder dimensions.
Section Web Height Composite

4 %
Number (in) Inertial (in )
1 33.25 58415.8 90
2 35.25 64852.5 100
3 37.25 71666.6 110.5
68
shrinkage for different composite moment of inertia and boundary conditions.

(psi) (psi)
Composite Moment of Inertia 90% 100% 110.5% 90% 100% 110.5%
121.21 121.38 130.32 131.25 131.52
Support)
130.84 132.07 155.14 155.35 155.58
Support)
131.40 132.59 145.35 145.81 146.30
Section)
210.30 210.30 210.30 210.30 210.30
Section)
different composite moment of inertia and boundary conditions.
69
different composite moment of inertia and boundary conditions.
Effect of Amount of Longitudinal Reinforcement
Increasing the area of longitudinal reinforcement increases the deck stresses for pin-
roller and pin –pin spans. There is no significant change in stress for other boundary
conditions. These results show that contrary to what is generally believed, increasing
the amount of reinforcement not only will not reduce deck stresses prior to cracking, but
also in some cases increases deck stresses.
Table 13 shows effect of 50% increase and decrease in rebar area on deck stresses
from 55-microstrain uniform shrinkage. 50% change in rebar area corresponds
approximately to increasing rebar number by 1 size (say from #4 to #5). For
comparison, a theoretical deck with zero reinforcement is also analyzed. These results
are shown in Table 14. Figures 45 and 46 show the results in graphical format.
70
different area of longitudinal reinforcement and boundary conditions.

(psi) (psi)
Percent of Original Reinforcement 50%Ar 100%A 50%Ar
150%Ar 100%Ar 150%Ar
Area r
121.21 122.30 129.97 131.25 131.85
Support)
130.84 131.53 155.10 155.35 155.61
Support)
131.40 132.14 145.36 145.81 146.26
Section)
210.30 210.30 210.30 210.30 210.30
Section)
zero area of longitudinal reinforcement and boundary conditions.

(psi) (psi)
Pin – Roller (Mid Section) 66.32 28.37
Pin – Pin (Mid Section) 65.62 33.30
Three Span Continuous (Over Support) 118.95 129.38
Two Span Continuous (Over Support) 129.45 154.88
Fixed – Roller (8%L From End Section) 129.91 144.92
Fixed – Pin (8%L From End Section) 183.85 210.74
Fixed – Fixed (8%L From End Section) 210.30 210.30
71
72
Effect of Distribution of Longitudinal Reinforcement
Changing the distribution pattern of total reinforcement area among top and bottom
reinforcement do not seem to change the deck stresses. In other words, for the same
amount of reinforcement in the section, increasing top or bottom reinforcement do not
change the deck stresses.
Table 15 shows the effect of longitudinal reinforcement distribution on deck stresses

produced by 55-microstrain uniform shrinkage. Keeping the total amount of
reinforcement constant, 25 percent of total reinforcement is moved from bottom to top
and vice versa. Figures 47 and 48 illustrate the effect of changing the pattern of
longitudinal reinforcement distribution on deck stresses.
different distribution of longitudinal reinforcement and boundary conditions.
73
different distribution of longitudinal reinforcement and boundary conditions.
shrinkage for different distribution of longitudinal reinforcement and boundary
conditions.

(psi) (psi)
Location of Reinforcement Increase Bottom None Top Bottom None Top
121.21 121.13 131.38 131.25 131.13
Support)
130.84 130.66 155.66 155.35 155.04
Support)
131.40 131.27 146.01 145.81 145.61
Section)
210.30 210.30 210.30 210.30 210.30
Section)
74
Effect of Girder Shoring During Construction
Temporary shoring during construction reduces the stresses in the concrete deck
significantly and even induces some compressive stress for pin-roller and pin-pin
boundary conditions. However, temporary shoring for other boundary condition does not
have beneficial effect on deck stresses as the dead load produces additional tensile
stresses in the deck upon removal of the shoring.
If the girder is shored during deck pouring and curing, the dead load of concrete is
carried by composite section after removal of shoring. Table 16 shows the dead load
stresses produced in the deck for simply supported girder (tension is positive).
Compressive stresses in the deck provide additional safety against deck cracking. In
other words, tensile stresses produced by concrete volume change should be much
larger for a shored construction compared to un-shored one in order to crack the
concrete deck. However, for other types of boundary conditions, the dead load induces
tensile stresses in some regions and may worsen the problem of cracking.
Table 16. Deck top and bottom stresses caused deck dead load.

(psi) (psi)
Pin – Roller (Mid Section) -175.43 -331.85
Pin – Pin (Mid Section) -185.07 -268.72
Effect of Pouring Sequence
Properly selected pouring sequence reduces the possibility of cracking by inducing

compressive stress in the deck. As a general rule, positive moment sections should be
placed first followed by negative moment sections.
Figure 49 shows the recommended pouring sequence for different boundary conditions.
In each case, the numbers on the top of the deck show the pouring sequence of deck
concrete and the length of pour is indicated at the bottom of girder. As it is illustrated in
Figure 49, the middle half of each span is placed first and then the remaining portion is
placed in the sequence shown. Tables 17 through 25 show the residual stresses in the
deck as a result of pouring sequence shown in Figure 49. These results show that
considerable stress is produced in the deck depending on the sequence of pouring. As
a general rule to minimize the tensile stresses and maximize the compressive strength
in the deck, in single span bridges, positive moment portions should be poured first,
followed by negative moment sections. In multi span bridges, the whole span should be
placed in one pour. Since this is not practical in most cases, the pouring sequence as
depicted in Figure 49 is suggested.
75
2 1 3 2 1 3 3 1 2
a) b) c)
3 1 2 3 1 2 3 1 4 2 5
d) e) f)
4 1 5 2 6 3 7
g)
Figure 49. Suggested pouring sequence and lengths for different boundary conditions.
Table 17. Residual top and bottom deck stresses caused by pouring sequence
(a) for pin-roller boundary condition.

Pin – Roller Boundary Condition
(psi) (psi)
After Pouring Section 1 2 3 1 2 3
End Section (8%L From Left
0 0 -2.06 0 0 -3.52
Support)
Middle Section 0 -17.12 -32.85 0 -37.17 -70.45
End Section (8%L From Right
0 0 0 0 0 0
Support)
(b) for pin-pin boundary condition.

Pin – Pin Boundary Condition
(psi) (psi)
0 0 -3.85 0 0 6.59
Support)
Middle Section 0 -22.94 -42.18 0 -31.85 -56.62
0 0 0 0 0 0
Support)
76
(c) for fixed-roller boundary condition.

Fixed - Roller Boundary Condition
(psi) (psi)
0 0 0 0 0 0
Support)
Middle Section 0 -11.54 -17.02 0 -22.94 -34.09
0 0 -1.57 0 0 -0.80
Support)
(d) for fixed-pin boundary condition.

Fixed - Pin Boundary Condition
(psi) (psi)
0 0 0 0 0 0
Support)
Middle Section 0 -19.67 -26.99 0 -25.97 -37.62
0 0 -1.43 0 0 0.39
Support)
(e) for fixed-fixed boundary condition.

Fixed - Fixed Boundary Condition
(psi) (psi)
0 0 0 0 0 0
Support)
Middle Section 0 11.80 4.49 0 9.60 -2.63
0 0 -3.58 0 0 -11.03
Support)
Table 22. Residual bottom deck stresses caused by pouring sequence (f) for two
span continuous girder.
Deck Bottom Stress

Two Span Continuous Condition (psi)
After Pouring Section 1 2 3 4 5

Middle of Left Span 0 5.34 -9.70 -9.83 -9.63
Over Interior Support 0 0 0 0 1.80
Middle of Right Span 0 0 4.73 8.34 5.52
77
Table 23. Residual top deck stresses caused by pouring sequence (f) for two
Deck Top Stress

Two Span Continuous Condition (psi)
After Pouring Section 1 2 3 4 5

Middle of Left Span 0 10.62 -18.59 -18.90 -18.62
Over Interior Support 0 0 0 0 4.83
Middle of Right Span 0 0 9.94 17.74 12.38
Table 24. Residual bottom deck stresses caused by pouring sequence (g) for
three span continuous girder.
Deck Bottom Stress

Three Span Continuous Condition (psi)
After Pouring Section 1 2 3 4 5 6 7

Middle of Left Span 0 7.87 6.07 2.66 1.28 1.71 1.40
Over Left Interior Support 0 0 0 0 0 0.65 0.21
Middle of Middle Span 0 0 8.26 8.96 7.97 6.97 8.05
Over right Interior Support 0 0 0 0 0 0 1.97
Middle of Right Span 0 0 0 -0.15 0.12 -1.16 -4.37
Table 25. Residual top deck stresses caused by pouring sequence (g) for three
Deck Top Stress

Three Span Continuous Condition (psi)
After Pouring Section 1 2 3 4 5 6 7

Middle of Left Span 0 14.47 11.18 5.27 2.80 3.62 3.04
Over Left Interior Support 0 0 0 0 0 1.69 0.48
Middle of Middle Span 0 0 15.35 16.53 14.73 12.99 14.73
Over right Interior Support 0 0 0 0 0 0 4.81
Middle of Right Span 0 0 0 -0.26 0.21 -8.01 -8.01
Effect of Shrinkage Profile
So far it was assumed that shrinkage is uniform across the deck depth. Shrinkage
profile (gradient of shrinkage through the depth of deck) changes stress profile in the
deck. Based on actual bridge properties, volume change gradient may increase or
decrease stresses in the deck.
Table 26 shows deck stresses produced by +2 and –2oF temperature gradients (11
microstrain) through the deck as shown in Figure 50. As seen in the Figure 51 deck
stress profile tends to follow the change in shrinkage profile in the deck. In other words,
increasing shrinkage in top causes an increase in the top stress. It seems desirable to
design the deck and girder to have uniform stress profile for uniform shrinkage profile.
This will reduce the adverse effect of shrinkage gradient through the section.
78
Table 26. Deck top and bottom stresses caused by different shrinkage profile in
deck for different boundary conditions.

(psi) (psi)
Volume Change Distribution Number 1 2 3 1 2 3
121.21 97.98 111.30 131.25 151.21
Support)
130.84 108.72 134.77 155.35 175.92
Support)
131.40 109.29 125.48 145.81 166.13
Section)
210.30 189.27 189.27 210.30 231.33
Section)
Figure 50. Shrinkage stress profile.
79
Figure 51. Deck top and bottom stresses for different boundary conditions and (a)
shrinkage profile 1, (b) shrinkage profile 2, (c) shrinkage profile 3.
80
Summary of Results of Two dimensional Analysis
Effect of several design factors on deck stresses is evaluated and the results are
presented in tables and figures. In summary:
1. Increasing boundary condition restraint of deck and girder system increases tensile
stress in the deck.
2. Span length does not affect the deck stresses considerably.
3. Increasing deck thickness reduces deck stresses for all except fixed-fixed boundary
conditions.
4. Increasing girder spacing will reduce deck stresses produced by volume change of
concrete in all but fixed-fixed deck girder system.
5. Increasing the ratio of girder to deck moment of inertia increases deck stresses for
all but fixed-fixed boundary condition
6. Increasing composite section moment of inertia increases deck stresses slightly for
all except fixed-fixed and fixed-pin boundary conditions
7. Increasing the area of longitudinal reinforcement increases the deck stresses for pin-
roller and pin –pin spans. There is no significant change in stress for other boundary
conditions.
8. Changing the distribution pattern of total reinforcement area among top and bottom
reinforcement does not change the deck stresses
9. Deck stresses for fixed-fixed supported deck and girder is only a function of volume
change and do not change with other parameters.
10. Temporary shoring during construction reduces tensile stresses in the concrete deck
significantly and induces some beneficial compressive stress for pin-roller and pin-
pin boundary conditions. However, shoring for other boundary condition does not
have beneficial effect on deck stresses as the dead load produces additional tensile
stresses in the deck.
11. Properly selected pouring sequence reduces possibility of cracking by inducing
compressive stress in the deck. As a general rule, positive moment sections should
be placed first followed by negative moment sections.
12. Volume change gradient through the section changes stress profile in the deck.
However based on actual bridge properties, this change in profile may increase or
decrease the stress in the deck.
Three Dimensional Model of Deck and Girder

The three dimensional model is used to investigate crack patterns in concrete bridge
deck. The analysis considers the bridge deck at early ages (less than three months) .To
be effective and useful the model must:
1. Capture important design characteristics of the bridge including geometry, boundary

conditions, and sectional properties
2. Be capable of predicting cracks
3. Be capable of modeling behavior of bridge after cracking
81
To achieve these objectives, a non-linear finite element model was developed using
ANSYS (V5.5, 1998).
Similar to the 2D model, the 3D bridge model uses geometric and design information
for one of the bridges surveyed (i.e., Hackensack Ave. over NJ Route 4 SB Bridge -
Table 4).
3D Model Details
Basic Assumptions: The following are the basic assumptions made in building the 3D
model:
1. Since the girders on the adjacent sides of the girder under consideration are equally
spaced, relative symmetry is assumed.
2. The middle girder of the bridge as shown in Figure 31 is modeled. So the girder web
is a plane of symmetry under uniform shrinkage, thermal load, and uniform gravity
load on the deck. In other words, this means that the girder has no displacement in
the direction normal to the plane of its web. Similarly, it is argued that the girder has
no rotation about its longitudinal axis passing the girder web plate. Due to this
symmetry, only a half of the girder and deck is modeled as shown in Figure 52.
3. As shown earlier in the chapter, the effect of adjacent portion of deck and girders
can be ignored.
4. Abutments and supports are considered rigid.
Deck and Girder Modeling: One girder of the bridge with the corresponding portion of
the deck is modeled in three dimensions. Due to symmetry of model, only half of deck
and girder are modeled. The bridge deck is modeled using solid elements capable of
cracking, while the girder is modeled using shell elements. The length of the span is
divided into 49 strips. In each strip the deck is modeled with 21 elements located in
three layers. The girder is modeled using 4 elements in each strip, 2 elements for the
top flange, one for the web and one for the bottom flange. Figure 52 shows the model.
The deck elements are slightly smaller over the girder.
It should be mentioned that there is a lack of compatibility between the quadratic shell
elements representing the girder with linear solid elements modeling the deck girder
boundary. However, the effect of this issue is not important at all because the fine mesh
used is well capable of representing the deformation shape(s) expected under the
loadings considered.
82
Figure 52. 3D finite element model.
Reinforcement Elements: Reinforcements are modeled with beam elements. Since the
moment of inertia of these elements is very small they basically act as truss elements
(truss elements produce analysis instability after cracking due to rotation at ends).
These elements are placed in two directions and four layers along the boundary of the
solid elements on a separate mesh of nodes. These four layers represents top and
bottom mesh of reinforcement in longitudinal and transverse directions. To simplify
modeling, spacing of truss elements is chosen to be equal to the spacing of the solid
elements. Since the spacing of the rebar elements is not exactly equal to the actual
reinforcement spacing per plans, the equivalent area of rebars are used for each layer
of rebar element to have the same amount of reinforcement in the section. Furthermore
to prevent bending of deck in transverse direction, moment of inertia of the transverse
rebars are increased to prevent this bending.
Rebar elements are connected to solid element representing the deck with nonlinear
springs modeling bond and slip of rebars in order to capture the behavior of the
concrete and reinforcement realistically. Modeling bond-slip is essential to capturing
post-cracking behavior of the concrete. Bond slip relationship proposed by Houde
(1973) is used in this study, which is defined as follow:
u = 1.96 E 6 × w − 2.35 E 9 × w 2 + 1.39 E12 × w 3 − 0.33E15 × w 4 (7)
Where u is bond stress in psi and w is slip in inches. Figure 53 shows a plot of this
equation.
83
800
700
600
Bond Stress (Psi)
500
400
300
200
100
0
0 0.0005 0.001 0.0015 0.002
Silp (in)
Figure 53. Bond-slip relationship (Houde, 1973).
Since finite element model includes this effect in a discrete manner, equivalent spring
constant should be defined as follow:
k = u × πd rebar l element (8)
Where d rebar is the bar diameter and l element is the element length.
Shear Stud Elements: Figure 54 shows shear stud details extracted from structural
plans for the bridge under consideration. Shear studs are modeled using nonlinear
spring elements. These elements connect nodes on the top of the girder to nodes on
the bottom of the deck (which are geometrically coincident).
Figure 54. Shear Connector Detail.
84
Behavior of shear studs modeled using these springs with the load slip model (Yam and
Chapman, 1968) shown in Figure 55. The referenced load slip behavior is obtained by
back calculation of the composite section behavior by Yam and Chapman (1968). In the
absence of substantial experimental data this relationship will be utilized.
7000
6000
5000
Load (lb)
4000
3000
2000
1000
0
0 0.01 0.02 0.03 0.04 0.05
Slip (in)
Figure 55. Shear Connector load slip behavior (Yam an Chapman, 1968).
Boundary Condition: The span modeled is simply supported. So the nodes on the
bottom flange of the girder at one end is restrained in three directions and those at the
other end is only restrained in the transverse and vertical directions. Also symmetric
boundary condition is applied on the plane of symmetry (i.e., girder web plane).
Additionally longitudinal rebar nodes are constrained to move with deck in Z and X
directions (See Figure 55 for directions) and transverse rebar nodes is constrained to
move with the deck in Z and Y directions. Deck bottom nodes in contact with the girder
top nodes are also constrained to move in the Z direction with these nodes.
Concrete Compressive Strength: According to NJDOT specifications, a fixed

compressive strength of 4500 psi is chosen. This is a good approximation, since the
deck is cured at lease 7 days and after 7 days the change in compressive strength is
not very large.
Concrete Modulus of Elasticity and Poisson’s Ratio: ACI specification proposes the
following equation for the value of modulus of elasticity (all values in psi)
E c = 57000 f c′ (9)
Poisson’s ratio for concrete at stresses lower than 0.7 f c′ falls within limits of 0.15 to 0.2
(Winter and Nilson, 1968). A value of 0.2 is used for the model and it is assumed
constant for all ages.
85
Concrete Tensile Strength: According to recommendation of ACI the modulus of rupture
is assumed equal to:
f r = 7.5 f c′ (10)
Steel Modulus of Elasticity and Poisson’s Ratio: The values of 29E6 psi and 0.3 will be
used for steel modulus of elasticity and Poisson’s ratio, repsectively.
Steel and Concrete Coefficient of Thermal Expansion: Winter and Nilson (1986) state
that the concrete coefficient of thermal expansion should be in the range of 4E-6 to 7E-6
in/in per degree of Fahrenheit. Khan, Cook, and Mitchell (1998) measure the values of
5.3e-6 to 5.5e-6 in/in per oF for maturing normal weight concrete. In this study the
values of 5.5E-6 and 6.25E-6 in/in per oF are assumed for concrete and steel,
respectively.
Concrete Stress-Strain Curve: Since the magnitudes of compressive stresses are well
within linear portion of the concrete strain stress curve a linear material model would be
considered for compression. A linear stress strain curve is also assumed for concrete in
tension before cracking.
Crack Modeling: Figure 56 shows the concrete material model in tension as defined by
ANSYS (V5.5, 1998). It is further assumed that the shear stress transfer coefficient is
0.5 for open and close cracks. This value basically accounts for the rebar dowel action
and aggregate interlocking at crack interface.
Figure 56. Concrete material model in tension.
Loading: The loading consists of three major components: drying shrinkage, thermal
shrinkage, and ambient temperature. Note that creep effects are not modeled. However,
considering the fact that the analyses are mostly concerned with behavior of the system
in early age (less than 3 month), the results are useful and the error is negligible.
86
Results of experiments conducted as a part of this study, as shown in Figure 57, shows
that after the first few days of pouring there is effectively no temperature difference
between ambient temperature and the bridge deck in NJ. Also, field measurements
show that hydration temperature rise produces no significant locked in residual stresses
in the bridge deck after the first few days of pouring. Based on these two results only
shrinkage-induced effects are considered in the 3D models and all models are
subjected to a uniform shrinkage in the deck.
35 50
40
30 Cocncrete Strain & 2 per. Mov. Avg.
30
25
Strain (microstrain)
20
Concrete
Temp (C)
20 10
Temperature
15 0
-10
10
-20
5
-30
Ambient Temperature
0 -40
9-Nov 11-Nov 13-Nov 15-Nov 17-Nov 19-Nov 21-Nov 23-Nov 25-Nov 27-Nov 29-Nov 1-Dec
Time
Figure 57. Measured concrete deck strain and temperature of a simply supported
bridge.
Results of Three Dimensional Analyses

Crack patterns, stress history, and deflection history has been studied for four different
types of boundary conditions. To study bridge response, a uniform shrinkage is applied
to the deck and several response parameters are reported.
Pin-Roller Boundary Condition
Figure 58 shows the deflected shape of the structure at the end of the analysis. The
cracks at mid span and at the end of the span are shown in Figure 59. Note that the
circles in this figure indicate the plane of cracks. It is shown that except for a small
portion at the ends, bridge deck develops transverse cracks as a result of applied
87
shrinkage. Deflection at the center of the span vs. the shrinkage strain at mid span is
shown in Figure 60. Downward deflection is negative. The sudden jump in deflection
(bounce back) at a strain around 0.0004 in/in indicates transverse cracking of the bridge
deck and stress relief.
Figure 58. Deformed shape of the girder and cracked deck at the end of analysis (pin-
roller boundary condition).
Figure 59. Deck cracks at end of span (bottom) and mid span (top) of the bridge at the
end of analysis(pin-roller case).
88
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-0.2
-0.4
Deflection (in)
-0.6
-0.8
-1
-1.2
-1.4
Shrinkage Strain
Figure 60. Mid span girder deflection vs. shrinkage strain (pin-roller case).
Figure 61 shows the concrete deck top and bottom stresses at mid span. It is noticeable
that the deck top stress is far less than tensile strength of concrete (500 psi). However,
it experiences the same jump as the deck bottom stress and the results show that the
element at the top has cracked. This indicates that in this particular case cracking has
started form bottom of the deck and propagates almost suddenly through the section. In
other words, cracks are full depth.
600
500
400
Stress (psi)
300
200
100
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-100
Shrinkage Strain
Deck top stress @ mid span Deck bottom stress @ mid span
Figure 61. Deck top and bottom stress over girder at mid span for pin-roller boundary
condition.
89
Deck top stresses at mid span and quarter span are compared in Figure 62. The
sudden decrease of stress around the strain of 0.0004 in/in again shows development
of cracking. Results show that both mid span and quarter span cracks develop at the
same time. Results show that cracks at mid span stay open after cracking. However,
due to stress redistribution cracks at quarter span close and as the graph shows
compressive stresses develop at this region due to downward deflection of the bridge.
300
250
200
150
Stress (psi)
100
50
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-50
-100
-150
Shrinkage Strain
Deck top stress @ mid span Deck top stress @ quarter span
Figure 62. Deck top stress over the girder at mid span and quarter span for pin-roller
boundary condition.
It is also observed that cracks are developed almost at the same time. Evaluation of the
entire data indicates that, in fact the entire bridge cracks at once (e.g. Figures 61, 62,
63). Figure 63 shows the deck top stress across the deck at mid span at three locations.
Location 1 is over girder, location 2 is midway between girder and edge of slab, and
location 3 is slab edge. This figure supports the fact that cracks propagate transversely
across the whole deck at the same time. This behavior is observed by Krauss and
Rogalla (1996) in their experimental study of Portland-Columbia Bridge in Pennsylvania.
They did not observe any cracks in their initial survey of the bridge after construction.
However, in their next survey after a few days, they noticed that bridge deck has
developed transverse cracks all over.
90
300
250
200
150
100
50
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-50
Shr i nkag e St r ain
Deck t op st ress @ mid span location 1 Deck t op stress @ mid span locat ion 2
Deck t op st ress @ mid span location 3
Figure 63. Deck top and bottom stress across the slab at mid span for pin-roller
boundary condition.
As it can be seen from Figure 59 not all elements cracked during the analysis. In mid
span, every other row of element developed full depth cracks across the slab.
Considering that the length of an element is 22 in and each element has 4 integration
points, which capture the cracking, crack spacing is 2.75 ft long. This spacing is the
spacing between full depth cracks that run across the slab. This spacing as shown in
Figure 59 reduces towards the ends of span.
Initiation of cracks can also be observed by sudden change in longitudinal

reinforcement stress as shown in Figure 64. The longitudinal stress in reinforcement at
mid span experiences a sudden increase followed by increasingly high stress with
increasing shrinkage. In fact, in an experimental study Frosch et al. (2002) have
observed yielding of longitudinal reinforcement as a result of cracking. Although the
reinforcement stress does not reach yielding in FE analysis, the high stress level
suggests that yielding might be possible in some cases.
91
30000
25000
20000
Stress (psi)
15000
10000
5000
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
Shrinkage Strain
Top rebar stress @ mid span Top rebar stress @ quarter span
boundary condition.
Pin-Pin Boundary Condition
The response of a pin-pin bridge to shrinkage strains is similar to that of a pin-roller

case. However, there are some differences that are discussed here. A diagram of
cracks at the end of an analysis shown in Figure 65. As it can be seen almost all of the
bridge developed full depth transverse cracks. Deflection history is shown in Figure 66.
The deflection curve is similar to that for a pin roller case but the maximum deflection is
lower indicating the increase in stiffness of the structure due the change in boundary
condition. There in not much difference in the level of strain that causes cracking.
Cracks develop around 0.0004 in/in shrinkage strain. This result was expected based
on the 2D finite element results.
92
Figure 65. Deck cracks at end of span (bottom) and mid span (top) of pin-pin bridge at
the end of analysis. Circles indicate cracks for each element.(pin-pin case).
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-0.2
-0.4
Deflection (in)
-0.6
-0.8
-1
-1.2
Shrinkage Strain
93
The stress histories at two points through the depth of the deck at mid span are shown
in Figure 67. Similar to the pin-roller case cracks are full depth and start from the
bottom. As it can be seen from Figure 68 stress redistribution results in development of
compressive stress in the top of the deck.
600
500
400
300
Stress (psi)
200
100
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-100
-200
-300
Shrinkage Strain
Deck top stress @ mid span Deck bottom stress @ mid span
Figure 67. Deck top and bottom stress over girder at mid span for pin-pin boundary
condition.
Figure 68 shows a comparison of the deck top stresses at mid span and quarter span,
while Figure 69 shows the stress distribution across the deck at mid-span. In this figure
location 1 is over girder, location 2 is midway between girder and edge of slab, and
location 3 is slab edge.
300
200
100
Stress (psi)
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-100
-200
-300
Shrinkage Strain
Deck top stress @ mid span Deck top stress @ quarter span
Figure 68. Deck top stress over the girder at mid span and quarter span for pin-pin
boundary condition.
94
300
200
100
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-100
-200
-300
Shr i nkag e St r ai n
Deck t op st ress @ mid span loact ion 1 Deck t op st ress @ mid span locat ion 2
Deck t op st ress @ mid span locat ion 3
Figure 69. Deck top and bottom stress across the slab at mid span for pin-pin boundary
condition.
As it can be seen from Figure 65, more full depth cracks are developed in this case. So,
the spacing between full depth cracks that run across the span is lower. This spacing is
estimated to be 1 ft based of the mesh size for FE model. However, for a better
estimate finer mesh is required. Longitudinal reinforcement stresses are displayed in
Figure 70. As shown in this figure, the top and bottom reinforcements develop high
stresses after cracking.
30000
25000
20000
Stress (psi)
15000
10000
5000
0
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
Shrinkage Strain
Top rebar stress @ mid span Top rebar stress @ quarter span
boundary condition.
95
Fixed-Roller Boundary Condition
The deformed shape of fixed-roller bridge is shown in Figure 71. Figure 72 shows the
cracks at the end of the analysis. Circles indicate plane of cracking in Figure 73. It is
observed that the entire bridge deck is cracked. While cracks are spread right to the
fixed end support, they do not appear near the roller end. Deflection history at mid-span
is shown in Figure 73. Unlike the previous cases there is no sudden jump in the
deflection curve but rather deflection re-bounces in several steps suggesting that cracks
do not happen all at once. The re-bounce starts around strain of 0.0002 in/in, which is
lower than previous cases but expected due to higher stiffness of a fixed-roller beam
compared to pin-roller or pin-pin case. Similarly, deflection values are far less than the
previous cases.
Figure 71. Deformed shape of the girder and cracked deck at the end of analysis (fixed-
roller boundary condition).
Figures 74 and 75 show the deck top and bottom stress at different locations along the
span. It is apparent from these graphs that transverse cracks start from fixed end and
as the shrinkage strain increases they spread towards the roller end. These cracks are
again full depth and as the results show extend across the slab. It is also noticeable that
cracks happen at a stress lower than tensile strength of concrete. This is due to multi-
axial nature of stresses at the time of cracking.
96
Figure 72. Deck cracks at fixed end of span (bottom left), roller end of span (right end)
and mid span (top) of fixed-roller bridge at the end of analysis.
0.00E+00
0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-5.00E-02
-1.00E-01
-1.50E-01
Deflection (in)
-2.00E-01
-2.50E-01
-3.00E-01
-3.50E-01
-4.00E-01
-4.50E-01
-5.00E-01
Shrinkage Strain
97
400
300
200
Stress (psi)
100
0
0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-100
-200
Strain
Deck top stress @ 1/3 span Deck top stress @ mid span Deck top stress @ 2/3 span
Figure 74. Deck top stress at 1/3, half and 2/3 of span from fixed end for fixed-roller
boundary condition.
500
400
300
200
Stress (psi)
100
0
0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
-100
-200
-300
Strain
Deck bottom stress @ 1/3 span Deck bottom stress @ mid span Deck bottom stress @ 2/3 span
Figure 75. Deck bottom stress at 1/3, half and 2/3 of span from fixed end for fixed-roller
boundary condition.
98
Figure 76 shows the reinforcement stress along the span. There is an increase in
reinforcement stress after cracking, which were observed in previous cases too.
20000
18000
16000
14000
Stress (psi)
12000
10000
8000
6000
4000
2000
0
0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007
Strain
Top rebar stress @ 1/3 span Top rebar stress @ mid span Top rebar stress @ 2/3 span
Figure 76. Top reinforcement stress at 1/3, half and 2/3 of span from fixed end for fixed-
roller boundary condition.
Fixed-Fixed Boundary Condition
The behavior of the structure before cracking in the case of a fixed-fixed bridge is
exactly similar to the classical example of two different materials connected together
and undergoing temperature difference. Since the deck and girder are fixed at both
ends deflection due to shrinkage is negligible and deck stress is equal to shrinkage
strain times the modulus of elasticity for concrete. This stress is the same all over the
slab. The moment that the stress level reaches the tensile strength of concrete every
element in the deck cracks. Figure 77 shows the stress history of a deck element. The
same pattern is observed for all other elements in the deck. It is noticeable that the level
of shrinkage strain when cracking occurs is considerably lower for this case. This has
also been shown using the 2D models and it is due to higher stiffness of the system.
99
600
500
400
Stress (psi)
300
200
100
0
0.0001 0.00012 0.00014 0.00016 0.00018 0.0002 0.00022 0.00024
Shrinkage Strain
Figure 77. Stress of deck elements for fixed-fixed boundary condition.
Based on FEA results, a bridge with fixed-fixed boundary condition undergoes the most
severe cracking while the level of shrinkage is the lowest. This means that fixed-fixed
bridges have higher potential for transverse deck cracking. A point that must be
explicitly considered in their design. An easy to use program that can facilitate
determination of tensile stresses in bridge decks is developed under this study. This is
discussed in Chapter six.
Summary of Results of Three Dimensional Analysis
3D analyses of a typical bridge with four different types of boundary conditions were
presented and discussed in light of crack pattern, stress and deflection time histories,
and distribution of stresses. The results can be summarized as follow:
1. Cracks develop suddenly all over the deck in bridges with pin-roller, pin-pin, and
fixed boundary conditions. However, for fixed roller boundary condition cracks
spread along the span as the shrinkage strain increases.
2. Cracks are full depth and run across the slab.
3. Crack spacing for simply supported bridges in estimated to be 2.75 ft and it reduces
as the rigidity of the boundary condition increases.
4. There is a sudden increase in reinforcement stress after cracking which may cause
rebars to yield.
5. Deflection of the bridge reduces as cracks develop and stresses are relieved.
100
A SIMPLE METHOD FOR ESTIMATING STRESSES CAUSING
TRANSVERSE CRACKING IN CONCRETE BRIDGE DECKS
This chapter describes a simple method to estimate stresses causing transverse
cracking in concrete bridge decks with full composite action between deck and girder.
A system of equations for a pin - roller supported composite girder is developed to
estimate the deck stresses due to volume change in concrete caused by shrinkage
and/or other effects. These equations are further extended to consider different
boundary conditions. The results of this method agree very well with the results of finite
element analyses. It is proposed that this practical method be used as a tool during
bridge design to examine concrete bridge deck against the possibility of transverse
cracking. To facilitate this, a MS Windows application is developed that provides a
simple mean for designers to perform AASHTO (3.12) checks on shrinkage and
temperature loading. Furthermore, an overview of factors causing the volume change in
bridge deck concrete and their magnitude is also presented herein to simplify the use of
this method.
Compatibility Equations
The basic method of solid mechanics is used to derive a set of equations for computing
internal stresses due to an assumed temperature profile. Figure 78 shows a composite
girder with bilinear temperature profile in the deck and girder. To compute internal
stresses caused by this temperature profile in a pin-roller supported composite deck-
girder system, the composite section is divided into four sections, each with linear
temperature changes. Sections are numbered from one to four beginning from the top
(Figure 79). Section 1 and 2 divide deck and section 3 and 4 divide girder. Internal
stresses transferred between the sections and between sections and reinforcements
are replaced by their resultant force and moments (Figure 79).
To calculate internal force and moment resultants, eight compatibility equations should
be solved simultaneously. These equations ensure the strain and curvature
compatibility among sections and reinforcements. Equations 1, 2, and 3 enforce strain
equality at the interface of the separated sections. Equations 4, 5, and 6 ensure
curvature equality in all sections and equations 8 and 9 ensure strain equality between
reinforcement and surrounding concrete matrix. The current form of the equations,
consistent with design practice, assumes that two layers of longitudinal reinforcements
exist. First layer of reinforcement is located in section one and the other layer is in
section two (Figure 79). However, the equations can be easily modified for more than
two layers of reinforcements.
101
Figure 78. Temperature profile along the section.
Figure 79. Compatibility forces and moments and temperature distribution.
Q1 F Fr1 Fd Fr1 (d t1 − CTop )

ε Bottom1 = α 1 (T2 − T0 ) + + 1 − + 1 b1 + =
E1 S b1 E1 A1 E1 A1 E1 S b1 E1 S b1
(11)
Q2 − Q1 F2 − F1 Fr2 F d + F1 d t 2 Fr2 (d b 2 − C Bottom )
ε Top 2 = α 2 (T3 − T0 ) + + − + 2 b2 −
E2 St 2 E 2 A2 E 2 A2 E2 S t 2 E2 St 2
102
Q2 − Q1 F2 − F1 Fr2 F d + F1d t 2 Fr2 (d b 2 − C Bottom )
ε Bottom 2 = α 2 (T4 − T0 ) + + − + 2 b2 − =
E2 Sb2 E 2 A2 E 2 A2 E2 Sb2 E2 S b2
(12)
Q3 − Q2 F3 − F2 F3 d b 3 + F2 d t 3
ε Top 3 = α 3 (T5 − T0 ) + + +
E3 S t 3 E3 A3 E3 S t 3
Q3 − Q2 F3 − F2 F3 d b 3 + F2 d t 3
ε Bottom 3 = α 3 (T6 − T0 ) + + + =
E3 S b3 E3 A3 E3 S b3
(13)
Q3 F Fd
ε Top 4 = α 4 (T7 − T0 ) − − 3 + 3 t4
E 4 S t 4 E 4 A4 E 4 S t 4
1 α 1 (T2 − T1 ) Q1 Fd Fr1 (d t1 − CTop )

= + + 1 b1 + =
R1 h1 E1 I 1 E1 I 1 E1 I 1
(14)
1 α 2 (T4 − T3 ) Q2 − Q1 F1 d t 2 + F2 d b 2 Fr2 (d b 2 − C Bottom )
= + + −
R2 h2 E2 I 2 E2 I 2 E2 I 2
1 α 2 (T4 − T3 ) Q2 − Q1 F1d t 2 + F2 d b 2 Fr2 (d b 2 − C Bottom )

= + + − =
R2 h2 E2 I 2 E2 I 2 E2 I 2
(15)
1 α 3 (T6 − T5 ) Q3 − Q2 F2 d t 3 + F3 d b 3
= + +
R3 h3 E3 I 3 E3 I 3
1 α 3 (T6 − T5 ) Q3 − Q2 F2 d t 3 + F3 d b3
= + + =
R3 h3 E3 I 3 E3 I 3
(16)
1 α 4 (T8 − T7 ) Q Fd
= − 3 + 3 t4
R4 h4 E4 I 4 E4 I 4
Fr1
ε Re bar1 = α r1 (Tr1 − T0 ) + = (17)
E r1 Ar1
103
Q1 F Fr1 Fd Fr1 (d t1 − CTop )
α 1 (Tr1 − T0 ) + + 1 − + 1 b1 +
E1 S r1 E1 A1 E1 A1 E1 S r1 E1 S r1
Fr2
ε Re bar1 = α r 2 (Tr 2 − T0 ) + =
E r 2 Ar 2
(18)
Q − Q1 F2 − F1 Fr2 F d + F1 d t 2 Fr2 (d b 2 − C Bottom )
α 2 (Tr 2 − T0 ) + 2 + − + 2 b2 −
E2 S r 2 E 2 A2 E 2 A2 E2 S r 2 E2 S r 2
In these equations ε Bottomi and ε Topi are strains at bottom and top of section i. Ri is the
curvature of section i, Fi and Qi are the force and moment resultant of stresses at
interface of section i and i+1, Ei , hi , Ai , I i , S ti , S bi , and α i are respectively modulus of
elasticity, height, area, moment of inertia, top section modulus, bottom section modulus
and coefficient of thermal expansion of section i. d bi and d ti are respectively the
distance from centroid to bottom and top of section. Fri , Sri ,α ri , C Bottom , and CTop
represent respectively force resultant of stresses in reinforcement layer i, section
modulus at level of reinforcement layer i, coefficient of thermal expansion of
reinforcement layer i, and bottom reinforcement cover and top reinforcement cover. Ti
and Tri define the temperature distribution in section as shown in Figure 79.
Upon calculating forces and moments required to satisfy compatibitity equations, the
deck stresses can be obtained by using the following equations:
F2 − Fr2 − F1 F2 d b 2 + Q2 − Q1 + F1d t 2 − Fr2 (d b 2 − C Bottom )
σ DeckBottom = + (19)
A2 Sb2
F1 − Fr1 F1d b1 + Q1 + Fr1 (d t1 − CTop )

σ DeckTop = + (20)
A1 S t1
In these equations σ DeckBottom and σ DeckTop are respectively deck bottom and top stresses.
Equations 11 through 20 can be applied to a cantilever system too. Since the same
representation of internal forces and the same compatibility equations (equations 11-18)
are used, similar system of equations characterizes the behavior of cantilever
composite beam. This results in identical stresses from equation 9 and 10 for pin-roller
and cantilever systems. However, due to boundary conditions deflection curves would
be different for these two cases.
104
Adequacy of Assumptions
There are two basic assumptions in deriving compatibility equations; namely (i) the
assumption that approximating the internal stresses along the boundary of sections and
reinforcements with resultant forces and moments is a good approximation, and (ii) the
assumption that deck is in plane stress condition.
Deformations and strains computed based on resultant forces and moments do not
capture the effect of distributed nature of internal stresses and consequently do not
result in exact solution. However, the deviation from exact solution is not significant. In
fact the accuracy of equations depends on nature of stress distribution along the
boundary of sections and reinforcements, which in turn depends on temperature profile.
To evaluate the accuracy of this assumption and for comparison of results, a 2-D Finite
Element (FE) model of a pin-roller supported girder and the tributary portion of the deck
with top and bottom reinforcements is developed using ANSYS FE package (1998)
(described completely in chapter 5). Deck and girder are modeled using plane stress
elements with different thickness and reinforcements are modeled using truss elements
attached over the boundary of deck elements (Figure 80). The FE model is developed
using an actual bridge dimensions and properties, which developed transverse deck
cracks. The bridge data shown in Table 27 are used as an example in this paper. Deck
and girder model of Table 27 are subjected to arbitrary 5.6oC (10oF) uniform
temperature decrease in the deck. Figure 81 shows the distribution of nodal forces in
longitudinal direction between the deck and the girder. Similarly, Figure 82 shows
distribution of internal nodal forces in a direction normal to the plane of the deck for the
same model. As it can be seen from these figures, internal nodal forces are
concentrated at both ends and justify resultant representation between sections 2 and 3
as shown in Figure 79. However based on beam theory, resultant representation of
internal stresses between sections 3 and 4 ignores the distributed nature of stresses.
This introduces deviation from exact solution. However as results in Table 28 show, this
deviation is not significant.
105
Figure 80. Finite element model of deck and girder (ANSYS V5.5, 1998).
Table 27. Design characteristic of modeled bridge.
Design Characteristic Value

Span Length 27.48-m (1082-inch)
Span Width 14.33-m (564-inch)
Girder Spacing 7 [email protected] (85-inch)
Deck Width per Girder 2.16-m (85-inch)
Deck thickness 21.6-cm (8.5-inch)
Top Longitudinal Bars #[email protected] (15-inch)
Bottom Longitudinal Bars #[email protected] (6-inch)
Top and Bottom Transverse Bars #6@18-cm (7”-inch)
Girder Bottom Flange Thickness 5-cm (2-inch)
Girder top Flange Thickness 3.17-cm (1.25-inch)
Girder Bottom Flange Width 61-cm (24-inch)
Girder Bottom Top Width 50.8-cm (20-inch)
Girder Web Thickness 2.5-cm (1-inch)
Girder Web Height 89.5-cm (35.25-inch)
Deck Top Cover 6.4-cm (2.5-inch)
Deck Bottom Cover 2.5-cm (1-inch)
106
150
100
Force in Longitudinal Direction
50
(KN)
0
0 20 40 60 80 100
-50
-100
-150
Length Ratio from One End (Percent)
Figure 81. Distribution of nodal forces in longitudinal direction for pin-roller deck girder
system between deck and girder.
20
15
Force in Normal Direction
10
5
(KN)
0
0 20 40 60 80 100
-5
-10
-15
-20
Length Ratio From One End (Percent)
Figure 82. Distribution of nodal forces in direction normal to deck for pin-roller deck
girder between deck and girder.
107
Table 28. Comparison of FEA and compatibility equations results for the pin-
roller supported deck-girder system.
Deck Bottom Deck Top Stresses

Stresses KPa (PSI) KPa (PSI)
Compatibility Equations 434.16 (62.97) 167.06 (24.23)
2D FEA 434.23 (62.98) 167.06 (24.23)
Another assumption in developing the set of equations is that the composite deck is
under plane stress condition. Table 29 shows the effect of plane stress vs. plane strain
assumption on deck stresses for the example bridge. This table shows that there is a
considerable difference between plane stress and plane strain assumptions.
Table 29. Comparison of compatibility equations for plane stress and plane strain
assumption.
Deck Bottom Stresses Deck Top Stresses

KPa (PSI) KPa (PSI)
Plane stress 434.16 (62.97) 160.17 (24.23)
Plane strain 669.00 (97.03) 213.12 (30.91)
A detailed 3-D FE analysis is performed to verify the plane stress assumption. The
results of this analysis support the plane stress case (see chapter 5 for details).
Results and Comparison to Finite Element Analysis

To evaluate the results obtained from equations 1-10, a 2-D Finite Element (FE) model
of Figure 80, as described earlier, is analyzed when subjected to 5.6oC (10oF) uniform
temperature drop in the deck. Table 28 presents deck stresses computed using the
compatibility equations and those for 2D finite element analysis. The results are almost
identical.
Other Boundary Conditions

Equations 1 through 10 estimate deck stresses in the case of pin-roller supported or
cantilever girder. Computation of stresses for other than pin-roller or cantilever
boundary condition utilizes the principle of superposition. Either a pin-roller supported or
a cantilever system can be chosen as base case. Any extra boundary conditions will be
considered as redundant boundary condition. The redundant boundary conditions
impose additional forces on the system to satisfy deformation compatibility. These
forces can be obtained using compatibility requirements. After calculating redundant
forces, deck stresses is calculated by addition of stresses estimated by equations 1
through 10 to the stresses caused by redundant forces. Validity of the final results is
guaranteed by the principal of superposition for elastic structures.
108
Figure 83 shows the application of the principle of superposition to different boundary
conditions. Numerical results of application of this method to the example bridge
section, (with different boundary conditions) subjected to a 5.6oC (10oF) uniform
temperature decrease in the deck are shown in Table 30. The equations used for
obtaining these results are presented in the following sections.
Figure 83. Superposition principle applied for finding deck Stresses in different
boundary conditions.
Table 30. Comparison of FEA and compatibility equations results for the different
boundary conditions.

Boundary Condition Equations FEA Error Equations FEA Error
KPa KPa (PSI) (%) KPa KPa (%)
(PSI) (PSI) (PSI)
429.06 429.00 0 203.60 201.05
Pin – Pin 1.2
(62.23) (62.22) (29.53) (29.16)
915.07 874.6
Fixed – Roller 4.6 (152.10) (150.71) 0.9
(132.72) (126.85)
Fixed – Fixed w/o Axial 754.70 743.19 754.56 743.19
1.5 1.5
Restraint (109.46) (107.79) (109.44) (107.79)
1449.97 1449.97 0 1449.97 1449.97
Fixed – Fixed 0
(210.30) (210.30) (210.30) (210.30)
914.93 867.43 1048.69 1041.18
Two Span Continuous 5.4 0.69
(132.7) (125.81) (152.1) (151.01)
818.96 799.79 872.32 864.53
Three Span Continuous 2.4 0.9
(118.78) (116.00) (126.52) (125.39)
Pin-Pin Boundary Condition
109
Figure 83-A shows this case and illustrates the application of superposition principle.
The value of the redundant force P, required to satisfy compatibility of deformation is:
( F2 d tg − Q3 )d bg F2 1 d bc
2
P = −[ − ][ + ] −1 (21)
E Girder I Girder AGirder E Girder Acomposite E Composite E Composite I Composite
In this equation AGirder , I Girder , and EGirder are respectively area, moment of inertia and
modulus of elasticity of girder, AComposite , I Composite , and E Composite are respectively area,
moment of inertia and modulus of elasticity of composite section, d bg and d tg are
respectively distance from centroid of the girder to top and bottom fiber of the girder,
and d bc and d tc are respectively distance from centroid of composite section to top and
bottom of section. This force, P, produces additional stresses at top and bottom of the
deck ( ∆σ DeckTop , ∆σ DeckBottom ):
1 d tc d bc
∆σ DeckTop = P[ − ] (22)
Acomposite I Composite
1 (d tc − t )d bc
∆σ DeckBottom = P[ − ] (23)
Acomposite I Composite
Where t is the deck thickness. Thus, the deck’s final stress is obtained from addition of
these stresses to the results of equations 9 and 10.
Fixed-Roller Boundary Condition
Figure 83-B shows this case and illustrates the application of the principal of
superposition. Force P required to produce zero displacement in horizontal direction at
the support is:
( F2 d tg − Q3 ) 3E Composite I Composite
P= × (24)
2 E Girder I Girder L
Where L is the span length. This force, P, produces additional stresses in the deck,
where:
PLd tc
∆σ DeckTop = (25)
I Composite
PL(d tc − t )
∆σ DeckBottom = (26)
I Composite
Thus, the deck’s final stresses are obtained from addition of these stresses to the
results of equations 9 and 10.
110
Fixed-Fixed Boundary Condition With Axial Force Release
Figure 83-C shows this case and illustrates the application of superposition principle.
Although cantilever system could be chosen as the base case but due to symmetry, it is
computationally easier to chose simply supported case as base case. The value of the
redundant moment M required to satisfy compatibility conditions is:
( F2 d tg1 − Q3 ) E Composite I Composite
M= (27)
E Girder I Girder
Moment M produces additional stress in the deck, where:
Md tc
∆σ DeckTop = (28)
I Composite
M (d tc − t )
∆σ DeckBottom = (29)
I Composite
Deck’s final stresses are obtained from addition of these stresses to the results of
equations 9 and 10.
Fixed-Fixed Boundary Condition
For the case of fixed-fixed boundary condition the deck stresses are simply computed
from the following equation:
σ Dec kTop ( Bottom ) = α Deck ∆TTop ( Bottom ) E Deck (30)
Where ∆T is the equivalent temperature change in the deck at top and bottom of deck.
Two Span Continuous System
Figure 83-D shows the application of the superposition method to the continuous beam
case. It is assumed here that there is no change in cross section in the two spans.
However, the method is general and the reader can apply this concept with necessary
modifications when sectional properties are different in each span.
Continuity moment M is calculated by:

( F2 d tg − Q3 )(3E Composite I Composite )
M= (31)
2 E Girder I Girder
Moment M produces additional stresses in the deck, which can be calculated using
equations 19 and 20.
111
Three Span Continuous System
It is assumed here again that there is no change in cross section in the three spans.
Continuity moments at 2 spans can be computed by solving the following equations,
which comes from compatibility:
( F2 d tg − Q3 )( L1 + L2 ) M 1 ( L1 + L2 ) M 2 L2
= +
2 E Girder I Girder 3E Composite I Composite 6 E Composite I Composite
(32)
( F2 d tg − Q3 )( L2 + L3 ) M 2 ( L 2 + L3 ) M 1 L2
= +
2 E Girder I Girder 3E Composite I Composite 6 E Composite I Composite
In these equations L1 , L2 , and L3 are span lengths and M 1 and M 2 are continuity
moment between spans. This gives:
( F2 d tg − Q3 )( L2 + L3 )(2 L1 + L2 ) E Composite I Composite
M1 = 3
E Girder I Girder (4 L1 L2 + 4 L1 L3 + 3L2 + 4 L2 L3 )
2
(33)
( F2 d tg − Q3 )( L2 + L1 )(2 L3 + L2 ) E Composite I Composite
M2 =3
E Girder I Girder (4 L1 L2 + 4 L1 L3 + 3L 2 + 4 L 2 L3 )
2
Based on the bridge data from Table 27 and for L1 = L2 = L3 = 1374cm (541in) , and an
arbitrary 10oF uniform temperature decrease in the deck the results of compatibility
equations and FE model are compared (Table 30).
Results show that deck stresses do not depend on length of span. Figure 84 and 85
shows comparison of the deck top and bottom stresses for different boundary
conditions.
1600
1400
1200
Stress (KN)
1000
800
600
400
200
0
Pin-Pin
Pin-Roller
Fixed-Pin
3 Span
2 Span
Fixed-
Fixed
w/o Axial
Fixed-
Fixed
Bending Compatibility Equations Finite Element Results
Figure 84. Comparison of deck bottom stresses for different boundary conditions.
112
1600
1400
1200
Stress (KNi)
1000
800
600
400
200
0
3 Span 2 Span Fixed- Fixed- Fixed-Pin Pin-Pin Pin-Roller
Fixed Fixed
w /o Axial
Restraint
Bending Compatibility Equations Finite Element Results
Figure 85. Comparison of deck bottom stresses for different boundary conditions.
Loading definition for simplified procedure

To investigate transverse deck cracking using developed equations appropriate loading
values and distributions should be chosen. These temperature distributions represent
effect of different thermal factors, which causes deck cracking. There are four major
different causes producing stress and strain in the system. These four causes are
autogenous shrinkage, drying shrinkage, hydration temperature rise, and temperature
gradient. Creep strains are not included among the causes. Excluding stress relief due
to creep strains results in higher stresses in the system. Due to early occurrence of
these cracks, creep strains seems to not contribute significantly and it is justifiable to
ignore creep strains. However, if long-term performance of bridge deck is being
investigated, creep effect should also be included.
Temperature profile and values in the section are the key parameters in estimating deck
stresses. For modeling purposes and uniformity in the equations, shrinkage strains will
be converted to equivalent temperature change (by dividing the strain by the coefficient
of thermal expansion) and it will be added to other thermal loadings. Procedures based
on design guidelines and literature to obtain reasonable values and profiles for each of
the four thermal loadings are presented.
Drying and Autogenous Shrinkage
Drying shrinkage is the result of water loss of hardened concrete. Volume of concrete
reduces as water withdraws from concrete. However, only part of (40-70 percent) the
113
shrinkage is recoverable with future wetting cycles. Both AASHTO (LRFD, 1986) and
ACI (Committee 207, 1986) recommend different equations for shrinkage computation.
Either of these equations can be used in evaluating shrinkage strains, although
AASHTO (LRFD, 1986) equation results in lower shrinkage strains. In absence of any
other data, engineering judgment should determine the use of one of these two
equations.
ACI (Committee 207, 1986) recommends the following equation for predicting drying
shrinkage strain of moist cured concrete:
t
ε sh ,t = K vs K h 0.78 × 10 −3 (34)
t + 35
Whereas AASHTO recommends:
t
ε sh ,t = K vs K h 0.51 × 10 −3 (35)
t + 35
In this formula ε sh,t is the shrinkage at time t, K vs , and K h are volume to surface
coefficient and humidity coefficient and t is the time after curing. Both codes provide
guidelines for choosing K values.
Due to small thickness of the bridge deck, shrinkage strains across the section are
considered uniform. This means that a temperature decrease equal to ε sh ,t / α Deck should
be applied to deck concrete to model shrinkage.
Autogenous shrinkage is the concrete shrinkage without loss of water. This kind of
shrinkage occurs at low w/c ratios and significantly increases with use of silica fume,
HRWRAs (High Range Water Reducing Admixtures) and finer cement. In the past, this
type of shrinkage was insignificant. However with the downward trend of w/c ratio in
concrete mixes, use of silica fume, use of finer cements, and widespread use of
HRWRAs this type of shrinkage has come into attention. If there is any indication that
this type of shrinkage may be involved, magnitude of this shrinkage should be
considered in similar manner.
Temperature Gradients
Daily temperature variations and solar radiation produces temperature differential

throughout the section. Stresses produced by this temperature differential add up to the
locked in stresses from different types of shrinkage. AASHTO (LRFD, 1986) provides
guidelines for estimating positive and negative temperature gradients through the depth.
The temperature gradient profile suggested by AASHTO (LRFD, 1986) for concrete
girder is shown in Figure 86. Where C shall be taken as:
• 30-cm (12.0-inch) for concrete superstructures that are 16.0” or more in depth
• For concrete sections shallower than 40-cm (16.0-inch), C shall be taken 10-cm
(4.0-inch) less than the actual depth
• 30-cm (12.0-inch) for steel superstructures
114
Based on the location of the bridge different values are suggested by AASHTO to define
the temperature profile (Table 31).
Figure 86. AASHTO temperature gradient profile for steel and concrete girders.
Table 31. Basis for temperature gradients (AASHTO LRFD 1998).

o o o o
Zone T1 C ( F) T1 C ( F)
1 30 (54) 7.8 (14)
2 25.6 (46) 6.7 (12)
3 22.8 (41) 6.1 (11)
4 21.1 (38) 5 (9)
Temperature Rise Due to Hydration
Concrete temperature rises after placement due to hydration. If the concrete is restraint,
part of temperature rise, subsequent cooling, and increase in modulus of elasticity
produces tensile stresses in concrete and may cause cracking few days after
placement. It is also possible that these locked in stresses add up to stresses produced
by other causes. To evaluate hydration temperature effects, the maximum temperature
rise during hydration should be estimated. After computing the maximum temperature
rise the structure should be analyzed under a temperature drop equal to peak
temperature minus setting temperature (worse case scenario). Setting temperature of
bridge is defined as actual air temperature averaged over 24-hour period immediately
preceding setting (AASHTO LRFD, 1986).
115
ACI report 207 (1986),which mainly discusses mass concrete, provides a method for
estimating the maximum temperature rise during hydration. ACI mentions, “The report
can be applied to normal structural concrete; however its application is not usually
warranted”. In the absence of any other data, it is suggested that ACI procedure can be
applied to estimate the maximum temperature rise
Windows Based Application

A user friendly MS Windows based application is prepared to perform the analysis
based on the input values from user. Figure 87 shows typical input and output windows
of the program. The program can be downloaded from http://web.njit.edu/~ala.
Figure 87. Typical input and output window of application.
116
CONCLUSIONS AND RECOMMENDATIONS
Many concrete bridge decks develop transverse cracking and most of these cracks
develop at early ages, many right after construction. These cracks are typically full-
depth and spaced 1-3 m (3-10 ft) apart along the length of the bridge. Transverse
cracks can accelerate corrosion of reinforcing steel, deteriorate deck concrete, possibly
cause damage to underneath components of the bridge, and damage bridge esthetic.
As a result of these adverse effects of transverse cracking, the maintenance costs will
increase and ultimately the service life of the bridge system will be shortened. There
have been many studies on the cause of transverse deck cracking over the past several
decades. However, the causes of transverse deck cracking are not yet fully understood
and the problem still exists. Previous studies were mostly focused on concrete mix
design and improvement through changes to construction practice to alleviate shrinkage
problem. In many instances there are major disagreements on the factors affecting
transverse cracking indicating the need for further research. As a part of this research
study 24 bridges were surveyed in the state of New Jersey. Material and mix design
values for these bridges were consistent with the recommendations reported in the
literature. Despite this fact, majority of these bridges surveyed have developed cracks
highlighting the importance of design factors, which was the main thrust of this study.
Cracks in concrete occur when a restraint mass of concrete tends to change volume.
Volume change in concrete depends mix design and construction procedures.
Restraint, which is basically due to composite action of deck and girder, depends on
design characteristics of the bridge such as continuity, relative deck to girder stiffness,
span length, girder spacing, amount of deck reinforcement, etc. To study these factors
several 2-D and 3-D linear and nonlinear finite element models were employed. Results
along with specific conclusions are presented in details in Chapter 5 and will not be
repeated here. However, based on these analytical work, survey of bridges, and
literature review the following conclusions can be highlighted. These are presented in
the form of specific recommendations for possible implementation, and are grouped
under the three categories of structural design factors, mix design, and constructions
practices.
Structural Design Factors: Recommendations

Specify an Upper Limit on Concrete Strength
Survey results indicate that the actual strength is much higher than the specified design
strength. Therefore, it is recommended that an upper limit on concrete strength be
required. This should not be viewed upon as discouragement to the use of HSC but
rather a requirement on design versus actual concrete strength. Also, when “open-early”
is not an issue, design should use low-early strength concrete.
117
Boundary Restraints Should be Consistent with Design
Construction practice should not introduce undue boundary restraint on the girders. For
example, for a span designed as simply supported, which is desirable under shrinkage
load, the girders should not be embedded in the end diaphragms or they should be
debonded.
Time-Dependent Loadings must be considered in Design of Bridges with

Integral Abutments
Integral abutments have many structural and maintenance benefits and are becoming
more popular around the nation including in the state of New Jersey. However, bridges
with integral abutments have a much higher tendency for transverse deck cracking.
Therefore, time and temperature dependent loadings must be explicitly considered in
their design. A simple and effective design tool to facilitate this aspect of design has
been developed under this study (see chapter six).
Minimize the Ratio of Girder/Deck Stiffness
Try to minimize the ratio of girder to deck stiffness through changes in deck thickness,
girder spacing, and girder moment of inertia. As have been shown, larger spacing,
flexible girders and thicker decks are preferred. Try to provide the required moment of
inertia with more contribution from the deck.
Employ more Flexible Superstructures
Analyses results indicate that more flexible superstructures have lower tendency for
deck cracking. Therefore, the design should employ a more flexible superstructure. This
objective can be pursued through two different venues. Under current practice, where
there is a deflection limit, the design manual should prevent design of an overly stiff
superstructure by putting a limit on the margin by which the deflection requirement is
satisfied. Currently strength requirements usually control the design, and deflection is
only checked and it is quite common for this requirement to be satisfied by a very large
margin (e.g., deflection equal to 1/1500 of span length as oppose to required 1/1000 of
span length). A second approach would be, consistent with AASHTO LRFD, to drop
deflection requirement under service load or at least to increase the limit.
Use Uniform Reinforcement Meshes
Uniform reinforcement meshes on top and bottom are recommended to control

cracking. Use of empirical design method is encouraged. Increasing the volume of
reinforcement above code requirement does not have any effect on cracking and is
discouraged.
118
Consider AASHTO Article 3.12
Current designs often do not explicitly consider time-dependent loadings as specified

under AASHTO article 3.12. Such a consideration is recommended. A simple windows
application tool (StrEstimate) has been developed under this study to facilitate
comparison of different designs and estimate deck stresses during the design stage.
Mix Design: Recommendations

Adopt a Restraint Shrinkage Test
Concrete mix to be used in bridge deck should be tested for cracking using one of
cracking tendency tests described in this report. Cements and mixes with poor results
should not be used in bridge deck.
Reduce cement content to 650-660 lb/yd3, and consider using fly ash.
Use AASHTO specification Type II cement for bridge deck construction.
Limit Water Cement Ratio
Limit water cement (w/c) ratio to 0.4-0.45. Make use of water reducers to reduce water
content. Consider w/c<0.4 with the use of water reducers.
Aggregate Size and Shape
• Use largest possible grain size as specified in ACI –318.

• Use crushed stone for coarse aggregate.
• Maximize aggregate content.
Consider using type K Shrinkage Compensating Concrete when available
Construction Practices: Recommendations
Employ the following pouring sequence
• Pour complete deck at one time whenever feasible within the limitation of the
maximum placement length based on drying shrinkage consideration.
• If multiple placements must be made and the bridge is composed of simple
spans, then place each span in one placement.
• If bridge is simple span, but cannot be placed in a single placement, divide the
deck longitudinally and make two placements.
119
• If the bridge is simple span and single placement cannot be made over full span
length, then place the center of span segment first and make this placement as
large as possible.
• If multiple placements must be made and the bridge is continuous span, then
place concrete in the center of positive moment region first and observe a 72-h
delay between placements.
• When deck construction joints are created, require priming existing interfaced
surfaces with a Primer/Bonding agent prior to placement of new concrete.
Wind, Weather and Concrete Temperature (Already included in the NJDOT

Specs)
Follow procedures in sections 501.12 and 501.17 of NJDOT Specs. Make use of
evaporation rate chart proposed by ACI. Cast the deck in mild temperatures. It is also
recommended that wind and humidity levels be recorded on the Inspection/Testing
datasheets.
Protection and curing (Already included in NJDOT Specs)
Follow procedures in section 501.17. Start curing immediately after finishing and cure at
least 7 consecutive calendar days. If “early-open” is not an issue consider 14-day wet
curing.
Use temporary shoring for simply supported girders when practical
Future Research Needs

In light of the results of this and prior studies and the recommendations made under this
study the following areas for possible future research work are identified.
Restraint Shrinkage Tests
There are two distinct shrinkage tests (bar test and ring test) that can be used in
determining shrinkage potential of deck concrete. A study should be conducted to
determine the suitability of one of these two tests. More importantly, although several
highway agencies are considering the use of these tests, there is still a great need for
additional research work to quantify the results of these tests.
Impact of Flexible Superstructure
One recommendation of this study is to employ a more flexible superstructure. It is

prudent to conduct additional research study on the implication of this recommendation
on serviceability requirements. Results of such a study will also be beneficial in
determining the impact of AASHTO-LRFD provision (or lack of) on deflection.
120
Study on Construction Practices
As it was mentioned before, many of the recommendations on mix design and

construction practice already exist in the most recent version of NJDOT Standard
Specifications for Road and Bridge Construction (NJDOT Specs.). A study to monitor
and compare performance of bridges built based on these specifications will be valuable
to our understanding of the causes of deck cracking and will help in identifying more
specifically the important construction and mix design factors.
Controlled Composite Action
As discussed before, there is an agreement among researchers that shrinkage (thermal

and drying) and a temperature gradient of restraint deck are the main causes of
cracking. The restraining effect is basically due to the composite design of the
superstructure. Composite construction reduces the size of girders, which means less
cost. While an economical design requires composite behavior under ultimate loads,
under initial portion of service loads (such as dead load) the girders alone are generally
capable of resisting the entire load. Thus, shrinkage cracking can be prevented if a
mechanism can be developed to prevent composite action during early ages (i.e., as
concrete shrinks), while it is activated for higher service load and under ultimate loading
condition. To achieve this objective (i.e., no composite action initially and full composite
action ultimately) the shear connectors can be wrapped in hyperelastic material of
carefully designed thickness, as shown in Figure 88.
Design of these materials, i.e., their thickness, modulus of elasticity, etc. should be
chosen very carefully to achieve the required controlled composite action (CCA). A
typical stress strain curve for hyperelastic materials is shown in Figure 89. Under low
level of stresses the material does not provide any resistance and it deforms easily.
Thus, when the shear connectors are wrapped by such a material the concrete deck
can shrink without any restraint. There might be a need for bond breaker between the
deck and girder to further reduce bonding. Upon development of shrinkage strains the
hyperelastic material will start to develop higher level of resistance and will ultimately
provide full composite action. Proper design will require development of a realistic
relationship between the shrinkage strain and the wrap thickness. A feasibility study to
investigate the viability of such a design is highly recommended since it has the
potential to entirely eliminate the problem of transverse cracking in concrete bridge
decks.
121
Figure 88. Details of controlled composite action connector.
Figure 89. Stress-strain curve for typical hyperelastic materials
122
REFRENCES
1. American Association of State Highway and Transportation Officials. AASHTO
LRFD Bridge Design Specification, 2nd Edition. Washington D.C., 1998
2. ACI committee 209. Prediction of Creep, Shrinkage, and Temperature Effects.
Detroit, 1978
3. ACI Committee 207. Shrinkage and Creep effects in Concrete. Detroit, 1986
4. ACI committee 308. Standard Practice for Curing Concrete. Detroit, 1986
5. American Concrete Institute. ACI Manual of Concrete Practice- Part 1:Materials and
general Properties of Concrete, Detroit, 1986
6. ANSYS. Computer Software, ANSYS Corporation, 1994-1998
7. K. Babaei and R.L. Purvis. “Premature Cracking of Concrete Bridge Decks: Cause
and Methos of prevention” Proceedings, 4th International Bridge Engineering
Conference, 1996
8. R.L. Blaine, H.T. Arni, D.H. Evans, M.R. Defore, J.R. Clifton and R.G. Methey.
“Interrelation between Cement and Concrete Properties” Building Research Division
of National Bureau of Standards, 1953
9. R. Breitenbucher and M. Mangold. “Minimization of Thermal Cracking in Concrete
Members at Early Ages” Thermal Cracking in Concrete at Early Ages, E&FN Spon,
London, pp.205-212, 1994
10. Burrows R, Visible and Invisible Cracks. ACI Monograph No. 11, 1999
11. R.W. Carlson. “Cracking of Concrete” presented at Joint meeting of the Boston
Society of Civil Engineers and The Designers’ Section, BSCE, Jan. 14, 1942
12. T.T.H. Cheng and D.W. Johnston. Incidence Assessment of Transverse Cracking in
Concrete Bridge Decks: Construction and Materials Considerations. Rep. No.
FHWA/NC/85002, 1, Dep. Of Civ. Engrg., North Carolina State University, Raleigh,
N.C., 1985
13. J.J. Chui and W.G. Dilger. “Temperature Stress and Cracking Due to Hydration
Heat”, Creep and Shrinkage of Concrete, E&FN Spon, London, pp.271-276, 1993
14. F.H. Dakhil, P.D. Cady and R.E. Carrier. “Cracking of Fresh Concrete as Related to
Reinforcement” ACI Struct. Journal, 72(8), pp. 421-428, 1975
15. J. Ducret and J. Lebet. “Measurements on a Composite Bridge: Effect of Concrete
Hydration” Proceedings of an Engineering Foundation Conference, American
Society of Civil Engineers, New York, 1996
16. J. Emanuel and J. L. Hulsey. “ Temperature distributions in composite bridges”
Journal of Structural Division, ASCE, Jan 1978, pp 65-78
17. C. French, L. Eppers, Q. Le, and J.F. Hajjar. “Transverse cracking in concrete
bridge decks” Transportation Research Record. n 1688 Nov 1999. p 21-29
18. R.J. Frosch, R.D. Radabaugh and D.T. Blackman. “Investigation of Transverse Deck
Cracking“ ASCE Structures Congress, Denver, April 2002
19. G.T. Halverson. “Troubleshooting concrete cracking during construction,”
Aberdeen’s Concrete Constr., 36(11), 811-816, 1991
123
20. J. Houde. Study of force-displacement relationship for the finite element analysis of
reinforced concrete. Report No. 72-3, Dept. of Civil Engineering and applied
mechanics, McGill University, Montreal, Dec. 1973.
21. M.A. Issa. “Investigation of cracking in concrete bridge decks at early ages” Journal
of Bridge Engineering, V 4 n 2 1999, p 116-124
22. A.A. Khan, W.D. Cook, and D. Mitchell. “Thermal properties and transient thermal
analysis of structural members during hydration” ACI Materials Journal, V95 n 3
May-Jun 1998, p 293-303
23. T. Kochanski, J. Parry, D. Pruess, L. Schuchardt, and J. Ziehr, J. Premature
Cracking of Bridge Decks Study. Wisconsin Dept. of Transportation, Madison,
Wisconsin, 1990
24. P.D. Krauss, and E.A. Rogalla. Transverse Cracking in Newly Constructed Bridge
Decks. NCHRP Report 380, Transportation Research Board, Washington, DC,
1996
25. Microsoft Access 2000, Computer Software, Microsoft Corporation, 1992-1999
26. Microsoft Excel 2000, Computer Software, Microsoft Corporation, 1985-1999
27. C. Meyers. Survey of Cracking on Underside of Classes B-1 and B-2 Concrete
Bridge Decks in District 4. Investigation 82-2, Division of Material and Research,
Missouri Highway and Transportation Department, Jefferson City, Mo., 1982
28. New Jersey Department of Transportation. 1998 Standard Specification for Road
and Bridge Construction and 1998 Supplemental Specification. Trenton, NJ, 1998
29. Portland Cement Association. Final Report - Durability of Bridge Decks – A
cooperative Study. Skokie, Illinois, 1970
30. H.B. Prenger. Bridge deck crackin, Research report. Rep. No. MD-93-04, Maryland
Dept. Of Transp., State Highway Administration, Baltimore, Md.,19., 1992
31. M. Paillere, M. Buil and J.J. Serrano. “Effect of Fiber Addition on the Autogenous
Shrinkage of Silica Fume Concrete” ACI Material Journal, V.86, No.2, Mar.-Apr., pp.
139-144, 1989
32. G.E. Ramey, A.R. Wolff, R.L. Wright. “Structural design actions to mitigate bridge
deck cracking” Practice Periodical on Structural Design & Construction, v 2 n 3 p
118-124, 1997
33. T.R. Schmitt, D. Darwin. “Effect of material properties on cracking in bridge decks”
Journal of Bridge Engineering, v 4 n 1, p 8-13, 1999
34. E. Tazawa, Y. Matsouka, S. Miyazawa, and S. Okamoto. “Effect of Autogenous
Shrinkage on Self Stress in Hardening Concrete” Thermal Cracking in Concrete at
Early Ages, E&FN Spon, London, pp.222-228, 1994
35. E. Tazawa and S. Miyazawa. “Experimental Study on Mechanism of Autogenous
Shrinkage of Concrete” Cement and Concrete Research, V.25, No.8, pp. 1633-1638,
1995
124
36. R. Sprinegnschmid and R. Breitenbucher “Cement with low crack Susceptibility”,
Proceedings of Advances in Cementitious Materials Conference, American Ceramic
Society, pp. 701-713, 1990
37. R. Sprinegnschmid, R. Breitenbucher and M. Mangold. “Thermal Cracking in
Concrete at Early Ages”, E7FN Spon, London, pp. 137-134, 1994
38. G.J. Verbeck and G.J. Foster. “Long-term study of cement performance in concrete”
ASTM Proceedings, Vol 50, pp 1235-1257, 1950
39. Winter and Nilson. Design of concrete structures. Mc-Graw Hill, 1986
40. L.C.P. Yam, and J.C. Chapman. “The inelastic behaviour of simply supported
composite beam of steel and concrete”, Proc. Istin Civ. Engrs, 1968, 41, Dec., 651-
683.
41. W. Zuk. “Thermal and Shrinkage Stresses in Composite Beams” Journal of ACI, Vol
58,p327-339, 1961
125
APPENDIX A: BRIDGE INFORMATION AND SURVEY
RESULTS
This appendix contains important structural design, material properties, concrete mix
design and some construction information, supplemented by the results of field surveys
of several bridges. Data presented here were collected from different sources and were
used in various parts of the study (see chapter 3).
Following this introduction, section A.2 describes major elements of the data for these
bridges and how they are grouped. Definitions of some data entry are also provided
when necessary. The final section (Section A.3) presents the data for bridges surveyed.
Organization
The data presented for each bridge is grouped into six parts. Part one contains general
information about the bridge. This information includes geographical location of the
bridge (i.e. county, township) and construction year, which are obtained from the
original list provided by NJDOT and subsequently entered into the survey forms.
Part two summarizes important design characteristics of the bridge. These data are
mostly derived from the structural plans of the bridge. Note that spans are numbered in
the direction of traffic when it is only in one direction on the bridge, otherwise span
number is explicitly identified. This part contains the following structural design related
data for each span:
• Number of spans
• Traffic direction
• Girder type
• Span type (i.e., continuous or simply supported at interior spans)
• Span length and width
• Framing information (i.e., spacing of girders in each span)
• Deck design information (i.e., rebar details, thickness, cover depth, and wearing
surface)
• Girder properties (i.e., area, depth, and moment of inertial)
• Shear stud spacing
In some cases where the data item is not constant through the range considered, like
span length in curved bridges or the moment of inertia when different girders are used
in one span, the range of data for that item is reported.
Part three presents material properties and mix design information based on NJDOT
inspection/testing datasheets. This information includes cement type, cement content,
water content, air content, w/c ratio, compressive strength, slump and admixtures. As
126
explained in chapter 3, these data represents an average value of the data item
considered for the bridge.
In part four construction related data, such as air temperature, concrete temperature at
the time of placement, and the month of casting are reported. These data are also
obtained from the NJDOT inspection/testing datasheets. Part five summarizes the
survey results, and finally part 6 presents observations and photos during the survey, if
any.
Bridge Information and Survey Results

Bridges are ordered based on the first four digits of NJDOT structure numbers, from the
lowest to the highest.
127
Structure Number 0206-165
1.General Information
Bridge Location Hackensack Avenue over NJ RT4
County Bergen County City Hackensack
Year Built 1998
2.Structural Design Information

Number of spans: 2
Direction of Traffic: Two ways
Girder type: Steel Girder
Span type: Simply Supported
Type Of Bearing: Steel Bearings
Skewness: None
North Bound South Bound

Span Number 1 2 1 2
Span Width 47’-0” 47’-0” 47’-0” 47’-0”
Span Length 90’-2” 90’-2” 90’-2” 90’-2”
Framing Information:
-North Bound (All Spans): 5 Girders@7’-6” + 2 Girders@6’-3”
-South Bound (All Spans): 7 Girders@7’-1”
Deck Slab Information:

1/2 1/2
Deck Thickness 8 ” Top Cover 2 ”
Wearing Surface Concrete, Sawcut Bottom Cover 1”
Longitudinal Rebar Information:

-North & South Bound (All Spans): #5@6”(B), #5@15”(T)
Transverse Rebar Information:
-North & South Bound (All Spans): #6@7”(T&B)
Girder Information:
North & South Bound
4 2
Span Number I (in ) A (in ) Depth (in)
1&2 26876.39 15.58 38.5
Shear Stud Information:

North & South Bound
Span Number Spacing (in) Number of Studs in Row
1&2(Midspan) 12 4
1&2(Support) 9 4
3.Material Properties & Mix Design Information

Cement Type: Blue Circle Atlantic Type II
3
Cement Content: 658 lb/yd
3
Water Content: 35 Gal/yd
W/C Ratio: 0.44
Air Content: 6.6%
Compressive Strength: 5174 Psi
Slump: 3.5”
Admixtures: Water Reducer
128
4.Construction Information
Air Temperature: 73 F
Concrete Temperature: 83 F
Month Of Casting: August-September
5.Survey results
Crack Information:
Span No. 1 2 1 2
Type Transverse/Other Transverse/Other Transverse/Other Transverse/Other
Location Center/End Center/End Center/End Center/End
Spacing 0’-2’ 0’-2’ 0’-2’ 0’-2’
Size Typical Typical Typical Typical
Survey Team Member: Ala
6. Observations & Photos

-None
129
Bridge Location RT21-I-78, RAMP FROM 21S TO I-78 WEST
County Essex Township Newark
Year Built 2000

Number of Spans: 8
Direction of Traffic: One way
Span type: Mixed (3 Continuous + 1 Simply supported+ 4 Continuous)
Skewness: Mild
Ramp
Span Number 1 2 3
Span Width 29’-6”~32’-4” 29’-6” 29’-6”~ 42’-5”
Span Length 165’-6” 158’-8” 143’-10”
Span Number 4 5 6
Span Width 42’-9”~82’-1” 38’-6” 38’-6”
Span Length 142’-7” 131’-6” 165’-0”
Span Number 7 8
Span Width 38’-6” 36’-`0”~38’-6”
Span Length 165’-0” 131’-2”
-Ramp:
-Span 1: 4 Girders @8’-2”
-Span 3: 4 Girders @8’-2”~12’-6”
-Span 4: 3 Girders @8’-2”~12’-9”+4 Girders@ 6’-4”~12’-10”

1/2” 1/2
Wearing Surface Concrete, Turf Drug Bottom Cover 1”

-All Spans Midspan: #5@8”(B), #5@15”(T)
1/2
-Continuous Spans Over Pier: #5@8”(B), #5@7 ”(T)
-All Spans EB: #6@6”(T&B)
Girder Information:
Ramp
4 2
1 (Midspan) 76151~105242 81.5~100.75 78~79
1 (support) 66278~165253 73.75~136.25 78~81
130
2 (Midspan) 61947~64896 70.75~72.75 78
2 (support) 110331~156253 102.25~136.25 79.5~81.5
3 (Midspan) 73450~77368 79.25~81 78
3 (support) 73450~165253 79.25~136.25 78~81.5
4 (Midspan) 72620~111092 81~109 78
4 (support) 62992~94680 74~96 78
5 (Midspan) 67623~99855 74.75~101.75 78~78.75
5 (support) 64896~200418 72.75~159.75 78~81.75
6 (Midspan) 67623~94739 74.75~99.63 78~78.5
6 (support) 136365~200418 119.75~159.75 80.5~81.75
7 (Midspan) 67623~94739 74.75~99.63 78~78.5
7 (support) 136365~183783 119.75~149.5 80.5~81.25
8 (Midspan) 67623~99855 74.75~101.75 78~78.75
8 (support) 67623~183783 74.75~149.5 78~81.25

Ramp
1 (Midspan) 19~22 3
1 (support) 16~18 3
2 (Midspan) 24 3
2 (support) 10 3
3 (Midspan) 24 3
3 (support) 17~19 3
4 (Midspan) 12~22 3
4 (support) 15~18 3
5 (MIDSPAN) 15 3
5 (support) 12~24 3
6 (Midspan) 15 3
6 (support) 24 3
7 (Midspan) 15 3
7 (support) 24 3
8 (Midspan) 15 3
8 (support) 24 3

Cement Type: Essroc Type II
3
3
Water Content: 32.6 Gal/yd
W/C Ratio: 0.41
Air Content: 5.5 %
Slump: 3.3”
Month Of Casting: July-October
5.Survey results
Crack Information:
Ramp
Span No. All Spans
131
Type Transverse
Location All over
Spacing >5’
Size Hair Line
Survey Team Member: Ala, Allyn, Ross and Raj

-Not open to traffic at the time of survey but used by construction vehicle
-Hairline Cracks Observed.
Figure 90. Crack on the deck (Bridge No 0713-151).
132
Figure 91. Construction vehicle traffic on the bridge (Bridge No. 0713-151).
Figure 92. Crack on the deck (Bridge No. 0713-151).
133
Figure 93. Crack extending into the parapet (Bridge No. 0713-151).
134
Figure 94. Crack on deck and parapet (Bridge No. 0173-151).
135
Bridge Location RT31 Over South Branch Raritan River
County Hunterdon County Township Clinton
Year Built 1996

Number of spans: 2
Girder type: Prestressed Concrete Girder
Skewness: Mild

Span Number 1 2 1 2
Span Width 49’-0” 49’-0” 52’-9” 52’-9”
Span Length 124’-10”~125’-1” 124’-11”~124’-10” 124’-10”~124’-9” 124’-10”~124’-7”
-North Bound:
-Span1: 6Girders@7’-11” + 1 Girder@Varies (7’-10”~6’-9”)
-Span2: 6Girders@7’-4” + 1 Girder@Varies (6”-9”~ 6’-1”)
-South Bound:
-Span1: 6Girders@8’-0” + 1 Girder@Varies (6’-6”~7’-2”)
-Span2: 6Girders@7’-10” + 1 Girder@Varies (7”-2”~ 8’-3”)

3/4 1/4 1/4
Deck Thickness 7 +1 ” Latex Concrete Top Cover 2 ”
Wearing Surface Latex Concrete, Sawcut Bottom Cover 1”

-North Bound (All Spans): #5@9”(B), #5@15”(T)
-South Bound (All Spans): #5@6”(B), #5@15”(T)
-North Bound (All Spans): #6@8”(T&B)
-South Bound (All Spans): #5@6”(T&B)
Girder Information:
North & South Bound
4 2
1&2 761550.4 1125 72

North & South Bound
1&2(Midspan) 12 2
1&2(Support) 6 2

3
3
136
W/C Ratio: 0.44
Air Content: 5.4%
Slump: 3.5”
Month Of Casting: October
5.Survey results
Crack Information:
Span No. 1 2 1 2
Type None None None None
Location - - - -
Spacing - - - -
Size - - - -
Survey Team Member: Ala, Rambod

-Cracks on side walk
-Crack on North bound Approach slab
Figure 95. A view from underneath the bridge (Bridge No. 1013-151).
137
Figure 96. Girder end condition at the abutment (Bridge No. 1013-151).
138
Bridge Location Alexander RD Over US 1
County Mercer County Township West Windsor
Year Built 1996

Number of Spans: 2
Skewness: None
East Bound West Bound

Span Number 1 2 1 2
Span Width 67’-10” 67’-10” 43’-9” 43’-9”
Span Length 95’-3” 95’-3” 95’-3” 95’-3”
-East Bound (All Spans): 9 Girders@7’-11” + 1 Girder@5’-3”
-West Bound (All Spans): 6 Girders@7’-11”

3/4 1/2

-All Spans EB & WB: #5@6”(B), #5@15”(T)
-All Spans EB & WB: #6@6”(T&B)
Girder Information:
East & West Bound
4 2
1&2 (Midspan) 38457.26 96.375 45.125~45.5
1&2 (Support) 24587.67 68.875 45.125~45.5

East & West Bound
1&2 (Midspan) 14” 4
1&2 (Support) 9” 4

Cement Type: White Hall Type II
3
3
W/C Ratio: 0.40
Air Content: 5.4%
Slump: 3.1”
Admixtures: Water Reducer & Retarder
139
Air Temperature: 56.9 F
Concrete Temperature: 72.6 F
Month Of Casting: April
5.Survey results
Crack Information:
East Bound West Bound
Span No. 1 2 1 2
Location - - - -
Spacing - - - -
Size - - - -
Survey Team Member: Ala, Rambod

-None
140
Bridge Location NJ Route 29 WB Over Watson’s Creek
County Mercer Township Hamilton
Year Built 1995

Number of Spans: 12
Span type: Mixed (1 Simply Supported + 1 Simply Supported+ 1 Simply Supported+ 3 Continuous+ 3
Continuous+ 3 Continuous)
Skewness: None
West Bound
Span Number 1 2 3 4
Span Width 59’-1”, 75’- 111’~95’-3” 95’-3”~90’-1” 90’-1”~88’-9”
11”~51’-11”
Span Length 88’-10” 90’ 90’ 90’
West Bound
Span Number 5 6 7 8
Span Width 88’-9”~87’-6” 87’-6”~86’-3” 86’-3”~84’-7” 84’-7”~82’-4”
Span Length 90’ 90’ 60’ 60’
West Bound
Span Number 9 10 11 12
Span Width 82’-4”~80’-6” 80’-6” 80’-6” 80’-6”
Span Length 60’ 90’ 90’ 88’-10”
-West Bound:
-Span 1: 8 Girders @7’-9” + 2 Girders @8’-5” + 6 Girders @9’~4’-11”
-Span 2: 10 Girders @8’-5” + 3 Girders @9’-8” ~5’-1”
-Span 10-11-12: 10 Girders @8’-5”

1/2” 1/4” 3/4
Deck Thickness 7 + 1 Latex Concrete Top Cover 2 ”

1/2
-All Spans WB Midspan: #5@7 ”(B), #5@15”(T)
1/2 1/2
-All Spans WB Pier: #5@7 ”(B), #5@7 ”(T)
-All Spans WB: #6@6”(T&B)
141
Girder Information:
West Bound
4 2
1-6 & 10-12 260034 789.0 54
7-8-9 125014 559.5 45

West Bound
1-6 & 10-12 (Midspan) 24” 2
1-6 &10-12(Support) 6” 2
7,8,9 (Midspan) 18” 2
7,8,9 (Support) 12” 2

Cement Type: Essroc type II
3
3
W/C Ratio: 0.38
Air Content: 5.5 %
Compressive Strength: Psi
Slump: 3”
Month Of Casting: November-December
5.Survey results
Crack Information:
West Bound
Span No. All Spans
Type None
Location -
Spacing -
Size -
Survey Team Member: Ala, Fred and Tom

-Continuity Joints cracked on this bridge but these joints did not cracked on the eastbound side bridge,
which is similar to this bridge. Cracks are observed on eastbound side deck while the westbound side
deck was un-cracked.
142
Figure 97. Cracked continuity joint (Bridge No. 1130-152).
Figure 98. Another cracked continuity joint (Bridge No.1130-152).
143
Figure 99. End condition in simply supported spans (Bridge No 1130-152).
144
Bridge Location NJ Route 29 EB Over Watson’s Creek
Year Built 1995

Number of Spans: 11
Span type: Mixed (2 Continuous + 3 Continuous+ 3 Continuous+ 3 Continuous)
Skewness: None
East Bound
Span Number 1 2 3
Span Width 55’-9”, 47’-9”~50’-7” 55’-9”, 50’-7”~52’-11” 55’-9”, 52’-11”~54’-7”
Span Length 88’-10” 90’ 90’
East Bound
Span Number 4 5 6-9
Span Width 55’-9”, 54’-7”~55’-6” 55’-9”, 55’-6”~55’-9” 55’-9”, 55’-9”
Span Length 90’ 90’ 90’
East Bound
Span Number 10 11
Span Width 55’-9”, 55’-9”~58’-8” 55’-9”, 58’-8”~71’-0”
Span Length 90’ 88’-10”
-West Bound:
-Span 1: 7 Girders @8’-6”, 4 Girders @8’-6” + 2 Girders @4’-6”~5’-9”
-Span 6-8: 7 Girders @8’-6”, 6 Girders @8’-6”
-Span 10: 7 Girders @8’-6”, 5 Girders @8’-6” + 1 Girders @9’-4”~10-2”

3/4 1/4 3/4

1/2
- All Spans EB Midspan: #5@7 ”(B), #5@15”(T)
1/2 1/2
- All Spans EB Pier: #5@7 ”(B), #5@7 ”(T)
- All Spans EB: #6@6”(T&B)
Girder Information:
East Bound
145
4 2
All 260034 789.0 54

East Bound
1-6 & 10-12 (Midspan) 24” 2
1-6 &10-12(Support) 6” 2

Cement Type: Essroc type II
3
3
W/C Ratio: 0.38
Air Content: 3.2 %
Slump: 5.6”
Admixtures: Water Reducer &Retarder
5.Survey results
Crack Information:
East Bound
Span No. All Spans
Type Transverse
Location All over
Spacing 3-4’
Size Typical

-Continuity Joints did not cracked but these joints cracked on the westbound side. Cracks are observed
on eastbound side.
146
Figure 100. Transverse cracks on the deck (Bridge No.1130-153).
Figure 101. Un-cracked continuity joint (Bridge No.1130-153).
147
Bridge Location RT NJ29 Freeway Over Ramps G, H, I, J and Conrail B Town
County Mercer County Township Hamilton
Year Built 1995

Number of spans: 2
Span type: Continuous
Skewness: Mild
South Bound North Bound

Span Number 1 2 1 2
Span Width 66’ 66’ 70’-5” 70’-5”
Span Length 151’ 98’ 98’ 151’
-South Bound (All Spans): 8 Girders@8’-9”
-North Bound (All Spans): 8 Girders@9’-3”

3/4 1/4 3/4

1/2
-Mid Span: #5@7 ”(B), #5@15”(T)
1/2
-Over Pier: #5@7 ”(T&B)
-All Spans: #6@6”(T&B)
Girder Information:
North Bound
4 2
1 (Support) 147906~256541 130.6~198.5 80.75~83
1 (Midspan) 147906~196700 130.6~163.5 80.75~81.75
2 (Support) 162400~256541 137.6~198.5 81~83
2 (Midspan) 159928~208896 137.6~170 81~82
South Bound
4 2
1 (Support) 91025~206425 111~165 68.75~82.00
1 (Midspan) 91025~159928 111~137 68.75~81.00
2 (Support) 84317~206425 105~165 68.5~82.00
2 (Midspan) 84317~137106 105~123 68.5~80.5

North & South Bound
Support 8~12 4
Midspan 8~23 4
148
3
3
W/C Ratio: 0.37
Air Content: 5.4%
Slump: 3.1”
5.Survey results
Crack Information:
Span No. 1 2 1 2
Location - - - -
Spacing - - - -
Size - - - -
Survey Team Member: Ala., Tom and Fred

-Approach slab cracked.
149
Bridge Location RT NJ29 Ramp H Over Ramp G and Conrail B Town
Year Built 1995

Number of spans: 2
Skewness: Severe
Ramp
Span Number 1 2
Span Width 44’-3” 44’-3”
Span Length 162’-0” 159’-9”
-All Spans: 5 Girders@9’-3”

1/2
Deck Thickness 9” Top Cover 2 ”

-Mid Span: #5@8”(B), #5@15”(T)
-Over Pier: #5@8”(B), #5@5”(T)
Girder Information:
Ramp
4 2
1 (Support) 194714~301013 173~220 75~84
1 (Midspan) 116694~301013 117~220 75~84
2 (Support) 194714~301013 173~220 75~84
2 (Midspan) 116694~301013 117~220 75~84

Ramp
All Spans 24 4

3
3
W/C Ratio: 0.37
Air Content: 5.6%
Slump: 3.4”
150
Month Of Casting: July
5.Survey results
Crack Information:
Ramp
Span No. 1 2
Type None None
Location - -
Spacing - -
Size - -

-Approach slab cracked.
151
Bridge Location RT NJ29 Ramp F Over Ramp G and Conrail B Town
Year Built 1995

Number of spans: 4
Direction of Traffic: 1
Span type: 3 Span Continuous (1,2, &3) + 1 Span Simply Supported (4)
Skewness: Mild
Ramp
Span Number 1 2 3 4
Span Width 33’-6” 33’-6” 33’-6” 33’-6”
Span Length 124’-11” 147’-11” 139’-11” 49’-11”
-All Spans: 5 Girders@7’-3”

1/2 1/2
Deck Thickness 8 ” Top Cover 2 ””

-Span 1-3 (Mid Span): #5@9”(B), #5@15”(T)
-Span 1-3 (Over Pier): #5@9”(B), #5@5”(T)
-Span 4: #5@9”(B), #5@15”(T)
-Span 1-3: #6@9”(T&B)
-Span 4: #6@7”(T&B)
Girder Information:
Ramp
4 2
1 (Support) 73512.66~118564.4 90.5~126.5 70
1 (Midspan) 73512.66 90.5 70
2 (Support) 118564.4 126.5 70
2 (Midspan) 73512.66 90.5 70
3 (Support) 73512.66~118564.4 90.5~126.5 70
3 (Midspan) 73512.66 90.5 70
4 9750~34371 47~52 36~70

Ramp
1(Midspan) 21 3
1(Support) 21~24 3
2(Midspan) 16 3
2(Support) 24 3
3(Midspan) 18 3
152
3(Support) 18~24 3
4 9~15 3

3
3
W/C Ratio: 0.37
Air Content: 5.8%
Slump: 3.4”
Month Of Casting: September
5.Survey results
Crack Information:
Ramp
Span No. 1 2 3 4
Location - - - -
Spacing - - - -
Size - - - -

Approach slab cracked.
153
Bridge Location Route I-295 South Bound Over RT I-195
Year Built 1995

Number of spans: 2
Type Of Bearing: Steel Bearing
Skewness: None
South Bound
Span Number 1 2
Span Width 97’-6” 97’-6”
Span Length 129’ 97’
-All Spans: 6 Girders@10’-3”+ 1 Girder@5’+ 4 Girder @8’

1/2” 1/4” 3/4”
Deck Thickness 8 + 1 Latex Concrete Top Cover 2
Wearing Surface Latex concrete, Sawcut Bottom Cover 1”

-All Spans: #5@15” (T), #6@6” (B)
Girder Information:
South Bound
4 2
1 (Support) 55001~62590 65~73.75 79
1 (Midspan) 64196~72198 72~81.75 79
2 (Support) 80990~100124 88.5~100 80
2 (Midspan) 106866~130627 104.5~119.5 81.5

South Bound
All Spans (Midspan) 15~18 4
All Spans (Support) 9~10 4

3
3
Water Content: 31.8 lb/yd
W/C Ratio: 0.38
Air Content: 5.6 %
Slump: 3.1”
154
Month Of Casting: November
5.Survey results
Crack Information:
Ramp
Span No. 1 2
Type None None
Location - -
Spacing - -
Size - -
Survey Team Member: Fred, Ala

-Approach slab cracked
155
Bridge Location Route 133 EB Over One Mile Road
County Mercer County
Year Built 1998

Number of Spans: 1
Girder type: Prestressed Concrete
Type Of Bearing: Elastomeric Pad
Skewness: None
East Bound
Span Number 1
Span Width 42’
Span Length 91’-11”
-East Bound (All Spans): 5 Girders@9’-3”

Deck Thickness 8.75” Top Cover 2.5”
Wearing Surface Concrete, Saw cut Bottom Cover 1”

-Longitudinal Bars (All Spans): #5@15 (T), #5@9 (B)
-Transverse Bars (All Spans EB): #6@7 (T&B)
Girder Information:
East Bound
4 2
1 540516 1053 63

East Bound
1 (Midspan) 24 6
1(Support) 12 6

3
3
W/C Ratio: 0.38
Air Content: 5.2 %
Slump: 3.5”
Admixtures: Water Reducer &Retarder
156
Air Temperature: 60F
Concrete Temperature: 78F
Month Of Casting: June-July
5.Survey results
Crack Information:
East Bound
Span No. 1
Type Transverse
Location All Over
Spacing 3’-5’
Size Wide
Survey Team Member: Frank, Fred, Ala, and Tom

-Cracks begin parallel to the bridge longitudinal axis at the ends and change direction and become
normal to the axis as it goes towards the center of span.
-Bridge end diaphragm cast around the girders.
-Approach slabs cracked.
Figure 102. Deck cracks at the bridge end (Bridge No. 1143-466).
157
Figure 103. Wide view of deck cracks at the bridge end (Bridge No. 1143-466).
Figure 104. Close view of a crack (Bridge No. 1143-166).
158
Bridge Location Route 133 WB Over One Mile Road
Year Built 1998

Number of Spans: 1
Girder type: Prestressed concrete
Span type: Simply supported
Type Of Bearing: Elastomeric Pads
Skewness: None
West Bound
Span Number 1
Span Width 42’


All Spans: #5@15 (T), #5@9 (B)
All Spans EB: #6@7 (T&B)
Girder Information:
West Bound
4 2
1 540516 1053 63

West Bound
1 (Midspan) 24 6
1 (Support) 12 6

3
3
W/C Ratio: 0.38
Air Content: 5 %
Slump: 3”
159
Month Of Casting: December- November
5.Survey results
Crack Information:
West Bound
Span No. 1
Type Transverse
Location All Over
Spacing 3’-5’
Size Wide
Survey Team Member: Fred, Ala, Tom, and Frank

160
Bridge Location Route 133 EB Over Rocky Brook
Year Built 1998

Number of Spans: 2
Skewness: None
East Bound
Span Number 1 2
Span Width 40’-5” 40’-5”


-All Spans EB Midspan: #5@15 (T) #5@9 (B)
-All Spans EB Pier: #8@5 (T) #7@9 (B)
-All Spans EB: #6@7 (T&B)
Girder Information:
East Bound
4 2
1&2 (Midspan) 540516 1053 63
1&2 (Support) 540516 1053 63

East Bound
1&2 (Midspan) 12 6
1&2 (Support) 6 6

3
3
W/C Ratio: 0.38
Air Content: 6.7%
Slump: 3.5”
161
Month Of Casting: October
5.Survey results
Crack Information:
East Bound
Span No. 1 2
Type Transverse Transverse
Location All over All over
Spacing 0-2’ 0-2’
Size Typical Typical

Figure 105. Girders embedded in end diaphragm (Bridge No.1143-168).
162
Figure 106. Continuity joint (Bridge No. 1143-168).
163
Bridge Location Route 133 WB Over Rocky Brook
Year Built 1998

Number of Spans: 2
Skewness: None
West Bound
Span Number 1 2
Span Width 40’-5” 40’-5”


All Spans WB Midspan: #5@15 (T) #5@9 (B)
All Spans WB Pier: #8@5 (T) #7@9 (B)
All Spans WB: #6@7 (T&B)
Girder Information:
West Bound
4 2
1&2 (Midspan) 540516 1053 63
1&2 (Support) 540516 1053 63

West Bound
1&2 (Midspan) 12 6
1&2 (Support) 6 6

3
3
W/C Ratio: 0.38
Air Content: 6%
Slump: 3.5”
164
Month Of Casting: November
5.Survey results
Crack Information:
West Bound
Span No. 1 2
Spacing 0-2’ 0-2’
Survey Team Member: Tom, Fred, Ala, and Frank

Figure 107. Transverse cracks on bridge deck (Bridge No. 1143-169).
165
Bridge Location Route 133 EB Over Route 130
Year Built 1998

Number of Spans: 2
Skewness: None
East Bound
Span Number 1 2
Span Width 52’-6” 52’-6”
Span Length 80’-6” 87’


-All Spans EB Midspan: #7@10”+#5@10”(T), #7@9”(B)
-All Spans EB Pier: #8@10” + #7@10” (T), #7@9” (B)
All Spans EB: #6@7” (T&B)
Girder Information:
East Bound
4 2
1&2 (Midspan) 260034 789 54
1&2 (Support) 260034 789 54

East Bound
1&2 (Midspan) 24 2
1&2 (Support) 12 2

3
3
W/C Ratio: 0.38
Air Content: 5.3 %
Slump: 4.0”
166
:
Air Temperature 76F
Month Of Casting: May
5.Survey results
Crack Information:
East Bound
Span No. 1 2
Spacing 3-5” 3-5”

Figure 108. Repaired transverse crack (Bridge No. 1143-170).
167
Figure 109. Close up view of repair patch (Bridge No. 1143-170).
168
Bridge Location Route 133 WB Over Route 130
Year Built 1998

Number of Spans: 2
Skewness: None
West Bound
Span Number 1 2
Span Width 52’-6” 52’-6”
Span Length 87’ 80’-6”


-All Spans WB Midspan: #7@10”+#5@10”(T), #7@9”(B)
-All Spans WB Pier: #8@10” + #7@10” (T), #7@9” (B)
-All Spans WB: #6@7” (T&B)
Girder Information:
West Bound
4 2
1&2 (Midspan) 260034 789 54
1&2 (Support) 260034 789 54

West Bound
1&2 (Midspan) 24 2
1&2 (Support) 12 2

3
3
W/C Ratio: 0.38
Air Content: 5%
Slump: 3.4”
169
Month Of Casting: June
5.Survey results
Crack Information:
West Bound
Span No. 1 2
Spacing 3-5” 3-5”

Figure 110. Cracks on the deck (Bridge No.1143-171).
170
Figure 111. Another marked crack on the deck (Bridge No.1143-171).
171
Bridge Location Route 133 EB Over North Main Street
Year Built 1998

Number of Spans: 1
Skewness: None
East Bound
Span Number 1
Span Width 42’
-East Bound (All Spans): 5 Girders@ 9’-3”

Deck Thickness 9.25” Top Cover 2.5””

-All Spans EB: #5@15”(T), #5@9”(B)
Girder Information:
East Bound
4 2
1 (Midspan) 260034 789 54
1 (Support) 260034 789 54

East Bound
1 (Midspan) 24 2
1 (Support) 12 2

3
3
W/C Ratio: 0.38
Air Content: 6.1 %
Slump: 3.4”
172
5.Survey results
Crack Information:
East Bound
Span No. 1
Type Transverse
Location All over
Spacing 3-5’
Size Typical
Survey Team Member: Ala, Fred, Tom, and Frank

Figure 112. Close up view of deck cracks at the ends (Bridge No.1143-172).
173
Bridge Location Route 133 WB Over North Main Street
Year Built 1998

Number of Spans: 1
Skewness: None
West Bound
Span Number 1
Span Width 42”


-All Spans EB: #5@15”(T), #5@9”(B)
Girder Information:
West Bound
4 2
1 (Midspan) 260034 789 54
1 (Support) 260034 789 54

West Bound
1 (Midspan) 24 2
1 (Support) 12 2

3
3
W/C Ratio: 0.38
Air Content: 6.7%
Compressive Strength:
Slump: 3.8”
174
5.Survey results
Crack Information:
West Bound
Span No. 1
Type Transverse
Location All over
Spacing 3-5’
Size Typical
Survey Team Member: Ala, Frank, Tom, and Frank

175
Bridge Location Route 133 EB Over Wyckoff ‘s Mill Road
Year Built 1998

Number of Spans: 1
Skewness: None
East Bound
Span Number 1
Span Width 42’

Deck Thickness : 8.75” Top Cover 2.5”

- All Spans EB: #5@15”(T), #5@9”(B)
-All Spans EB): #6@7 (T&B)
Girder Information:
West Bound
4 2
1 (Midspan) 260034 789 54
1 (Support) 260034 789 54

West Bound
1 (Midspan) 24 2
1 (Support) 12 2

3
3
W/C Ratio: 0.38
Air Content: 5.1%
Slump: 3.9
176
5.Survey results
Crack Information:
East Bound
Span No. 1
Type Transverse
Location All over
Spacing 3-5’
Size Typical
Survey Team Member: Tom, Ala, Fred, and Frank

177
Bridge Location Route 133 WB Over Wyckoff’s Mill Road
Year Built 1998

Number of Spans: 1
Skewness: None
West Bound
Span Number 1
Span Width 42’


-All Spans WB: #5@15”(T), #5@9”(B)
-All Spans WB: #6@7 (T&B)
Girder Information:
West Bound
4 2
1 (Midspan) 260034 789 54
1 (Support) 260034 789 54

West Bound
1 (Midspan) 24 2
1 (Support) 12 2

3
3
W/C Ratio: 0.38
Air Content: 5.9%
Slump: 3.1”
178
5.Survey results
Crack Information:
West Bound
Span No. 1
Type Transverse
Location All over
Spacing 3-5’
Size Typical

179
Bridge Location Route 133 EB Over One NJ Turnpike
Year Built 1998

Number of Spans: 3
Span type: Mixed (1 Simply supported +2 Continuous)
Skewness: None
East Bound
Span Number 1 2 3
Span Width 41’ 41’ 41’
Span Length 94’ 131’-3” 124’
-East Bound (All Spans): 5 Girders@9’


-All Spans EB Midspan: #5@15(T), #5@9(B)
-EB Span 2&3 Over Pier: #6@7(T&B)
Girder Information:
East Bound
4 2
All (Midspan) 540516 1053 63
All (Support) 540516 1053 63

East Bound
All (Midspan) 24 6
1 (Support) 12 6
2&3 (Support) 6 6

3
3
W/C Ratio: 0.38
Air Content: 5.9%
Slump: 3.4”
180
Admixtures: Retarder & Water Reducer
Air Temperature: 79.5F
Month Of Casting: August
5.Survey results
Crack Information:
East Bound
Span No. 1 2 3
Type Transverse Transverse Transverse
Location All over All over All over
Spacing 3-5’ 3-5’ 3-5’
Size Typical Typical Typical

Figure 113. End diaphragm (Bridge No.1143-176).
181
Figure 114. A view from underneath the bridge (Bridge No.1143-176).
Figure 115. Transverse deck cracks on Bridge No.1143-176 (NJ Turnpike in the
background).
182
Bridge Location Route 133 WB Over NJ Turnpike
Year Built 1998

Number of Spans: 3
Span type: Mixed (2 Continuous + 1 Simply supported )
Skewness: None
West Bound
Span Number 1 2 3
Span Width 41’ 41’ 41’
Span Length 124’ 131’-3” 94’
-West Bound (All Spans): 5 Girders@9’


-All Spans WB Midspan: #5@15(T), #5@9(B)
-WB Span 1&2 Over Pier: #6@7(T&B)
-All Spans WB: #6@7 (T&B)
Girder Information:
West Bound
4 2
All (Midspan) 540516 1053 63
All (Support) 540516 1053 63

West Bound
All (Midspan) 24 6
1&2 (Support) 6 6
3 (Support) 12 6

3
3
W/C Ratio: 0.38
Air Content: 6.2%
Slump: 3.5”
183
Admixtures: Retarder & Water Reducer
Month Of Casting: August
5.Survey results
Crack Information:
West Bound
Span No. 1 2 3
Type Transverse Transverse Transverse
Location All over All over All over
Spacing 3-5’ 3-5’ 3-5’
Size Typical Typical Typical

Figure 116. Transverse (and longitudinal) cracks on deck (Bridge No.1143-177).
184
Bridge Location Whitehead Road Over AMTRAK
Year Built 1999
2. Structural Design Information

Number of Spans: 3
Skewness: None
East & West Bound

Span Number 1 2 3
Span Width 41’-0” 41’-0” 41’-0”
Span Length 103’-0” 175’-3” 103’-0”
-East & West Bound (All Spans): 6 Girders@11’-0”

1/2” 1/2
Deck Thickness 8 Top Cover 2 ”

- All Spans EB & WB Midspan: #5@10”(B), #5@15”(T)
- All Spans EB & WB Pier: #5@10”(B), #[email protected]”(T)
- All Spans EB & WB: #5@7”(T&B)
Girder Information:
East & West Bound
4 2
1&3 (Midspan) 41206 90 50.5
1&3 (Support) 34576~65792 78~126 50~52
2 (Midspan) 41207 90 50.5
2 (Support) 65792 126 52

East & West Bound
1&3 (Midspan) 18” 4
1&3 (Support) 14”~36” 4
2 (Midspan) 18” 4
2 (Support) 36” 4

3
3
W/C Ratio: 0.37
185
Air Content: 5.8%
Slump: 3.4”
:
Concrete Temperature 79.0 F
Month Of Casting: June
5.Survey results
Crack Information:
East & West Bound
Span No. 1 2 3
Type None Transverse/Other Transverse/Other
Location - All Over All Over
Spacing - - -
Size - - -

-Lots of seals are on the deck.
Figure 117. A view of the bridge (Bridge No.1149-168).
186
Bridge Location RT35 Over NAVESINK River
County Monmouth Township Red Bank
Year Built 2000

Number of Spans: 9
Span type: Mixed (3 Continuous + 3 Continuous + 3 Continuous)
Skewness: None
North Bound & South Bound

Span Number All
Span Width 89’-3”
- North Bound & South Bound:-All Spans: 10 Girders @7’-10”

1/2 1/2
Deck Thickness 8 Top Cover 2 ”
1/2
Wearing Surface Concrete, Sawcut Bottom Cover 1

-All Spans Midspan: #5@6”(B), #5@15”(T)
-Continuous Spans Over Pier: #5@6”(B), #5@15”+#8@15(T)
-All Spans EB: #6@6”(T&B)
Girder Information:
North Bound & South Bound
4 2
All 761550 1125 72

Ramp
All (Midspan) 6 6
All (Support) 23 6

3
3
W/C Ratio: 0.38
Air Content: 5.3 %
:
Compressive Strength 6711 Psi
:
Slump 3.1”
Admixtures: Water Reducer & Corrosion inhibitor
187
:
Air Temperature 56 F
:
Concrete Temperature 73 F
:
Month Of Casting March-June
5.Survey results
Crack Information:
Ramp
Span No. All Spans
Type Transverse
Location All over
Spacing 3’-5’
Size Hair Line
Survey Team Member: Tom, Anthony, and Jennifer

-Most of the cracks are on the northbound side, which has been built earlier
188

Final_Report

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Final_Report

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Final_Report

Uploaded by

Copyright:

Available Formats

FHWA-NJ-2002-19

Cause and Control of Transverse Cracking in Concrete

Dr. M Ala Saadeghvaziri Rambod Hadidi

Department of Civil and Environmental Engineering

NJDOT Research Project Manager

“The content of this report reflects the views of the author(s)

7. A ut hor (s) 8. Performing Organi zati on Report No.

15. Supplementary N otes

1 7 . Key W ords 1 8 . Dist ribution Stat ement

Co ncr ete , D ec k, Cr ac ki ng, Cr ac ks, T r a n s ver se,

U ncl as s i fi ed U ncl as s i fi ed 2 0203

DISCLAIMER STATEMENT ...........................................................................................II

FORM DOT 1700.7 .......................................................................................................III

TABLE OF CONTENTS ................................................................................................. V

LIST OF FIGURES...................................................................................................... VIII

LIST OF TABLES ........................................................................................................ XII

LIST OF SYMBOLS .................................................................................................... XIV

APPENDIX A: BRIDGE INFORMATION AND SURVEY RESULTS ..........................126

Structural Design Factors

Mix Design (Material)

• As specified in NJDOT Specs, ensure that curing starts immediately after

Training and Implementation

Future Research Needs

Figure 1. Corrosion due to transverse deck cracks in a bridge in New Jersey.

• Literature review and evaluation of previous research recommendations along

Finally, a simplified method to evaluate stresses in concrete bridge deck due to

Causes of Transverse Deck Cracking

Figure 3. Mechanism of transverse cracking. (a) Concrete is poured. (b) Concrete

Volume change in concrete bridge deck, as influenced by the properties of concrete, is

Material and Mix Design Factors

• 360 kg/m3 (611 lb/m3) (Iwoa DOT, 1986)

• 0.48 (PCA, 1970)

• 50 + 12 mm (2 + ½ in.) (PCA, 1970)

Effect of different types of admixtures on cracking is not completely understood yet.

• Maximum concrete placement temperature of 27oC (80oF) (PCA, 1970)

• Temperature difference of at least 12oC (22oF) for at least 24 hours is

Curing has a pronounced effect on the properties of hardened concrete such as

Pour length and Sequence

Vibration of Fresh Concrete

Sufficient vibration of concrete is essential to good concrete. Issa (1999) considers

Structural Design Factors

Girder Type, Boundary condition, and Spacing

Stud Configuration and Properties

• Minimum of 38 mm. (1.5 in.) (PCA, 1970)

Reinforcement Type, Spacing, Size and Distribution

Reinforcing detail is of paramount importance in controlling cracks in concrete

• Limiting the size of deck reinforcement to No. 5

Vibration and Impact Characteristics

Perfetti et al. (1985) found no relationship between frequency of vibration of

Many studies have considered material factors. However, there is no general

There seems to be more agreement among different researchers on the effect of

A promising approach to reduce deck cracking is the use of shrinkage compensating

Objectives and Scope

Figure 6. Geographical distribution of bridges surveyed.

Table 1. Construction data for bridges surveyed.

• Field survey forms

Field Survey Form

Degree of Skewness (quantify if possible): Longitudinal

Deck Surface: End(s)