USF Libraries
USF Digital Collections

Deterioration process and deck failure mechanism of Florida's precast deck panel bridges

MISSING IMAGE

Material Information

Title:
Deterioration process and deck failure mechanism of Florida's precast deck panel bridges
Physical Description:
Book
Language:
English
Creator:
Gualtero, Ivan A
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
concrete cracks
finite element analysis
forensic study
shear fatigue
differential shrinkage
Dissertations, Academic -- Civil Engineering -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: During the late 70's and early 80's, several precast deck panel bridges were constructed in Florida. These utilize prestressed precast panels as stay-in-place forms and are designed to act compositely with a cast-in-place deck which is poured subsequently. Such bridges offer advantages of quicker construction and lower costs. However, several such bridges built in Florida developed extensive cracking and spalling. Following localized failures, the Florida Department of Transportation has decided to replace all 127 precast panel deck bridges in Districts 1 and 7. Since deck replacement is contingent on funding, it is necessary to develop a rational procedure to decide the order in which they are replaced. This requires a better undertanding of the deterioration process and failure mechanism in such bridge decks.The methodology used in this study was to first analyze in detail 5 cases of sudden localized deck failures to identify the causes of the failures and any common factors in the failed bridges. Also, forensic studies were conducted on eight bridges scheduled for deck replacements during 2003 and 2004. In these studies it was possible to investigate in detail the condition of the deck at different stages of deterioration. Based on the information collected, a deck failure model was developed.
Thesis:
Thesis (M.S.C.E.)--University of South Florida, 2004.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Ivan A. Gualtero.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 173 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001488424
oclc - 56801154
notis - AJT5718
usfldc doi - E14-SFE0000466
usfldc handle - e14.466
System ID:
SFS0025158:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001488424
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 041012s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000466
035
(OCoLC)56801154
9
AJT5718
b SE
SFE0000466
040
FHM
c FHM
090
TA145
1 100
Gualtero, Ivan A.
0 245
Deterioration process and deck failure mechanism of Florida's precast deck panel bridges
h [electronic resource] /
by Ivan A. Gualtero.
260
[Tampa, Fla.] :
University of South Florida,
2004.
502
Thesis (M.S.C.E.)--University of South Florida, 2004.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 173 pages.
520
ABSTRACT: During the late 70's and early 80's, several precast deck panel bridges were constructed in Florida. These utilize prestressed precast panels as stay-in-place forms and are designed to act compositely with a cast-in-place deck which is poured subsequently. Such bridges offer advantages of quicker construction and lower costs. However, several such bridges built in Florida developed extensive cracking and spalling. Following localized failures, the Florida Department of Transportation has decided to replace all 127 precast panel deck bridges in Districts 1 and 7. Since deck replacement is contingent on funding, it is necessary to develop a rational procedure to decide the order in which they are replaced. This requires a better undertanding of the deterioration process and failure mechanism in such bridge decks.The methodology used in this study was to first analyze in detail 5 cases of sudden localized deck failures to identify the causes of the failures and any common factors in the failed bridges. Also, forensic studies were conducted on eight bridges scheduled for deck replacements during 2003 and 2004. In these studies it was possible to investigate in detail the condition of the deck at different stages of deterioration. Based on the information collected, a deck failure model was developed.
590
Adviser: Sen, Rajan.
653
concrete cracks.
finite element analysis.
forensic study.
shear fatigue.
differential shrinkage.
690
Dissertations, Academic
z USF
x Civil Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.466



PAGE 1

Deterioration Process and Deck Failure Mechanism of Florida’s Precast Deck Panel Bridges by Ivan A. Gualtero A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Civil Engineering Department of Civil & Environmental Engineering College of Engineering University of South Florida Major Professor: Rajan Sen, Ph.D. Gray Mullins, Ph.D. Ashraf Ayoub, Ph.D. Date of Approval: September 17, 2004 Keywords: Concrete Cracks, Differential Shrinkage, Shear Fatigue, Forensic Study, Finite Element Analysis. Copyright 2004, Ivan A. Gualtero

PAGE 2

DEDICATION This work is dedicated to my parents and Tere. Tha nk you for all of the support that you have given me in my academic and professio nal pursuits.

PAGE 3

i TABLE OF CONTENTS LIST OF TABLES v LIST OF FIGURES vii ABSTRACT xi CHAPTER 1. INTRODUCTION 1 1.1 Precast Deck Panel System 1 1.2 Deck Panel System Construction Details 2 1.2.1 Types of Precast Panels 2 1.2.2 Types of Panel Bearings 3 1.3 Floridas Deck Panel Bridges 4 1.4 Use of Deck Panel Construction in Other States 6 CHAPTER 2. DECK PANEL BRIDGE PERFORMANCE ON FDOT DISTRICTS 1 & 7 8 2.1 Introduction 8 2.2 Deck Underside Deficiencies 9 2.2.1 Bottom Transverse Cracks 9 2.2.2 Panel Corner Crack 10 2.2.3 Panel Spalls and Delaminations 11 2.3 Deck Top Deficiencies 11 2.3.1 Failed Repairs 11 2.3.2 Deck Top Spalls 12 2.3.3 Deck Top Longitudinal and Transverse Cracking 13 CHAPTER 3. LOCALIZED FAILURES 14 3.1 Introduction 14 3.2 I-75 North Bound Over Bee Ridge Road, Bridge #170146 14 3.2.1 Failure Details 16 3.2.1.1 Newspaper Account 17 3.2.2 Analysis 19 3.2.2.1 Inspection Reports 19

PAGE 4

ii 3.2.3 Environmental Conditions 21 3.2.4 Punching Shear 21 3.2.5 Conclusions 24 3.3 I-75 NB Over Clark Rd Bridge #170086 25 3.3.1 Failure Details 26 3.3.1.1 Newspaper Account 27 3.3.2 Analysis 29 3.3.2.1 Inspection Reports 29 3.3.3 Environmental Conditions 31 3.3.4 Punching Shear 32 3.3.5 Conclusions 33 3.4 I-75 SB Over Clark Rd Bridge #170085 34 3.4.1 Failure Details 34 3.4.1.1 Newspaper Account 35 3.4.2 Analysis 36 3.4.2.1 Inspection Reports 36 3.4.3 Environmental Conditions 39 3.4.4 Punching Shear 40 3.4.5 Conclusions 40 3.5 CrossTown Viaduct over Downtown Tampa, Bridge #1003 32 Span 38 41 3.5.1 Failure Details 43 3.5.1.1 Newspaper Account 45 3.5.2 Analysis 45 3.5.2.1 Inspection Reports 45 3.5.3 Environmental Conditions 48 3.5.4 Punching Shear 49 3.5.5 Conclusions 49 3.6 CrossTown Viaduct over Downtown Tampa, Bridge #1003 32 Span 70 50 3.6.1 Failure Details 50 3.6.1.1 Newspaper Account 52 3.6.2 Analysis 53 3.6.2.1 Inspection Reports 53 3.6.3 Environmental Conditions 57 3.6.4 Punching Shear 58 3.6.5 Conclusions 59 3.7 Summary and Conclusions 60 3.7.1 Failure Trend 60 3.7.2 Environmental Factors 62 3.7.3 Failure Location 62 3.6.4 Bridge Characteristics 63 CHAPTER 4. FORENSIC INVESTIGATION 64 4.1 Introduction 64 4.2 Objectives 65 4.3 I-75 NB and SB over Moccasin Wallow Rd. (Br idges #130079, #130078) 66

PAGE 5

iii 4.3.1 Bridge Details 66 4.3.2 Inspection Method 68 4.3.3 Findings 71 4.3.3.1 No Deck Surface Cracking 73 4.3.3.2 Deck Surface Longitudinal Cracking 74 4.3.3.3 Deck Surface Transverse Cracking 75 4.3.3.4 Additional Longitudinal Cracking 76 4.3.3.5 Deck Spalling and Delamination 77 4.3.3.6 Findings on Panel Bearings 78 4.3.3.7 Findings on Core Examination 78 4.4 I-75 NB over N Toledo Blade Blvd. (Bridge # 170140) 79 4.4.1 Bridge Details 79 4.4.2 Inspection Method 81 4.4.3 Findings 81 4.5 I-75 SB over CSX R/R. (Bridge #130075) 84 4.5.1 Bridge Details 84 4.5.2 Inspection Method 86 4.5.3 Findings 86 4.6 I-75 NB over US 92. (Bridge #100415) 88 4.6.1 Bridge Details 88 4.6.2 Inspection Method 91 4.6.3 Findings 92 4.6.3.1 Deteriorated M1 Repair and Walking Spall s 92 4.6.3.2 Deck Panel Bearing 94 4.7 I-75 NB over Sligh Ave & Ramp D-1 (Bridge # 100398) 95 4.7.1 Bridge Details 95 4.7.2 Inspection Method 97 4.7.3 Findings 98 4.8 I-75 NB over Ramp B-1 (Bridge #100417) 100 4.8.1 Bridge Details 100 4.8.2 Findings 102 4.9 I-75 NB Over SR 64 (Bridge #130085) 103 4.9.1 Bridge Details 104 4.9.2 Findings 106 4.10 Study Summary 108 CHAPTER 5. FAILURE MECHANISM 111 5.1 Introduction 111 5.2 Deck Failure Mechanism Model 111 5.2.1 Stage #1 Initial Condition 111 5.2.2 Stage #2 Longitudinal / Transverse Cracki ng 112 5.2.3 Stage #3 Shear Failure Longitudinal Crack ing 113 5.2.4 Stage #4 First Spall 116 5.2.5 Stage #5 Spall Increase, Then Spall Patch 118 5.2.6 Stage #6 New Spalling Plus Spall Increase 119

PAGE 6

iv 5.2.7 Stage #7 M1 Repair 120 5.2.8 Stage #8 Shear Failure Cracking Adjacent to an M1 Repair 123 5.2.9 Stage #9 Spalling Adjacent to an M1 Repai r 124 5.2.10 Stage #10 Cracking on M1 Repair and Adja cent Spalling Increase 125 5.2.11 Stage #11 Adjacent Spall Patch 126 5.2.12 Stage #12 Additional Adjacent Spalling 1 27 5.2.13 Stage #13 Deck Localized Failure 129 5.3 Summary 130 CHAPTER 6. SUMMARY AND CONCLUSIONS 131 6.1 Summary 131 6.2 Localized Failures 132 6.2.1 Failure Trend 132 6.2.2 Environmental Factors 132 6.2.3 Failure Location 133 6.2.4 Bridge Characteristics 133 6.3 Forensic Investigation 134 6.4 Deterioration Model 135 6.5 Recommendations for Bridge Deck Replacement Prioritization 135 6.6 Future Work 136 REFERENCES 138 APPENDICES 141 APPENDIX A: PUNCHING SHEAR CALCULATIONS 142 APPENDIX B: CORE EVALUATION I-75 OVER MOCCASIN WALL OW RD 152

PAGE 7

v LIST OF TABLES Table 3.1 Localized Deck Failures 14 Table 3.2 Bridge #170146 16 Table.3.3 Excerpts from Inspection Reports (Bridge #170146) 20 Table 3.4 Punching Shear Resistance Bridge #170146 23 Table 3.5 Bridge #170086 25 Table 3.6 Excerpts from Inspection Reports (Bridge #170086) 30 Table 3.7 Punching Shear Resistance Bridge #170086 33 Table 3.8 Bridge #170085 Details 34 Table 3.9 Excerpts from Inspection Reports (Bridge #170085) 38 Table 3.10 Bridge #100332 Details 42 Table 3.11 Excerpts from Inspection Reports (Bridge #100332 Span 38) 46 Table 3.12 Excerpts from Inspection Reports (Bridge #100332 Span 70) 54 Table 3.13 Excerpts from Monthly Inspection Reports (Bridge #100332) 57 Table 3.14 Punching Shear Resistance Bridge #100332 Span 70 59 Table 3.15 Inspection Record 61 Table 3.16 Failure Comparison 62 Table 4.1 Forensic Studies 64 Table 4.2 Bridges #130078 and #130079 67 Table.4.3 Bridge #170140 80 Table 4.4 Bridge #130075 84 Table 4.5 Bridge #100415 89 Table 4.6 Bridge #100398 96 Table 4.7 Bridge #100417 100 Table 4.8 Bridge #130085 104

PAGE 8

vi Table 4.9 Bearing and Deck Condition Summar y 108 Table 4.10 Forensic Study Summary 109 Table 6.1 Failure Summary 133 Table 6.2 Deck Replacement Prioritization Approach 136 Table B.1 Core Details of Deck Section # 1 153 Table B.2 Core Details of Deck Section # 3 155 Table B.3 Core Details of Deck Section # 4 156 Table B.4 Core Details of Deck Section # 5 158

PAGE 9

vii LIST OF FIGURES Figure 1.1 Typical Cross Section View 2 Figure 1.2 Precast Deck Panel. Section A-A Fig. 1.1 3 Figure 1.3 Types of Panel Bearing 3 Figure 1.4 Construction Details of Florida’s Precas t Panel Decks 5 Figure 1.5 Precast Panel Prestressing Strands Confi gurations 6 Figure 2.1 Bottom Transverse Crack 10 Figure 2.2 Panel Corner Crack 10 Figure 2.3 Bottom Panel Spalls and Delaminations 11 Figure 2.4 Failed Repair (Walking Spall) 12 Figure 2.5 Deck Top Spall 13 Figure 2.6 Deck Top Longitudinal and Transverse Cra ck 13 Figure 3.1 Cross Section View of Bridge #170146 – M ain Span 15 Figure 3.2 Composite Deck Section 16 Figure 3.3 Location of Failed Panel, Bridge #170146 (I-75 NB) 18 Figure 3.4 Sarasota Precipitation 22 Figure 3.5 Cross Section View of Bridge #170086 26 Figure 3.6 Location of Failed Panel, Bridge #170086 28 Figure 3.7 View of Failed Panel Bridge#170086 29 Figure 3.8 Sarasota Precipitation (Oct 27 – Nov 27 / 2000) 32 Figure 3.9 Cross Section View of Bridge #170085 35 Figure 3.10 Location of Failed Panel Bridge #17008 5 36 Figure 3.11 Deck Deficiency Six Months Before Failu re, Bridge #170085 38 Figure 3.12 Sarasota Precipitation (Nov 20 – Dec 20 / 2000) 40 Figure 3.13 Cross Section View of Bridge #100332, Span 38 42

PAGE 10

viii Figure 3.14 Localized Deck Failure. Bridge #1003 32, Span 38 43 Figure 3.15 Location of Failed Panel, Bridge #10 0332, Span 38 44 Figure 3.16 Deck Thickness Measurements and Deta ils of Failed Section, Bridge #100332, Span 38 48 Figure 3.17 Tampa International Airport (Sep 02 – Oct 02 2002) 49 Figure 3.18 Cross Section View of Bridge #100332 Span 70 50 Figure 3.19 Localized Deck Failure, Bridge #1003 32, Span 70 51 Figure 3.20 Location of Failed Panel, Bridge #10 0332 Span 70 52 Figure 3.21 Deck Spall, 23 days before failure, Bridge #100332, Span 70 57 Figure 3.22 Tampa International Airport Precipit ation (Aug 6 – Sep 5 2003) 58 Figure 3.23 Simplified Deck Deterioration Proces s 60 Figure 4.1 Cross Section View of Bridge #130078 68 Figure 4.2 Composite Deck Section 68 Figure 4.3 Removed Deck Sections from SB Bridge #13 0079 69 Figure 4.4 Marked Deck Sections Removed From NB Bri dge #130078 70 Figure 4.5 Coring of Marked Sections 70 Figure 4.6 Overview of Findings from Bridge #130079 72 Figure 4.7 No Deck surface Cracking 73 Figure 4.8 Development of Deck Surface Longitudinal Crack 74 Figure 4.9 Deck Surface Transverse Crack 75 Figure 4.10 Additional Longitudinal Cracking 76 Figure 4.11 Development of a Deck Surface Spall 77 Figure 4.12 Crumbled Concrete in Top of a Diag onal Crack. (Core 1-3) 79 Figure 4.13 Cross Section View of Bridge #170140 80 Figure 4.14 View of Bridge #170140 81 Figure 4.15 Findings Overview, Bridge #170140 82 Figure 4.16 Retrieved Panels from Bridge #170140 83 Figure 4.17 Cross Section View of Bridge #130075 85 Figure 4.18 Deck Overview of Bridge #130075 85 Figure 4.19 Inspection Methods Bridge #130075 86 Figure 4.20 Deck Cross Section View over Girder # 3 87

PAGE 11

ix Figure 4.21 Panel Bearing Condition, over Girder # 3 88 Figure 4.22 Cross Section View of Bridge #100415 span 2 90 Figure 4.23 Bridge #100415 Span 2, Prior to Deck Removal 90 Figure 4.24 Cross Section View of Cut Pattern on Bridge #100415 91 Figure 4.25 Examination of a Deteriorated Deck S ection on Bridge #100415 93 Figure 4.26 Panel Bearing Examination on Bridge #100415 94 Figure 4.27 Cross Section View of Bridge #100398 96 Figure 4.28 Deck Overview of Bridge #100398 97 Figure 4.29 Inspection Methods Bridge #100398 97 Figure 4.30 Panel Bearing Examination Bridge #10 0398 98 Figure 4.31 Vertical and Longitudinal Cracks Bri dge #100398 99 Figure 4.32 Findings Overview Bridge #100398 99 Figure 4.33 Cross Section View of Bridge #100417 span 2 101 Figure 4.34 Deck Before Removal Bridge # 100417 (Bays 4-6) 101 Figure 4.35 Panel Bearing Examination Bridge #10 0417 102 Figure 4.36 Panel Bearing Examination Bridge #10 0417 103 Figure 4.37 Findings Overview Bridge #100417 103 Figure 4.38 Cross Section View of Bridge #130085 span 2 105 Figure 4.39 Bridge #130085 Prior to Deck Removal 105 Figure 4.40 Bridge #130085 Original Panel Bearin g Detail 106 Figure 4.41 Bridge #130085 Bearing Detail after Epoxy Repair 107 Figure 4.42 Bridge #130085 Panel Bearing Details 107 Figure 4.43 Surface Longitudinal Crack Bridge #1 30085 108 Figure 5.1 Deterioration Stage #2 112 Figure 5.2 Effect of Vertical Crack Shape in She ar Reduction 113 Figure 5.3 Shear Failures for Different Degrees of Shear Reduction 114 Figure 5.4 Deterioration Stage #3 115 Figure 5.5 Examples of Deterioration Stage #3 11 6 Figure 5.6 Deterioration Stage #4 117 Figure 5.7 Deterioration Stage #5 118 Figure 5.8 Deterioration Stage #6 120

PAGE 12

x Figure 5.9 M1 Repair Procedure (Stage #7) 121 Figure 5.10 Deterioration Stage #7 122 Figure 5.11 Deterioration Stage #8 123 Figure 5.12 Deterioration Stage #9 124 Figure 5.13 Deterioration Stage #10 125 Figure 5.14 Deterioration Stage #11 126 Figure 5.15 Deterioration Stage #12 128 Figure 5.16 Example of Deterioration Stage #12 1 29 Figure 5.17 Deterioration Stage #13 130 Figure 5.18 Failure Summary 130 Figure A.1 Shear Failure Detail (Corner) 143 Figure A.2 Shear Failure Detail (No Composite and Corner) 145 Figure A.3 Shear Failure Detail (Edge) 147 Figure A.4 Shear Failure Detail (No Composite Edge) 150 Figure B.1 I-75NB over Moccasin Wallow Bridge 152

PAGE 13

xi DETERIORATION PROCESS AND DECK FAILURE MECHANISM OF FLORIDA’S PRECAST DECK PANEL BRIDGES Ivan Gualtero ABSTRACT During the late 70’s and early 80’s, several pre cast deck panel bridges were constructed in Florida. These utilize prestre ssed precast panels as stay-in-place forms and are designed to act compositely with a ca st-in-place deck which is poured subsequently. Such bridges offer advantages of qui cker construction and lower costs. However, several such bridges built in Florida deve loped extensive cracking and spalling. Following localized failures, the Florida Departmen t of Transportation have decided to replace all 127 precast panel deck bridges in Distr icts 1 and 7. Since deck replacement is contingent on funding, it is necessary to develop a rational procedure to decide the order in which they are replaced. This requires a better undertanding of the deterioration process and failure mechanism in such bridge decks. The methodology used in this study was to first analyze in detail 5 cases of sudden lo calized deck failures to identify the causes of the failures and any common factors in th e failed bridges. Also forensic studies were conducted on eight bridges scheduled for deck replacements during 2003 and 2004. In these studies it was possible to investigate in detail the condition of the deck at different stages of deterioration. Based on the inf ormation collected, a deck failure model was developed.

PAGE 14

1 CHAPTER 1. INTRODUCTION 1.1 Precast Deck Panel System Precast deck panel systems started as an option to reduce bridge construction cost and time by eliminating most of the field formwork needed, and reducing the amount of cast-in-place concrete to be placed in the deck. Th is system is basically a precast prestressed concrete panel that spans between bridg e girders serving as a support for the cast-in-place topping. When the topping concrete se ts, it acts compositely with the panel in resisting subsequent dead and live load. This deck construction system was first introduced in the early fifties in the Illinois highway system. In the years following, it s use was limited due to questions and uncertainties about its performance, typically gene rated because of the innovative nature of this construction system. In the seventies, Depa rtments of Transportation in several states such as Florida, Texas and Pennsylvania cond ucted extensive research to find answers to these questions [14]. Following encouraging and positive results from di fferent researches at that time, the panel deck construction system was finally acce pted and incorporated into the American Association of State Highway and Transport ation Officials (AASHTO) specification [1]. This, as well as its inherent ec onomy led to wide spread use of this deck construction system in highways including some majo r interstate networks which were built at the time.

PAGE 15

2 1.2 Deck Panel System Construction Details As mentioned earlier, this deck construction metho d consists of a precast prestressed concrete panel that spans between the g irders of the bridge (see Fig. 1.1). Cast-in-Place Concrete Precast Panel Panel Bearing Girder A Girder A Figure 1.1 Typical Cross Section View For this construction system, alternate constructi on details are available; basically different panel details and different types of pane l bearing. In the early stages of the introduction of this construction method, there wer e no standard construction details, so different states used different details. 1.2.1 Types of Precast Panels There are two different types of panels: panel wit h ribs and panel without ribs (flat panel surface). The panels without ribs were used mainly in the initial years of introduction of this system. It was then found that a flat panel surface could lead to bond problems between the cast-in-place concrete and the panel. With the introduction of ribs in the panel, the bond between the panel and the to pping concrete, was substantially improved.

PAGE 16

3 Traffic direction Precast Deck Panel Cast in place Concrete Prestressing Strands Figure 1.2. Precast Deck Panel. Section A-A Fig. 1.1 Another construction detail that changed in precas t panels was the length of the prestressing strand – whether it extended or stoppe d at the end of the panel. The extension is typically 3 inches. The idea of extend ing the strands came in an effort to obtain better control on shrinkage cracks and to im prove the composite action between the vertical face of the panel and the topping conc rete over the bridge girder. Different studies on the strand extensions have shown differe nt results regarding the benefit of doing this [15]. This may be the reason why this co nstruction detail was not used on Florida’s bridges. 1.2.2 Types of Panel Bearings Based on the structural behavior panel bearings ma y be classified as (1) positive panel bearing and (2) negative panel bearing. Negative Panel BearingShear Connector Soft Bearing Material (Fiberboard) Positive Panel Bearing Panel Panel Panel Panel Figure 1.3. Types of Panel Bearing

PAGE 17

4 As shown in Fig. 1.3, in the positive bearing case the panel overhangs a strip of soft bearing material (fiberboard). The overhang is at least 1 in. from the interior vertical face of the strip and the vertical face of the panel. Then the topping concrete is poured with special care over the girders to assure that concrete is placed under the precast panel. After the topping concrete is set, t he panel is not longer supported by the fiberboard strip, but by the concrete underneath th e panel. In the case of negative bearing (Fig. 1.3.), the f iberboard strip has the same width as the part of the panel that is supported by the f lange of the girder. In other words, there is no concrete under the support of panel after the topping concrete is placed. This is called negative bearing because after the topping c oncrete is set the panel is no longer supported by the fiberboard but by the topping conc rete on top of the panel. This type of panel bearing makes deck construction easier. Origi nally it was thought to have no negative effects on the structural deck behavior, b ut recent studies have shown that this is not the case. 1.3 Florida’s Deck Panel Bridges Florida DOT has used has used different constructi on details for deck panel bridges throughout the years, including both positi ve and negative panel bearings. In order to find the exact type of deck panel desi gn used in each bridge, an extensive search was conducted at the FDOT District 1 and 7 maintenance office. In the search over a hundred bridge plans for deck panel s ystems were examined to obtain construction details. It was found that in almost a ll the bridges a full depth cast in place concrete deck instead of the deck panels was shown. Only in a few cases “as built” plans were found that showed deck panel construction deta ils. This is shown in Figs. 1.4 1.5. As shown in Fig. 1.4, the type of bearing used in these bridges is negative bearing with the soft bearing material covering the entire support surface of the panel. The recommended width of the bearing strip ranged from 1in. to 1 in. and the thickness

PAGE 18

5 varied from ” minimum to 1 ” maximum. The materia l used as bearing material was fiberboard ( board composed of wood chips bonded together with r esin ). Also it is noticeable that there no strand extensions in the p anels. 3/16@6" 3 4" Min. 1 1 2" Min. 3/8" Prestressing Strands 6" 6" 2/8@6" 1 4 Min 2 1 2" 1" Varies (7" to 8") Figure 1.4 Construction Details of Florida’s Precast Panel De cks Plan view Isometric of Shear Isometric of Shear Tie Bearing Material. Min. Width 1” Max. Width – 1 ” Min. Thickness ” Max. Thickness 1 ” Cast in Place Slab Precast Panel Turn Hooks in to C Beam L Concrete Girder Steel Girder Welded Shear Connectors A) SHEAR CONNECTOR DETAILS B) DECK CROSSECTION VIEW C) PANEL BEARING DETAILS

PAGE 19

6 Regarding the panels, they usually have ribs. These ribs were 6 in. wide and 1 in. high and were spaced at 6 in. intervals. The panel thick ness was usually 3 in. (including ribs), the panel length 8 ft and the panel length ( span) varied from 5 ft – 9 ft depending on the girder configuration. The prestressing strands were typically located under the panel ribs leaving a minimum clearance of in from the b ottom. The amount and the distribution of the prestressing strands depended o n each deck design. (See Fig. 1.5) Figure 1.5 Precast Panel Prestressing Strands Configurations 1.4 Use of Deck Panel Construction in Other States In a detail research about the use of deck panel construction in other states [5] it was found that Texas is only one state where this c onstruction system is widely used. Almost 85 % of bridges in Texas use panel decks. Al so these bridges have exhibited a performance comparable to full depth cast in place decks. Only longitudinal and transverse cracking have been observed on few occas ions, but never sudden deck failures as in Florida. Tot al # of strands

PAGE 20

7 Texas DOT Deck Panel Design Specifications [16]. (1) Panels at end of spans must have #3 bars extending into CIP portion (2) Panels to be supported at least 1/4 in. above the g irder so that mortar can flow under the panels to provide positive bearing under live loads. (3) Polystyrene foam (Dow PL 300 Glue) used instead of fiberboard, available up to 4 in. thick. (4) Panel overhangs bearing is 1 in. minimum. Texas does prohibit the use of panel decks for cert ain applications: (1) Curved steel girder bridges : Texas DOT’s Bridge Design Engineer prefers to have a monolithic deck on these units because of t he complicated interaction between the deck, the curved girders, and the diap hragms. (2) Bridge widening : Panel decks are not allowed in the bay adjacent t o the existing structure because it is usually not possible to se t the panels properly on the existing structure. It can be used on the other gi rders when the widening involves multiple girders. (3) Phased construction : Panel decks are not often allowed in the bay adja cent to the previously placed deck because it is difficult to install a header form that leaves enough room for the panels to be set properly on t he girders from the earlier stages. (4) Steel girders with narrow flanges: Girders with flanges less than 12 inches wide make panel deck use difficult because the shear st uds conflict with the panels. Standard details allow shear studs to be skewed ac ross the flange width to facilitate the use of panels where sufficient flan ge width is available.

PAGE 21

8 CHAPTER 2. DECK PANEL BRIDGE PERFORMANCE ON FDOT DI STRICTS 1 & 7 2.1 Introduction In Florida many precast deck panel bridges were bui lt during the period from 1980 to 1984, mostly in the Interstate Highway System, spec ifically in FDOT District 1 and 7. Of the 120 odd bridges, about 95% are loca ted on the Interstate (I-75). In most of the bridges constructed in Florida, aft er 2 or 3 years of its construction, the deck started to exhibit unusual longitudinal cr acking on the deck surface. As a result FDOT funded research to determine how this early cr acking would affect service life of the bridge and its maintenance, and to identify met hods that could reduce the deterioration [11]. From these studies the FDOT cam e up with a repair method to improve the structural behavior of the bridges and stop the deterioration. This method consists of removal of the fiberboard bearing and i ts replacement by non shrink epoxy. In theory this method works, but in practice deteriora tion has continued in most of the bridges. This it thought to be due to poor workmans hip on the repairs and the high degree of difficulty required to place epoxy in the narrow space between the edge of the panel and the top of the girder (see Fig. 1.3). In this chapter typical deficiencies found in deck panel bridges will be described in detail. This information was collected from brid ges in FDOT District 1 and 7. It was obtained from official FDOT bridge inspection repor ts.

PAGE 22

9 Some of the typical deficiencies described in his chapter have resulted in sudden localized deck failures. Such failures are describe d in detail in Chapter 3. The typical structural deficiencies found in the p recast deck panel bridges may be divided in two groups: 1. Deck underside (Precast panel) deficiencies: Transverse Cracks Corner cracks Delaminations Spalls 2. Deck top deficiencies: Failed repairs of spalls Spalls Delaminations Transverse and longitudinal cracks.. 2.2 Deck Underside Deficiencies 2.2.1 Bottom Transverse Cracks This is a crack that appears in the bottom of the deck panels, transverse to the traffic direction. The average crack width is about 0.5 mm. It also has been found that this crack tends to run between two strands. This c racking can be found in the midspan, as well as close to the piers. Also it seems more l ikely to happen in steel girder bridges.

PAGE 23

10 Figure 2.1 Bottom Transverse Crack 2.2.2 Panel Corner Crack This crack is not as common as the transverse panel crack, the average crack width is about 1mm, this crack tends to be in a 45 degree an gle and seldom is larger than 2 ft. Figure 2.2 Panel Corner Crack Bridge Steel Girder Transverse Panel Crack Fiber Board Bearing Transverse Panel Joint Panel Transverse Joint Prestressed Concrete Girder Panel Corner Crack

PAGE 24

11 2.2.3 Panel Spalls and Delaminations The occurrence of panel spall is not as common as o ther panel deficiencies. A panel spall can be found anywhere on the panel underside surfac e there is no trend regarding its location. The regular spall size is 3 to 6 in. In case of delaminations, they are located almost always near the panel supports (see Fig. 2.3 ). It has been found that delamination seems to occur more frequently in steel girder brid ges, than concrete girder bridges. The following picture was taken in I-75 over Alafia riv er (Bridge #100358 -59), one of the first bridges to exhibit problems in the deck in FD OT District 1. Figure 2.3. Bottom Panel Spalls and Delaminations 2.3 Deck Top Deficiencies 2.3.1 Failed Repairs Failed repairs are basically caused by the walking spall effect. This is when a deck spall is repaired (removing the adjacent concr ete and placing epoxy or new concrete), and after a few days a new spall appears right next to the repair (see Fig 2.4) This new adjacent spalling also causes deterioratio n of the old repair. Depending on how Concrete Girder Fiberboard bearing replaced by epoxy Panel Underside Panel spall Panel Delamination

PAGE 25

12 this deficiency is treated, it can lead to sudden d eck failures. This is a very common deficiency in Florida’s precast deck panel bridges. Figure 2.4 Failed Repair (Walking Spall) 2.3.2 Deck Top Spalls There are 2 different types of spalls that can be found in the deck surface. The first type is the spall that is related to the deck concrete quality and bridge age, this can occur anywhere on the deck, and since is not relate d with the type of construction it can be found in any concrete bridge deck. The second ty pe of spall is directly related with the deck panel construction. See Fig. 2.5. The typical spall occurs between 2 longitudinal cracks, and under the wheel path. The spall sizes v ary depending the age of the spall, and the trend is to keep growing in the longitudinal di rection if not special repair is done. Concrete Spall Repair Walking Spall Walking Spall

PAGE 26

13 Figure 2.5. Deck Top Spall 2.3.3 Deck Top Longitudinal and Transverse Cracking Longitudinal cracks are very common in precast deck panels; this type of cracking is present in almost 90% of the deck panel bridges in FDOT Districts 1 and 7. It has been found that the longitudinal cracks are always located over the edges of the girders. Transverse cracks are not as common as longitudinal cracks. It has been observed that this type of crack is always located over the transverse panel joints. Figure 2.6. Deck Top Longitudinal and Transverse Cracks Deck Top Traffic direction Longitudinal Longitudinal Cracks over the edges of the girder Transverse Crack over panel joints

PAGE 27

14 CHAPTER 3. LOCALIZED FAILURES 3.1 Introduction Between 2000 and 2003, localized failures occurred in five panel bridges in Districts 1 and 7 (see Table 3.1. This chapter summ arizes relevant information relating to these failures with the intent of identifying under lying trends, if any, for subsequent use in combination with information obtained in Chapter 5 to develop a rational deterioration and failure mechanism of these bridges. Table 3.1 Localized Deck Failures Bridge # District Failure Date Bridge Location 170146 1 2/12/2000 Sarasota, I-75 NB Over Bee Ridge Rd 170086 1 11/27/2000 Sarasota, I-75 NB Over Clark Rd 170085 1 12/20/2000 Sarasota, I-75 SB Over Clark Rd 100332 7 10/02/2002 Tampa, Crosstown Viaduct WB Span 38 100332 7 9/05/2002 Tampa, Crosstown Viaduct WB Span 70 In the following sections descriptions and analyse s of each localized failure are presented in the same order as their listing in Tab le 3.1 in Sections 3.2-3.6. A summary of the principal findings is included in Section 3.7. 3.2 I-75 North Bound Over Bee Ridge Road, Bridge #170146 This 3-span bridge located in Sarasota, FL was bui lt in 1981 and was 19 years old when it failed in February 2000. It has two 36 ft s econdary spans ( span 1, span 3 ) and a 118 ft 8 in. main span ( span 2 ) to make the total bridge length 190 ft 8 in. The shorter

PAGE 28

15 spans were built using two AASHTO Type IV girders o n the outside and five AASHTO Type II girders on the inside all spaced 8 ft 10 in apart. In the main span, fifteen AASHTO Type IV girders are spaced at 4 ft 4 1/4 in. or 4 ft 4 5/16 in. on centers as shown in Fig. 3.1. The deck has a 7 in. thick concrete slab with the precast panel component being either 2- in. or 3- in. (at the rib-section) thic k as shown in Fig. 3.2. This panel thickness is typical for all the deck panel bridges in this area. The specified compressive strength of concrete for the precast panel is 5,000 psi. It is 3,000 psi for the cast in place concrete slab. Additional information regarding dec k panel construction may be found in Chapter 1. Varies 4'-4 1 4 4'-4 1 4 4'-4 5 16 4'-4 1 4 4'-4 5 16 4'-4 5 16 4'-4 5 16 4'-4 5 16 4'-4 1 4 4'-4 5 16 4'-4 1 4 4'-4 1 4 Varies 4'-4 5 16 4'-4 5 16 12' 12' 12' 12' 6' 10' Bay 1Bay 2Bay 3Bay 4 Bay 8 Bay 7 Bay 6 Bay 5 Bay 12 Bay 11 Bay 10 Bay 9 Bay 14 Bay 13 Figure 3.1 Cross Section View of Bridge #170146 – Main Span The bridge has four 12 ft wide lanes, and 6 ft or 10 ft wide shoulders as shown in Fig. 3.1. There is an auxiliary lane that merges wi th traffic entering the interstate from Bee Ridge Road. The average daily traffic (ADT) in the bridge during the year 2000 was 34,000 [27]. Thirty percent of the ADT was truck tr affic (ADTT). Details are summarized in Table 3.2. 18 in. x 24 in. Deck Failure *Girder AASHTO Type IV

PAGE 29

16 Table 3.2 Bridge #170146 Bridge #170146 Characteristics Year Built 1981 Number of Spans 3 Lanes on Structure 4 ADT 34,000 Percent Truck (ADTT) 30% Deck Condition Rating (1999) 6 (Satisfactory) Composite Slab Thickness 7 in. Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type AASHTO Type II and IV National Bridge Invent ory (1999) 3/16@6" 3 4" Min. 1 1 2" Min. 2 1 2" 3/8" Prestressing Strands 6" 6" 2/8@6" 1 4 Min 7" 1" Traffic Direction Figure 3.2 Composite Deck Section 3.2.1 Failure Details Localized failure occurred suddenly in the main sp an on the morning of Saturday, February 12, 2000. A hole formed in a panel that wa s estimated to be about “ two feet square ” [6]. A newspaper account made it 3 square ft 18 in. x 24 in. [25]. However, no photographs of the damage are available. Fig. 3.3 shows the location of the failed panel ta ken from reference [6]. It was redrawn to clarify the details. Failure occurred in “ Span 2, Bay 10 at the edge of panel 13 ” [6]. This location is also identified in Fig.3.1 as coinciding with the placement of the right truck wheel in the slow lane (lane 1) close to the face of a girder.

PAGE 30

17 Ref. 6 also noted the following “ On the deck surface numerous asphalt and concrete type spall repairs had been performed over the years extending south from the hole about six more feet. From that point extending approximately fifteen additional feet, M-1 type repairs have been made. This consisted of asphaltic type material about 18 in. wide…” 3.2.1.1 Newspaper Account In view of the limited information available, news paper accounts of the failures were also reviewed. Two articles were printed in th e local newspaper, Sarasota Herald Tribune [21, 25]. The first article [25] was published on February 13 2000 with the headline “ Fallen asphalt closes lanes: a large pothole has developed again in the I-75 over pass at Bee Ridge .” The newspaper account stated “ No one was injured from the falling debris, but this is the second time in three months that a large pothole has developed in the overpass ”…. FDOT crews last had problems with the overpass afte r a motorist saw a 18 in. hole in the south bound center lane in October ”. No records of this 18 in. hole could be found. A follow-up article [21] was published on February 15, 2000 with the headline “ FDOT will have I-75 hole fixed soon ”. The article stated that “ Workers should be finished patching a hole in the northbound Intersta te 75 overpass at Bee Ridge Road on Wednesday [February 16] according to the Florida Department of Transporta tion”.. FDOT spokesman Marsha Burke stated “ It’s old and is going to require maintenance. It’s something that just happens with older bridges ”

PAGE 31

18 Lane 3 Lane 2 M1 Repair *Not To Scale Panel Support by Timber Panels Replaced with C.I.P concrete Panel Replaced with C.I.P concrete M1 ReparBay 5 End Bent 4 End Bent 1 Pier 3 Pier 2Lane 1 Aux. Lane N Bay 1 Bay 2 Bay 3 Bay 4 Bay 6 Bay 7 Span 3 Span 2 Span 1 36' 118'-8" 66'-9" 10' 12' 12' 12' 12' 6' 1 2 3 4 5Bay 1 Bay 2 Bay 3 Bay 4 Bay 5 Bay 6 Bay 7 Bay 8 Bay 9 Bay 10 Bay Bay 12 Bay 13 Bay 14Bay 1 Bay 2 Bay 3 Bay 4 Bay 5 Bay 6 Bay 7 2 3 4 5 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 36' 18in x 24in Panel Failure (02/12/2000) Figure 3.3 Location of Failed Panel, Bridge 170146 (I-75 NB) [6]

PAGE 32

19 3.2.2 Analysis Analyses were carried out to identify the likely c ause of failure. The starting point of the investigations was a review of inspection re ports and environmental factors. Additionally, a simplified code-based [2] punching shear analysis was carried out to provide a measure of the magnitude of the failure l oad. These are briefly described in Sections 3.2.2.1-3.2.2.3. 3.2.2.1 Inspection Reports To help determine underlying trends, five consecu tive inspection reports covering the period from 1992 to 1999 were reviewed. The fin al inspection in this sequence was carried out on November 24, 1999 less than 3 months before failure occurred on February 12, 2000. Scanned excerpts from the relevant sectio ns of the inspection report are included in Table 3.3. The earliest report (May ’92) notes the presence o f Class 1 (0-1/64th in.) longitudinal cracks along inside girders. The botto m had “ occasional ” transverse cracks with efflorescence that had not changed since Jan 1 984. Mention is made of spalls in span 3 adjacent to a previously patched area and span 2 (right travel lane where the failure occurred). This information is more or less repeate d in the next two reports (Dec ’94 and Dec ’95). In the report prepared in Nov’ 97 dimensi ons of the spall in the right travel lane (6 ft 6 in. x 6 in.) are given. The inspector is al so critical of the use of asphalt (“inappropriate material”) for repair since it is “ respalling around the edges”.

PAGE 33

20 Table 3.3 Excerpts from Inspection Reports (Bridge #170146) FDOT Bridge Inspection Report (Deck) 11/24/99 11/05/97 12/13/95 12/13/94 05/04/92

PAGE 34

21 The final inspection report (Nov ’99) classifies the deck rating as ‘6’ (satisfactory) with a condition state of 3 since th e combined area of distress between 2% to 10% of total deck area. The longitudinal cracks described in all previous reports are mentioned though now there were “ random minor transverse cracks with efflorescence ”. Details of cracking in Bay 4 of Span 3 and transver se cracking in Bay 5 of Span 3 are mentioned. Significantly, no reference is made as to the cond ition of the deck in Span 2, right lane (where failure actually occurred). This had be en identified in the four previous reports from 1992-1997 shown underlined in Table 3. 3. 3.2.3 Environmental Conditions It had been speculated that rainfall can be a cont ributory factor towards failure. Fig. 3.4 shows the distribution of rainfall for Sar asota in the period from Jan 12-Feb 12 2000 [17]. In the week immediately preceding failu re there was no rainfall. However, there was significant (over 1 in.) rainfall 2 weeks earlier on Jan 24. For the record, on the day of the failure, the tem perature varied from a minimum of 55F to a maximum 80F [17]. 3.2.4 Punching Shear An estimate of the punching shear resistance can be obtained using code specified formula [2]. The analysis is approximate since avai lable information is limited, e.g. the exact location of the punching failure in the deck is unknown. Only the panel where failure occurred was shown in the sketch (Fig. 3.3) included in the consultant’s emergency report [6].

PAGE 35

22 Sarasota Precipitation (1/12/2000 to 2/12/2000) 0.00 0.25 0.50 0.75 1.00 1.251 2-J a n 14-Jan 1 6 J an 18-J a n 2 0 -Jan 2 2-J a n 24-Jan 2 6-J a n 28-Jan 3 0 -Jan 1-F e b 3-Fe b 5 Fe b 7-Feb 9 Fe b 11-Febin Figure 3.4 Sarasota Precipitation Two extreme cases are analyzed (1) full composite action and, (2) no composite action. For both cases, the wheel load (rectangular footprint, 10 in. x 20 in. [1]) is positioned at the critical section adjacent to the girder as shown in Fig. 3.1. Full composite action refers to the case where the wheel load is resisted by the entire 7 in. thick concrete slab (Fig. 3.2). This provides an up per bound on the maximum shear resistance. A lower bound on the shear resistance i s provided when due to spalling and subsequent temporary repairs using flexible, asphal t-type material, the entire load is resisted by the precast, prestressed panel. In the analysis, the failure plane is assumed to be unaffected by the differing compressive strength s of the CIP (3000 psi) and precast prestressed panel (5000 psi). Inspection reports indicated that cracking develop ed along both the longitudinal and transverse edges of the panel. The fiberboard b earing does not transfer loads to the girder and therefore, shear resistance was only pro vided by the two uncracked surfaces that extended half the effective depth away, 0.5de, from the wheel for the assumed 450 failure surface. Calculation of the punching shear load for both cases is summarized in Table 3.4. Complete calculations are shown in Appen dix A.

PAGE 36

23 In Table 3.4, b0 signifies the failure perimeter as defined in Ref. 2. For the noncomposite case, the minimum depth of the precast pa nel is used to calculate the effective depth. The calculation shows that the failure load varies from 15.3 kips to 56.3 kips. The former load is smaller than the AASHTO design wheel load without the impact factor. The dramatic reduction in punching shear resistance in the absence of contribution from the cast-in-place slab provides an explanation as t o why failure occurred. Table 3.4 Punching Shear Resistance Bridge # 170146 Load Case Punching Shear Resistance* Full Composite Action r 5 H E D U r $ Y H U D J H Y D O X H V 6 W U D Q G 9 H U W L F D O & U D F N r ) L E H U E R D U G % H D U L Q J T ire Contact area: b=20in l = 10 in CIP de = 4 in. b0 = 34 in. VCIP= 29.8 kips PANEL de = 2.56 in. (ave) VPANEL = 26.5 kips VTOTAL = 56.3 kips No Composite Action !U L E" !S D Q H O" & R Q W D F W D U H D 6 S D O O ) L E H U E R D U G E H D U L Q J !U L E" !S D Q H O" 6 W U D Q G T ire Contact area: bs=20in ls = 10 in CIP VCIP= 0 kips PANEL de = 2.06 in. (min) b0 = 32.1 in. VPANEL = 15.3 kips VTOTAL = 15.3 kips

PAGE 37

24 Table 3.4 (Continued) Effective Depth š 3 U H V W U H V V L Q J 6 W U D Q G V š # 0 L Q š # 7 U D I I L F L U H F W L R Q ASSUMPTIONS 1) Failure plane unaffected by the presence of higher compressive s trength of the precast deck. 2) Fiberboard does not transfer loads. Shear resistance of crac ked transverse and longitudinal panel boundaries are neglected See Appendix A for detailed calculations. 3.2.5 Conclusions Inspection reports indicate that longitudinal refl ective cracks formed along the girder lines but remained dormant for over 10 years (1984-1994). Subsequently, there was more transverse cracking, spalling, repair and failure of re-repair culminating in localized failure. The dormant period suggests that failure may have been due to cumulative shear fatigue. Also, loads in the slower right lane could also have been lower. However, no information on the distribution of truc k traffic over lanes is available. Simplified analysis indicated that regions of the deck where the cast-in-place slab did not resist any load could fail under design loa ds (Table 3.4). A review of the inspection records indicates that barring the final inspection, all four previous inspections had commented on the span where failure eventually occurred. Environmental factors may have played a role. Sustained rainfall could ha ve led to bond degradation between concrete and reinforcement thereby lowering the she ar capacity. Such effect would be limited to the cast-in-place slab.

PAGE 38

25 3.3 I-75 NB Over Clark Rd Bridge #170086 This four span bridge also located in Sarasota was built in 1980 and was 20 years old at the time of failure. It has two 88 ft 3 in. spans ( span 2, span 3 ) and two 32 ft 6 in. secondary spans ( span 1, span 4 ) for a total bridge length of 241 ft 6 in. The sho rter spans use two AASHTO Type IV girders on the outside and f ive AASHTO Type II girders on the inside. These girders were all spaced 8 ft 10 i n. apart as shown in Fig. 3.5. The two longer spans use seven AASHTO Type IV girders also spaced 8 ft 10 in. apart. The composite slab is 7 in. thick. No specific de tails are available. However, they are likely to be similar to that shown in Fig. 3.2. The specified compressive strength of concrete for the precast panel is 5,000 psi and is 3,000 psi for the cast in place concrete slab. More details regarding deck panel constructio n may be found in Chapter 1. The bridge has three 12 ft lanes, and two 10 ft wi de shoulders as shown in Fig. 3.5. The average daily traffic (ADT) in the bridge during the year 2000 was 34,000 [27]. Thirty percent of the ADT was truck traffic. Detail s are summarized in Table 3.5. These are identical to that for the previous bridge. Table 3.5 Bridge #170086 Bridge #170086 Characteristics Year Built 1980 Number of Spans 4 Lanes on Structure 3 ADT [3.1] 34,000 ADTT [3.1] 30% Deck Condition Rating (2000) 7 (Good) Composite Slab Thickness 7 in Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type AASHTO Type II and IV

PAGE 39

26 3.3.1 Failure Details Localized punching shear occurred late morning, Mo nday November 27, 2000. According to the consultant’s emergency response re port [7], failure occurred in span 4 (secondary span), bay 6, on the right lane where a 60 in. by 36 in. gaping hole developed near end bent 5 (Fig. 3.6). The report stated “ Half of the end panel adjacent to the expansion joint had been replaced at some previous time. This hole was the result of the failure of the remaining half of that panel ”. A photograph of the failed bridge panel obtained f rom the Sarasota Herald [4] is shown in Fig. 3.7. The entire concrete in the faile d corner region was missing and debris can be seen lying on the road below. Some of the re inforcement had deformed plastically though none appear to be broken. However, the prest ressing strands were ruptured. The location of the failed panel in span 4 is identifie d in the sketch provided in the consultant’s report. As before, it has been re-draw n for clarity. This location is also identified in Fig. 3.5 as coinciding with the place ment of the right truck wheel in the slow lane (Lane 1) close to the face of a girder. Figure 3.5 Cross Section View of Bridge #170086 Bridge # 170186 Secondary Span 8'-10" 8'-10" 2'-9" 8'-10" 9'-10 1 2 2'-9" 8'-10" 8'-10" 8'-10" 9'-10 1 2 12' 12' 12' Panel Failure (Right Lane)

PAGE 40

27 3.3.1.1 Newspaper Account Two articles related to the failure were reported in the local newspaper, Sarasota Herald Tribune [4, 22]. The first article [4] published on November 28, 20 00 with the headline “ Hole opens up in bridge on I-75 at State Road 72 ”. It noted that the hole that opened up was within “ a week after a state crew made repairs on the same spot ”. The Florida Highway Patrol reported that there were “ no injuries or vehicle damage… ”.

PAGE 41

28 88'-3" 32'-6" 88'-3" 32' -6" 5 1 3 2 2 3 4 56 7 89 1011 1 11 10 9 8 7 6 5 4 3 2 1 2 3 4 5 Panel 1 Bay 6 Bay 5Bay 4 Bay 3Bay 2Bay 1*Not To Scale 60 in x 36 in Panel Failure 11/27/00 Bay Replaced 11-29-00 M1 RepairPoor Condition M1 Repair Poor Condition M1 Repair Good Condition 10' 10' 53' Lane 3 Lane 2 Lane 1 12' 12' 12' Span 4 Span 3 End Bent 5 Pier 4 Pier 3N Span 2 Pier 2 End Bent 1Span 1 Figure 3.6 Location of Failed Panel, Bridge 170086 [7]

PAGE 42

29 The second article published the following day [22 ] made the following observation “ The DOT offers assurances of daily checks and close inspections every 45 days, but those haven’t predicted these failures. M ore effort and funds are needed immediately to make these bridges safe as soon as p ossible – before lives are lost. If protecting public safety requires shifting prioriti es or obtaining emergency funding, so be it. ” Figure 3.7 View of Failed Panel Bridge #170086 (Courtesy Sarasota Herald) [4] 3.3.2 Analysis 3.3.2.1 Inspection Reports Table 3.6 contains relevant scanned excerpts from the last five inspection reports over the period Jan ’93 to May ’00. The last report (May ’00) refers to the deck condition about six months prior to failure on Nov 27 ’00. Failed Panel Replaced panel segment

PAGE 43

30 The first three reports over the period Jan 93 to Jun 96 are quite similar. Longitudinal cracks formed first along the girder lines followed by occasional transverse cracks at panel joints. As for the previous bridge (Table 3.3), the inspectors found no significant change over the 11 year period from May 85 to Jun 96. Table 3.6 Excerpts from Inspection Reports (Bridge #170086) FDOT Bridge Inspection Report (Deck) 05/08/00 05/04/98 06/19/96 08/24/94

PAGE 44

31 Table 3.6 (Continued) 01/04/93 However, significant deterioration was observed in the next inspection carried out in May ’98. Instead of “ occasional ” cracks reported earlier, longitudinal and transve rse cracks had developed “ throughout ”. There was also severe cracking of repairs and sp alls around the edge of the repair. The cracks were as w ide as 1/8 in. (class 5). Mention is also made of deck repair over a large region about 26 ft x 4 ft at abutment 5. The description is not clear to tie it to eventual fail ure (see Fig. 3.7). In the final report (May ’00), top deck cracking i s described as “minor”. This suggests that deficiencies identified earlier had b een repaired. A small spall (4 in. x 4 in. x 0.2 in.) is mentioned as occurring at the “ center of the west lane, 3m (10 ft) from the abutment 5 joint ”. The actual failure occurred at approximately the same location but in the east lane 3.3.3 Environmental Conditions Fig. 3.8 shows the distribution of rainfall for Sa rasota in the period from Oct 27Nov 27 2000 [17]. In the week immediately precedin g failure there was about 0.68 in. of rain. It rained on 24th and 25th just 2 days before failure occurred. In this insta nce, rainfall may have been a factor. For the record, on the day of the failure, the temperature varied from a minimum of 53F to a maximum 72F.

PAGE 45

32 Sarasota Precipitation (10/27/2000 to 11/27/2000) 0.00 0.10 0.20 0.30 0.40 0.502 7Oc t 2 9-Oct 3 1Oc t 2 -No v 4 N o v 6 -No v 8 N o v 10-Nov 1 2No v 14-Nov 1 6No v 18-Nov 2 0-N o v 22 Nov 2 4-N o v 26 No vin Figure 3.8 Sarasota Precipitation (Oct 27 – Nov 27 / 2000) 3.3.4 Punching Shear According to the consultant’s report cited earlier, “ half of the end panel adjacent to the expansion joint had been replaced at some pr evious time. This hole was the result of the failure of the remaining half of that panel ” [7]. The panel section that was replaced is marked in Fig. 3.7. Assuming that no shear transfer was possible at t he joint between the old and new panel, and reflective transverse cracking on th e other side of the panel, the resistance of the slab is by one-way, not two-way shear. This “beam shear” type resistance is given by 2 d b fw c '. Table 3.6 shows an estimate of the shear resistan ce taking bw as 36 in. (the estimated unfailed length of a panel) with an avera ge effective depth d of 2.56 in. Only the case where there is no composite action is cons idered since it gives lower loads. The calculated resistance is 13 kips smaller than the d esign load. The extent of the failed region is believed to be much greater in this failu re because of the joint between the old and new panel (see Fig. 3.7, Table 3.7). Deck Failure

PAGE 46

33 Table 3.7 Punching Shear Resistance Bridge # 170086 Load Case Shear Resistance Girder Top Girder TopPrecast PanelPrecast Panel Precast Panel Precast Panel Section of the panel previously replaced 10" 1'-8" End of the bridge Wheel Contact area 8' Construction Joint Panel Joint 2' 3' Plan View bw= 36in d = 2.56in (ave) f’c = 5000 psi VPANEL = 2 f’cbwd VPANEL =13 kips VTOTAL = 13 kips 3.3.5 Conclusions A number of factors were responsible for this unus ual failure. The most important of these was the joint between a panel segment – re paired and old adjacent to an expansion joint (Fig. 3.7). In addition, there was heavy rainfall prior to failure that may have been a contributory factor by degrading the bo nd between concrete and steel. Unfortunately, there are too many unknowns to arriv e at any definite conclusion. The last inspection report six months prior to fai lure, mentions a spall close to the eventual failure location excepting that the west r ather than the east lane was mentioned. It also noted damage to repaired areas in the form of cracking and spalling. The newspaper account stated that failure occurred at t he same spot where temporary repairs had been carried out a week earlier. The shear fail ure load (Table 3.7) indicates that the deck could fail under design loads for this conditi on.

PAGE 47

34 3.4 I-75 SB Over Clark Rd Bridge #170085 This 4-span bridge is identical to the one describ ed on Section 3.3 and was also constructed the same year. A cross-section view is given in Fig. 3.9 while Table 3.8 provides a summary of relevant bridge details (this is identical to Table 3.5). 3.4.1 Failure Details Localized failure occurred early morning on Wednes day December 20, 2000. According to the emergency response report [8] fail ure occurred in the first panel, bay 2 in span 3 adjacent to bent 3. The hole that punche d right through the panel was estimated to be about 18 in. x 18 in. Fig. 3.10 shows the loc ation of the failed panel. This is taken from reference [8] but was re-drawn for clarity. No photos of the localized failure are available. From the cross section view Fig. 3.9 it can be see n that failure again occurred in the right lane close to the panel support (girder f ace). Table 3.8 Bridge #170085 Details Bridge #170085 Characteristics Year Built 1980 Number of Spans 4 Lanes on Structure 3 ADT (2000)* 34,000 Percent Truck ADTT 30% Deck Condition Rating (2000)* 7 (Good) Composite Slab Thickness 7 in Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type AASHTO Type II and IV From National B ridge Inventory (2000)

PAGE 48

35 8'-10" Bridge # 170085 Main Span 8'-10" 9'-10 1 2 2'-9" 8'-10" 8'-10" 8'-10" 8'-10" 9'-10 1 2 2'-9" Figure 3.9 Cross Section View of Bridge #170085 3.4.1.1 Newspaper Account Three articles regarding the failure were reported in the local newspaper, Sarasota Herald Tribune [23,24,26]. Of these only the first and last had r elevant information. The first article published on December 21, 2000 [ 26] stated that the hole was discovered at 7 am and that no one was injured. The reported size of the hole is 3 ft x 5 ft – same as in the previous bridge – possibly a mista ke. The reporter quotes FDOT spokesman Gene O’Dell who said “ We just had our consultant inspect the Clark Road bridge two weeks ago and they said it was fine ”. The last article published on December 23, 2000 [23] stated that the damage had been repai red and the bridge was opened to traffic. Mention was also made that a consultant wa s inspecting the bridge decks every 45 days and FDOT employees check them out once a month to “ see if there are any bad cracks, anything that will create a hole ” (O’Dell’s quote). 18 in x 18 in Panel Failure (Right Lane)

PAGE 49

36 53' 32'-6" 30' Asphalt Patch Along Center of Center Lane Numerous ConsecutiveM1 Repairs Asphalt Patch (To fail on 12/20/00) 1 Pier 31 Bay 5 2 3 4 Span 4End Bent 5 Pier 4Span 3 N 12' 12' 12' Lane 1 Lane 2Lane 3 10' 10' Bay 1 Bay 2 Bay 3 Bay 4 Bay 6 Panel 5 121110 98 7 65 4 3 2 88'-3" Figure 3.10 Location of Failed Panel Bridge #170085 [8] 3.4.2 Analysis 3.4.2.1 Inspection Reports The five inspections preceding the localized fail ure were carried out on the same dates as the previous bridge (Table 3.6) in Jan ’93 Aug ’94, Jun ’96, May ’98 and May ’00. The last inspection was completed about 7 mont hs prior to the localized failure that occurred on Dec 21 ’00. Scanned excerpts from the c omplete reports are summarized in Table 3.9.

PAGE 50

37 The first two reports over the period Jan ’93 to A ug ’94 are quite similar to that for the previous bridge. Longitudinal cracks occurr ed first with occasional transverse cracks at panel joints. The inspectors state that t he cracks first noted in the report dated May 1985 “ appear to show no change ”. The next inspection carried out in Jun ’96 reporte d more deterioration. All four spans contained longitudinal cracks along the beam lines with transverse cracking at the panel joint. Span 3 (where failure eventually occur red) had developed three spalls in the left lane ranging from 10 in. x 6 in. x 0.4 in. to 30 in. x 6 in. x 1 in. A fairly large 6 ft 6 in. x 6 in. x 1 in. spall had also developed in t he center lane. In addition, patched areas in the left lane had cracked. This was expected to spall in the future. Aside from longitudinal and transverse cracking in all spans, the inspection carried out in May ’98 mentions that damage reporte d previously in span 3 had been repaired. However, cracks (up to 1/16 in.) and dela mination had occurred in the repairs along the “ west edge pavement stripe ”. A delamination area 20 in. x 12 in. surrounding an asphalt patch at midspan in span 4 in the same r egion (west edge pavement stripe) had formed. In the final report (May ’00), top decking crackin g is described as “minor”. The delamination in the middle of span 4 reported in th e previous report had not grown in size. Fig. 3.11, scanned from the photo addendum of this inspection report, shows “c oncrete and asphalt patches throughout spall span 3 1m x 50mm with exposed steel ”. The deficiency shown here happens to be at the exac t location where failure occurred six months later. The deck condition rating of was give n as 7, and the condition state of the bridge was reported as 2. None of the reports descr ibe the underside of the deck. This suggests there was no cracking or efflorescence.

PAGE 51

38 Figure 3.11 Deck Deficiency Six Months Before Failure, Bridge #170085 Table 3.9 Excerpts from Inspection Reports (Bridge #170085) FDOT Bridge Inspection Report (Deck) 05/08/99 05/04/98

PAGE 52

39 Table 3.9 (Continued) FDOT Bridge Inspection Report (Deck) 06/19/96 08/24/94 01/ 04/93 From the cross section view Fig. 3.9 we can see that the failure occurred again in the right lane and close to the panel support girder face. 3.4.3 Environmental Conditions Fig. 3.12 shows the distribution of rainfall for Sarasota in the period from Nov 20Dec 20 2000 [17]. In the ten days immediately preceding failure it rained on six occasions. It rained 0.05 in. the day before failure occurred. In this instance, rainfall may have been a factor. For the record, on the day of the failure, the temperature varied from a minimum of 38F to a maximum 58F.

PAGE 53

40 Sarasota Precipitation (11/20/2000 to 12/20/2000) 0.00 0.10 0.20 0.30 0.40 0.5020N ov 2 2 Nov 24-Nov 2 6 N o v 28 Nov 3 0 N o v 2-D e c 4-Dec 6-D e c 8-Dec 1 0 D e c 12-Dec 1 4 D e c 16 Dec 18D e c 20 Decin Figure. 3.12 Sarasota Precipitation (Nov 20 – Dec 20 / 2000) 3.4.4 Punching Shear As the geometry and the material properties in the deck were identical to that in the previous bridges, the calculated punching shear failure load is also identical. The lower bound for the failure load is calculated to b e 15.3 kips which is smaller than the design wheel load. See Table 3.4 for details. 3.4.5 Conclusions The failure in this bridge was very similar to tha t in the first bridge (Section 3.2). Shear fatigue may have been responsible for failure The failure load was estimated to be 15.3 kips (Table 3.4). The last inspection report s tated that repairs had started to crack. As all three bridges failed in the same region, fau lty construction was undoubtedly a factor though precise faults cannot be pinpointed a t this time. Deck Failure

PAGE 54

41 3.5 CrossTown Viaduct over Downtown Tampa, Bridge #100332 Span 38 This 9,600 ft bridge is the longest deck panel bri dge in the area. It has a total of 91 spans, of which 24 were built in 1975 using a fulldepth cast in place concrete slab. The remaining 67 spans were built in 1980 using precast deck panels. Two of 67 spans used steel girders (average span 170 ft) while the rest used prestress girders (average span 80 ft). This was one of the first deck panel bridges b uilt on a main highway in District 7. Span 38, where the failure occurred, was built us ing prestressed concrete girders. Its span length was 47 ft. The bridge section was 2 2 year old at the time of the failure. The composite slab was 7 in. thick with the precas t panel thickness varying between 2- in. or 3- in. (at the rib-section) as shown in Fig. 3.2. This panel thickness is typical for all deck panel bridges in this area. The specified compressive strength of concrete used for the precast panel is 5,000 psi. I t is 3,000 psi for the cast in place concrete slab. More details regarding deck panel co nstruction may be found in Chapter 1. The bridge has two 12 ft lanes. The right shoulder is 8 ft wide and the left shoulder is only 4 ft wide, as shown in Fig. 3.13. The average daily traffic (ADT) during 2002 was 23,000 [27]. Eight percent of the ADT was truck traffic (Table 3.10).

PAGE 55

42 Figure 3.13 Cross Section View of Bridge# 100332, Span 38 Table 3.10 Bridge #100332 Details Bridge #100332 Characteristics Year Built 1975 (spans 1 – 24) 1980 (spans 25 – 91) Number of spans 91 Lanes on Structure 2 ADT (2002) [17] 23,000 Percent Truck ADTT [17] 8% Deck Condition Rating Span 38 (2001) 5 (Fair) Deck Condition Rating Span 70 (2003) 5 (Fair) Composite Slab Thickness 7 in. Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type (Span 38) AASHTO Type IV 5 ft 3 in by 2ft 6 in Deck Failure

PAGE 56

43 3.5.1 Failure Details This failure was first noticed early morning on We dnesday, October 2 2002. It was located on the right lane close to mid-span. A gaping 5 ft 3 in. by 2 ft 6 in. hole formed (see Fig. 3.14). The same figure shows photo s of the failed region and its underside two days prior to failure. Staining of th e underside is visible. The concrete and repair material separated from the reinforcement wh ich did not rupture. Fig. 3.15 provides a sketch showing the failure location on t he deck. Figure 3.14 Localized Deck Failure. Bridge #100332, Span 38 2 Days Before Failure Deck Failure Top View Underside View 2 Days Before Failure Deck Failure

PAGE 57

44 This damage was repaired by demolishing the whole bay where the failure occurred and placing a new deck using full depth ca st in place concrete. 4' 12' 12' 8' 38'-9" Lane 2 Lane 1 53' *Not To ScalePier 38Span 38 Pier 37 Bay 3Bay 4 Bay 2 Bay 1 N 72'-3 1 16 Precast Panel 5 ft 3 in by 2ft 6 in Deck Failure Figure 3.15 Location of Failed Panel, Bridge #100332 Span 38

PAGE 58

45 3.5.1.1 Newspaper Account No account of the failure was published in the loc al newspaper. 3.5.2 Analysis 3.5.2.1 Inspection Reports The five inspections preceding the localized failu re were carried out over eight years in May ’93, May ’95, Aug. ’97, Aug. ’99 and A ug. '01. The reports provide information on the entire bridge and for this reaso n there is minimal information relating to span 38 where the failure occurred. In the last biennual inspection completed in Aug '01 approximately 14 months before the failure occu rred on October 2 2002, the deck was given a condition rating of 5 (Fair) and a conditio n state of 2. No significant deficiencies relating to span 38 were documented. Scanned excerp ts from these inspection reports are summarized in Table 3.11 for completeness. Because of widespread deterioration of the bridge it was continuously monitoring by FDOT. Information from these monthly inspections provide invaluable information on the progression of degradation leading to failure. Deterioration of the section that eventually faile d was first reported on July 31 2002 as 30 in. x 20 in. concrete delamination ". This was determined on the basis of a "hammer test" in which the suspected region is hit with a hammer and a hollow sound detected. By August 19 2002 the delamination had ch anged to a 48 in. by 10 in. spall. This spall was temporarily patched at that time. At the next inspection on September 30 2002, the patch was found to have failed. In additi on, the extent of the spall had increased to 48 in. by 30 in. by 1.5 in deep (See Fig. 3.14). Temporary repairs were again carried out and the patch repaired. Two days later, this ne w patch failed and a 48 in. by 30 in. gaping hole developed at the site as shown in Fig. 3.14.

PAGE 59

46 A USF research team visited the bridge one day after failure. Measurements taken at the site and from retrieved debris indicated that the deck was thinner than its nominal thickness. It was found to be 6-3/8 in. not 7 in. as specified in the plans stamped "as built" (see Fig. 3.16). Table 3.11 Excerpts from Inspection Reports (Bridge #100332 Span 38) FDOT Bridge Inspection Report (Deck) 08/29/01

PAGE 60

47 Table 3.11 (Continued) FDOT Bridge Inspection Report (Deck) 08/31/99 08/26/97 05/15/95

PAGE 61

48 Table 3.11 (Continued) FDOT Bridge Inspection Report (Deck) 05/20/93 Figure 3.16 Deck Thickness Measurements and Details of Failed Section. Bridge #100332, Span 38 3.5.3 Environmental Conditions Precipitation readings at Tampa International Airport (6 miles from the bridge) for a period of one month before the localized deck failure are shown in Fig. 3.17. It may seen that that there was continuous rain over five days with a rainfall of 0.55 in. one week before the failure. However, no rain occurred 4 days before failure, For the record, on the day of the failure, the temperature varied from a minimum of 75F to a maximum 88F.

PAGE 62

49 Tampa International Airport Precipitation (9/12/2002 to 10/12/2002) 0.00 0.25 0.50 0.75 1.00 1.25 1.502 Sep 4-Sep 6 Sep 8-Sep 1 0 -Sep 12-S e p 14-Sep 1 6-S e p 18-Sep 2 0-S e p 22-Sep 2 4-S e p 26-Sep 2 8 -Sep 30-S e p 2 Octin Figure. 3.17 Tampa International Airport (Sep 02 – Oct 02 2002) 3.5.4 Punching Shear Although the deck was found to be thinner than its nominal value (see Fig. 3.15), the panel thickness was the same. In view of this, the lowerbound value of punching shear would still be the same 15.3 kips. For deta ils see Table 3.4. 3.5.5 Conclusions The biennual inspection data just provides a snaps hot on the condition of the bridge and is therefore not always very useful. Con tinuous monitoring data indicated that delaminations led to large spalls. If flexible mate rials are used for repairs, they are unable to transfer wheel loads to the adjoining slab becau se of their low stiffness and localized failure can occurr at loads below the design load ( Table 3.4). Measurements indicated that the thickness of the deck could be smaller tha n nominal dimensions at specified locations. Rainfall could have been a contributory factor in this case. Deck Failure

PAGE 63

50 3.6 CrossTown Viaduct over Downtown Tampa. Bridge #100332 Span 70 This deck failure case occurred in the same bridge described in Section 3.5 but in span 70, (see Table 3.10 for general details). Span 70 is 65.5 ft long, and is built using type III AASHTO prestressed concrete girders. The g irders are spaced center to center 6 ft 5 in. 4' 6' Full depth deck replacement Localized Deck Failure Type III Girder Bridge # 100332 Span 70 Figure 3.18 Cross Section View of Bridge #100332, Span 70 3.6.1 Failure Details Localized punching shear occurred in early morning Friday September 5 2003. The failure was located close to the midspan and in the right lane. The failure region measured by the USF research team was estimated to be about 2 ft by 3 ft. Photos of the failed section are shown in Fig. 3.1 9. A sketch showing the location of the failure in the deck is shownin Fig. 3.20. In itial spalling ahead of an M1 repair extended into the repair itself. Under subesequent loading, rebars were exposed in the spalled region. The concrete ultimately separated f rom the steel due to the impact of repeated wheel loads and a void formed. December 2002 Septemer 2003

PAGE 64

51 Fig. 3.19 has three photos. The main photo is a cl ose-up plan view of the damage from the top of the deck. Note that the rebars are not broken nor plastically deformed. Small sections of concrete just separated from the reinforcement. One of the prestressing strands can be seen to be intact. A second photo pr ovides an overview of the deck. The third photo shows the extent of the opening in the deck from the underside. Water staining is clearly visible. This failure was repai red by demolishing the whole bay where the failure occurred, and the one adjacent in the l eft lane, and placing a new deck using full depth cast in place concrete. Figure 3.19 Localized Deck Failure, Bridge #100332, Span 70

PAGE 65

52 3.6.1.1 Newspaper Account Two articles regarding the failure were reported in the local newspaper, Tampa Tribune [19,20]. The first article [19] published on September 9 2003 with the headline Small Hole Paves Commuters' Way To A Traffic Jam , makes reference to the large delays users are facing due to the deck failure. It also offered an explanation as to why the hole developed Floridas endless down pours opened a small hole in a bridge on the Lee Roy Selmon Expressway on Friday, creating a huge mess for morning rush hour commuters that wont improve until Sunday . Precast Panel Varies 42' Pier 69 8' Pier 70 M1 Repair Span 70 N Full Depth Deck Replacemen 65' *Not To Scale Spall M1 Repair Deteriorated M1 Repair 2 ft by 3ft Deck Failure 9/5/2003 Lane 1 Lane 2 12' 4' Figure 3.20 Location of Failed Panel, Bridge #100332 Span 70

PAGE 66

53 The second article [20] published on September 10, 2003 with the headline “ Time Catches Up With Expressway ”. It stated that “The 3-square-foot hole was the site of an earlier temporary patch ”. Pat McCue, executive director of the local expre ssway authority was quoted as saying “ Truck traffic caused the layers to separate and cra ck in spots. Rainwater seeped into the cracks and, forced outward by the weight of traffic, crumbled the concrete, leaving a gaping hole ”. Ben Muns, the expressway authority's chief engineer was quoted as saying T here's just no telling when the next one [hole] wil l be" 3.6.2 Analysis 3.6.2.1 Inspection Reports The same five inspections reviewed for the previou s failure in Span 38 describe the condition of the bridge over the eight year per iod from May ’93 to Aug. '01. As mentioned earlier, the reports provide information on the entire bridge and there is limited information relating to span 70 where failu re occurred. Scanned excerpts from these inspection reports are summarized in Table 3. 12 for completeness. Additional monthy inspections (Table 3.13) noted t hat deterioration of the section that eventually failed was first observed on August 12 2003. It was described as a new 2' x 1' x 1" spall and delamination area with exposed steel. This had not been observed in the previous inspection carried out a month early o n July 10. Fig. 3.21 provides a photographic record of the ev ents leading to failure. The first photo, A shows a 3 ft x 1 ft x 1.5 in spall that w as observed 14 months prior to failure. The second photo, B, shows M1 repair carried out 8 months prior to failure. The last photo, C shows a spall developing ahead of the M1 r epair taken 23 days before failure on August 12, 2001. The next picture in the sequence can be seen in Fig. 3.19 where the failed section can be seen.

PAGE 67

54 Table 3.12 Excerpts from Inspection Reports (Bridge #100332 Span 70) FDOT Bridge Inspection Report (Deck) 8/29/03 08/29/01

PAGE 68

55 Table 3.12 (Continued) FDOT Bridge Inspection Report (Deck) 08/31/99 08/26/97 05/15/9 5

PAGE 69

56 Table 3.12 (Continued) FDOT Bridge Inspection Report (Deck) 05/20/93 B A Bridge # 100332 Date: 7/31/02 Span 70, lane 1 Spall, 3 ft x 12 in 1.5 in, with exposed steel. INCREASE.

PAGE 70

57 Figure 3.21 Deck Spall Bridge # 100332, Span 70. A) Initial spall 14 months before failure B) M1 repair over initial spall 8 months before failure, C) Spall next to the M1 repair, 23 days before failure Table 3.13 Excerpts from Monthly Inspection Reports (Bridge #100332) 8/12/03 7/10/03 3.6.3 Environmental Conditions The precipitation readings at Tampa International Airport (6 miles from the bridge) over a one month period prior to failure are shown in Fig. 3.22. Total rainfall one week before failure was about 1.1 inches. Two days before failure, rainfall of 0.8 in. was registered, 0.3 in. rain fell on the day of the failure. Thus, rain may have been a factor in C No reference to the spot that failed

PAGE 71

58 degrading the concrete reinforcement bond that led to concrete pieces separating from the steel and creating a void in the deck. For the record, on the day of the failure, the temperature varied from a minimum of 74F to a maximum 79F. 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.506A ug 8A ug 1 0 -A u g 1 2 -A u g 1 4 -A u g 1 6 -A u g 1 8 -Aug 20-Aug 22-Aug 24-Aug 26-Aug 28-Aug 3 0-Aug 1 Se p 3S ep 5S epin Figure 3.22 Tampa International Airport Precipitation (Aug 6 Sep 5 2003) 3.6.4 Punching Shear Analysis The localized failure occurred within the panel. Consequently, two-way shear resistance was provide by three edges Table 3.14 summarizes the calculated punching shear values for the two extreme cases full composite and panel slab only. It may be seen that the value of the failure load is higher in this case (21.7 kips vs 15.3 kips, Table 3.4). A photograph of the underside of the panel shows water damage and longitudinal cracking within the assumed region providing resistance. Thus, assumption of support from three surfaces is perhaps on the optimistic side in this situation. Deck Failure Tampa International Airport Precipitation (8/6/2003 to 9/5/2003)

PAGE 72

59 Table 3.14 Punching Shear Resistance Bridge # 100332 Span 7 0 Load Case (Panel Edge) Punching Shear Resistance* Full Composite Action r 5 H E D U r $ Y H U D J H Y D O X H V 6 W U D Q G 9 H U W L F D O & U D F N r ) L E H U E R D U G % H D U L Q J T ire Contact area: b=20in l = 10 in CIP de = 4 in. b0 = 58 in. VCIP= 50.8 kips PANEL de = 2.56 in. (ave) VPANEL = 41.4 kips VTOTAL = 92.2 kips No Composite Action !U L E" !S D Q H O" !U L E" !S D Q H O" ) L E H U E R D U G E H D U L Q J 6 W U D Q G & R Q W D F W D U H D T ire Contact area: b=20in l = 10 in CIP VCIP= 0 kips PANEL de = 2.06 in. (min) b0 = 54.12 in. VPANEL = 21.7 kips VTOTAL = 21.7 kips See Appendix A for detailed calculations. 3.6.5 Conclusions Biennual inspection records were of limited value. However, monthly inspection records for this bridge provides a photographic rec ord of the sequence in which failure occurs (see Fig. 3.21 and Fig. 3.19). Again, failur e was due to re-repair regions. Punching shear failure loads assuming resistance was provide d from three surfaces overestimated the failure load. The condition of the underside of the bridge, especially if it shows signs of water stains may indicate impending localized fa ilure. For this bridge, rainfall was a contributory factor as there was a fair amount of r ain just prior to failure (see Fig. 3.22).

PAGE 73

60 3.7 Summary and Conclusions This chapter provided detailed information on five localized failures in panel deck bridges that occurred over the period between Febru ary 2000 and September 2003. These occurred at two locations Sarasota and Tampa. One other failure was mentioned in the local newspaper (Section 3.2.1.1) in bridge #170146 but no records of this could be found. The primary goal of this chapter was to identify u nderlying trends that led to failure in order to develop a rational deterioratio n and failure mechanism of these bridges. To this end, attention was focussed on where failur es occurred, inspection and environmental information. The principal conclusion s are summarized below: 3.7.1 Failure Trend National Bridge Inventory deck condition rating (T able 3.15) was found to be a poor indicator for predicting panel deck failures. All bridges that failed were rated between 5 (satisfactory) to 7 (good). Inspection re cords give a periodic snapshot on the condition of the bridge. Whereas biennial inspectio n data were generally unable to predict failure, monthly inspection records were fa r more successful in tracking problems that led to failure (see Table 3.15, Figs. 3.21/3.2 0). Based on the information provided in the inspection records for the five failures, the s equence leading to failure may be summarized as shown in Fig. 3.23. Figure 3.23 Simplified Deck Deterioration Process Longitudinal/ Transverse Class 1 cracking – U nchanged for up to 10 years Delamination/ Spalling / Repair Repair deterioration / Re-repair / FAILURE Variable

PAGE 74

61 The simplified model indicates that longitudinal c racks first develop along the girder lines. This is followed by occasional reflec tive transverse cracking. Such defects appear within 5 years of construction. These cracks may not change for nearly 10 years (Tables 3.3, 3.6, 3.9) after which there is more wi despread transverse cracking. Longitudinal and transverse cracking result in spal ling, delamination that require repair. In most cases, such damage occurs in regions where the panel is improperly supported on fiberboard. Depending on the materials and quality of the repair the deck can perform poorly or satifactorily. Where deck repairs are com bined with proper panel bearing, e.g. by injecting epoxy, repairs are satisfactory. Where this is not carried out, and repairs are limited to surface repairs, there is progressive de gradation (Fig. 3.21/3.20) which can lead to failure. In several instances, failures occurred at locations where temporary repairs had not been replaced. Simplified calculations show that punching failure s could result at loads below the design wheel load. This assumed the cast-in-pla ce deck to provide no resistance and the panel to be supported on fiberboard with well d eveloped cracking along the transverse and longitudinal panel boundaries. The f ailure load was calculated to be around 15 kips (Table 3.4). Otherwise, failure load s were nearly four times higher. Table 3.15 Inspection Record Bridge # Conditon Rating Last Inspection # of Rainfall events in past 7 days Comments 170146 6 (Satisfactory) 3 months 0 Not identified 170086 7 (Good) 6 months 2 (0.68 in) Not identified 170085 7 (Good) 7 months 4 (0.2 in.) Identified 100332 5 (Fair) 2 days 2 (0.55 in.) Identified 100332 5 (Fair) 23 days 3 (1.1 in.) Identified

PAGE 75

62 3.7.2 Environmental Factors In four out of the five cases there was rainfall p rior to failure (Table 3.15). The most severe rainfall preceded the last failure (1.1 in.). Also, photos of the underside of the bridges that failed show water stains (see Figs. 3. 7, 3.14, 3.19). The exact role of rainwater is not known. However, given that the con crete in the deck separates cleanly from the reinforcement (e.g. Fig. 3.19), it probabl y adversely affects bond and degrades the cohesiveness of the cement paste. Thus, it is r easonable to conclude that rainfall accelerates existing damage that can result in fail ure. 3.7.3 Failure Location All failures occurred under the wheel loads applie d close to the face of the girders where initial longitudinal cracks developed. Also in all five cases, the failure occurred in the right lane, i.e. slow lane (Table 3.16). Failur e was generally in the edge or corner panels whose boundaries developed reflective longit udinal and transverse cracking. Table 3.16 Failure Comparison Bridge # Year Built Age at Failure (yrs) ADT (ADTT) Failure Size Location in Panel Comment 170146 1981 19 34,000 (30%) 18 in x 24 in Edge or Corner? Failure at M1 repair 170086 1980 20 34,000 (30%) 36 in x 60 in Corner Support Patch repair 170085 1980 20 34,000 (30%) 18 in x 18 in Corner Failure adjacent to M1 repair 100332 1980 22 23,000 (8%) 48 in x 30 in Near corner Asphalt Patch 100332 1980 23 23,000 (8%) 24 in x 36 in Edge Failed M1 repair with flexible patch material National Bridge Inventory condition rating given in the bridge inspection prior to the deck failure

PAGE 76

63 3.7.4 Bridge Characteristics All failures occurred in bridges where the deck wa s nominally 7 in. thick. No failures occurred in deck panel bridges with thicke r slabs. The ADTT varied between 830% (Table 3.16). Also it may be noted that the failures occurred in two twin bridges (NB and SB 170086, 170085), and in a bridge adjacent to these two (170146). It is very likely that these three bridges were built with similar defects by the same contractor. The other two cases also occurred in the same bridge (100332 span s 38 and 70).

PAGE 77

64 CHAPTER 4. FORENSIC INVESTIGATION 4.1 Introduction In the previous chapter, five reported failures we re investigated with a view towards identifying underlying trends that could be used to predict future failures. This chapter describes on-site investigations that were carried out to pursue the same objective: to gain enhanced understanding of the de gradation process. In the study, several panel deck bridges scheduled for replacemen t during 2003-2004 and located within easy driving of the USF campus were investig ated. A list of these bridges is given in Table 4.1. Table 4.1 Forensic Studies Bridge # District Built Study Date Bridge Location 130078 1 1981 6/03 I-75 SB over Moccasin Wallow Rd (Manat ee County) 130079 1 1981 6/03 I-75 NB over Moccasin Wallow Rd (Man atee County) 170140 1 1981 1/04 I-75 NB over Toledo Blade Blvd (S arasota County) 130075 1 1981 5/04 1-75 SB over CSR R/R (Manatee County) 100415 7 1983 6/04 I-75NB over US 92 (Hillsborough C ounty) 100398 7 1984 6/04 I-75NB over Sligh & Ramp D-1 (Hill sborough County) 100417 7 1983 7/04 I-75NB over Ramp B-1 (Hillsboroug h County) 130085 1 1981 8/04 I-75NB over SR-64 (Sarasota Count y) The bridges included in this forensic study were s cheduled for a complete deck replacement for a variety of reasons not necessaril y related to the state of disrepair. As a result, both badly deteriorated and those not so ba dly deteriorated decks were investigated. This made it possible to investigate the condition of the decks at different stages of deterioration.

PAGE 78

65 The aim of the investigation was to compile a phot ographic record of the deterioration that could be used in developing a ra tional failure model. Forensic inspection methods were designed to obtain maximum information with minimal disruption to the contractor. The specific informat ion of interest is summarized in Section 4.2. Self-standing sections relating to each bridge in Table 4.1 is presented in Sections 4.3-4.10. The investigations reported could not have been ca rried out without the cooperation and unconditional assistance of the dec k replacement contractors: Zep Constructions Inc. and AIM Engineering & Surveying. 4.2 Objectives The main objective was to obtain first hand eviden ce on actual deck deterioration in order to get a better understanding of how defic iencies are initiated and how they propagate in typical deck panel bridges. Specific information of interest was for identifyi ng conditions that resulted in: 1. No deck cracking. 2. Longitudinal deck surface cracks. 3. Transverse deck surface cracks. 4. Deck surface spalling including “walking” spal ls. 5. Deficient M1 repairs. 6. Underside longitudinal and transverse panel cr acking. 7. Condition of fiberboard bearing. 8. Effect of epoxy panel bearing. 9. Effect of different wheel locations. Not all the information could be retrieved from a single bridge given that they were in different states of disrepair. In the sect ions that follow the same basic format will

PAGE 79

66 be followed: a description of the bridge that was r eplaced followed by the inspection method used and the principal findings. 4.3 I-75 NB and SB over Moccasin Wallow Rd. (Bridge s #130079, #130078) The replacement of the deck in these twin bridges w as carried out in June 2003 by Zep Constructions. In three weeks, the existing pan el deck was removed and replaced by a full-depth cast in place concrete slab. In all a deck area of 35,680 sq. ft was replaced. 4.3.1 Bridge Details The I-75 NB and SB bridges over Moccasin Wallow in District 1 are located in Manatee County, a few miles north of the I-75 I-2 75 intersection. These 3-span bridges were built in 1981 and were in service for nearly 2 3 years before replacement. Each bridge has two approximately 100 ft. long main span s ( span 2, span 3 ) and two 45 ft long secondary spans ( span 1, span 4 ). The total length is about 290 ft. In the north bound bridge, the shorter spans were built using two AASHTO Type IV girders on the outside and six AASHTO Type II gi rders on the inside all spaced 9 ft 3 1/2 in. apart. For the main span, nine AASHTO Type IV girders are spaced at 8 ft 1 1/2 in on centers as shown in Fig. 4.1. In the south bound bridge, the shorter spans use two AASHTO Type IV girders on the outside and five AASH TO Type II girders on the inside all spaced 8 ft 10 in. apart. In the main span, sev en AASHTO Type IV girders are spaced at 8 ft 10 in on centers. The deck had a 7.5 in. thick concrete slab with th e precast panel component being either 2- in. or 3- in. (at the rib-section) thic k as shown in Fig. 4.2. The specified compressive strength of concrete for the precast pa nel is 5,000 psi. It is 3,000 psi for the cast in place concrete slab.

PAGE 80

67 The bridge has three 12 ft wide lanes, and 10 ft w ide shoulders as shown in Fig. 4.1. There is an auxiliary lane that merges with tr affic entering the interstate from I-275. These dimensions and the bridge cross-section are t ypical of all panel deck bridges in Districts 1 and 7 excepting that the deck thickness (7.5 in.) is slightly greater than the 7 in. norm. In general the deck was in reasonable condition in both bridges with typical longitudinal and transverse cracking. Some regions had deteriorated and both M1 Repairs and spalling were present. Table 4.2 Bridges #130078 and #130079 [27] Bridge #130078 (SB) Bridge #130079 (NB) Year Built 1981 1981 Number of Spans 4 4 Lanes on Structure 3 4 ADT (2003) 26,500 27,000 Percent Truck (ADTT) 30% 30% Composite Slab Thickness 7- in. 7- in Precast Panel Thickness 2- in (panel) 3- in (ribs) 2- i n (panel) 3- in (ribs) Girder Type AASHTO Type II and IV AASHTO Type II and IV Deck Condition Rating (2003) 7 (Good Condition) 7 (Good Condition)

PAGE 81

68 Figure 4.2 Composite Deck Section 4.3.2 Inspection Method As several panels (Fig. 4.3) had already been removed when I was able to access the site, three different procedures were to optimize the investigation. This involved (1) examination of already removed panels from the southbound bridge, (2) inspection of panels that had been identified prior to their removal and (3) inspection of panels in-place after adjoining panels had been removed. The last scenario provides the best information but is rarely possible since it can interfere with the contractors work. Thru Traffic Lanes Type IV Girders 8'-1 1 2 12' 8'-1 1 2 12' 10' 1'-4 1 2 1 2 8'-1 1 2 Bay 1 12' 8'-1 1 2 Bay 2 Bay 3 8'-1 1 2 Bay 4 8'-1 1 2 12' Bay 6 8'-1 1 2 Bay 5 Bay 7 10' 1'-4 1 2 2'-10 1 2 8'-1 1 2 Bay 8Aux. Lane Lane 1Lane 2 Lane 3(Main Span Deck Cross Section) Figure 4.1 Cross Section View of Bridge #130078 3/16@6" 1 1 2" 2 1 2" 3/8" Prestressing Strands 6" 6" 2/8@6" 1 4 Min 75" 1" Traffic Direction

PAGE 82

69 A quick visual inspection was conducted to identif y panels that exhibited typical deficiencies and a detailed was done with all the i nformation collected in the form of photographs, sketches and field notes. Care was tak en to isolate existing deficiencies from those induced by the removal process. Figure 4.3 Removed Deck Sections from SB Bridge #130079 This was undertaken for the east (right) half of t he northbound bridge. Following a quick inspection of the deck regions of special i nterest were identified (Fig. 4.4). These included sections with well defined typical deficie ncies as well those with no apparent defects. A total of six panel sections were marked and removed. The average dimension of these sections was 8 ft by 10 ft. The contractor removed the marked sections taking e xtra care to minimize additional damage and then stored them at an assign ed place for subsequent detailed inspection.

PAGE 83

70 Figure 4.4 Marked Deck Sections Removed From NB Bridge#13 0078 Inspection of these marked sections included detai led visual examination, crack survey of the deck surface and the cross section, a nd extraction of concrete cores (Fig. 4.5) from locations of special interest. A total of 15 cores were taken from the 6 deck sections. Please refer to Appendix B for detailed information on these deck cores. Figure 4.5 Coring of Marked Sections To eliminate any doubt that the crack patterns were induced by the removal process, insitu inspection was instituted wherein deck secti ons were examined prior to their removal. This provided authentic information on cra ck propagation through the thickness

PAGE 84

71 of the deck. As mentioned earlier, this was possib le when adjacent sections had already been removed to allow access to the vertical faces of the section. This inspection confirmed that the condition of the deck deficienci es was unaffected by the removal process. 4.3.3 Findings Fig. 4.6 is a schematic drawing highlighting some of the findings. It provides details of their location in the deck cross section and also cross-refers to figure numbers where photographs of the particular deficiencies ar e provided. In the following sections, detailed information is provided for each of the following findings some of which are shown in Fig. 4.6. These are: 1. No Deck Surface Cracking 2. Longitudinal Deck Surface Cracking 3. Transverse Deck Surface Cracking 4. Additional Longitudinal Cracking 5. Deck Spalling and Delamination.

PAGE 85

72 Separation at vertical panel face Small vertical crack No surface cracks Diagonal Crack Vertical Crack *Separation at vertical panel face Longitudinal surface crack Bay 2 8'-1 1 2 Inspected Bays 12' 2'-10 1 2 8'-1 1 2 10' 7 1 2 Bay 1 Main Span 12' Bay 5 8'-1 1 2 8'-1 1 2 Bay 3Bay 4 Thru Traffic Lanes Figure 4.6 Overview of Findings from Bridge # 130079 Diagonal Crack Vertical Crack Longitudinal crack at the surface Severe cracking and spalling See Fig. 4.7 See Fig. 4.10 See Fig. 4.8 See Fig. 4.11

PAGE 86

73 4.3.3.1 No Deck Surface Cracking Fig. 4.7 shows a retrieved panel with no surface cracking. The location of the prestressed girder support and the bearing pad has been drawn to provide better understanding. Figure 4.7 No Deck Surface Cracking Inspection of Fig. 4.7 shows that there is separation of the precast panel from the cast-in-place concrete slab possibly due to long term differential creep and shrinkage movement. This separation is of about 5 mm wide. There is a vertical crack emanating from the corner of the panel that does not propagate all the way to the deck surface. This could because the effect of creep and differential shrinkage was lower for this case, e.g. lower effective prestress, smaller age difference between casting of the panel and CIP deck. Precast Panel C.I.P N N o o C C r r a a c c k k s s P P a a n n e e l l S S i i d d e e F F a a c c e e S e e p a r r a a t i o n V e e r t t i c c a a l C r a a ck P P r e e s t r e s s s s e e d G G i i r r d e e r P P a a n n e e l l B B e e a a r r i i n n g g

PAGE 87

74 4.3.3.2 Deck Surface Longitudinal Cracking This is the most common deficiency observed in precast deck panel bridges found in almost all the deck panel bridges. Fig. 4.8 shows how a typical longitudinal crack develops. This picture was taken with the panel in place in the bridge after the adjoining panel had been removed. The prestressed girder shown is the actual girder which supported the panel. The fiberboard bearing support is also visible. Figure 4.8 Development of Deck Surface Longitudinal Crack Inspection of Fig. 4.8 shows clear separation of the vertical interface between the panel and the cast in place slab, i.e. the face of the panel completely debonded from the cast in place concrete. A vertical crack emanates from the top corner of the precast panel and propagates to the top of the deck. This pattern is replicated along the entire edge of the pane creating reflective longitudinal cracking on the deck surface. P r r e s s t t r e e s e d d G G i i r r d e e r Typical Longitudinal crack Bonded interface Vertical crack Separation at panel face Saw Cut

PAGE 88

75 It is important to recognize that this type of cracking can even be found on the shoulders of the bridge where live load is minimal. Thus, this type of cracking is not related to live load. Also for this case and for all the sections inspected it was found that a very good bonded interface existed between the top face of the panel and the cast in place concrete. This indicates that composite action under bending loads. 4.3.3.3 Deck Surface Transverse Cracking Transverse deck surface cracking is not as common as longitudinal cracking. In most cases this is a hairline crack and it tends to remain stable without causing any further damage. In the forensic examination it can only be detected from sections that have been removed (Fig. 4.9). Figure 4.9 Deck Surface Transverse Crack Traffic Direction Transverse Crack Panel Panel

PAGE 89

76 In Fig. 4.9 the location of the two adjoining panel s has been drawn to provide better understanding. Cracking emanates at the join t and eventually propagates to the deck surface. Thus, it is a reflective crack that m aps the location of the transverse panel joint on the deck surface. Where it does not reach the top surface, no cracking is visible. 4.3.3.4 Additional Longitudinal Cracking In addition to the typical longitudinal crack runni ng over the edge of the panels (See 4.3..3.3) another type of longitudinal crackin g was found. This crack runs about 4 in. parallel to typical longitudinal cracks (Fig. 4.10) Figure 4.10 Additional Longitudinal Cracking This additional longitudinal cracking is caused by a bifurcation of the vertical crack emanating from the corner of the precast pane l. It propagates at an angle of less than 45 degrees to reach the deck surface, generati ng an additional longitudinal crack on the deck surface. Additional Longitudinal cracking Typical Longitudinal cracking Precast deck panel

PAGE 90

77 Precast Panel Prestressed Girder Diagonal Cracking Severe Cracking And Spalling Vertical Cracking Panel Face Separation Panel Bearing Wheel Load This type of crack is not as common as the typical longitudinal crack; it is only found in localized regions of the deck whereas othe r cracks tend to occur along the entire span. Also it was found that this additional cracki ng only occurs when a wheel load is located close to the panel support (see Fig. 4.6). 4.3.3.5 Deck Spalling and Delamination This is one of the most important deficiencies in deck panel bridges. In the previous chapter examples are provided where sudden localize d deck failures occurred at sites where temporary spalling repairs had been carried o ut. The deck section analyzed was removed from span 3, bay 3 from the north bound bridge (see bridge cross section detail, Fig. 4.6). The spall was located right under the wheel load with the wheel load positioned at the ed ge of the girder. Figure 4.11 Development of a Deck Surface Spall

PAGE 91

78 In this specific case, the spalled area studied wa s located next to an existing M1 repair. This is a common deficiency in deck panel b ridges; it is also known as a “walking spall” because it always occurs next to a spall pat ch or repair. Fig. 4.11 is a photograph of the retrieved panel. A prestressed girder is drawn to provide contextual reference. Inspection of Fig 4.1 1 show s that it has all of the cracks described earlier, i.e. panel separation, vertical crack, diagonal crack, but with an increase in width of the cracks and additionally more diagon al cracks under the spalled area. The longitudinal and diagonal cracking causes the concrete surface to break up into small pieces that can be easily detached from the deck by traffic creating the spall. Deck deterioration starts to accelerate due to the impact of the wheel loads on the spall. 4.3.3.6 Findings on Panel Bearings Regarding the precast panel’s bearing it was found that the bridge was built using only fiberboard to support the panels (negative bea ring), but recently only in some areas of the bridge the fiberboard bearing has been remov ed and replaced by epoxy. The replacement of the fiberboard by epoxy was recommen ded in a previous research study [11], as a method to reduce future deterioration of the bridge deck, but exactly in the spots where the major deterioration was found, the fiberboard had not been replaced by epoxy. 4.3.3.7 Findings on Core Examination Most of the cracks found on the cores show signs o f water and dust infiltration ( Cores 1-3, 1-4, 5-4, 5-6, 5-7) (Fig. 4.12). In the case of vertical cracks, some of them show these signs only over half the depth indicatin g that the prestressed slab was uncracked. But when the section deteriorated, infil tration occurred over the entire deck depth (Cores 5-6, 5-7 ). In most of cases (cores 1-3, 1-4), concrete at the top of diagonal crack was crumbled and showed signs of water infiltration.

PAGE 92

79 Figure 4.12 Crumbled Concrete in Top of a Diagonal Crack. ( Core 1-3) M1 repairs debonded only near the panel edges in t he vertical direction as well as at its horizontal interface with the cast-in-place slab (core 1-1, 1-2, 5-4, 5-5 ). Along the longitudinal interface away from the panel edge the re was no debonding (cores 5-3) The depth of the deck, measured at each core locat ion, varied from 7 in to 8 in. See Appendix B for a detail description of each core. 4.4 I-75 NB over N Toledo Blade Blvd. (Bridge #170 140) The deck replacement of this bridge was performed i n January 2004 by Zep Constructions Inc. Fort Myers FL. At that time the bridge was widened and an additional 12 ft lane added on the left side. 4.4.1 Bridge Details This 3-span bridge located in Sarasota County, FL was built in 1981. Its deck was in service for 23 years before replacement. The bri dge has a main span ( span 2 ) of 107 ft 8 in. and two 41 ft secondary spans ( span 1, span 3 ). Its overall length is 189 ft 8 in. Diagonal Crack Core 1-3 Panel Crumbled Concrete C.I.P

PAGE 93

80 The shorter spans were built using two AASHTO Type IV girders on the outside and three AASHTO Type II girders on the inside all spaced 9 ft 3 in. apart. In the main span, seven AASHTO Type IV girders are spaced at 6 ft 2 in on centers as shown in Fig. 4.12. The deck has a 7.5 in. thick concrete slab with the precast panel component being either 2in. or 3in. (at the rib-section) thick as shown in Fig. 4.2. The specified compressive strength of concrete for the precast panel is 5,000 psi. It is 3,000 psi for the cast in place concrete slab. Table 4.3 Bridge #170140 Bridge #170140 Characteristics Year Built 1981 Number of Spans 3 Lanes on Structure 2 ADT (2003) 19,000 Percent Truck (ADTT) 30% Composite Slab Thickness 7in. Precast Panel Thickness 2in (panel) 3in (ribs) Girder Type AASHTO Type II and IV Deck Condition Rating (2003) 7 (Good Condition) 12' 6'-2" 12' 10 2' -101 2" 6'-2" 6'-2" 6'-2" 6'-2" 6'-2" Bridge Number 170140 Main Span 2' -101 2" 7 1 2 6' Figure 4.13 Cross Section View of Bridge #170140 Bays covered in the study Bay 6 Bay 5 Bay 4 Bay 3 Bay 2 Bay 1

PAGE 94

81 The bridge has two 12 ft wide lanes, a 10 ft wide shoulder on the right of the traffic, and a 6 ft shoulder on the left, as shown in Fig. 4.13. It is in District 1 and is located 30 miles south of Sarasota. The part of the bridge where the study was conduct ed was in apparent good condition. It only exhibited typical longitudinal a nd some transverse cracking. No previous repairs were found on the deck. 4.4.2 Inspection Method The methodology used for this bridge was the same as the one used on the I-75 NB over Moccasin Wallow. Deck sections of special i nterest were marked for careful removal and subsequent detailed inspection (Fig. 4. 14). However, no cores were extracted from the deck sections. A) Marked section B) Removed section Figure 4.14 View of Bridge #170140 4.4.3 Findings An examination of the retrieved panels confirmed th e findings from the previous bridge. Longitudinal cracks emanated from the corne r of the prestressed panel and propagated through the slab thickness to emerge as visible cracks (Fig. 4.16a, 4.16b). Additional parallel cracking due to divergence of t he crack emanating from the panel

PAGE 95

82 corner was also observed. However, the parallel cracking on the deck surface only appeared intermittently as shown in Fig. 4.16c. 2' -101 2" 1'-4 1 2 12' 6'-2 6'-2" 6'-2" 10' Figure 4.15 Findings Overview, Bridge # 170140 As before, there was separation between the precast panel and the cast-in-place slab at its vertical interface. This was suspected to be due to long term creep and shrinkage as stated earlier. An example was also found of the deck panel being supported by epoxy instead of fiberboard. In this instance, the extent of the longitudinal cracking was reduced (Fig. 4.16d). Overall, there were no dramatic new findings, simply confirmation of what was found earlier. crack at the surface Diagonal Crack Vertical Crack Separation at verical panel face No surface cracks Small vertical crack Vertical Crack *Separation at vertical panel face Longitudinal surface crack See Fig. 4.16(a) See Fig. 4.16(c) See Fig. 4.16(b) See Fig. 4.16(d)

PAGE 96

83 (a) Longitudinal Cracking (b) Longitudinal Cracking (c) Additional Longitudinal Cracking (d) Epoxy Support Figure 4.16 Retrieved Panels from Bridge #170140 Panel C.I.P Longitudinal Crack Panel C.I.P Longitudinal Cracks Diagonal Crack C.I.P Girder Epoxy Bearing Repair Separation Vertical Crack Panel C.I.P Longitudinal Crack

PAGE 97

84 4.5 I-75 SB over CSX R/R. (Bridge #130075) The deck replacement of this bridge was performed in May 2004 by Zep Constructions Inc. Fort Myers FL. 4.5.1 Bridge Details This 3-span bridge located in Sarasota County, FL was built in 1981. Its deck was in service for 23 years before replacement. The bri dge has a main span ( span 2 ) of 79 ft 2 in and two 45 ft 5 in secondary spans ( span 1, span 3 ). Its overall length is 170 ft. The shorter spans were built using two AASHTO Typ e III girders on the outside and five AASHTO Type II girders on the inside all s paced 8 ft 10 in. apart as shown in Fig. 4.16. For the main span, nine AASHTO Type III girders are spaced at 6 ft 7 in on centers. The deck has a 7.5 in. thick concrete slab with th e precast panel component being either 2- in. or 3- in. (at the rib-section) thic k as shown in Fig. 4.2. The specified compressive strength of concrete for the precast pa nel is 5,000 psi. It is 3,000 psi for the cast in place concrete slab. Table 4.4 Bridge #130075 [27] Bridge #130075 Characteristics Year Built 1981 Number of Spans 3 Lanes on Structure 3 ADT (2003) 36,500 Percent Truck (ADTT) 30% Composite Slab Thickness 7- in. Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type AASHTO Type II and III Deck Condition Rating (2003) 5 (Fair Condition)

PAGE 98

85 Bridge Number 130075 Secondary SpanBay 6 Bay 5 Bay 4 Bay 3 Bay 2 Bay 1 2'-10 1 2" 8'-10" 8'-10" 8'-10" 8'-10" 8'-10" 8'-10" 2'-10 1 2" 12' 10' 10' 12' Lane 2Lane 3 12' Lane 1 Figure 4.17 Cross Section View of Bridge #130075 The bridge has three 12 ft wide lanes, and 10 ft w ide shoulders, as shown in Fig. 4.17. It is in District 1 and is located 2 miles no rth of Ellenton. From the inspection performed before the deck remo val, typical longitudinal and transverse cracking plus various M1 repairs, some o f them stable and some unstable. (Fig 4.18) were found. Figure 4.18 Deck Overview of Bridge #130075 Bays covered in the study

PAGE 99

86 4.5.2 Inspection Method This bridge was not part of the original investiga tion. Access was arranged at the last minute when much of the deck had already been removed. In view of this a different approach had to be employed. In the modified approach there was no time for mar king sections and then having them carefully removed by the contractor. Consequen tly, it was necessary to perform a quick inspection to locate and document major deck deficiencies. Following this inspection, each deck section was inspected in plac e after the adjoining section had been removed (Fig. 4.19a). Also deck sections that had b een removed were also inspected, Fig. 4.19b. Figure 4.19 Inspection Methods Bridge #130075 4.5.3 Findings As in previous examinations, panel face separation, and vertical cracking other typical cracking described in detail earlier were d etected (see Fig. 4.20). New information relating to panel support was found. Fig. 4.20 is a view of a section of the panel deck and the prestressed girder. The panel on the left is supported by epoxy while the s ection on the right is on fiberboard Thus, the replacement was partial and not over the entire deck as recommended in a (a) (b)

PAGE 100

87 previous research study [14]. Inspection of Fig. 4. 20 shows that when epoxy was used to replace the fiberboard bearing it only penetrated o ver approximately one third the bearing width leaving a region that was unsupported (bay 3) Note the emergence of a vertical crack from this unsupported region. A similar crack appears from the edge of the fiberboard support on the right (bay 2). This was m ore heavily loaded and required an M1 repair. The divergence of the vertical crack causes separation of the interface between the M1 repair and the panel that cannot act compositely under flexural loading. The deterioration is more severe in bay 2 because of a combination of heavier loads and fiberboard supports. Figure 4.20 Deck Cross Section View over Girder # 3 Panel Panel M1 Repair C.I.P Saw Cut C.I.P Girder Top Epoxy bearing replacement Fiberboard bearing Induced Damage Separation Bay 3 Bay 2 Void under the panel

PAGE 101

88 Figure 4.21 Panel Bearing Condition, over Girder # 3 Due to the examination of the deck sections before removal, it was possible to prove that the typical cracking found in this bridg e and in previous cases is not caused by the deck removal process. 4.6 I-75 NB over US 92. (Bridge #100415) The deck replacement of this bridge was performed i n June 2004 by AIM Engineering & Surveying. This was conducted simulta neously with two other deck panel bridges (#100398 #100417) that are part of I-75I4 interchange. 4.6.1 Bridge Details This 3-span bridge located in Hillsborough County, FL was built in 1983. Its deck was in service for nearly 21 years before replaceme nt. The bridge has a main span ( span Girder Top Bay 3 Bay 2 Fiberboard Bearing (No epoxy repair) Water stains Under deck top deterioration Epoxy bearing replacement Remains of Fiberboard

PAGE 102

89 2 ) of 107 ft. and two 40 ft 7in secondary spans ( span 1, span 3 ). Its overall length is 188 ft 2 in. Only the main span ( span 2 ) was built using precast deck panels, the other sp ans being built using full depth cast in place concrete The shorter spans used two AASHTO Type IV girders on the outside and five AASHTO Type II girders on the inside all spaced 10 ft 1 in. apart. The main span has ten AASHTO Type IV girders spaced about 6 ft 9 in on ce nters as shown in Fig. 4.21. The deck has a 7.5 in. thick concrete slab with th e precast panel component being either 2- in. or 3- in. (at the rib-section) thic k as shown in Fig. 4.2. The specified compressive strength of concrete for the precast pa nel is 5,000 psi. It is 3,000 psi for the cast in place concrete slab. Table 4.5 Bridge #100415 Bridge #100415 Characteristics Year Built 1983 Number of Spans 3 (only span 2 –deck panel) Lanes on Structure 4 ADT (2003) 43,000 Percent Truck (ADTT) 30% Composite Slab Thickness 7- in. Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type AASHTO Type II and IV Deck Condition Rating (2003) 5 (Fair Condition)

PAGE 103

90 12' 6'-9 3 16 Bay 9 6'-9 3 16 6' Type IV Girders Thru Traffic Lanes Bay 8Bay 7Bay 6 Lane 3 Lane 2 Lane 1 Aux. Lane 12' 12' 12' 6'-9 3 16 6'-9 1 8 6'-9 3 16 6'-9 1 8 6'-9 3 16 6'-9 3 16 6'-9 3 16 10' Bay 5*Bay 4*Bay 3*Bay 2*Bay 1* Figure 4.22 Cross Section View of Bridge #100415 span 2 This bridge has three main lanes, a 12 ft wide aux iliary lane, and two shoulders – one 6 ft wide and the other 10 ft as shown in Fig. 4.22. It is in District 7. Figure 4.23 Bridge #100415 Span 2, Prior to Deck Removal The forensic study was conducted on span 2, bays 1 to 5. This section of the bridge exhibited longitudinal and some transverse c racking typical of deck panel construction. Also along bay 5, there were two dete riorated M1 repairs and several walking spall patches as shown in Fig. 4.23. This f igure also identifies the cut patterns used by the contractor. Bays covered on the study (1 – 5) M1 Repairs and walking spalls Cut patterns

PAGE 104

91 4.6.2 Inspection Method The intent was to follow the same procedure used i n earlier forensic studies. However, the contractor used a different cut patter n (Fig. 4.24) and therefore regions of greatest interest (the supported edges of the panel along the girder lines) were not included in the removed section. Therefore analysis focused mainly on the deck sections that were left on the top of the girders. These pro vided information on the bearing support provided to the panels. Before the deck was removed, a detailed inspection was conducted to document the deficiencies and to determine their exact locat ion so that their position could be identified in the remaining deck section on the top of the girders. Special interest was placed on assessing the condition of the panel bear ings along the bridge deck. The cut pattern used in this case (Fig 4.24) helped to prov ide a detailed and unaltered view on the deck bearing in most of the deck. The panel section s removed were also inspected but not much information was obtained from them. 3 U H F D V W 3 D Q H O Figure 4.24 Cross Section View of Cut Pattern on Bridge #100415

PAGE 105

92 4.6.3 Findings 4.6.3.1 Deteriorated M1 Repair and Walking Spalls Fig. 4.25 shows the location of the M1 repair and the walking spalls in bays 4 and 5 in span 2. There are four numbered locations 1-4 in the plan view. These identify elevation views of the supporting girder and deck s ection after the panel had been cut out. The top left figure marked 1 shows the support for the panel at the M1 repair location. Note the longitudinal delamination in the cast-in-p lace (CIP) slab near the top. The figure marked 2 is a view of the panel after it was remove d and placed on temporary barrier supports. The patch repairs and regions adjacent to it separated readily indicating loss of bond. The figure marked 3 is same as the one marked 1 except that it is located at a deteriorated region. A hammer top can be easily ins erted indicating lack of bearing support and separation (also shown in the figure ma rked 4 where the concrete was removed. Separation of the vertical face (not visib le) is also marked.

PAGE 106

93 Figure 4.25 Examination of a Deteriorated Deck Section on Bridg e #100415 2 Spall patches Traffic Direction Delaminated area Delamination Crack, under rebar # 6 Rebar Settlement Patch 1 Spall patches Movement Delamination cracks Panel C.I.P No panel bearing Missing fiber boar bearing Panel face separation Less than ” of concrete under panel edge 3 4 M1 Repair Consecutive spall patches 1 3 & 4 2 View Orientation Traffic Direction Bay 5 Bay 4

PAGE 107

94 4.6.3.2 Deck Panel Bearing Figure 4.26 Panel Bearing Examination on Bridge #100415 Remaining panel edge 1 Fiber Board Void under panel edge Missing fiber boar bearing Panel face separation Less than of concrete under panel edge 4 3 Spall Patch Remaining panel edge C.I.P. 1 Fiber Board 2 Concrete bearing under panel edge Remaining panel edge C.I.P 1 Fiber Board Void under panel edge 2 1 4 2 3 1 Previous location of deterioration (Fig. 4.#) Bay 5 Bay 4

PAGE 108

95 Fig. 4.26 shows support for the panels at various l ocations along the bridge. Four pictures reflecting bearing locations marked 1-4 in the plan view are shown. The figure marked 1 shows a region where the slab was supporte d by 1 in. of fiberboard and 2 in. of concrete. This was unexpected from the Crosstown co nstruction drawings and from previous research that indicated that the fiberboar d was placed at the ends leaving no room for concrete to penetrate under the panel. Onl y vertical cracking was present in the panel with no delamination. Unfortunately, it canno t be seen because it was saw cut. The figures marked 2 and 3 show alternate locations whe re the concrete was unable to penetrate below the panel. The last figure, marked 4, also appearing in Fig. 4.25, shows lack of support that led to cracking and spalling o f the deck. Thus, this figure provides evidence on the role of the bearing support on the performance of the deck. 4.7 I-75 NB over Sligh Ave & Ramp D-1 (Bridge #100 398) The deck replacement of this bridge was performed in June 2004 by AIM Engineering & Surveying. This deck replacement was conducted simultaneously with two other deck panel bridges (#100415 #100417) that are part of I-75I-4 interchange. 4.7.1 Bridge Details This 5-span bridge located in Hillsborough County, FL was built in 1984. Its deck was in service for nearly 20 years before replaceme nt. The span lengths are as follows: Span 1 (south) 35 ft, Spans 2 and 3, 82 ft, Span 4, 54 ft 10 in, and Span 5 107 ft. Its overall length is 360 ft 10 in. Span 1 has two AASHTO Type IV girders on the outs ide and six AASHTO Type II girders on the inside all spaced 10 ft 2 in. apa rt. Spans 2, 3 and 4, all have the same configuration as span 1 but used only girder type I V. The longest span (span 5) has eleven AASHTO Type IV girders spaced about 6 ft 7 in on centers as shown in Fig. 4.27.

PAGE 109

96 Table 4.6 Bridge #100398 Bridge #100398 Characteristics Year Built 1984 Number of Spans 5 Lanes on Structure 4 ADT (2003) 43,000 Percent Truck (ADTT) 30% Composite Slab Thickness 7- in. Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type AASHTO Type II and IV Deck Condition Rating (2003) 5 (Fair Condition) The deck has a 7.5 in. thick concrete slab with th e precast panel component being either 2- in. or 3- in. (at the rib-section) thic k as shown in Fig. 4.2. The specified compressive strength of concrete for the precast pa nel is 5,000 psi. It is 3,000 psi for the cast in place concrete slab. Bridge Number100398 Spans 2, 3, 4 9'-8 5 16 12' Bay 7Bay 6 Bay 5 2'-10 1 2 Varies Varies 9'-8 5 16 6' 12' Bay 4 9'-8 5 16 2'-10 1 2 9'-8 5 16 9'-8 5 16 9'-8 5 16 10' 12' 12' Bay 3*Bay 2* Bay 1* 2'-10 1 2 2'-10 1 2 Bay 10Bay 9Bay 8Bay 7Bay 6Bay 5 Bridge Number100398 Spans 5 Varies 6'-7 1 2 6'-7 1 2 6'-7 1 2 6'-7 1 2 6'-7 1 2 6'-7 1 2 6'-7 1 2 6'-7 1 2 Varies 6'-7 1 2 10' 6' 12' 12' 12' 12' Bay 4*Bay 3*Bay 2*Bay 1* Figure 4.27 Cross Section View of Bridge #100398 Bays covered on the study (1 – 3) Bays covered on the study (1 – 4)

PAGE 110

97 After the half of the bridge to be replaced was cl osed, a detailed inspection of the bridge deck was conducted. During this inspection, no major deterioration was found, only typical longitudinal and some transverse crack s typical of deck panel bridges. Also no signs of previous repairs were found. (Fig. 4.28 ). Figure 4.28 Deck Overview of Bridge #100398 4.7.2 Inspection Method Bearing in mind that in this bridge, the contracto r did not use the same cut pattern as the ones used in the previous bridge (Fig. 4.24), ( they used a cut pattern similar to the one used on Moccasin Wallow Bridge, (Fig. 4.8), plus th e fact that this bridge deck only exhibited random cracking but no major deficiencies a different inspection method was used. Figure 4.29 Inspection Methods Bridge #100398 (a) (b) Deck Section to Be Replaced

PAGE 111

98 First, random sections previously removed were ins pected to identify panel face separation, vertical cracking or other type of typi cal internal deterioration (Fig. 4.29a). Then a detailed inspection of the panel bearing ove r the edges of the girders was conducted (Fig. 4.29b). 4.7.3 Findings The most important finding was related to panel be aring support. It was found that the type of panel bearing used in this bridge was c ompletely different, to the ones believed to be used in all deck panel bridges in Fl orida. Here a positive bearing was provided by a layer of grout placed next to a 1 in. fiberboard strip (Fig. 4.30). This system provided a stiff support for the panel. Soft fiberboard bearing is known to be responsible for premature deterioration of Florida’ s deck panel bridges [11]. Keeping in mind that this bridge was built on 1984 it is likely that in this bridge the panel bearing detail was changed to a positive bear ing to prevent deterioration that had been observed in panel deck bridges built earlier i n this area. This is the main reason why this bridge deck did not exhibit major deterioratio n after 20 years of service. Figure 4.30 Panel Bearing Examination Bridge #100398 Saw-cut Fiberboard Grout Girder Edge Fiberboard Grout Girder Top Girder Edge C.I.P

PAGE 112

99 Despite the use of positive panel bearing in this bridge, typical deterioration such as panel face separation and vertical cracking was obs erved (Fig. 4.31). This proves that this kind of cracking is not related to the type of bear ing used to support the panel. Positive bearing only prevents the occurrence of additional shear cracking that causes spalling in the deck surface, and may lead to sudden failures. Figure 4.31 Vertical and Longitudinal Cracks Bridge #100398 Fig. 4.32 provides a summary of all the findings o f the forensic investigation of bridge #100398. It shows the panel bearing detail u sed, and the typical panel face separation and vertical cracking found on in almost all the deck sections inspected on this bridge. Figure 4.32 Findings Overview Bridge #100398 C.I.P Panel Panel C.I.P Girder Pane Pane Deck Top 1” Fiberboard 2” wide Grout layer 1” Fiberboard Crack Separation

PAGE 113

100 4.8 I-75 NB over Ramp B-1 (Bridge #100417) The deck replacement of this bridge was performed i n June 2004 by AIM Engineering & Surveying. This deck replacement was conducted simultaneously with two other deck panel bridges (#100398 and #100415) that are part of I-75I-4 interchange. 4.8.1 Bridge Details This 3-span bridge located in Hillsborough County, FL was built in 1983. Its deck was in service for nearly 21 years before replaceme nt. The bridge has a main span ( span 2 ) of 107 ft. and two 40 ft 7in secondary spans ( span 1, span 3 ). Its overall length is 160 ft 6 in. The shorter spans used two AASHTO Type IV girders on the outside and four AASHTO Type II girders on the inside all spaced 10 ft 7-3/16 in. apart, except the center two spaced at 10 ft 7-1/4 in. The main span has sev en AASHTO Type IV girders spaced about 8 ft 10 in on centers as shown in Fig. 4.33. The deck has a 7.5 in. thick concrete slab with th e precast panel component being either 2- in. or 3- in. (at the rib-section) thic k as shown in Fig. 4.2. The specified compressive strength of concrete for the precast pa nel is 5,000 psi. It is 3,000 psi for the cast in place concrete slab. Table 4.7 Bridge #100417 Bridge #100417 Characteristics Year Built 1983 Number of Spans 3 Lanes on Structure 3 ADT (2001) 48,290 Percent Truck (ADTT) 30% Composite Slab Thickness 7- in. Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type AASHTO Type II and IV Deck Condition Rating (2003) 5 (Fair Condition)

PAGE 114

101 8'-10" 12' Bay 6Bay 5Bay 4 2'-10 1 2 8'-10" 8'-10" 10' 12' Bay 3Bay 2Bay 1 8'-10" 2'-10 1 2 8'-10" 8'-10" 12' 10' Figure 4.33 Cross Section View of Bridge #100417 span 2 This bridge has three main lanes, a 12 ft wide aux iliary lane, and two shoulders – both 10 ft wide as shown in Fig. 4.33. This bridge is part of District 7. Figure 4.34 Deck Before Removal Bridge # 100417 (Bays 4-6) After the half of the bridge to be replaced was cl osed, a detailed inspection of the bridge deck was conducted. During this inspection o nly typical longitudinal cracks along the entire deck and some transverse cracks were fou nd. Both are typical of deck panel construction. Also no signs of previous repairs wer e found (Fig. 4.34). Bays covered on the study (4-6) Typical Longitudinal Crack

PAGE 115

102 4.8.2 Findings In this bridge was found exactly the same type of panel bearing detail as in the previous bridge (Section 4.7). Here a positive bear ing was provided by a layer of grout placed next to a 1 in. fiberboard strip (Fig. 4.35) This system provided a stiff support for the panel. Soft fiberboard bearing is known to be r esponsible for premature deterioration of Florida’s deck panel bridges [11]. The use of positive panel bearing in this bridge i s believed to be the reason why it did not exhibit major deterioration after 20 years of service. Figure 4.35 Panel Bearing Examination Bridge #100417 It was also found the longitudinal cracks observed in the inspection conducted, were caused, as found in all the pervious studies, due t o separation of the vertical face of the panel (Fig. 4.36). It is again proven that the pres ence of deck surface longitudinal cracking is related only to the use of precast pane ls regardless the type of panel bearing used, whereas the occurrence of spalls or additiona l cracking directly linked to the type of bearing used (negative bearing). 2” Grout Girder Edge 1” Fiberboard Panel C.I.P 1.5” Grout Girder Edge 1” Fiberboard

PAGE 116

103 Figure 4.36 Panel Bearing Examination Bridge #100417 Fig. 4.37 provides a summary of all the findings of the forensic investigation of bridge #100417. It shows the panel bearing detail used, and the typical panel face separation and vertical cracking found on in almost all the deck sections inspected on this bridge. Figure 4.37 Findings Overview Bridge #100417 4.9 I-75 NB Over SR 64 (Bridge #130085) The deck replacement of this bridge was performed in August 2004 by Zep Constructions Inc. Fort Myers FL. In about three weeks, the existing panel deck was removed and replaced by a full-depth cast in place concrete slab. Girder Panel Panel Deck Top 1 Fiberboard 1 to 2 wide Grout layer 1 Fiberboard Crack Separation Panel C.I.P *Joint between panel and diaphragm C.I.P Panel Vertical face Panel*

PAGE 117

104 4.9.1 Bridge Details This 4-span bridge located in Manatee County, FL D istrict 1 was built in 1981. Its deck was in service for nearly 23 years before repl acement. The bridge has two main spans of 108 ft 10 in. with two secondary spans of 43 ft. Its overall length is 303 ft 8 in. The shorter spans used two AASHTO Type IV beams o n the outside and six AASHTO Type II beams on the inside all spaced 8 ft 8 9/16 in. apart. The main span has eleven AASHTO Type IV beams spaced about 6 ft 9 in on centers as shown in Fig. 4.38 The deck has a 7.5 in. thick concrete slab with th e precast panel component being either 2- in. or 3- in. (at the rib-section) thic k as shown in Fig. 4.2. The specified compressive strength of concrete for the precast pa nel is 5,000 psi. It is 3,000 psi for the cast in place concrete slab. Table 4.8 Bridge #130085 Bridge #130085 Characteristics Year Built 1981 Number of Spans 4 Lanes on Structure 4 ADT (2001) 25,000 Percent Truck (ADTT) 10% Composite Slab Thickness 7- in. Precast Panel Thickness 2- in (panel) 3- in (ribs) Girder Type AASHTO Type II and IV Deck Condition Rating (2003) 7 (Good condition )

PAGE 118

105 12' 12' 12' 12' 6' 2'-10 9 16 "6'-1 3 16 "6'-1 3 16 "6'-1 3 16 "6'-1 3 16 Bay 1 Bay 10 Bay 9 Bay 8 Bay 7 Bay 6 Bay 5 Bay 4 Bay 3 Type IV Girders Bay 2 Thru Traffic Lanes Aux. Lane 2'-10 9 16 6'-1 3 16 6'-1 3 16 6'-1 3 16 6'-1 3 16 6'-1 3 16 6'-1 3 16 6'-1 3 16 6'-1 3 16 6'-1 3 16 6'-1 3 16 2'-10 9 16 10' 6' 12' 12' 12' 12' Figure 4.38 Cross Section View of Bridge #130085 span 2 This bridge has three main lanes, a 12 ft wide aux iliary lane, and two shoulders – one 6 ft wide and the other 10 ft, as shown in Fig. 4.38. The forensic study was conducted on spans 2 to 4, bays 1 to 5. This section of the bridge and the rest of the bridge, exhibited only l ongitudinal and some transverse cracking typical of deck panel construction. No spa lls or previous deck repairs were noticed on this bridge (Fig 4.39). Figure 4.39 Bridge # 130085 Prior to Deck Removal Bays covered on the study (1 – 5)

PAGE 119

106 4.9.2 Findings In this bridge was found that the panels were init ially placed over a 1 wide fiberboard strip, and a 2 in wide grout layer (Fig. 4.40), providing positive bearing support to the panels. But also was found that the fiberboard was later replaced by epoxy, as recommended in a previous research study [14] as a way to stop actual or to prevent future deterioration on deck panel bridges. The panel bearings were inspected in a large area of the deck, and it was found that some spots where the precast panel wasn’t long enou gh to reach the grout, so those panels were only supported initially by the fiberbo ard and now by the epoxy repair (Fig. 4.40 4.41). Figure 4.40 Bridge # 130085 Original Panel Bearing Detail 1 1/2” Fiberboard 1 1/2” Fiberboard 1” to 2” wide Grout layer Panel too short Panel Deck Top

PAGE 120

107 Figure 4.41 Bridge # 130085 Bearing Detail after Epoxy Repair Figure 4.42 Bridge # 130085 Panel Bearing Details Despite the use of positive panel bearing in this bridge, typical deterioration such as panel face separation, vertical cracking were detec ted (Fig. 4.43). The same cracking has been found in bridges with negative panel bearing. This proves that it is not related to the type of bearing used to support the panel. Panel Panel Deck Top Epoxy Epoxy 1” to 2” wide Grout layer See Fig. 4.41(b) See Fig. 4.41(a) Panel too short Grout Girder Edge Epoxy Girder Edge Epoxy 3” Panel Support 1 1/2” Panel Support a) b)

PAGE 121

108 Figure 4.43 Surface Longitudinal Crack Bridge # 130085 As in bridges (100398 and 100417) the relative goo d condition of this bridge can be linked to the fact that the this bridge was built u sing positive (grout) bearing for the precast panels, plus the fact that the fiberboard w as later replaced by epoxy, that could have helped to prevent deterioration in spots were the panel was supported only by the fiberboard (Fig. 4.42(b)). 4.10 Study Summary The following table (Table 4.9) summarizes all the different types of panel bearings found in the bridges covered on the forens ic study, and the link between the type of bearing and the condition of the deck. a) b) Panel Panel C.I.P C.I.P Vertical Crack Longitudinal Crack Partial Vertical Crack Separation Separation

PAGE 122

109 Table 4.9 Bearing and Deck Condition Summary Bridge # Study Date Built Bearing Type Comments 130078 & 130079 6/03 1981 Originally Fiberboard Bearing, Now Partial Epoxy Bearing Replacement. Major deficiencies found in spots were no epoxy replacement was conducted. Maybe because fiberboard was too thin to be removed and too thin the space for epoxy to be placed. 170140 1/04 1981 Originally Fiberboard Bearing, Now Full Epoxy Bearing Replacement. No major deficiencies found. Thick fiberboard therefore full epoxy bearing replacement. 130075 5/04 1981 Originally Fiberboard Bearing, Now Partial Epoxy Bearing Replacement. Major deficiencies found in spots were no epoxy replacement was conducted. 100415 6/04 1983 Fiberboard with panel overhang, to allow concrete under the panel supports. Major deficiencies found in spots were no concrete went under the panel. Or the panel was to short to provide the overhang. 100398 6/04 1984 Fiberboard Plus Grout Bearing. Bridge deck in good condition, only longitudinal and transverse cracking. 100417 7/04 1983 Fiberboard Plus Grout Bearing. Bridge deck in good condition, only longitudinal and transverse cracking. 130085 8/04 1981 Originally Fiberboard Plus Grout Bearing. Now Full Epoxy Bearing Replacement plus grout. Bridge deck in good condition. In some sports the panel was too short to reach the grout and was only supported by the fiberboard (now epoxy). Even though it was originally thought that all the deck panel bridges in FDOT Districts 1 and 7 were built supported only by fibe rboard (negative bearing), it was found that 4 out of 7 bridges in the study had some kind of positive panel bearing -grout or concrete-. And all the major deterioration is linke d to negative bearing created due to original design or construction inaccuracy. Table 4.10 provides a summary of the most significa nt findings in each bridge included in the study. These findings cover all the typical deck top deficiencies in deck panel bridges.

PAGE 123

110 Table 4.10 Forensic Study Summary Bridge # Study Date New Finding Verification 130078 & 130079 6/03 No Deck Surface Cracking. Longitudinal Deck Surface Cracking. Transverse Deck Surface Cracking. Additional Longitudinal Cracking. Deck Spalling and Delamination. 170140 1/04 Epoxy Panel Bearing Repair Condition. Longitu dinal Deck Surface Cracking. No Deck Surface Cracking. Additional Longitudinal Cracking. 130075 5/04 Epoxy Bearing Repair Cond. Cracking on M1 Repair Longitudinal Deck Surface Cracking. 100415 6/04 Different types of bearing (Positive Negative) Longitudinal Deck Surface Cracking. Deck Spalling and Delamination. 100398 6/04 Positive Panel Bearings Longitudinal Deck Surface Cracking. 100417 7/04 Longitudinal Deck Surface Cracking. Positive Panel Bearings 130085 8/04 Grout + Epoxy Bearing Repair. Longitudinal Deck Surface Cracking From the previous studies it was possible to obtain valuable information regarding the deterioration of deck panel bridges. This infor mation would have been very difficult to obtain only from lab tests.

PAGE 124

111 CHAPTER 5. FAILURE MECHANISM 5.1 Introduction From the information obtained from the analysis of failed bridges (Ch. 3) and forensic investigations, it is possible to develop a model that identifies the progression in deterioration that can potentially lead to localize d failure. This is described in the following sections. 5.2 Deck Failure Mechanism Model 5.2.1 Stage #1 Initial Condition The first stage is the initial condition of the br idge after being built. At this point we can identify two main groups of parameters that can affect long term performance of the bridge deck, these are: 1. Design Parameters – Relative easy to quantify a. Type of deck design Deck panel, or full depth cast in place concrete d eck b. Type of panel bearing Positive panel bearing (grout, concrete), Negative bearing (fiber board) c. Deck geometry Deck thickness, beam spacing, beam type, span leng th d. Deck material properties Concrete f’c, water cement ratio e. Traffic volume Average daily traffic (ADT), Average daily truck tr affic (ADTT), Actual and estimated future values. f. Lane placement Location of the wheel path relative to the edge of the girders.

PAGE 125

112 2. Construction quality parameters Very difficult to quantify – a. Deck thickness accuracy – it was found that in some cases the deck thickness was 10% smaller than the design value. An d this type of deck is very sensitive to reductions in the slab thickness. b. Top steel rebar cover When the top steel rebar is very close to the su rface of the deck chances of delamination and spalling ar e greater. c. Concrete properties Actual water cement content ratio, concrete curing process, f’c value before the bridge was opened to traffic, capacity of the concrete to resist the environment, actual f’c valu es. d. Real panel bearing condition; It was found that poor workmanship can significantly affect the real condition of the pane l bearing. 5.2.2 Stage #2 Longitudinal/ Transverse Cracking The second stage is the occurrence of longitudinal cracks over the edges of the girders. This is the most common type of cracking i n deck panel bridges and starts early. This crack is mainly the result of creep induced by prestressing forces in the panel, and the differential shrinkage between the cast in plac e concrete and the deck panel (Fig. 5.1). Following the formation of longitudinal cracking sp oradic transverse cracks can also develop due only to differential shrinkage. Positive Moment Tension Cast in place concrete shrinkage Delaminated Panel Face Prestress Creep and Shrinkage Induced Creep Shrinkage Figure 5.1 Deterioration Stage #2

PAGE 126

113 5.2.3 Stage #3. Shear Failure Longitudinal Cracking The third stage is the occurrence of additional lon gitudinal cracking on the deck surface, parallel to the cracking described previou sly. This additional cracking is the result of a shear failure of the cast in place conc rete. This type of cracking is the first sign of future deck deterioration. This type of cracking is related only to panels supported on negative bearing. When the deck panels are supported only by fiberbo ard and no strand extension has been provided in the panel face, all the shear loads on this region have to be supported by the cast in place section in top of th e panel edge, instead of being transferred by the entire composite section The shear failure that causes this crack occurs i n part due to the reduction of the shear capacity in the -already overstressedcast p lace concrete slab. This shear reduction is cause by the vertical cracking described in sect ion 5.2.3 It was also found that this reduction is affected by the shape of the vertical crack. Fig. 5.2 relates the shape of the crack to shear reduction. Reductions are higher whe n the crack extends towards the girder edge, and smaller when the crack extends towards th e center of the girder. Figure 5.2 Effect of Vertical Crack Shape in Shear Reduction High Reduction Low Reduction

PAGE 127

114 From the forensic examinations, two different type s of shear failures were identified. The first type of failure occurs when the cast in place concrete still has some capacity to transmit shear according to the shape o f the vertical crack (Fig. 5.2). This shear failure is manifested by the appearance of a diagonal cracking emanating from the corner of the precast panel and that propagates at an angle of less than 45 degrees. In most of the cases this crack reaches the surface ge nerating additional longitudinal cracks ( Fig. 5.3 (a)). Figure 5.3 Shear Failures for Different Degrees of Shear Reduc tion a) Low Re duction of C.I.P Shear Capacity Compressed Fiberboard Longitudinal cracks Load Panel Panel Diagonal Shear Crack Girder *Exaggerated Deformation b) High Reduction o f C.I.P Shear Capacity Longitudinal cracks Compressed Fiberboard Girder Rebar Delamination Panel Panel Load *Exaggerated Deformation

PAGE 128

115 The second type of shear failure occurs when due t o the vertical crack, the cast in place concrete has lost all its shear capacity. In this case, the shear is transmitted only by the steel rebar by dowel action. When the shear loa d is too high, the rebar acts like a “crowbar” in to concrete on top of the rebar, creat ing delamination in that area. (See Fig. 5.3(b)). Since this crack is related to shear, it is more l ikely to occur in the cases where the wheel loads are located close to the support of the panels; this is the load location that provides the highest shear value in the section of interest. L D J R Q D O & U D F N L Q J 9 H U W L F D O & U D F N $ G G L W L R Q D O & U D F N L Q J H O D P L Q D W H G 3 D Q H O ) D F H ) L E H U E R D U G % H D U L Q J / R Q J L W X G L Q D O & U D F N Figure 5.4 Deterioration Stage #3

PAGE 129

116 Figure 5.5 Examples of Deterioration Stage #3 5.2.4 Stage #4 First Spall After the occurrence of the second parallel crack, the concrete trapped between the two cracks is already internally cracked and st arts to crumble. As a result, a spall develops. At this stage, a new parameter is introdu ced, the effect of the rainwater forced inside the cracks by vehicles. Although this is dif ficult to quantify, bridge inspectors have observed this phenomenon over the years, Initial Longitudinal cracking (Over girder edges) Wheel path Additional longitudinal cracking Wheel path Additional Longitudinal cracking Typical Longitudinal cracking Precast deck panel No additional Cracking (Far from wheel path L)

PAGE 130

117 ) L U V W 6 S D O O 9 H U W L F D O & U D F N H O D P L Q D W H G 3 D Q H O ) D F H L D J R Q D O & U D F N L Q J ) L E H U E R D U G % H D U L Q J $ G G L W L R Q D O & U D F N L Q J 6 S D O O / R Q J L W X G L Q D O & U D F N Figure 5.6 Deterioration Stage #4 Initial longitudinal crack Initiation of a spall Shear failure crack

PAGE 131

118 5.2.5 Stage #5 Spall Increase, Then Spall Patch After the occurrence of the first spall in stage# 4 it will keep increasing in size basically due to the effect of the impact of the wh eels at the edges of the existing spall. The maximum size of the spall and additional deteri oration of the deck depends on how long it is left unrepaired. Usually the spalls are patched before they reach a relatively large size. For the majority of the cases, the repa ir consists of a temporary patch using a flexible material. 9 H U W L F D O & U D F N 6 S D O O 3 D W F K H O D P L Q D W H G 3 D Q H O ) D F H & U D F N L Q J Q F U H D V H!) D L O H G V H F W L R Q" 3 D Q H O 6 H W W O H P H Q W !0 R Y H P H Q W" !5 H G L V W U L E X W L R Q R I 6 W U H V H V V W R W K H D G M D F H Q W V H F W L R Q V" 6 S D O O 3 D W F K / R Q J L W X G L Q D O & U D F N Figure 5.7 Deterioration Stage #5

PAGE 132

119 5.2.6 Stage #6 New Spalling Plus Spall Increase Depending on all the different factors mentioned e arlier, new spalls can appear in the areas adjacent to the repaired spall after some time. Note that after the spall is created, the residual shear capacity of that region is almost zero, even after it has been patched, therefore, the shear that was to be suppor ted by that region now has to be redistributed to sections adjacent to the spall. Th is creates additional stresses in that region, and accelerates its deterioration generatin g new spalls. 1 H Z V S D O O L Q J D S S H D U V L Q D G M D F H Q W V H F W L R Q V 1 H Z 6 S D O O V / R Q J L W X G L Q D O & U D F N 6 S D O O 3 D W F K Figure 5.8 Deterioration Stage #6

PAGE 133

120 Figure 5.8 (Continued) 5.2.7 Stage #7 M1 Repair Generally after several patch and re-patch process es an M1 repair is done in the affected area. An M1 repair consists of the removal of all the patched, spalled, and unsound concrete section, and it is replaced by rep air material. To do this, the edges of the section to be removed are cut and the concrete inside is removed using a jack hammer. Usually the intent is to remove cast in pla ce concrete as close as possible to the deck panel surface. The opened surface is then clea ned and the removed concrete is replaced with different types of high strength epox y materials. And in some cases the fiberboard bearing is replaced by epoxy. The durability of the M1 repair and the condition of the deck area around it depends of the following parameters: 1. Time period between spall, spall repair, and M1 repair 2. Possible internal damage to the panel ind uced from previous stages 3. Possible internal damage to the panel induced from removal o f cast in place concrete 4. Bonding between the old concrete and the repair material 5. Stress redistribution to adjacent areas (after removal of the damaged cast in place concret e that deck region is no longer transferring shear to the supports, so that shear i s redistributed to the transverse edges of Patched spalls (Asphalt) New Spall Spall increase

PAGE 134

121 the repair) (6)Repair Material – (7) are the panel shear connectors embedded in the M1 Repair (8) How quickly was the repaired section ope ned to traffic. And finally the most important: (9) was the fiberboard removed and repla ced with non shrink epoxy? 5 H P R Y H G 7 R S & D V W L Q 3 O D F H & R Q F U H W H H F N 3 D Q H O 7 R S 6 X U I D F H 6 D Z & X W 6 K H D U 6 W U H V V & R Q F H Q W U D W L R Q U H J L R Q X H W R 6 W U H V V 5 H G L V W U L E X W L R Q & D V W L Q S O D F H F R Q F U H W H R Y H U J L U G H U W R S ) L E H U E R D U G % H D U L Q J 3 U H F D V W H F N 3 D Q H O 6 K H D U 6 W U H V V & R Q F H Q W U D W L R Q U H J L R Q X H W R 6 W U H V V 5 H G L V W U L E X W L R Q 5 H P R Y H G 7 R S & D V W L Q 3 O D F H & R Q F U H W H 3 D Q H O 6 H W W O H P H Q W G X H W R O L Y H O R D G V S U H V H Q W L Q W K H G H F N G X U L Q J U H S D L U S U R F H V V 5 H P R Y H G 7 R S & D V W L Q 3 O D F H & R Q F U H W H Figure 5.9 M1 Repair Procedure (Stage #7)

PAGE 135

122 0 H F N 5 H S D L U 6 K H D U 6 W U H V V & R Q F H Q W U D W L R Q U H J L R Q 0 H F N 5 H S D L U ) L E H U E R D U G % H D U L Q J 3 U H F D V W H F N 3 D Q H O 6 K H D U 6 W U H V V & R Q F H Q W U D W L R Q U H J L R Q 5 H P R Y H G 7 R S & D V W L Q 3 O D F H & R Q F U H W H Figure 5.10 Deterioration Stage #7 M1 Repair

PAGE 136

123 5.2.8 Stage #8 Shear Failure Cracking Adjacent to a n M1 Repair Assuming that the panel bearing wasn’t replaced by epoxy, or that it was replaced but due to construction problems there is no full s upport of the panel by the epoxy, the deck area adjacent to an M1 repair, starts the det erioration process again with the appearance of the additional parallel cracking – sh ear failure cracking described in (section 5.2.3). The parameters that affect the occ urrence of this additional deterioration case are the same mentioned for stages #3 and #7. L D J R Q D O & U D F N L Q J 9 H U W L F D O & U D F N $ G G L W L R Q D O & U D F N L Q J H O D P L Q D W H G 3 D Q H O ) D F H ) L E H U E R D U G % H D U L Q J 0 H F N 5 H S D L U 6 K H D U 6 W U H V V & R Q F H Q W U D W L R Q U H J L R Q Figure 5.11 Deterioration Stage #8

PAGE 137

124 5.2.9 Stage #9 Spalling Adjacent to an M1 Repair After the occurrence of additional longitudinal cr acking, a new spall develops and it follows the same mechanism mentioned in stage #4 6 S D O O Q F U H D F H V G X H W R W K H ( I I H F W R I W K H : K H H O V 9 H U W L F D O & U D F N H O D P L Q D W H G 3 D Q H O ) D F H 3 D Q H O 6 H W W O H P H Q W!' X H W R I D L O X U H R I W R S V H F W L R Q"' L D J R Q D O & U D F N L Q J 0 H F N 5 H S D L U 1 H Z 6 S D O O ) L E H U E R D U G % H D U L Q J 1 H Z 6 S D O O 3 U H F D V W H F N 3 D Q H O 5 H P R Y H G 7 R S & D V W L Q 3 O D F H & R Q F U H W H Figure 5.12 Deterioration Stage #9 M1 Repair Typical longitudinal crack Additional longitudinal crack Spall

PAGE 138

125 5.2.10 Stage #10 Cracking on M1 Repair and Adjace nt Spalling Increase When the spall mentioned in stage #9 is not patche d quickly, is very likely that the M1 repair can be fractured due to the constant impact of the wheels over it. Impact can also cause growth of adjacent spalls, and delam ination between the panel surface and the M1 repair. 9 H U W L F D O & U D F N 6 S D O O Q F U H D F H V G X H W R W K H ( I I H F W R I W K H : K H H O V H O D P L Q D W H G 3 D Q H O ) D F H 3 D Q H O 6 H W W O H P H Q W!' X H W R I D L O X U H R I W R S V H F W L R Q"& U D F N L Q J 5 H S D L U & U D F N L Q J % L J J H U 6 S D O O 6 L G H H O D P L Q D W L R Q 0 H F N 5 H S D L U ) L E H U E R D U G % H D U L Q J 1 H Z 6 S D O O 3 U H F D V W H F N 3 D Q H O 5 H P R Y H G 7 R S & D V W L Q 3 O D F H & R Q F U H W H Figure 5.13 Deterioration Stage #10 Typical Longitudinal Crack Typical Transverse Crack Additional Cracking M1 Repair Additional Cracking Spall

PAGE 139

126 5.2.11 Stage #11 Adjacent Spall Patch Right after the spall is noticed and depending on its size, it is patched. Usually a quick temporary repair is done in most cases using flexible material. Then what we have here is a fractured and delaminated M1 repair, plus a flexible material patch. The structural capacity of this deck section is very li mited generating redistribution of stresses to adjacent areas. Also in the case where no positi ve bearing has been provided to the panel, it will experience settlement every time the deck section is loaded due to the lack of stiff support. 6 S D O O 3 D W F K !) O H [ L E O H 0 D W H U L D O" H O D P L Q D W H G 3 D Q H O ) D F H 3 D Q H O 6 H W W O H P H Q W!' X H W R I D L O X U H R I W R S V H F W L R Q"& U D F N L Q J 5 H S D L U & U D F N L Q J % L J J H U 6 S D O O 6 L G H H O D P L Q D W L R Q 0 H F N 5 H S D L U ) L E H U E R D U G % H D U L Q J 6 S D O O 3 D W F K 3 U H F D V W H F N 3 D Q H O 5 H P R Y H G 7 R S & D V W L Q 3 O D F H & R Q F U H W H Figure 5.14 Deterioration Stage #11

PAGE 140

127 In some cases where the fiberboard is deteriorated or is missing, the panel can even touch the top of the girder every time it defl ects. Due to the dynamic nature of the wheel loads, panel movement can generate a pulse be tween the panel and the top of the girder, creating a hammering action and introducing new stresses in the panel. The parameters that affect this stage are (1) Time period between spall beginning, and spall patch, (2) patch material, (3) lack of bo nd between repair and panel top, (4) degree of disrepair of the precast panel and the M1 repair. 5.2.12 Stage #12 Additional Adjacent Spalling After stage #12, since the structural capacity of the section is not restored deterioration of the deck surface will continue and can generate new spalls adjacent to the previous patch, and next to the edges of the M1 rep air. At this point, the deterioration of the deck panel is accelerated by the effect of whee l loads applied over small chunks of concrete over the panel. This concentrates the whee l load over a very small region of the panel surface instead of distributing it over the e ntire deck section. As a result large stresses are generated in the panel which increases the probability of the occurrence of a punching shear failure of the panel.

PAGE 141

128 6 S D O O 3 D W F K !) O H [ L E O H 0 D W H U L D O" H O D P L Q D W H G 3 D Q H O ) D F H 3 D Q H O 6 H W W O H P H Q W!' X H W R I D L O X U H R I W R S V H F W L R Q" & U D F N L Q J 5 H S D L U & U D F N L Q J 6 S D O O 3 D W F K $ G G L W L R Q D O 6 S D O O L Q J 0 H F N 5 H S D L U ) L E H U E R D U G % H D U L Q J 6 S D O O 3 D W F K 3 U H F D V W H F N 3 D Q H O 5 H P R Y H G 7 R S & D V W L Q 3 O D F H & R Q F U H W H Figure 5.15 Deterioration Stage #12 Fractured M1 Repair Patch Additional spall Longitudinal Cracking

PAGE 142

129 5.2.13 Stage #13 Deck Localized Failure After experiencing all the previous deterioration stages, a localized failure is likely to occur. When this happens, the top steel b ar is the only structural element that prevents the occurrence of the failure of the entir e bay. 0 H F N 5 H S D L U 5 H S D L U & U D F N L Q J 6 S D O O 3 D W F K $ G G L W L R Q D O 6 S D O O L Q J Figure 5.16 Deterioration Stage #13

PAGE 143

130 Figure 5.17 Example of Deterioration Stage #13 5.3 Summary The deck deterioration model described in this chap ter can be summarized in the following failure tree. Figure 5.18 Deck Failure Tree Longitudinal Crack Parallel Longitudinal Crack (Shear) Spalling Spall increase (Stress redistribution) Spall Patching New Spalling M1 Repair M1 Walking Spalling Spall Patching Spall Increase & M1 deterioration Panel Failure M1 Repair Failed M1 repair Flexible Patch Spall

PAGE 144

131 CHAPTER 6. SUMMARY AND CONCLUSIONS 6.1 Summary The goal of this study was to identify the deterior ation process and failure mechanism of deck panel bridges in Florida. Typical deficiencies on deck panel bridges are des cribed in Chapter 2. This information was collected from FDOT’s deck inspecti on reports available since the construction of the bridges to date. Between 2000 and 2003, localized failures occurred in five panel bridges in Districts 1 and 7. Relevant information relating to these failures was collected and analyzed with the intent of identifying underlying trends. Forensic investigations were carried out on seven deck panel bridges scheduled for replacement during 2003-2004 and located within easy driving of the USF campus. The objective was to obtain first hand evidence of deterioration and to gain understanding of the degradation process. All the information collected was used to develop a deck failure model, and to identify the parameters that affect the structural behavior of the deck.

PAGE 145

132 6.2 Localized Failures 6.2.1 Failure Trend National Bridge Inventory deck condition rating w as found to be a poor indicator for predicting panel deck failures. All bridges tha t failed were rated between 5 (satisfactory) to 7 (good). Inspection records give a periodic snapshot on the condition of the bridge. Whereas biennial inspection data were g enerally unable to predict failure, monthly inspection records were far more successful in tracking problems that led to failure. Simplified calculations show that punching failure s could result at loads below the design wheel load. This assumed the cast-in-pla ce deck to provide no resistance and the panel to be supported on fiberboard with well d eveloped cracking along the transverse and longitudinal panel boundaries. The f ailure load was calculated to be around 15 kips (Table 3.4). Otherwise, failure load s were nearly four times higher. 6.2.2 Environmental Factors In four out of the five cases there was rainfall p rior to failure. The most severe rainfall preceded the last failure (1.1 in.). The e xact role of rainwater is not known. However, given that the concrete in the deck separa tes cleanly from the reinforcement, it probably adversely affects bond and degrades the co hesiveness of the cement paste. Thus, it is reasonable to conclude that rainfall accelera tes existing damage that can result in failure.

PAGE 146

133 6.2.3 Failure Location All failures occurred under the wheel loads applie d close to the face of the girders where initial longitudinal cracks developed. Also in all five cases, the failure occurred in the right lane, i.e. slow lane (Table 3.16). Failur e was generally at the edge or corner panels whose boundaries developed reflective longit udinal and transverse cracking. Table 6.1 Failure Summary Bridge # Year Built Age at Failure (yrs) ADT (ADTT) Failure Size Location in Panel Comment 170146 1981 19 34,000 (30%) 18 in x 24 in Edge or Corner? Failure at M1 repair 170086 1980 20 34,000 (30%) 36 in x 60 in Corner Support Patch repair 170085 1980 20 34,000 (30%) 18 in x 18 in Corner Failure adjacent to M1 repair 100332 1980 22 23,000 (8%) 48 in x 30 in Near corner Asphalt Patch 100332 1980 23 23,000 (8%) 24 in x 36 in Edge Failed M1 repair with flexible patch material National Bridge Inventory condition rating given in the bridge inspection prior to the deck failure 6.2.4 Bridge Characteristics All failures occurred in bridges where the deck wa s nominally 7 in. thick. No failures occurred in deck panel bridges with thicke r slabs. The ADTT varied between 830% (Table 6.1). Also it may be noted that the failures occurred in two twin bridges (NB and SB 170086, 170085), and in a bridge adjacent to these two (170146). It is very likely that these three bridges were built with similar defects by the same contractor. The other two cases also occurred in the same bridge (100332 span s 38 and 70).

PAGE 147

134 6.3 Forensic Investigation The most important conclusion is that the lack of positive panel bearing is clearly the main factor responsible for the occurrence of m ajor deck deterioration such as delamination, spalling, failing repairs, and in the worst case localized punch-thru deck failures. The lack of positive bearing can occur du e to two main reasons: 1. When the initial deck design indicates the use of o nly fiberboard as bearing material for the panels (Fig.1.3). 2. Or in the case where positive bearing is specified in the design, but due to construction deficiencies the panel may not be prop erly supported over stiff material. (Fig. 4.25 – 4.26). Not all the deck panel bridges in FDOT Districts 1 and 7 were built using negative panel bearing (panel supported by fiberboa rd only) as originally thought. Four out of seven bridges covered in the study where bui lt using positive panel bearings. The occurrence of deck surface longitudinal and tr ansverse cracking is not related to the type of panel bearing, positive or negative. This can be found in both types of bearing. This type of cracking has proven to remain stable through the years in bridges with positive panel bearing. Three common factors were found in all the deterior ated decks: 1. Lack of stiff support for the deck panels (negative bearing) 2. Wheel loads close to the supports (creating maximum shear stresses) 3. Vertical crack (due to creep and shrinkage) that re duces the shear capacity of the cast in place concrete.

PAGE 148

135 6.4 Deterioration Model Even though a deterioration model and a deck failu re mechanism for deck panel bridges was successfully developed in this study, i t is still difficult to accurately predict the future condition of different deck panel bridge s using the model because most of the factors that influence how fast the deterioration c an occur are very difficult to quantify. i.e. 1. Construction details: Type of panel bearing us ed (positive – negative), deck thickness (in most of the bridges real “as built” plans are not available). 2. Deck construction quality: Actual deck thickn ess, concrete quality (this can influence the formation of creep and shrinkag e cracks), concrete cover for top rebars, panel length (does it have the right length to be supported by the grout) for positive bearing construction. 3. How is the shear capacity of the deck affected after the appearance of creep and shrinkage cracks?( see Fig. 5.2). This is ver y difficult to quantify. 4. Specifications and quality of previous deck rep airs (epoxy panel bearing replacement, M1 repairs, spall patches. WHAT DOES THIS MEAN 6.5 Recommendations for Bridge Deck Replacement Pr ioritization As mentioned in 6.4, to develop an efficient deck replacement prioritization plan for deck panel bridges, based on the prediction of the future structural behavior is not feasible. The recommendation we can give after conducting th is study is to conduct a more in-depth search in order to quantify the two most i mportant factors that affects the condition of the bridge decks. These are the type o f panel bearing detail used in each bridge (positive – negative bearing) and deck thick ness. Even though searches were conducted at FDOT District 1 and 7 Bridge Maintenan ce office, these details were not found. These may be available in construction recor ds archive at another FDOT office.

PAGE 149

136 After identifying those parameters, this is the re commended approach for deck replacement prioritization: Table 6.2 Deck Replacement Prioritization Approach Priority Initial Bearing detail (Fig. 1.3) Epoxy bearing replacement Actual deck condition Comments High Negative No Heavy deterioration Replace the entire deck Negative Not in the entire deck, or too thin < “(Fig. 4.20) Heavy deterioration Replace the entire deck Negative Yes and more than ” thick (Fig. 4.16 d) Deterioration in isolated locations only Replace the entire deck Positive No Deterioration in isolated locations only (Fig 4.40) If feasible, replace only deteriorated deck sections covering entire bay Negative Yes and more than ” thick (Fig. 4.16 d) Only typical cracking Replacement may not be required, unless new deterioration appears Low Positive Yes and more than ” thick (Fig. 4.41) Only typical cracking Replacement may not be required, Assume also typical longitudinal and transverse c racking. 6.6 Future Work In order to fully validate the deterioration model developed in this study, each deterioration stage should be analyzed in detail us ing finite element analysis. The objective is to obtain additional information about the model that could not be obtained from forensic examination.

PAGE 150

137 The numerical analysis of the model will be conduc ted by Dr. Niranjan Pai, Postdoctoral Fellow at the University of South Flor ida, as part of an FDOT research project. This will lead to the development of a dec k replacement prioritization scheme for the panel deck bridges remaining in FDOT Districts 1 and 7.

PAGE 151

138 REFERENCES [1] AASHTO (2001). LRFD Bridge Design Specifications -U.S. Units, 17th Edition, Washington, D.C., Section 3.6.1.2.5. [2] ACI 318-2002, American Concrete Institute, Farmingt on Hills, MI. [3] American Association of State Highway and Transpor tation Officials, Standard Specifications for Highway Bridges, 12th Edition, 1977. [4] Bebe B. (2000). “ Hole opens up in bridge on I-75 at State Road 72 ”. Article dated Nov. 28, Sarasota Herald Tribune. [5] Deshmukh, Ganesh, “Replacement Prioritization of Pr ecast Deck Panel Bridges in Florida” Master Thesis, Department of Civil and Env ironmental Engineering, University of South Florida, April 2004. [6] Englert, J. (2000). Emergency Response Letter to Mr Pepe Garcia dated Feb. 16, E. C. Driver and Associates, Tampa, FL. [7] Englert, J. (2000). Letter to Mr. Hamid Kashani on Bridge No 170086 dated Nov. 29, p.8. E. C. Driver and Associates, Tampa, FL. [8] Englert, J. (2000). Emergency Response Letter to Mr Pepe Garcia dated Dec. 21, E. C. Driver and Associates, Tampa, FL. [9] Fagundo, F.E., Callis, E.G., Hays, C.O., “Study of Cracking of I-75 Composite Deck Bridge Over Peace River,” U49F, Department of Civil Engineering, Engineering and Industrial Experiment Station, Univ ersity of Florida, Gainesville, August 1982, pp. 202. [10] Fagundo, F.E., Hays, C.O., and Richardson, J.M., “S tudy of Composite Bridge Decks in Florida,” Final Report No.U69F, Department of Civil Engineering, Engineering and Industrial Experiment Station, Univ ersity of Florida, Gainesville, July 1983.

PAGE 152

139 [11] Fagundo, F.E., Hays, C.O., Tabatabhai, H, H., and S oongswang, K., “The Effect of Crack Development and Propagation on the Mainten ance Requirements of Precast Deck Bridges,” Final Report No.U88, Departm ent of Civil Engineering, Engineering and Industrial Experiment Station, Univ ersity of Florida, Gainesville, July 1984, pp 152. [12] Florida Department of Transportation, Bridge Inspec tors Field Guide, Structural Elements, March 13, 2000, pp.133. [13] Graddy, J.C., Kim, J., Whitt, J.H., Burns, N.H., a nd Klinger, R.E., “Punching Shear Behavior of Bridge Decks Under Fatigue Lo ading,” ACI Structural Journal, V.99, No.3, May-June 2002, pp.257-266. [14] Hays, C.O., and Tabatabhai, H., “Summary of Researc h on Florida Precast Panel Bridges,” Structures and Materials Research report No.85-1, Department of Civil Engineering, Engineering and Industrial Experiment Station, University of Florida, Gainesville, August 1985, pp.77. [15] Klinger, R.E., and Bieschke, L.A., “The Effect of Transverse Strand Extensions on the Behavior of Precast Prestressed Panel B ridges,” Research Report No. 303-1F, Center for Transportation Research, the Un iversity of Texas, Austin, June 1982. [16] Merill, B.D., “Texas Use of Precast Concrete Stay i n Place Forms for Bridge Decks”, Texas DOT, Bridge Conference, 2002, pp.1-19 [17] NOAA's Forecast Systems Laboratory, “Hourly/Daily R ain Data” http://precip.fsl.noaa.gov/hourly_precip.html [18] O’brien Eugene & Keogh Damien (1999). Bridge deck analysis. London: E & FN Spon. [19] Sloan, J. (2003). Small Hole Paves Commuters' Way To A Traffic Jam" Article dated Sep. 9, Tampa Tribune. [20] Sloan, J. (2003). “ Time Catches Up With Expressway ”. Article dated Sep. 10, Tampa Tribune. [21] Staff Reporter. (2000). “FDOT will have I-75 hole f ix soon”. Article dated Feb. 15, Sarasota Herald Tribune. [22] Staff Reporter. (2000). “Falling bridges I-75’s cru mbling overpasses need to be made safer”. Article dated Nov. 29, Sarasota Heral d Tribune.

PAGE 153

140 [23] Staff Reporter. (2000). “ Troubled bridges I-75 overpasses warrant emergency attention ”. Article dated Dec. 23, Sarasota Herald Tribune. [24] Staff Reporter. (2000). “State orders repair of I-7 5 overpasses”. Article dated Dec. 23, Sarasota Herald Tribune. [25] Sullivan, J. (2000). “Fallen asphalt closes lanes: a large pothole has developed again in the I-75 overpass at Bee Ridge.”. Article dated Feb. 13, 2000. Sarasota Herald Tribune. [26] Sullivan, J. (2000). “3rd hole opens on Clark Road Overpass.”. Article dated Dec. 21, Sarasota Herald Tribune. [27] U.S. Department of Transportation, Federal Highway Administration. (2003). National Bridge Inventory.

PAGE 154

141 APPENDICES

PAGE 155

142 APPENDIX A: PUNCHING SHEAR CALCULATIONS ASSUMPTIONS 1. Failure plane assumed to be linear. 2. Failure plane unaffected by the presence of higher compressive strength of the precast deck. 3. Prestressed panel assumed to be reinforced concrete for shear calculations. NOTE: Tire Contact area: in l in b 10 20 = = Determination of shear strength: As per ACI 11.12.2 .1 Shear strength of concrete Vc is smallest of the following Vc1 = d b fo c c' 4 2 +b (Equation 11-33) Vc2 = d b f b do c o s' 2 +a (Equation 11-34) Vc3 = 4 d b fo c' (Equat ion 11-35) Where, bo = punching shear area at distance d/2 from the fac e of the loaded area c = ratio of long side to short side of the concentr ated area c = 2 10 20 = s = 20 (corner) s = 40 (center)

PAGE 156

143 Appendix A, (Continued) Case 1 Full Composite Action (Corner) Figure A.1 Shear Failure Detail (Corner) Cast in place slab psi 3000 'fCIP c = in 4 dCIP = n + + n + = 2 20 CIP CIPd l d b b n + + n + = 2 4 10 2 4 200b in b 340 =

PAGE 157

144 Appendix A, (Continued) () () () () () () 35) 11 (Eq. 8. 29 4 34 3000 4 4 34) 11 (Eq. 9. 49 4 34 3000 2 34 4 40 2 33) 11 (Eq. 8. 29 4 34 3000 2 4 2 4 23 3 0 3 2 2 0 0 2 1 1 0 1kip V V d b f V kip V V d b f b d V kip V V d b f Vc c CIP CIP c c c c CIP CIP c CIP s c c c CIP CIP c c c= = = = n + = n n + = = n + = n n + =a b S hear strength of cast in place slab kip VCIP c8. 29_ = -Precast deck panel psi 5000 'fpan c = in 56.2 dpan = in b b d d l d d b bpan CIP pan CIP56. 36 2 56.2 2 4 10 2 56.2 2 4 20 2 2 2 20 0 0= n + n + + n + n + = n n + n + + n n + n + = () () () () () () 35) 11 (Eq. 5. 26 56.2 56. 36 5000 4 4 34) 11 (Eq. 7. 31 56.2 56. 36 5000 2 56. 36 56.2 40 2 33) 11 (Eq. 5. 26 56.2 56. 36 5000 2 4 2 4 23 3 0 3 2 2 0 0 2 1 1 0 1kip V V d b f V kip V V d b f b d V kip V V d b f Vc c pan pan c c c c pan pan c pan s c c c pan pan c c c= = = = n + = n n + = = n + = n n + =a b

PAGE 158

145 Appendix A, (Continued) Shea r strength of pre-cast panel kip Vpanel c5. 26_ = Total composite deck punching shear st rength CIP c panel c compV V V + = kip Vcomp3. 56 = Case 2. No Composite Action (Corner) Figure A.2 Shear Failure Detail (No Composite Corner) 2c = b psi 5000 'fpan c = Min in dpan06.2 = 20s = a (corner)

PAGE 159

146 Appendix A, (Continued) in b b d b d l bpan pan06. 32 2 06.2 20 2 06.3 10 2 20 0 0= n + + n + = n n + + n n + = () () () () () () 35) 11 (Eq. 7. 18 06.2 06. 32 5000 4 4 34) 11 (Eq. 3. 15 06.3 06. 32 5000 2 06. 32 06.2 20 2 33) 11 (Eq. 7. 18 06.2 06. 32 5000 2 4 2 4 23 3 0 3 2 2 0 0 2 1 1 0 1kip V V d b f V kip V psi V d b f b d V kip V V d b f Vc c pan pan c c c c pan pan c pan s c c c pan pan c c c= = = = n + = n n + = = n + = n n + =a b Shear strength of pre-cast panel kip Vrib c3. 15_ =

PAGE 160

147 Appendix A, (Continued) Case 3 Full Composite Action (Edge) Figure A.3 Shear Failure Detail (Edge) Cast in place slab psi 3000 'fCIP c = in 4 dCIP = n + + n + = 2 2 2 20CIP CIPd l d b b n + + n + = 2 4 2 10 2 4 20 20b in b 580 =

PAGE 161

148 Appendix A, (Continued) () () () () () () 35) 11 (Eq. 8. 50 4 58 3000 4 4 34) 11 (Eq. 4. 60 4 58 3000 2 58 4 40 2 33) 11 (Eq. 8. 50 4 58 3000 2 4 2 4 23 3 0 3 2 2 0 0 2 1 1 0 1kip V V d b f V kip V V d b f b d V kip V V d b f Vc c CIP CIP c c c c CIP CIP c CIP s c c c CIP CIP c c c= = = = n + = n n + = = n + = n n + =a b S hear strength of cast in place slab kip VCIP c1. 49_ = -Precast deck panel psi 5000 'fpan c = in 56.2 dpan = in b b d d l d d b bpan CIP pan CIP12. 63 2 56.2 2 2 4 2 10 2 56.2 2 4 20 2 2 2 2 2 2 2 20 0 0= n + n + + n + n + = n n + n + + n n + n + =

PAGE 162

149 Appendix A, (Continued) () () () () () () 35) 11 (Eq. 7. 45 56.2 12. 63 5000 4 4 34) 11 (Eq. 4. 41 56.2 12. 63 5000 2 12. 63 56.2 40 2 33) 11 (Eq. 7. 45 56.2 12. 63 5000 2 4 2 4 23 3 0 3 2 2 0 0 2 1 1 0 1kip V V d b f V kip V V d b f b d V kip V V d b f Vc c pan pan c c c c pan pan c pan s c c c pan pan c c c= = = = n + = n n + = = n + = n n + =a b Shea r strength of pre-cast panel kip Vpanel c4. 41_ = Total composite deck punching shear st rength CIP c panel c compV V V + = kip Vcomp2. 92 =

PAGE 163

150 Appendix A, (Continued) Case 4. No Composite Action (Edge) Figure A.4 Shear Failure Detail (No Composite Edge) 2c = b psi 5000 'fpan c = Min in dpan06.2 = 20s = a (corner) in b b d b d l bpan pan12. 54 2 06.2 20 2 2 06.3 2 10 2 2 2 20 0 0= n + + n + = n n + + n n + =

PAGE 164

151 Appendix A, (Continued) () () () () () () 35) 11 (Eq. 5. 31 06.2 06. 32 5000 4 4 34) 11 (Eq. 7. 21 06.3 06. 32 5000 2 06. 32 06.2 20 2 33) 11 (Eq. 5. 31 06.2 06. 32 5000 2 4 2 4 23 3 0 3 2 2 0 0 2 1 1 0 1kip V V d b f V kip V psi V d b f b d V kip V V d b f Vc c pan pan c c c c pan pan c pan s c c c pan pan c c c= = = = n + = n n + = = n + = n n + =a b Shear strength of pre-cast panel kip Vrib c7. 21_ =

PAGE 165

152 APPENDIX B: CORE EVALUATION I-75 OVER MOCCASIN WALL OW RD Location of retrieved deck sections used for coring 99'-9" 99'-9" 45'-3" Span 4 Span 31 2 3 Span 25 4 45'-3" Span 16 Figure B.1 I-75NB over Moccasin Wallow Bridge Deck sections # 2 and # 6 were rejected for the study due to heavy damage incurred during the removal process

PAGE 166

153 Appendix B, (Continued) Table B.1 Core Details of Deck Section # 1 Core Details of Deck Section # 1 Delaminated repair joint Cracking, delamination and spalling Wheel Paths M1 spall repair Girder1-1 1-3 1-4 1-2 Girder CORE DESCRIPTION This section intercepts a crack and and an M1 repair. It is located about 1 ft from the supported edge as shown in the sketch. The core was extracted as two pieces with the panel completely separated from the cast in place slab. The M1 repair was completely debonded from the cast-in-place slab. Signs of water going thru the interface of the M1 repair and signs of rebar corrosion were also present. This core was taken from the M1 repair section as indicated in the sketch above. The core was extracted in two pieces with the M1 repair completely Debonded at its interface with the cast in place concrete. The total thickness was 7 5/8 in with the M1 repair being 3 5/8 in, the cast-in-place slab 1 in and the prestressed panel 3 in. Separation Panel C.I.P 1-1 M1Repair 1-2 C.I.P P anel M1 Repair De-bonding

PAGE 167

154 Appendix B, (Continued) Table B.1 (Continued) CORE DESCRIPTION This core was taken between two parallel cracks close to the edge of the panel support. There was no debonding at the interface between the CIP slab and the panel. However, there was diagonal separation at the top (1/2 in at one end to 2 in at the other end ). The concrete in this section was four small pieces and signs of water infiltrating the crumbled concrete were present. Total core thickness 7 5/8 in The core was adjacent to 1-3 but was closer to the support. In this case, there was also no separation at the panel/CIP interface and a diagonal crack with the same slope (1/2 in at one end and 2 in. at the other end was present). However, the top segment was cracked but not in four pieces. Signs of water infiltrating the diagonal crack were found. Total core thickness 7 5/8 in 1-4 Panel De-bonded Top section C.I.P Diagonal Crack 1-3 Panel Crumbled Concrete C.I.P Diagonal Crack

PAGE 168

155 Appendix B, (Continued) Table B.2 Core Details of Deck Section # 3 Core Details of Deck Section # 3 Girder Wheel Paths M1 Repair Girder3-1Panel Midspan (No deficiencies) CORE DESCRIPTION This core was taken at the middle of the panel where there was no deterioration. No deterioration was detected in this core. The bond between the CIP slab and the precast panel was excellent. Total core height 7 1/2 in Panel C.I.P 3-1

PAGE 169

156 Appendix B, (Continued) Table B.3 Core Details of Deck Section # 4 Core Details of Deck Section # 4 Transverse crack over transverse panel joint Transverse and longitudinal cracks intersection4-2No Longitudinal cracking (Deck top over panel edge) Girder Random transverse crack4-4 4-3 4-1 Wheel Paths Girder CORE DESCRIPTION This core was taken at the intersection of a longitudinal and transverse crack as shown in the sketch above. The longitudinal crack extends all the way from the panel through the CIP slab. The transverse crack extends 2 in. below the top slab along the transverse panel joint. Despite the cracking, the concrete between the cracks is not in small pieces amd there are no signs of spalling or delamination on the deck surface. Total core height is 8 in This core was taken over a transverse joint. A hairline crack extends all the way from the top surface to the transverse panel joint The bottom part of the core (panel joint), was damaged during the extraction process. 4-1 Transverse Crack Transverse Joint Panel Reflective Crack Longitudinal Crack 4-2 Transverse Crack Reflective Crack Panel Transverse joint Deck Surface

PAGE 170

157 Appendix B, (Continued) Table B.3 (Continued) CORE DESCRIPTION This core was taken over the edge of the panel where there was no surface cracking. There is separation between the vertical face of the panel and the CIP slab possibly due to creep and shrinkage. There is also a hairline vertical crack emanating from the corner that extends 1 in upwards into the CIP slab. Total core height 8 1/2 in This core was taken at a section near the edge of the panel where there was a transverse crack. The crack extended all the way from the prestressing strand to the steel surface. The surface crack was an isolated crack with a length of less than 2 ft. Total core height 8 1/2 in 4-4 Panel Vertical Crack Panel C.I.P Vertical Crack No Crack Face Separation 4-3 Pre stressing Strand

PAGE 171

158 Appendix B, (Continued) Table B.4 Core Details of Deck Section # 5 Core Details of Deck Section # 5 5-5 M1 Repair Joint M1 Repair Joint5-6 5-4 5-2 Spall Patch Girder Typical longitudinal crack Wheel Paths Girder5-7 5-3 M1 Repair Joint CORE DESCRIPTION This core was rejected due to heavy damage incurred during extraction. This core was taken from an epoxy repaired region between near two parallel cracks where it intercepted one of them. There was excellent bonding between the epoxy material and the CIP slab. The core was broken 2 in from the top during the extraction process. This core has the mark of a shear connector embedded between the panel and the cast in place concrete. Total core height 8 in 5-2 Epoxy Patch C.I.P Panel Shear connector mark 5-1

PAGE 172

159 Appendix B, (Continued) Table B.4 (Continued) CORE DESCRIPTION This core was taken from the edge of an M1 repair. There was good bond between the M1 repair and existing concrete. This core was also broken in half during the extraction process. Total core height 7 in. This core was taken at a transverse joint for an M1 repair. There was no bond between the M1 repair and the existing concrete. The concrete adjacent to the vertical repair joint was crumbled, and had signs of water infiltration. The panel vertical face easily separated from the adjacent cast in place concrete. Not all the pieces of the core could be retrieved. This core was taken adjacent to 5-4 but some distance away from the edge. It shows de-bonding between the M1 repair and the adjacent cast in place concrete. This core broke at the vertical edge of the M1 repair during the extraction process. Total core height 8 in. Good bond Panel C.I.P 5-3 M1Repair 5-5 M1Repair Panel De-bonding 5-4 M1Repair Crumbled Concrete Panel Vertical Face C.I.P

PAGE 173

160 Appendix B, (Continued) Table B.4 (Continued) CORE DESCRIPTION This core was taken adjacent to 5-2 near the edge of the panel where it crossed a longitudinal crack. It shows a vertical reflective crack extending over the entire depth of the core separating the precast panel from the CIP slab. There are signs of water and dust infiltration.. The top surface includes a partial epoxy patch which is bonded to the CIP slab. The penetration of the epoxy penetrating below has prevented the top surface from crumbling. Total core height 8 in. This core was taken along a longitudinal crack located at the opposite supported edge of the panel from the previous cores. The same vertical crack detected in core 5-6 occurred but there was no additional damage. There were signs of water penetration in the CIP slab. Total core height 8 1/4 in. C.I.P 5-6 Panel Vertical Crack Epoxy Patch Crumbled Concrete Separation At Panel Face 5-7 Panel C.I.P Vertical Crack Longitudinal Crack C.I.P Diagonal Crack