Citation
Performance of thin bonded overlays on asphalt and concrete surfaces

Material Information

Title:
Performance of thin bonded overlays on asphalt and concrete surfaces
Creator:
Young, Logan Michael
Publication Date:
Language:
English
Physical Description:
xiv, 276 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Pavements -- Overlays -- Evaluation ( lcsh )
Asphalt ( lcsh )
Concrete ( lcsh )
Asphalt ( fast )
Concrete ( fast )
Pavements -- Overlays -- Evaluation ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 271-276).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Logan Michael Young.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
761024997 ( OCLC )
ocn761024997
Classification:
LD1193.E53 2011m Y68 ( lcc )

Full Text
PERFORMANCE OF THIN BONDED OVERLAYS
ON ASPHALT AND CONCRETE SURFACES
by
Logan Michael Young
B.S., University of Colorado Denver, 2010
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
2011


Copyright 2011
by Logan Michael Youn
All rights reserved.


This thesis for the Master of Science
degree by
Logan M. Young
has been approved
by


Young, Logan Michael (MS, Structural, Civil Engineering Department)
Performance of Thin Bonded Overlays on Asphalt and Concrete Surfaces
Thesis directed by Associate Professor Stephan A. Durham, Ph.D.
ABSTRACT
Protection of bridge decks from corrosion by chloride ions has been an important
topic since the first steel reinforced concrete bridge decks were placed. The Colorado
Department of Transportation has recently begun to experiment with thin bonded
overlays in an effort to protect their bridge decks from deterioration by chemical
attack.
This thesis is the evaluation of two thin bonded epoxy overlays, SafeLane
marketed by Cargill, and Flexogrid developed by PolyCarb. SafeLane is advertised as
an anti-skid/anti-icing overlay that stores deicing chemicals for release during winter
events. Flexogrid is an anti-skid overlay. These two products were compared on the
basis of physical properties, including: mean texture depth, surface friction, bond
strength, ability to stop chloride intrusion, and anti-icing properties, as well as traffic
safety, and cost.
Results have shown that both overlays work as intended. Mean texture depth
and friction testing have shown that they both provide a durable wearing surface with
good traction. All the SafeLane bond tests exceeded the required 250 psi (1.72 MPa).
Flexogrid also had high bond strengths, but had varied failure modes. Permeability


and chloride testing of the underlying concrete decks indicated that both overlays
work well to protect bridge decks from chloride ingress. Permeability was high, but
the chloride counts did not increase with age. The anti-icing property of SafeLane is
effective when pre-charged with deicing chemicals, but seems ineffective with excess
moisture. The three sites evaluated in this study indicate a reduction in crashes, but
further study is required. Final results indicate that CDOT should consider using thin
bonded overlays for future rehabilitation and new construction; however, additional
long-term testing should be performed to evaluate the durability of these overlay
systems.
This abstract accurately represents the content of the candidate's thesis. I recommend
it's publication.
Signed
Stephan A. Durham


ACKNOWLEDGEMENT
Behind any large piece of research there are countless others who provide
information, guidance, and support. Many people contributed to this research and
they have all provided valuable input.
First and foremost, the author wishes to sincerely thank Dr. Stephan A. Durham.
Without him, this thesis wmuld not have been possible. It was his proposal that
allow ed this thesis to exist, and he has spent countless hours providing assistance.
Thank you to the committee members Dr. Nien-Yin Chang, and Dr. Brian Brady for
taking their time to review this work. As well as to Dr. Bruce Janson for his help with
traffic related analysis.
The author also w ishes to sincerely thank the Colorado Department of Transportation
for the opportunity' this thesis provided, as well as its employees: Aziz Khan, Ali
Harajili, and Patrick Kropp for their research support. Skip Outcalt and Dave Weld
for their testing expertise. David Bourget and his team for traffic data support. Kenny
Neb and the Region 4 Maintenance Crew out of Hudson, CO for their installation
support. As well as Scott Roalfs and Tim Dunn for their assistance in the CDOT
laboratories.
A special thank you to Jim Anderson and Anthony Hensley from Cargill, Inc. for
their knowdedgeable product support and assistance with testing.


Figures
Tables
TABLE OF CONTENTS
x
xiii
Chapter:
1. Introduction...................................................................1
2. Background Research.....................................................4
2.1 Summary of Previous Work...............................................5
2.1.1 Colorado Department of Transportation..................................5
2.1.2 University of Colorado Denver...........................................9
3. Literature Review......................
3.1 Why Thin Bonded Overlays?..........
3.2 Polymer Concrete Overlays..........
3.2.1 Premixed Polymer Concrete Overlays
3.2.2 Broom and Screed Overlays..........
3.3 Flydraulic Concrete Overlays..........
3.3.1 Low-Slump Dense Concrete...........
3.3.2 Silica Fume Concrete...............
3.3.3 Latex Modified Concrete............
3.3.4 Internally Sealed Concrete Overlay....
3.3.5 Conductive Concrete Overlays.......
3.4 Asphalt Concrete...................
3.5 Testing of Overlays................
3.5.1 Texture Depth......................
3.5.2 Surface Friction...................
3.5.3 Bond Strength......................
3.5.4 Permeability.......................
17
17
19
21
24
33
34
36
37
40
43
44
48
48
52
64
72
vi


3.5.5 Chloride Content........................................................79
4. Problem Statement.......................................................83
5. Experimental Plan.......................................................86
5.1 Project Locations.......................................................86
5.1.1 1-76 and Weld County Road 53............................................86
5.1.2 Parker Road and 1-225...................................................90
5.1.3 1-25 and 1-225..........................................................92
5.2 Testing of Thin Bonded Overlays.........................................95
5.2.1 Sand Patch Test Method..................................................96
5.2.2 Friction Skid Test Method...............................................98
5.2.3 Bond Test Method.......................................................102
5.2.4 Rapid Chloride Permeability Test Method................................107
5.2.5 Chloride Content Test Method..........................................111
6. Overlay Application Process...........................................115
6.1 SafeLane Application..................................................115
6.1.1 Surface Preparation for SafeLane.......................................115
6.1.2 Overlay Application...................................................118
6.2 Flexogrid Application.................................................122
6.2.1 Surface Preparation....................................................124
6.2.3 Flexogrid Installation.................................................125
7. Experimental Results...................................................129
7.0. 1 1-76 / WCR-53..........................................................129
7.0. 2 Parker Rd. /1-225..................................................... 130
7.0. 3 1-25 /1-225............................................................132
7.1 Problems During Installation and Testing...............................134
7.2 Sand Patch Test.......................................................138
7.3 Friction Tests.........................................................147
vii


7.4 Bond Tests..........................................................154
7.5 Rapid Chloride Permeability Test....................................162
7.6 Chloride Content....................................................169
7.7 Anti-Icing Properties...............................................178
8. Traffic Safety......................................................188
8.1 1-76 / WCR-53.......................................................190
8.2 Parker Rd. /1-225.................................................. 192
8.3 1-25 /1-225.........................................................196
8.4 Comparison of Site Crash Data.......................................199
9. Cost Analysis.......................................................201
9.1 1-76/WCR-53.........................................................204
9.2 Parker Rd. /1-225...................................................206
9.3 1-25 / 1-225........................................................208
9.4 Site Cost Comparison................................................211
10. Conclusions.........................................................213
10.1 Texture Depth.......................................................214
10.2 Surface Friction....................................................214
10.3 Bond Strength.......................................................215
10.4 Permeability........................................................215
10.5 Chloride Content....................................................216
10.6 Anti-Icing Properties...............................................216
10.7 Traffic Safety......................................................217
10.8 Installation Cost...................................................217
10.9 Recommendations.....................................................218
viii


Appendix:
A.......................................................................219
B.......................................................................223
C.......................................................................228
D.......................................................................230
E.......................................................................244
F.......................................................................250
G.......................................................................254
Bibliography............................................................271
IX


10
14
15
46
47
47
49
53
54
56
57
59
60
61
63
64
65
66
67
74
77
81
87
88
88
89
90
91
92
93
93
94
95
97
FIGURES
States That Responded to the DOT Survey...........
Bond Test Setup for Asphalt and Concrete..........
Bond Test Results.................................
Membrane Layers...................................
Division of Preformed Waterproofing Systems.......
Division of Liquid Waterproofing Systems..........
Surface Texture and Surface Characteristics.......
Locked-Wheel Friction Test System.................
Locked-Wheel Test Tires...........................
British Pendulum Tester...........................
Dynamic Friction Tester...........................
Friction Numbers of Polymer Overlays..............
Skid Results from WSDOT Epoxy Overlays............
Skid Results from WSDOT MMA Overlays..............
Mn/'DOT Skid Numbers for SafeLane.................
Wearing of SafeLane After 26 Months...............
Typical Bond Test Setup According to ASTM Cl583.
Bond Test Failure Modes According to ASTM Cl583
Bond Strength of Different Polymer Overlays.......
RCPT Basic Setup..................................
Permeability of Polymer Overlays Over Time........
Typical Core Test Locations.......................
1-76 / WCR-53 in Relation to Denver, CO...........
1-76 / WCR-53in Relation to Keenesburg, CO........
Aerial View of 1-76 / WCR-53......................
Weather Station at 1-76 / WCR-53..................
FP2000 Deck Sensor................................
Parker Rd. /1-225 in Relation to Denver...........
Aerial View of Parker Rd. /1-225..................
1-25 /1-225 in Relation to Denver, CO.............
Aerial View of 1-25/ 1-225........................
1-25 /1-225 Spray System..........................
Sensor Masts on 1-25 /1-225.......................
Sand Patch Test Process...........................
x


.98
100
101
102
103
105
106
107
108
109
110
112
114
116
117
119
120
121
123
124
126
126
127
135
136
137
141
145
147
148
150
153
155
156
157
158
159
160
163
Locked-Wheel Friction Test System......................
British Pendulum Tester................................
British Pendulum Test Process..........................
James Bond 007 Tester..................................
Bond Test Epoxy Application............................
Bond Test Disk Application.............................
Bond Test in Progress..................................
Failure of the Asphalt Under SafeLane Overlay..........
Samples in Flooded Desiccation Chamber Under Vacuum
RCPT Preparation.......................................
RCPT Leak Testing......................................
Chloride Test Sample Preparation.......................
Chloride Content Titration.............................
SafeLane Application Flow Chart........................
SafeLane Surface Preparation...........................
Epoxy Mixing and Spreading Devices.....................
SafeLane Aggregate Being Placed by Hand................
SafeLane Overlay Cleaning..............................
Flexogrid Application Flow Chart.......................
Flexogrid Installation Preparation.....................
Flexogrid Machine......................................
Flexogrid Machine Aggregate Spreader...................
Flexogrid Aggregate Being Spread by Hand...............
Crack Filling with SafeLane............................
SafeLane Cracking......................................
Skid Resistance Testing on 1-76 / WCR-53...............
Mean Texture Depth of 1-76 WCR-53....................
Mean Texture Depth of Parker Rd. /1-225................
Mean Texture Depth of 1-25 /1-225 .....................
1-76 / WCR-53 Skid Resistance Numbers Over Time........
Parker Rd. /1-225 Skid Resistance Numbers Over Time...
1-25 /1-225 Skid Resistance Numbers Over Time..........
1-76 / WCR-53 Bond Strengths Compared to 250 PSI.......
Asphalt Bond Test Failure Modes........................
Parker Rd. / 1-225 Bond Strengths Compared to 250 PSI..
Test Cores From Parker Road / 1-225 Bond Testing.......
1-25 /1-225 Bond Strengths Compared to 250 PSI.........
Test Cores from 1-25 /1-225 Bond Testing...............
1-76 / WCR-53 Permeability Classification..............
xi


165
166
168
171
172
174
175
176
177
181
182
183
184
186
191
193
194
195
197
198
199
219
223
228
230
234
239
244
245
247
250
254
259
263
269
Parker Rd. / 1-225 Permeability Classification..........
Cracked Permeability Sample R7..........................
1-25 / 1-225 Permeability Classification................
March 2011 1-76 / WCR-53 Chloride Content...............
March 2011 1-76 / WCR-53 Chloride vs. Penneability......
March 2011 Parker Rd. / 1-225 Chloride Content..........
March 2011 Parker Rd. /1-225 Chloride vs. Permeability....
March 2011 1-25 /1-225 Chloride Content.............'...
1-25 / 1-225 Chloride vs. Permeability..................
SafeLane With Ice Warning Condition.....................
SafeLane Maintaining Ice Watch Conditions...............
Surface Conditions Around 15 F on 2/8/2011 at 10:26 AM
SafeLane Pre-Charge Effectiveness.......................
1-25 /1-225 Spray System Activation.....................
Total Crashes on 1-76 / WCR-53..........................
Weather Related Crashes on Parker Rd. /1-225............
Non-Weather Related Crashes on Parker Rd. /1-225........
Total Crashes on Parker Rd. /1-225......................
Weather Related Crashes on 1-25 / 1-225.................
All Crashes on 1-25 /1-225 .............................
All Crashes on 1-25 /1-225 .............................
Appendix
DOT Survey..............................
Sample Data.............................
Colorado Procedure Laboratory 2104....
1-76 / WCR-53 Crash Data................
Parker Rd. /1-225 Crash Data............
1-25 /1-225 Crash Data..................
1-76 / WCR-53 Structure Plans...........
Parker Rd. /1-225 Structure Plans.......
1-25 / 1-225 Structure Plans............
Parker Rd. / 1-225 and 1-25 /1-225 Bid Data
Cargill SafeLane HDX Overlay............
Unitex SmartBond........................
PolyCarb Mark-163 Flexogrid.............
ESI Spray System........................
xn


TABLES
Table:
3.2.2.1.1
3.2.2.1.2
3.2.2.1.3
3.2.2.1.4
3.2.2.2.1
3.2.2.2.2
3.2.2.2.3
3.2.2.2.4
3.2.2.2.5
3.5.2.4.1
3.5.2.4.2
3.5.3.1.1
3.5.3.1.2
3.5.3.1.3
3.5.3.1.4
3.5.4.2.1
3.5.4.3.1
3.5.4.3.2
3.5.4.3.3
3.5.5.1:
5.2.5.1:
7.0. 1.1:
7.0. 2.1:
7.0. 3.1:
7.2.1.1:
7.2.1.2:
7.2.1.3:
7.2.2.1:
72.2.2:
7.22.3:
72.2.4:
7.2.3.1:
7.3.1.1:
Binder requirements for SafeLane HDX.......
Cure times of Unitex SmartBond.............
Unitex SmartBond resin properties..........
Gradation of SafeLane HDX aggregate........
Cured properties of Mark-163 urethane epoxy
Approximate gel and cure times.............
Set times for Mark-163 epoxy...............
Breakdown of Mark-371 by weight............
Gradation of Mark-371 aggregate............
Skid results from VDOT trial overlays......
VDOT skid numbers for SafeLane.............
Bond strength of hydraulic concrete overlays .
Bond tests for WSDOT epoxy overlays........
Bond tests for WSDOT MMA overlays..........
VDOT bond test results for SafeLane........
Permeability classification................
RCPT results for hydraulic concrete overlays.
WSDOT overlay permeability results.........
VDOT permeability values of SafeLane.......
ACI 318 Chloride Limits of Concrete........
27
27
28
28
30
31
31
32
32
59
62
68
69
70
71
75
78
78
79
82
Sample size corresponding to CL range............................113
1-76 physical core dimensions....................................130
Average thickness of Parker Rd. /1-225 SafeLane overlay..........131
Average thickness of 1-25 /1-225 Flexogrid overlay...............133
June 2010 asphalt sand patch results.............................139
June 2010 SafeLane sand patch results............................139
March 2011 SafeLane sand patch...................................140
June 2010 concrete sand patch results............................142
May 2010 right lane SafeLane sand patch results..................143
May 2010 left lane SafeLane sand patch results...................143
March 2011 right lane SafeLane sand patch results................144
March 2011 Flexogrid sand patch test results on 1-25 /1-225...... 146
June 2010 portable skid test results on 1-76 / WCR-53............148
Xlll


149
150
152
152
155
157
159
162
164
168
170
170
203
205
207
209
210
211
Ribbed fixed wheel skid numbers on Parker Rd. /1-225.......
Smooth fixed wheel skid numbers on Parker Rd. /1-225.......
Ribbed fixed wheel skid numbers on 1-25 /1-225 ............
Smooth fixed wheel skid numbers on 1-25 /1-225 ............
1-76 / WCR-53 bond strength results........................
Parker Rd. / 1-225 bond strength results...................
Bond test results from 1-25 /1-225.........................
October 2009 1-76 / WCR-53 permeability results............
March 2011 Parker Rd. /1-225 permeability results..........
March 2011 1-25 /1-225 permeability results................
October 2009 1-76 / WCR-53 chloride content................
March 2011 1-76 / WCR-53 chloride content..................
Bid item list with associated contractor costs.............
Installation cost of the 1-76 / WCR-53 overlay.............
Installation cost of the Parker Rd. /1-225 overlay.........
Installation cost of the 1-25 / 1-225 overlay w/o spray system
Installation cost of the 1-25 / 1-225 overlay w/spray system...
Comparison of cost per area for each site..................
xiv


1. Introduction
Since the first highways were built, engineers have been continually attempting to
protect bridge structural systems from chemical attack, particularly by deicing salts.
The concept of a surface overlay is a fairly old one, and has been in extensive use
since the first modem roads required rehabilitation. It was not until the latter half of
the 20th century that overlays started to be featured on newer construction; as a
means of protecting the bridge structural components. The Colorado Department of
Transportation (CDOT) has long history of research into different bituminous and
concrete pavements, however thin bonded overlays are a newer concept for CDOT.
Colorado is a state that receives a wide range of weather, from blizzard snow-
storms to desert heat, and the roadways need to be able to handle everything from
freezing ice subjected to deicing salts, to high heat which can lead to stripping and
cracking. Initial research by CDOT into overlays has been sparse, with the majority
being concentrated on different variations of white topping, the process by which
existing flexible or rigid pavements are overlaid with 2 to 8 inches (5.08 to 20.32 cm)
of concrete. With the advances in polymer materials in last 40 years, and the
development of concrete admixtures, many different options exist to protect bridge
structures either through new construction or rehabilitation.


The focus of this thesis was to evaluate the performance of two epoxy-based
thin-bonded overlays. The overall goal was to further CDOT's expertise in the use of
thin bonded overlays in order to increase safety and reduce maintenance costs. This
was accomplished by tw o methods: conducting a survey of current literature on the
subject, and physical testing of three overlay installations.
A survey with regards to the current state of the art of thin bonded overlays
was conducted and their use in Colorado was first examined. CDOT provides the
majority of their research reports online and these were checked for relevant
information. A second source of information on overlays was from the University' of
Colorado Denver. Very little research exists from either one of these organizations /
institutions. In order to gain a better understanding, a literature review was conducted
in which the types, functions, materials, and costs of different thin bonded overlays in
use by other state DOTs was examined.
This thesis seeks to determine real w'orld performance of thin bonded overlays
by evaluating two products, SafeLane from Cargill Inc. and Flexogrid from PolyCarb,
Inc., on three Colorado bridge decks. An experimental plan was developed to
examine the performance of these overlays and included five specific physical and
chemical tests: (1) mean texture depth, (2) friction, (3) bond strength, (4)
permeability, and (5) chloride. In addition, data was gathered through


instrumentation, crash reports, and installation bids to examine the overlays
performance in regards to anti-icing capabilities, traffic safety, and cost.
Based on the study findings, the performance of the both overlay systems is
promising; however, it is recommend that a long-term evaluation of the overlays be
performed in order to determine the durability and wear resistance of the thin bonded
overlays.
This thesis is broken down into several chapters. Chapter 2 discuses
background research with regards to thin bonded overlays in Colorado. Chapter 3 is a
review of current literature, focusing on different types of thin bonded overlays as
well as methods to test them. Chapter 4 is the problem statement. Chapter 5 is an
overview of the three sites that were studied. Chapter 6 focuses on the installation of
SafeLane and Flexogrid. Chapter 7 is the physical testing results. Chapter 8 is an
analysis of the crash data. Chapter 9 is the breakdown of the installation costs.
Chapter 10 contains the conclusions and recommendations.
3


2. Background Research
Research on thin bonded overlays in the State of Colorado is fairly small. The two
largest research organizations that have performed research on bonded overlays in the
state are the Colorado Department of Transportation, and the University of Colorado
Denver.
The Colorado Department of Transportation's research into thin bonded
overlays has mainly been in the area of whitetopping. Since 1998, two separate
papers have been published on the topic of whitetopping in Colorado. The first report,
CDOT-DTD-R-98-1, formed the basis for the design and usage of whitetopping by
CDOT. The second report, CDOT-DTD-R-2004-12, investigated the instrumentation
of, revised the design of, and increased the predicted performance of whitetopping
from the 1998 report.
The University of Colorado Denver research into thin bonded overlays
consisted of a single thesis by Bindel on the effects of surface treatment of pavements
w ith respect to bond strength of overlays. The thesis title, "Effects of Concrete and
Asphalt Surface Treatment on Bond Strength of Thin-Bonded Overlays" took a close
look at the adhesive properties of the binder used in the SafeLane system. In addition,
Binders thesis included findings from a survey of Departments of Transportation
(DOT) which asked about usage of anti-icing/anti-skid thin bonded overlays.
4


2.1 Summary of Previous Work
2.1.1 Colorado Department of Transportation
2.1.1.1 C DOT-DT D-R-98-1
The Colorado Department of Transportation report CDOT-DTD-R-98-1, "Guidelines
For The Thickness Design of Bonded Whitetopping Pavements in the State of
Colorado" had the stated goal of developing guidelines for the design of whitetopping
overlays. The Portland Cement Association has published basic design techniques for
thin whitetopping (TWT), and these were used as a design basis for the paper.
In order to validate finite element models, CDOT installed three different
TWT overlays at different sites throughout the state. The objective of these three sites
was to:
determine the critical load location of whitetopping pavements
study the effects of different AC surface preparation techniques
measure the response of whitetopping pavements under traffic loading
evaluate interface bonding of strength between the concrete and
asphalt layers
investigate the effect of pavement age on load-induced stresses
calibrate the theoretical with measured stresses to develop thickness
design guidelines
5


Two of the sites were constructed in the summer of 1996, while the third was
constructed the summer of 1997. All three sites were tested at 28 days, and again at 1
year.
The first site was Santa Fe Drive in Denver, CO. One 500 foot (152.4 m)
section had 5 inch (12.7 cm) whitetopping layer placed on 4 inches (10.16) of asphalt.
Another 500 foot (152.4 m) section had a 4 inch (10.16) whitetopping layer placed on
5 inches (12.7 cm) of asphalt. Joints were cut to 1/3 depth across the overlay and tie
bars were used for longitudinal reinforcement. The asphalt had no special preparation.
The second site was on State Road 119 in Longmont, CO. This section used
different concrete thicknesses. Three different asphalt preparation schemes were used,
including: placing new asphalt prior to whitetopping, no preparation prior to
placement, and 1.5 inches (3.81 cm) milled prior to placement. Concrete thicknesses
for these sections were 5 inches (12.7 cm), 4.5 inches (11.43 cm), and 6 inches (15.24
cm), respectively. Tie bars were used for longitudinal reinforcement. All asphalt
surfaces were washed prior to placement.
The final site was on US Route 287 near Lamar, CO. Six different sections
were tested at this site. In all cases, the asphalt was milled 7 inches (17.78 cm) down
prior to whitetopping. A 6 inch (15.24 cm) whitetopping was applied to the sections
with dowel spacing between 6 (1.83 m) and 12 feet (3.66 m).
At each site, instrumentation in the form of strain gauges were installed prior
to whitetopping to monitor loading conditions. The Santa Fe site used surface gauges
6


to augment the strain gauges. Thermocouples were also installed to monitor pavement
temperature during curing.
Analysis of the collected data and results provided several conclusions and
insights into whitetopping. It was found that the asphalt and concrete bond is essential
to performance of the overlay. In addition, the whitetopping overlays behave as
partially bonded systems, and should be designed as such. The partial bond increases
strain in the bottom half of the concrete, which needs to be accounted for in the
design. Surface treatments also change the strain induced on the concrete; milled
surfaces produced lower strains than un-milled asphalt, and newdy placed asphalt
should not be milled. A partial bonded overlay has about 15% less strain than a fully
bonded overlay. Finally, at least 5 inches (12.7 cm) of asphalt should be in place as a
base for whitetopping. The results from CDOT-DTD-R-98-1 showed that the
procedures outline a first generation design method, and that additional research
should be conducted (Tarr, 1998).
2.1.1.2 CDOT-DTD-R-2004-12
The additional research mentioned in CDOT-DTD-R-98-1 came in the form of report
CDOT-DTD-R-2004-12, "Instrumentation and Field Testing of Thin Whitetopping
Pavement in Colorado and Revision of the Existing Colorado Thin White Topping
Procedure." This report was conducted and written by two of the same primary
authors as the original 1998 report.
7


One of the concerns from the 1998 report that was addressed in the 2004
report was the minimum asphalt thickness of 5 inches (12.7 cm). It is expressed in the
2004 report that many of the roadways to be rehabilitated by use of whitetopping
already are near the 5 inch (12.7 cm) minimum, and that the overlay may be
ineffective that close to asphalt base minimums. The 1998 study focused solely on
rehabilitation of roadways using an overlay. The 2004 report expanded on this by
evaluating a new roadway construction that utilized whitetopping technology.
In 2001,4 miles (6.44 km) of Wadsworth Boulevard near Denver, CO, was
reconstructed using thin whitetopping. Load testing of this new overlay occurred in
2001 and 2003, the results of which were used to update, verify, and expand the 1998
report results. From this data, a second generation design procedure was established.
As a result of the 2004 study and in addition to the 1998 study, several new
recommendations were made. It was found that optimal joint spacing was 6 feet (1.83
m) in both directions of the slab. At less than 4 feet (1.22 m), temperature has an
adverse effect on the strain induced in the slab. Whitetopping slabs tied to roadway
shoulders exhibit less strain. Dowels are not necessary at control joints unless the
underlying asphalt starts to deteriorate, at which point they provide additional stress
relief. Bond strength should be continually monitored on all test sections. The report
concludes that additional long-term testing of whitetopping should be performed. The
8


original 1998 design guidelines, updated by the 2004 report provided the second
generation design for whitetopping overlays (Sheehan, 2004).
2.1.2 University of Colorado Denver
Bindel's research on how surface preparation affects bond strength was the only
research the University of Colorado Denver has performed on bonded overlays.
Bindel performed in-depth background and literature research prior to designing and
conducting bond tests. A DOT survey was sent to materials and bridge engineers, and
yielded interesting responses in relation to which states are using and testing thin
bonded overlays. In addition, it provided insight into what their experiences have
been with thin bonded overlays. Using information from the background research,
literature review, and DOT survey, a test plan was created in which the bond between
epoxy overlays, and asphalt or concrete surfaces was tested.
2.1.2.1 Departments of Transportation Survey
The state department of transportation survey was developed to investigate which
states have used anti-icing/anti-skid thin bonded overlays, and their levels of success
with them. A web-based survey was created that asked a variety of questions which
9


mainly focused on performance, installation experience, and products in use. A copy
of the DOT survey is provided in Appendix A.
The survey was targeted at materials engineers within each state DOT.
Twenty-four out of the fifty state DOTs responded to the survey. Some states sent
multiple responses from both bridge and materials engineers. A total of 30 responses
were recorded. Figure 2.1.2.1.1 shows the states that responded.
Figure 2.1.2.1.1: States That Responded to the DOT Survey (Bindel, 2010)
10


The first question was about who was responding, but the second question
asked was whether or not the state DOT used an anti-icing thin bonded overlay such
as SafeLane. Seven states said yes, including Vermont, Wisconsin, Idaho, New York.
Mississippi, Minnesota, North Dakota, and Kentucky. Of particular note is
Wisconsin, who reported that they had 30 applications of SafeLane since 1999.
The third question was if the state DOT has recorded a reduction in crashes
with regards to their SafeLane installations. Only Minnesota and Wisconsin
responded affirmatively that SafeLane reduced the crashes on their bridge decks. The
other DOT's did not have this information available.
Question four examined typical skid numbers seen on thin bonded overlays by
the DOTs. Around half the respondents reported it was greater than 35, which was
classified as acceptable.
The fifth question was whether or not a reduction in the amount of deicing
chemicals was seen in association with the use of anti-icing overlays. Only one state
reported a 20% decrease in the amount used. The other states reported no change, or
that they did not use the anti-icing properties of the overlay.
The next question was regarding failures of anti-icing overlays. Wisconsin
reported that they had issues with mixing SafeLane prior to application and


Minnesota reported that they had a decrease in friction, with a corresponding increase
in the number of accidents after using SafeLane for several years.
Question seven asked whether any states created specifications for anti-icing
overlays. Only two responded that they did. The others stated that they are still in trial
phases.
The eighth question asked which other products were in use by state DOTs.
They responded with Flexolith 216, Flexogrid, Degussa, and Transpo T-48.
The final question was if states had placed SafeLane on an asphalt surface.
Only two reported that they had.
From the DOT survey it was found that there is fairly little use of anti-
icing/anti-skid thin bonded overlays, especially on asphalt surfaces. States that had
used epoxy-based thin-bonded overlays reported acceptable skid numbers, and a
decrease in the amount of crashes. Some issues were reported with durability, but
they seem to be localized to specific installations. Most states had not run into any
significant issues, barring installation problems (Bindel, 2010).
2.1.2.2 Bond Testing
Bond testing was conducted on both on asphalt and concrete surfaces. Two asphalt
samples were tested in which the surface was sand blasted or smooth. These samples
12


were acquired from older roadways with cooperation of the City and County of
Denver. The concrete samples were created specifically for testing and conformed to
CDOT Class H specifications. Concrete sample surfaces were hand troweled, sand
blasted, tined. or roughened. One set of tests that are of direct interest to this thesis
was the testing of a full layer of SafeLane. Two samples from each concrete surface
preparation category contained SafeLane, while the rest of the samples only had a
layer of Unitex Pro-Poxy Type III DOT, which is the binder that the SafeLane system
uses. Altogether, 21 samples were tested. The goal was to determine a correlation
between surface preparation and bond strength.
Since all testing was conducted in the University of Colorado Material
Laboratory, a unique test setup was devised to allow the samples to be precisely
tested on the MTS 20-kip (89 kN) hydraulic testing machine. The samples were
specially prepared to allow them to be tested in the machine. Since the concrete
samples were created in-lab, they were cast with threaded dowels inside them so they
could be bolted to the test setup. The asphalt samples were bolted with C-channel uni-
struts. Figure 2.1.2.2.1 shows the sample blocks bolted in place and the bond test
ready to be run.
13


Figure 2.1.2.2.1: Bond Test Setup for Asphalt and Concrete (Bindel, 2010)
Each sample was core drilled to a depth of 0.5 inches (1.27 cm) beneath the concrete
surface by a 2 inch (5.08 cm) diameter core barrel in the center of the 12 x 12x6 inch
(30.48 x 30.48 x 15.24 cm) block. A 2 inch (5.08 cm) pipe cap was epoxied to the
center of the drilled area. For the concrete samples, the blocks were placed into the
test setup, bolted to the bottom plate, and then the top plate was bolted onto the block.
A 2 inch (5.08 cm) diameter pipe was threaded onto the pipe cap. The top end of the 2
inch (5.08 cm) pipe was connected to the MTS machine by another pipe cap and
threaded rod. The asphalt samples were bolted to the test stand by use of C-channel
uni-strut.
The results from these bond tests were not encouraging and show several
problem areas related to the test setup. Figure 2.1.2.2.2 shows the results.
14


Epoxy
1 2
Surface Type Surface T reatment Ultimate Failure (psi) Failure Type Ultimate Failure (psi) Failure Type
Concrete Troweled 73 Cohesive Concrete 208 Cohesive Concrete
Tined 219 Cohesive Concrete 182 Cohesive Concrete
Roughened 162 Cohesive Concrete 87 Cohesive Concrete
Sand Blasted 159 Cohesive Concrete Cohesive Concrete
85aF - m
10CF 144 Adhesive
115F Adhesive ..
Asphalt Smooth 90 Cohesive Ashpalt
Sand Blasted - Cohesive Ashpalt
Epoxy and Aggregate
1 2
Surface Type Surface Treatment Ultimate Failure (psi) Failure Type Ultimate Failure (psi) Failure Type
Concrete T roweied 214 Cohesive Concrete 263 Adhesive
Tmed 192 Cohesive Concrete 105 Cohesive Concrete
Roughened 277 Cohesive Concrete 224 Conestve Concrete
Sand 8lasted 146 Cohesive Concrete Cohesive Concrete
Figure 2.1.2.2.2: Bond Test Results (Bindel, 2010)
Bindel reported several problems related to both the sample preparation and test
process. Of the 21 samples, 4 failed during preloading, a process in which a small
amount of force (~10 lbs./44.5 N) is placed on the sample prior to running the full
test. An additional sample had the pipe cap placed wrong and could not be attached to
the testing machine. Out of the 16 remaining samples, 14 showed failures in the
concrete or asphalt. Unfortunately, several of the results were below the generally
accepted minimum of 250 (1.72 MPa) psi bond strength for polymer bonded overlays.
Cargill uses the 250 psi (1.72 MPa) as its minimum bond strength for SafeLane.
Although the majority of the bond strengths were below 250 psi (1.72 MPa), the
failure mode of all the successful tests occurred in the asphalt or concrete layer,
15


which shows that the Unitex epoxy was at least as strong as the surface it was adhered
too. Bindel recommended additional testing with a larger sample size to get a true
representative set of data (Bindel. 2010).
16


3. Literature Review
3.1 Why Thin Bonded Overlays?
In 2010, the Federal Highway Administration (FHWA) carried out an update of its
extensive National Bridge Inventory (NBI) which monitors the location and structural
condition of bridges across the United States. The FHWA has recorded over 604,493
bridges into the NBI and approximately 146,636 of those bridges are either
structurally deficit or obsolete (FHW'A, 2011). In Colorado alone, 1,399 of its 8,506
bridges fall under one of the categories above. Of particular concern are the 578
bridges that are deemed structurally deficit (FHWA, 2011). While the FHWA terms
structurally deficit as needing replacement, it does not necessarily mean that the
bridge is unsafe, it may only be highly deteriorated.
One of the largest problems facing transportation departments around the
United States is the deterioration of bridge decks and structural members. Especially
in colder climates where the use of harsh deicing chemicals causes rapid erosion of
steel reinforcement, engineers have spent a considerable amount of time trying to
determine effective and economical ways to combat the deterioration of bridges.
Bonded overlays allow engineers to place a wearing surface that not only provides
anti-skid properties, but prevents chloride intrusion into the bridge structure by
effectively sealing the bridge deck. Whth the mix of these properties, DOTs can be
17


proactive in their maintenance of bridges, thus saving costs and increasing safety in
the long term.
It is the combination of properties that make thin bonded overlays an
attractive option for use on bridge decks. While other overlays may offer the same or
better individual properties, very few offer the combination of properties that make
thin bonded overlays exceptional. Namely, the combinations of curing time, cost,
anti-skid properties, waterproofing, and deicing or anti-icing properties. Thin bonded
overlays can combine each of these properties into a unique wearing surface that not
only protects the bridge superstructure, but also increases safety.
An additional aspect, unique to some thin bonded overlays, is the reduced
dead load on a structure by replacing an older, heavier overlay system with a thinner
system. Thin bonded overlays also allow a roadway to receive a new surface without
raising the deck height appreciably, an important consideration when clearances on
older overpasses are already close to specified limits. By far the majority of overlays
are geared towards either an anti-skid or waterproofing role. However, newer
overlays are starting to incorporate anti-icing roles into their functionality, all while
providing a multi-year wearing surface.
Thin bonded overlays generally fall into one of 3 major categories; polymer
based, concrete based, or asphalt based. There are many different types of overlays
within each of these categories and several of them will be examined here. In general,
18


polymer concrete overlays are classified as an overlay that uses a polymer as the
binding agent. Concrete and asphalt overlays are defined similarly; each uses their
respective materials as their binders.
3.2 Polymer Concrete Overlays
Polymer concrete overlays first appeared around the 1950's. These overlays were
simply coal tar spread on a roadway with a fine aggregate broadcast across it. These
coal tar epoxies as they were commonly called, performed poorly and had a fairly
short lifespan. Polymer concrete overlays received an upgrade in the 1960's with the
development of polyester resins, epoxy resins, and methyl methacrylate (ACI, 1994).
These new formulations allowed overlays to last longer and provided a much
improved wearing surface. However they suffered from being too brittle, especially
when exposed to ultra-violet radiation. They also suffered de-bonding problems due
to thermal incompatibility with the underlying concrete (ACI, 1994). This leads to
cracking and eventual failure. Moisture tolerant polymers were introduced by the
1970's which allowed the overlays to seal the bridge decks, thus creating a longer
term wearing solution that also protected the bridge deck. Throughout the 1980's and
90's, much more "flexible" polymers came into use which increased durability,
leading to an even longer lasting overlay. These are the same types of epoxies that are
19


used in polymer concrete today, and are generally characterized by their lower
modulus of elasticity and higher elongation tolerances (White et al., 1997).
Polymer overlays are uniquely different from the other two types since they
add very little material to the surface of the bridge deck. This is advantageous, since it
reduces the dead load on the bridge. This is ideal for older bridges, but even newer
ones benefit from the economy of lighter sections. Polymer overlays have the added
advantage of sealing the structural bridge deck from moisture and chemicals. Dinitz
et al. (2010) compared polysulfide epoxy overlays to asphalt and concrete thin
bonded overlays and noted that the air voids (up to 5%) in asphalt overlays contribute
to freeze/thaw damage, while the epoxy overlays seal the deck surface, preventing
this phenomenon. This helps to prolong bridge superstructure life as well as to
provide a surface that can be removed and replaced if excessive wearing occurs. The
biggest disadvantage to polymer overlays is that surface preparation is extremely
critical. Even the slightest moisture or debris on the surface can cause de-bonding of
the overlay.
The most common types of binding agents are epoxies, coal tar modified
epoxy, polyester, methyl methacrylate, and polyurethanes. Several factors affect
polymer overlays. Of utmost importance is surface preparation, followed by material
selection. Some polymers can react to alkalis in the concrete and saponify, becoming
little more than a soapy mixture (Scarpinato, 1996). Heating and cooling of the
20


overlay can lead to differential shrinkage between the binder and structural surface
underneath. The stresses induced by these temperature changes can lead to de-
bonding of the overlay. Some of the first epoxies suffered from age hardening under
ultraviolet light. However, newer epoxies based on specially modified polymers have
a high strength and can resist UV age hardening making them ideal for bridge decks
(Dinitz et al., 2010).
Polymer overlays can be divided into two categories depending on how they
are applied. Stenko et al. (2001) describe the two methods as the broom and screed
method, and the slurry or premixed method. For most bridges, polymer overlays are
installed on site, however for new construction using pre-cast bridge deck panels,
epoxy overlays can be installed at the pre-cast plant.
3.2.1 Premixed Polymer Concrete Overlays
Premixed Polymer Concrete Overlays (PMPCO) includes any type of overlay that is
installed using the slurry or premixed method (Maggenti, 2001). Typical installation
procedure for a slurry or premixed type polymer overlay is very similar to traditional
concrete methods. Most PMPCO's are mixed onsite in a special mixer, then placed on
the deck surface by hand or using traditional slip form methods. Finishing is often
performed by trowel, float and vibrating screed, with final texturing performed by


tining. Some methyl methacrylate overlays use an aggregate broadcast as a final
texturing step, while most polyester overlays use tining.
A typical premixed polymer concrete mix consists of a binder, silica and
basalt sand as the fine aggregate, gravel as the coarse aggregate, and admixtures to
improve various resin properties. An initiator is added during mixing to start the
curing process (Ozolin, 2007).
The premixed method is considered faster and less labor intensive since it
combines all the separate steps of the broom and screed into one step. Many different
types of polymer binders use the premixed placement, however, only two types are
the most commonly seen: polyester resin, and methyl methacrylate (MMA).
3.2.1.1 Polyester Polymer Concrete
Some of the first uses of polyester polymer concrete (PPC) was in 1960's for use in
pipes since the low permeability and high resistance to chemical attack make it ideal
to transport liquids (Lang, 2005). Some of the first polyester concrete overlays were
placed in California around 1983. Initially these were lightly traveled alpine roads,
but by 1984-86 PPC overlays were seen on high traffic roads with good results
(Glauz, 1993).
Polyester polymer concrete has many desirable properties for use as a
pavement. It has very low permeability, excellent skid resistance depending on finish,
and cures quickly with rapid strength development. PPC overlays are generally
22


designed to be between 1/2" to 1" (1.27 to 2.54 cm) thick, depending on underlying
surface conditions and can also have iron ore coke added to increase conductivity for
cathodic protection or to act as a heating element for deicing. It can also be installed
in a single pass, similar to traditional concrete. Expected life of PPC overlays ranges
from 15 to 20 years and depends highly on the mixture, as well as surface
preparation, both of which are extremely critical. (Sprinkel, 1990)
The cost for PPC varies greatly on who is placing it. Caltrans has extensive
experience with PPC overlays and can place them at very low cost, between $ 1 and
$3 per square foot ($10.76 to $32.29 per square meter), depending on project size
(Caltrans, 2011). Oregon DOT recently started experimenting with PPC overlays and
placed one for $4.44 per square foot ($47.79 per square meter) (ODOT, 2011).
3.2.1.2 Methyl Methacrylate Concrete
The other type of premixed polymer concrete uses a methyl methacrylate (MMA)
binder instead of polyester. Early MMA overlays had excessively fast cure times that
precluded spreading aggregate as a finish. This was fixed with better formulations of
the MMA binders and initiators.
Methyl methacrylate overlays are very similar to polyester overlays, but are
generally harder and have a compressively stronger surface that has excellent long
term friction characteristics. Due to the stronger characteristics, MMA overlays are
23


sometimes broadcast with additional aggregate as a finish instead of tining. Although
MMA overlays wear better, epoxy based overlays tend to have superior bond
characteristics, a feature that is very important for polymer overlays. To increase
adhesive properties, MMA overlays are preceded by a MMA sealer which acts as a
tack coat for the overlay (Wilson, 1995).
Oregon DOT has placed two different experimental MMA overlays in 1982
for $8.70 per square foot ($93.65 per square meter). However they estimated that
larger overlays placed on a regular basis would cost around $6.40 per square
foot($69.89 per square meter) (Quinn, 1984). WSDOT placed some larger overlays in
2007 for $7.20 per square foot ($77.50 per square meter) (WSDOT, 2007). Montana
DOT has also experimented with MMA overlays and installed one in 1995 for $4.89
per square foot ($52.64 per square meter) (Johnson, 1997).
3.2.2 Broom and Screed Overlays
The broom and screed method first applies the binder, and then applies the aggregate
by either manual or automatic broadcasting (Ozolin, 2007). Broom and screed
typically has two layers of binder and aggregate, although three is specified by some
manufacturers. The majority of epoxy based overlays utilize the broom and screed
method. This is because epoxy overlays tend to use an aggregate as the traction
surface instead of a filler as polyester and MMA polymer concretes do. This allows
24


the total overlay thickness to be thinner, usually around 3/8" to 3/4" (0.9525 cm to
1.905 cm) versus 1/2" to 1" (1.27 to 2.54 cm) for the other polymer overlay types.
Epoxy overlays offer the same benefits as other polymer overlays, often in a thinner
section.
3.2.2.1 SafeLane
SafeLane is an epoxy based polymer concrete overlay that was specifically developed
to fill the role of an anti-icing/anti-skid overlay (Cargill, 2007). It utilizes a patented
combination of aggregate and binder to obtain these properties, while providing
protection for bridge decks. The aggregate itself is a proprietary product that is
specially prepared to absorb deicing chemicals.
Cargill Incorporated is the licensed patent holder for both the SafeLane
system and the specialty aggregate. It is a private multinational corporation that
mainly specializes in the agricultural and food related business. They also have
branches in industrial, energy and pharmaceutical areas. SafeLane is an extension of
their winter maintenance line of products which mainly includes de-icing salts.
The original inventor of SafeLane is Dr. Russ Alger. Dr. Alger was the
director of the Institute of Snow Research at Michigan Technological University. The
Institute of Snow Research is part of the Keweenaw Research Center. Development
25


of SafeLane occurred over a period of 10 years until the right combination of
aggregate and binder was found. SafeLane technology was licensed to Cargill
Incorporated in 2003 (Perischetti, 2007).
SafeLane is a two-part anti-icing polymer surface overlay. It utilizes a
specialty aggregate bonded to a pavement surface by an epoxy binder. Currently,
there are two types of SafeLane products available (Cargill, 2007). The first, and
more commonplace is Cargill SafeLane HDX Overlay, which is the main product
used for roadways and bridges. The other product, Cargill SafeLane CA-48 Overlay
is used for parking lots, garages, and sidewalks. Cargill is seeking approval for using
SafeLane CA-48 on airport taxiways and parking aprons as well. They both use the
same aggregate material; however SafeLane CA-48 uses a smaller aggregate size.
The epoxy binder used in SafeLane must be approved by Cargill. Cargill
specifies an "epoxy resin base and hardener that is a modified Type III, two
component system which meets requirements given by ASTM-C-881, Grade 1,
Classes B & C." Table 3.2.2.1.1 shows the binder requirements from the technical
specifications for SafeLane HDX (Cargill, 2010).
26


Table 3.2.2.1.1: Binder requirements for SafeLane HDX (Cargill, 2010)
Property Requirement Test Method
m tils 15 to 45 ntir at 75 F124: C5 A5TM C831 !5C m, sample in paper cuii
Tsnsile Strength 2.000 to 5.000 psi at 7 days ASTV DS3B
Tensile Elongation 40% to 80% al ? days A3TV D638
VlKOSitY 7 to 25 poises ASTV D2333
Minimum, Compressive Strength at 3 fious 1-COOpslat 75 F (24s Cl ASTMC579
Minimum Compressive Strength at 24 mu's 5.000 BSi at 75 F (24= Q ASTVC579
tiSnmiurr. Adhesive Strength at 24 noire 250 psi ai 75 F124' Q ACS503R
While the above general requirements are for any type of epoxy, the type most
often used and recommended by Cargill is Unitex, Inc. SmartBond. Unitex describes
SmartBond as, "a solvent-free, moisture insensitive, 100% solids, low modulus, two
component bonding agent." The cure times for SmartBond are shown in Table
3.2.2.1.2 while the properties of SmartBond are shown in Table 3.2.2.1.3 from the
Unitex SmartBond Technical Specifications.
Table 3.2.2.1.2: Cure times of Unitex SmartBond (Unitex, 2004)
Minimum Curing Times of SMARTBOND
Average Temperaties of Owertay Component & Substrate
60-64F 65-69T 70-74F 75-79F 80-84F 85+F
16-18C 19-2rC 22-23C 24-26C 27-29C 29+cC
Min.
Cure
Time 6-5 hrs 5hrs 4 hrs 3hrs 3 hrs 3his
27


Table 3.2.2.1.3: Unitex SmartBond resin properties (Unitex, 2004)
LABORATORY TESTS RESULTS A STM C881 SPECIFICATIONS
Mix Ratio 1 :1 by volume None
D 695 Compressive Modulus 64.820 130,000 maximum
D 638 Tensile Strength 2,610 psi None
D 638 Tensile Elonqation 50% 30% minimum
C 882 Bond Strenqth (14 dav cure) 3,470 psi 1.500 psi minimum
D 570 Water Absorption 0.19% 1 0% maximum
0 2471 Gel Time 15 Minutes -
D 2393 Brookfield Vise. RV3 (a) 20 rpms 1425 cds 2000 cps maximum
D 2240 Shore D Hardness 69 None
C 883 Effective Shiinkaqe Pass Pass
C 884 Thermal Compatibility Pass None
AASHTO T-277 Chloride Ion Permeability 0 9 coulombs None
The SafeLane system uses a dolomite aggregate that has been specially
prepared by a proprietary method. Dolomite is one of two minerals that can make up
limestone; the other being calcite. By definition, dolomite is a limestone that contains
more than 90% of the mineral dolomite. The mineral dolomite is composed of
calcium, magnesium, and carbonate in the formulation of CaMglCO.^. While no
specific data is given, Table 3.2.2.1.4 shows the Cargill specified aggregate
gradations (Goodman, 1993).
Table 3.2.2.1.4: Gradation of SafeLane HDX aggregate (Cargill, 2010)
Gradation % Passlnq
3/8" (9 5 m m 98-190
#4 <4 75 mm} 50-80
#8 (2 36 mm} 0-15
28


SafeLane tends to be a more expensive overlay to install due to its current
experimental nature. While many different SafeLane sites exist throughout the United
States, the majority of these are DOTs testing the product. Several states reported
fairly high installation costs, including North Dakota DOT which installed two
experimental overlays in 2002 for between $10.96 and $12.72 per square foot
($117.97 to $136.92 per square meter) (Mastel, 2002). In addition, Virginia DOT
installed two overlays for $6.00 per square foot ($64.58 per square meter) in 2005,
however this was only material costs and does not include traffic control or other
associated costs (Izzepi, 2010).
3.2.2.2 Flexogrid
Flexogrid is a polymer concrete overlay system for use on bridge decks. It uses an
aggregate and two-part epoxy binder to create a thin bonded overlay and is developed
and marketed as a lightweight, anti-skid, durable wearing surface (PolyCarb, 2009).
Flexogrid was developed by PolyCarb, Inc. in the 1980s as a way to provide an anti-
skid surface that can protect bridge decks from chemical intrusion.
PolyCarb, Inc., the distributor of the Mark-163 Flexogrid system, is a branch
of Dow Formulated Systems, a business unit of Dow Chemical. Dow Chemical is a
multinational corporation whose main focus is the development and manufacturing of
29


chemical based products and systems. The specialty of PolyCarb, Inc. is the
development and marketing of epoxy based pavement and flooring systems.
Flexogrid consists of the Mark-371 aggregate and Mark-163 epoxy. The
Mark-163 urethane epoxy is a two part amber colored epoxy that is supplied by
PolyCarb for the Flexogrid system. It is 100% solids and is mixed in a 2:1 ratio of
parts A and B. Urethane epoxy provides a strong yet flexible binder that can resist
cracks from flexing as well as temperature changes in underlying materials. Table
3.2.2.2.1 below shows the properties of the cured Mark-163 epoxy.
Table 3.2.2.2.1: Cured properties of Mark-163 urethane epoxy (PolyCarb, 2009)
Adhesion of Concrete 100% Failure ASTM D-4541 /ACI-503R
Shore D Hardness 70 5 ASTM D2240-75
Compressive Strength 48.3 62.1 MPa (7,000 9,000 psi) ASTM C-109
Tensile Strength >17.2 MPa (>2,500 psi) ASTM D638-82
Tensile Elongation 30 10 ASTM D638-82
Water Absorption Max. 0.20% ASTM D-570
Abrasion Resistance Wear Index 75-85 milligrams ASTM C-501
CS-17 Wheel, 1,000 cycle, 1,000 gms
Flexural creep at low temperature 0.165 mm, min. California Test 419
Total movement in 7 days (.0065 inches, min.)
Flexural Yield Strength > 5,000 psi ASTM D-790
30


As with all epoxies, curing time varies with ambient and surface temperature.
Flexogrid should not be installed below 50F (10 C) to prevent installation and
durability problems. Table 3.2.2.2.2 shows the approximate curing times, while Table
3.2.2.2.3 shows the maximum amount of time between epoxy application and
broadcast of the aggregate.
Table 3.2.2.2.2: Approximate gel and cure times (PolyCarb, 2009)
Gel Time 25C (75 2F) 22-31 minutes (100 gms.)
Gel Time 25C (75 2F)
(with aggregate) 1.5 hours
Initial Set 25C (75 2F) 6 hours
Final Cure 25C (75 2F) 48 hours-7 days
Table 3.2.2.2.3: Set times for Mark-163 epoxy (PolyCarb, 2009)
90F (32C) 10 minutes
80F (27C) 15 minutes
70F (21C) 20 minutes
60F (16C) 25 minutes
50F (10C) 35 minutes
The aggregate used with the Mark-163 epoxy can vary, but the recommended Mark-
371 aggregate is basalt quartzite granite. Table 3.2.2.2.4 shows the following
breakdown of Mark-371 by weight. Gradation of the aggregate is shown in Table
3.2.2.2.5.
31


Table 3.2.2.2.4: Breakdown of Mark-371 by weight (PolyCarb, 2009)
Component Percent by weight
Si02 75.03
A1203 11.49
Fe203 3.57
CaO 2.84
MgO 1.59
Na20 2.58
K.20 0.99
Combined alkali 1.11
Ignition loss 0.72
Table 3.2.2.2.5: Gradation of Mark-371 aggregate (PolyCarb, 2009)
US Standard Sieve Size Percent passing by Weight
No. 6 100%
No. 10 10-35%
No. 20 0-3%
Due to the amount of time it has been around and the number of installations
of Flexogrid, it is a relatively inexpensive epoxy overlay to install. Several DOT's
around the United States have had experiences with Flexogrid. Alabama installed
Flexogrid on an 1-20 bridge in 1993 for $5.33 per square foot ($57.37 per square
meter) (Ramey, 2003). Iowa DOT installed Flexogrid in 1986 for approximately
32


$5.12 per square foot ($55.11 per square meter) (Adam, 2001). More recently. North
Dakota installed Flexogrid in 2008 for an estimated $5.27 per square foot ($56.73 per
square meter) (Mastel, 2009).
3.3 Hydraulic Concrete Overlays
The first documented use of any overlay dates back to 1918 when the city of Terra
Haute, IN, used a layer of concrete on top of their existing asphalt road. This was
relatively unheard of at the time, today it is commonly known as whitetopping.
Through the 1950's and 1960's whitetopping was used rather sparingly and almost
always in a rehabilitation or capacity upgrade role. It was relatively unheard of to use
whitetopping as a preventative measure for bridge decks. It was not until the 1980's
that overlays started to become common place. By the early 1990's whitetopping had
become much more widespread and had evolved into ultra-thin whitetopping (UTW),
a 2" to 4" (5.08 to 10.16 cm) overlay and thin-whitetopping (TWT) a 4" to 8" (10.16
to 20.32 cm) overlay (United States, 2011).
Thin bonded concrete overlays are typically defined as an overlay that is a 1"
to 4" (2.54 to 10.16 cm) concrete wearing surface on top of a structural deck. Several
different types of concrete are used in thin bonded overlays including: low-slump
dense concrete (LSDC), silica fume concrete (SFC) also known as a micro silica
modified concrete (MS or MSC), and latex-modified concrete (LMC). Shahrooz, et
33


al. (2000) mentions a super-dense plasticized concrete (SDC) that is used in overlays
(Rainey et al., 2003).
The goal of any of these concrete mixtures is to provide a concrete that resists
chloride penetration while providing a durable wearing surface for traffic. By
increasing density, you decrease the porosity of the concrete thus reducing the
amount of moisture and chemicals able to penetrate the bridge deck. In addition to
variations in concrete mixtures, some concrete overlays utilize an internally sealed
layer that combines the wearing and strength of concrete with the water resistance of
wax polymers. Although these concrete mixtures are effective at stopping chlorides
leeching through, they still present problems with freeze/thaw. Conductive concrete
overlays have been examined in which the concrete itself becomes a resistive heating
element to prevent icing.
3.3.1 Low-Slump Dense Concrete
Low-slump dense concrete was first seen in the 1960's in Iowa and Kansas. Early
LSDC overlays were approximately 1.25" (3.175 cm) thick, although now 2" (5.08
cm) overlays are the standard and can be used either as a new overlay, or as a
rehabilitation overlay. While LSDC overlays use a relatively old technology, it has
been highly effective for its cost. Some of the original overlays were in service for 20
to 25 years before needing replacement. Typical design life is around 25 years.
34


Typical LSDC overlays utilize a water to cement ratio of around 0.32. This
creates a strong concrete with low permeability to resist chloride penetration. Surface
roughness is gained by tining or texturing the concrete. De-icing is provided by
chemical methods, although other methods could be utilized.
New York DOT conducted a study on 50 different LSDC overlays and found
that 0.84% of the total deck area had any damage due to corrosion, delamination,
spalling, or patching. Half of that area was around joints indicating that the overlays
themselves were showing adequate performance. A survey of 153 LSDC overlays in
1991 by the Strategic Highway Research Program found that service life performance
depends highly on the deck preparation. Sandblasting combined with removal of
chloride saturated concrete was deemed the most effective method at prolonging
service life (Russell, 2004).
Minnesota DOT performed an extensive review of LSDC overlays,
particularly for bridge decks, and has found that they are able to place LSDC overlays
for approximately $5 per square foot ($53.82 per square meter). For newer overlays
going to bid, the engineers cost is typically around $4 per square foot ($43.06 per
square meter). These costs are attributed to the fact that Mn/DOT places a fairly large
volume of overlays, and that materials are fairly plentiful in the region. They also
note that their bridge construction costs are cheaper than other states. The cost of $5
per square foot ($53.82 per square meter) is a good base for installing a LSDC
overlay (Rowekamp, 2004).
35


3.3.2 Silica Fume Concrete
Silica fume concrete (SFC) was originally developed around the 1950s; however it
was not until the development of high-range water reducing admixtures (HRWA) that
placement of silica fume concrete became practical (Whiting et al., 1998). Ohio and
New York were the first US states to experiment with silica fiime concrete in overlays
and full depth bridge decks. Ohio installed their first full depth bridge deck in 1987
while New York only placed an overlay on 1-90.
Silica fume concretes have become a popular choice due to the increase in
density, strength, and durability provided to concrete. Early placements of SFC
overlays had problems with finishing and autogenous shrinking, or rapid drying of the
concrete. The curing and finishing of silica fume concrete is especially important
because of the lower water / cement ratio. In the plastic state before the concrete has
fully hardened lack of water leads to desiccation and cracking, especially on the
surface where moisture can evaporate quickly. The increased density of the concrete
reduces pores found in normal concrete, making the evaporated water harder to
replace. Cracking can also occur in cured concrete, since the increased density
prevents water from being absorbed by the gel matrix. In normal concrete, this water
helps to offset the effects of autogenous shrinkage, but in silica fiime concrete this
effect becomes greater due to less water. Overlays placed today rarely have this
problem due to the understanding of how density affects the hydration of concrete.
36


Typically, silica fume replaces cement at rates between 5% and 12% in the
concrete mixture. Whiting et al. (1998) found that between 6% and 8% is the optimal
replacement range for bridge decks. Past this range, a plateau effect was seen on
permeability. Greater strengths from increased silica fume content are not necessarily
useful for a bridge deck overlay, nor economical. They also recommended immediate
wet curing for at least 7 days; however, this is generally standard industry practice
today.
Silica fume overlays tend to be slightly more expensive than normal concrete
overlays due to the additional admixture costs. In 1994, Virginia DOT placed two
silica fume concrete overlays; one regular, and one high early strength for $2.68, and
$3.30 per square foot,($28.85 and $35.52 per square meter) respectively (Sprinkel,
2000). Montana DOT installed a trial silica fume overlay in 1996 for only $2.44 per
square foot ($26.26 per square meter) (Johnson, 1997).
3.3.3 Latex Modified Concrete
While traditional concrete methods have been around for almost a century it was not
until 1956 that Dow Chemical Company started experimenting with latex modified
concrete. Michigan Highway Department in cooperation with Dow Chemical placed
the first latex modified concrete overlay in 1958. Since about the 1960's, the majority'
of latex concrete's utilize one of three types of latex; styrene-butadiene rubber,
37


polyacrylic ester, or polyvinylidene chloride-vinyl chloride (Ramakrishnan, 1992).
The typical polymer compound used is styrene butadiene latex, a mixture that
separates in water to leave behind latex solids in the concrete matrix after curing.
Latex-modified concrete (LMC) is a concrete mixture that utilizes a latex
polymer in the mixture. Ramakrishnan (1992) defines latex as, "a stable dispersion of
organic polymer particles in an aqueous surfactant solution... All latexes used in
concretes are classified as nonionic since they carry no extra positive or negative
electron charges in their molecular configuration. This is necessary, since negative or
positively charged latex would adversely affect concrete and embedded reinforcing
steel. During the production of latex, surfactants are added to the latex mixture to
modify the surface tension and stability of the latex. These surfactants also provide
key benefits to concrete including acting as water reducing admixtures which allows
higher strengths to be achieved without significant losses in workability
(Ramakrishnan, 1992).
Latex modified concretes have higher than normal tensile, flexure, and bond
strengths as well as increased freeze/thaw durability with a low permeability
(Kuhlman, 1985). Research performed by Ramakrishnan (1992) shows that adding a
latex polymer to concrete increase the flexure strength by as much as 150 to 200%.
This is due to the latex film that forms in the concrete mixture during curing. The
latex bonds the two sides of the micro cracks together thus helping to prevent crack
propagation and applying a portion of its own tensile strength to the concrete. These
38


same principals help increase the impact resistance, toughness, and abrasive
resistance of the concrete all of which are valuable properties for a bridge deck
overlay.
The latex in the voids helps to seal the concrete to moisture. Testing of LMC
bridge deck overlays throughout the US have shown that LMC overlays all exhibit
very' low (less than 1000 coulombs) chloride permeability, even after multiple years.
Freeze-thaw resistance also increases, as does resistance to scaling. The Indiana State
Highway Commission tested LMC overlays for resistance to scaling and found that
even after 50 freeze/thaw cycles, no scaling had occurred (Ramakrishnan, 1992).
In terms of workability, LMC's behave like normal concretes that have had
water reducing admixtures added. Due to this effect, the compressive strength of
LMC's is usually higher due to a reduced water cement ratio.
Typical LMC overlays are 1.25" to 2" (3.175 cm to 5.08 cm) thick and can be
found on every type of project from new construction to major rehabilitations. The
higher adhesive properties of LMC make it an ideal material to be used for
rehabilitation projects. Anecdotal evidence from several state DOT's show that LMC
overlays last upwards of 20 years or more. Several of these overlays were placed with
high and very-high early strength concretes and allowed traffic on the overlay within
24 hours instead of the standard 4-7 days. Testing after one year showed these
overlays were still providing excellent wear resistance.
39


Latex Modified Concrete overlays are a relatively cheap overlay to install due
to the fact that it is mainly concrete with a latex admixture. VDOT has pretty
extensive experience with LMC overlays, and placed two in 1994. One was a
standard styrene-butadiene LMC that was placed for $3.00 per square foot ($32.29
per square meter), while another was a LMC modified for high early strength that was
placed for $3.70 per square foot ($39.83 per square meter) (Sprinkel, 2000). In
addition, North Carolina DOT has placed many LMC overlays. Several placed in
2010 cost $4.20 per square foot ($45.21 per square meter) (NCDOT, 2010).
3.3.4 Internally Sealed Concrete Overlay
Internally sealed concrete overlays are a fairly simple concept in which wax beads
placed in the concrete during mixing are later melted using an external heat source.
The theory is that the melted wax will flow into the pores of the concrete, sealing it.
The concept of adding wax beads to concrete is credited to the Monsanto Research
Corporation under a contract with the Federal Highway Administration (FHWA).
Original development took place in the 1970's, and several states through the FHWA
Demonstration Project 49 participated in field testing this new type of overlay
(Toney, 1987).The majority of State DOT'S that participated in the FHWA study used
their normal bridge deck overlay concrete with the wax replacing a portion of the fine
aggregate in a 1:1 ratio. The wax itself is a 25%/75% blend of montan and paraffin
40


waxes. Montan wax has a melting point of 180 F (82 C), while paraffin is closer to
147 F (64 C). It was reported by several DOT's that the wax modified concrete
does have lower compressive strength, but still above the 4000 psi (27.6 MPa)
required by some states. Air entrainment is recommended for additional freeze/thaw
resistance. Thorough mixing of the concrete is required to ensure even dispersal of
the wax beads. Typical thickness of the concrete overlays is approximately 2" (5.1
cm) (Tyson, 1978).
The most important aspect of internally sealed concrete overlays is the curing
and heating. Sufficient heat has to be put into the concrete to allow the bottom layer
to reach the approximately 185 F (85 C) to melt the wax all the way through the
overlay. However, heat cannot be applied too quickly, otherwise excessive spalling
and cracking in the concrete due to rapid evaporation of moisture and shrinkage will
occur. Washington State DOT evaluated four different methods of curing including
hot air, electric fired infrared, electric infrared, and electric blanket. The first three
methods caused spalling and cracking. Electric blankets provided enough control and
power to properly heat the concrete, and are the recommended heating method.
Heating is usually conducted over 7 to 10 hours depending on weather and overlay
thickness. This process is repeated for several days to ensure that all the wax has
melted and filled the pores (Tyson, 1978).
41


Results from several DOT studies all show similar trends. The first is that
internally sealed concrete overlays are effective at stopping chloride penetration of
concrete. Second is that cracking of the concrete overlay during the heating procedure
is commonplace. Varied methods can be used to reduce and prevent this cracking, but
it is unlikely to completely reduce the cracks that develop.
Internally sealed concrete overlays can cost a bit more than regular concrete
overlays to install. Oklahoma DOT mentioned that they were paying an extra $1.03
per square foot ($11.09 per square meter) for internally sealed concrete overlays
versus normal concrete overlays. Idaho DOT reported that it cost them twice as much
to place versus normal overlays (Toney, 1987).
Internally sealed concrete overlays are effective at stopping chloride
penetration and protecting bridge structures, but their higher cost and their
installation-critical performance make them a less popular option than other overlay
types. No recent internally sealed concrete overlays have been placed, and it is likely
this trend w ill continue until further research or newer heating methods are developed
(Tyson, 1978).
42


3.3.5 Conductive Concrete Overlays
Deicing a roadway surface by using heat is not a new idea, but it does have a new
twist in the form of conductive concrete. Two general methods are known for deicing
a bridge deck using heat. The first is to embed resistive heating elements into the
concrete and generate heat by running current though the elements. The second is to
turn the concrete into the resistive element. Yehia et al. (2000) defined conductive
concrete as, "...a cementitious admixture containing electrically conductive
components to attain a high and stable electrical conductivity." They also mention
that conductive concrete has also been used as anti-static flooring, electromagnetic
shielding, and cathodic protection of steel. The power provided to the concrete can be
either AC or DC with results being similar.
Conductive concrete is fairly simple in its composition. Generally, a standard
concrete mix is used, and then steel fibers and shavings are added to provide the
conductivity. Yehia et al. (1998) have also found that conductive concrete can be
made with conductive aggregate such as black carbon or furnace coke; however less
strength is developed due to a higher amount of water needed for the aggregates.
Yehia et al. (2000) developed a concrete mix that used 20% steel shavings and
1.5% steel fibers by volume, which met ASTM and AASHTO specifications. They
cast a 3.5" (8.89 cm) slab using this mix and tested the de-icing, and anti-icing
properties. It was found that the conductive concrete overlay performs both rolls
adequately. Their results show that the cost to run the conductive overlay per hour
43


was approximately $0,056 to $0,074 per square foot ($0.60 to $0.80 per square
meter). On larger bridges, this could be prohibitively expensive.
3,4 Asphalt Concrete
Asphalt Concrete, or bituminous pavement, is a bridge deck overlay and roadway
material that is in widespread use. Asphalt is typically made of a petroleum binder
and aggregate, although different materials have been added to improve certain
qualities. Although asphalt pavement is not a thin bonded overlay, they are often
installed on top of waterproofing membranes, which are very effective at sealing
bridge decks. These membranes will be examined here.
Asphalt has been in use as a pavement and sealer since ancient times, when
natural deposits were found and used for paving with bricks, or to seal water basins. It
was not until the 1860's that it started to see use in the United States, originally as
sidewalks, then as roadways. Most of these early asphalts were from natural sources,
although just after the turn of the century refined petroleum binders became more
common. Around the 1940's-50's is when the more modem types of asphalt started to
become widespread, primarily with investment from the war effort as well as the state
highway system (National Asphalt Pavement Association, 2011).
While asphalt is a very economical overlay due to economies of scale, it lacks
the moisture sealing capabilities and resistance to certain chemicals that other
44


overlays have. Several methods have been developed to deal with these deficiencies.
The simplest method is to add a water resistant membrane between the asphalt
overlay and the bridge deck.
The concept of protecting a bridge deck by applying a new overlay is
relatively old; it was not until 1972 when the FHWA introduced polices requiring the
protection of bridge superstructures from deicing chemicals. As a result of these
polices many state DOT's looked for the fastest solutions to protect their bridge
decks. Many had few to choose from, and the waterproofing membrane became
commonplace. Today, the use of these membranes with concrete overlays is
somewhat declining, mostly due to newer technologies. However, waterproofing
membranes are still used extensively on bridges that have any sort of asphalt wearing
surfaces.
While the membrane is referred to as a single layer, there are in fact multiple
layers that make up a membrane system. At the base is the concrete bridge deck. A
thin primer layer is first applied. This generally consists of a synthetic rubber, but can
also be a resin type primer. The purpose of the primer layer is to provide greater
adhesion of the membrane system to the concrete bridge deck. After the primer layer,
some systems use an adhesive layer and/or ventilation layer. The adhesive layer
provides additional bonding if necessary, while the ventilation layer provides a
method for chemical vapors to disperse, preventing blistering. The actual
45


waterproofing membrane layer is next, followed by protection layer to help prevent
damage to the membrane during construction. The final layer consists of a tack coat,
which helps to bond the overlay or asphalt to the membrane and bridge deck. The
overlay itself may be composed of a single or double layer of material, depending on
the type used. Figure 3.4.1 shows a graphical breakdown of the possible layers in
waterproofing membrane.
surface course of
bituminous concrete
base course of
bituminous concrete
o
i>
/
-tack coat
protection
/ / board
/ /
/ membrane
/ ventilating
/
* = = = *=? -adhesive
mm&// y la>'cr
- primer
y

o
concrete deck slab


o
-V

Figure 3.4.1: Membrane Layers (Manning, 1995)
The membranes themselves can be divided into two categories. Either they are
a preformed system that is applied as a solid roll or sheet, or they are a liquid system
that is applied by squeegee or in a similar manner. Further division of these depends
46


on the material of the overlay. Figures 3.4.2 and 3.4.3 show the breakdown of these
two divisions.
PREFORMED WATERPROOFING SYSTEM
Asphalt-Impregnated Polymer Elastomer i i Asphalt-Laminated
Fabric | Board
butyl poiyisoprene
bituminized
polyethylene
laminated
polyethylene
.. r.............. i
chlorosullonated ethylene
polyethylene propylene
ethylene vinyl polymer
acetate plasticized
polyvinyl chlonde
Figure 3.4.2: Division of Preformed Waterproofing Systems (Manning, 1995)
LIQUID WATERPROOFING SYSTEM
Bituminous
solutions.-'
compositions
refined
Resinous
mastic polyurethane
asphalt
elastomer
modified
I
epoxy acrylic
PMMA PMMA
urethane
modified
blended
bitumen
solutions
polymer
modified
compositions
elastomer/
carborundum
modified
coal tar coal far
modified modified
(unfilled! (mineral fillerl
elastomer/
coal tar
mod il ted
fast cure/
elastomer
modified
coal tar
and urethane
modified
Figure 3.4.3: Division of Liquid Waterproofing Systems (Manning, 1995)
47


Due to its widespread use, asphalt concrete is cheap to place. Typical DOT
costs are listed between $1.00 and $3.00 per square foot ($10.76 and $32.29 per
square meter) depending on binder properties and desired thickness. In some ways the
costs can be even lower since the equipment and personnel with experience are
common enough that special training is not required. Waterproofing membranes cost
an additional $1.00 to $2.00 per square foot ($10.76 to $21.53 per square meter)
(CDOT, 2011).
3.5 Testing of Overlays
The primary way to determine the performance of any roadway surface is through
testing. Over time, numerous test methods to determine different roadway properties
have been developed. The majority of the physical properties tested are: texture
depth, surface friction, bond strength, permeability, and chloride content. Each one of
these properties has several different tests that can be conducted, each with different
accuracies and precisions. The most common methods and the theory behind them are
examined and discussed.
3.5.1 Texture Depth
The primary influence on pavement-tire surface interaction is pavement texture. The
surface texture affects among other things; friction, noise, rolling resistance, and tire
48


wear. The texture itself is the result of aggregate texture and gradation, pavement
finish, and surface wearing. Surface texture is defined by two different characteristics,
the amplitude, and wavelength (ACPA, 2007). The amplitude is the vertical
measurement from the highest texture point, to the lowest. The wavelength is how
often amplitude is repeated. Each has an effect on tire-pavement interaction. The size
of the amplitude and wavelength can be used to divide pavement texture in four
different categories: microtexture, macrotexture, megatexture, and roughness. Figure
3.5.1.1 shows the different texture classifications and how they influence different
properties.
Texture 0.1 mil 1 mil 10 mil 0.1 in. 1 in. 1ft 10 ft 100 ft
Wavelength 1 pm 10pm 100pm 1mm 10mm 100mm 1m 10m 100m
PI ARC
Category
*
W
c
01
3
5=
s
K
X
\J
01
u
n
c
£
>
10
a.
Macmtextwre
1
I
J
tilde Quality {Smooth****! rs
.Vehicle Wear a
UVWhiCtft NOilft
Splach and Spray
Wet WeatfcW Friction
Dry Weather Friction
Key:
Good
Figure 3.5.1.1: Surface Texture and Surface Characteristics (ACPA, 2007)
49


Microtexture is defined as having a wavelength less than 0.02" (0.508 mm),
and amplitude less than 0.008" (0.2 mm). It is primarily responsible for stopping
ability on dry pavement. It has some stopping ability on wet pavement, but only when
vehicle speeds are less than 50 mph (80 kph). There is some difficulty in directly
measuring microtexture due to its small size.
Macrotexture is 0.02" to 2" long (0.508 mm to 5.08 cm),with an amplitude of
0.004" to 0.8" (0.102 mm to 2.03 cm). It is the primary determinant of stopping
power in the presence of water or ice on the roadway. This is because the larger peaks
of the texture in this range sit above the water or ice level, and still provide traction.
Megatexture is between 2" and 20" (5.08 cm to 50.8 cm), with an amplitude
between 0.004" and 2" (0.102 mm to 5.08 cm). Megatexture has no real effect on
traction, but does affect vehicle wear. Texture variations in this range are typically the
results of certain construction methods and practices. Unevenness is simply anything
larger than megatexture and affects vehicles in the same way.
There are two broad methods for measuring surface texture. The first is used
to find the mean texture depth (MTD) and is roughly used to measure macrotexture.
The second is to find the mean profile depth (MPD), and is used to measure
macrotexture through unevenness.
50


The mean texture depth is measured using a volumetric method commonly
referred to as the sand patch test, which is described in detail in ASTM E965. It uses
a known volume of sand spread on the roadway surface to determine the mean depth.
The known volume of sand is spread in a circle on the roadway until the peaks in the
texture are level with the sand. The diameter of the circle is measured multiple times,
and an average is used to calculate the mean texture depth based on the known
volume of sand.
The methods to find the mean profile depth differ depending on the type of
equipment being used. To measure longitudinal road profiles, the oldest systems used
physical displacements of multiple wheels to generate a profile graph of the roadway
surface. The newest systems use lasers to directly measure the surface profile. (ACPA
2007)
Measuring texture depth is important for two reasons. The first is that for
larger wavelengths, the surface ride can be determined, as well as problem spots in
underlying roadbeds. The second is that comparing the mean texture depth over time
helps to determine the durability of the road surface. Many thin bonded overlays use
an aggregate wearing course that wears down over time. An overlay that sees a
constant decrease in its mean texture depth will most often need replacing sooner than
an overlay that is maintaining its mean texture depth. The mean texture depth depends
51


highly on the type of overlay; the change in mean texture depth is what is important,
not the magnitude.
3.5.2 Surface Friction
Surface friction (or skid resistance) is the force that develops at the pavement-tire
interface to resist movements of the tire due to acceleration, deceleration (braking),
and lateral forces (sliding). The higher the surface friction, the more resistance to
slipping and sliding a vehicle has, and the safer a pavement surface is. By measuring
surface friction over time, the performance of a roadway can be tracked. It is expected
that a pavements surface friction will decrease over time, primarily due to wearing.
Pavements that have a surface friction that decreases too quickly become unsafe and
uneconomical, since they require replacement faster.
Several different factors determine the surface friction of a pavement. The two
key ones are texture, and moisture. During dry weather, the microtexture provides the
primary surface resistance. However, when the roadway becomes wet or icy, the
microtexture is typically flooded and the macrotexture becomes the primary
resistance surface (ACPA, 2007).
Moisture on roadways is typically found in the form of liquid water or ice.
Water acts as a lubricant at the pavement-tire interaction, thus decreasing the
52


resistance to movement of the tire. Ice acts in much the same way; it freezes to the
texture of the surface thus completely negating the pavement-tire interface and
functionally reducing the resistance to movement to zero (Caltrans, 2008).
It makes sense that any new overlay installation has its surface friction
monitored. Typically, baseline values are obtained from the original surface prior to
overlay placement. These values are then compared to the overlay values just after
installation and at different intervals later on. It is desired that the overlay will
maintain a higher surface friction than traditional pavement over the course of its
service life, and several different methods are available to measure this.
3.5.2.1 Locked-Wheel Friction Test
The traditional method of measuring surface friction is using the locked-wheel trailer
with either a ribbed or smooth tire. This method was first standardized as ASTM
E274 in 1965. The locked-wheel test system simulates an emergency braking scenario
by using a trailer with a locked wheel towed behind a truck. Figure 3.5.2.1 shows the
trailer and test truck.
Figure 3.5.2.1.1: Locked-YVheel Friction Test System (ICC, 2006)
53


The force imparted on the locked wheel at a constant speed, usually 40 mph (65 kph),
is measured and corresponding friction number is computed. The friction number is
determined based upon the speed and weight of the trailer. It is calculated as 100
times the force to pull the trailer at the specified speed, divided by the weight of the
wheel load. Most systems use water sprays to simulate emergency braking in wet
weather. Both ribbed and smooth tires are used because they respond to the pavement
differently. Figure 3.5.2.1.2 (a) shows a ribbed test tire, while (b) shows the smooth
test tire.
Figure 3.5.2.1.2: Locked-Wheel Test Tires (a) Ribbed, and (b) Smooth
The ribbed tire is fairly insensitive to water on the roadway and thus is a good
indication of microtexture performance. The ribbed tires that are used have much
larger flow channels for water than normal roadway tires, and often correspond
poorly to real world vehicle performance during wet or icy conditions. Smooth tires
make up for this deficiency by being more sensitive to macrotexture. The
54


microtexture becomes slippery with moisture, forcing the smooth tire to grip the
macrotexture. Because of this, smooth tire friction numbers are typically less than that
of ribbed tires on the same roadway surface. The measured friction numbers should
ideally be above 35 (Caltrans, 2008). Less than 35 usually mean the roadway
becomes dangerous under wet conditions.
Two newer variations of the locked-wheel friction test exist. The fixed and
variable slip systems attempt to simulate the use of anti-lock brakes. For the fixed slip
system, this is performed by allowing the test wheel to experience limited slip, or
slight movement on the order of 10-20%. The variable slip system automatically
changes the slip percentage through a pre-defined set of levels during the test process.
Neither system is in widespread use, most likely due to cost.
A variation of the locked-wheel test system is the side force tester. Side force
testers are used to determine the ability of a vehicle to maintain control in curve and
conform to ASTM E670. The test wheel is free to rotate, but is fixed in a plane, at a
yaw angle to the direction of motion. The side force that is imparted on the tire
perpendicular to its rotation is measured. These systems take continuous
measurements instead of spot measurements at certain time intervals (Caltrans, 2008).
55


3.5.2.2 British Pendulum Tester
The British Pendulum Tester (BPT) is another method of measuring the
friction of a roadway surface. It was originally developed to measure pavement
friction in a laboratory setting, but has been finding use in the field due to its compact
size and ability to measure smaller roadway lengths. Because the pendulum arm
moves slowly, about 6 mph, the BPT is a good indicator of pavement microtexture,
and thus dry weather performance (Caltrans, 2008). Figure 3.5.2.2 shows a pendulum
tester.
Figure 3.5.2.2: British Pendulum Tester
The BPT operates on the conservation of energy principle. The swing arm is released
and strikes the pavement surface at a designated point. The pavement surface absorbs
56


some of the kinetic energy of the swing arm, slowing it down. As the contact pad on
the swing arm breaks contact with the surface, it goes into a recovery stroke and
pushes an indicator arm up a scale. The more speed at which the arm recovers from
the pavement contact, the higher the indicator ami is pushed up the scale. The scale
correlates the friction of the surface with the British Pendulum Number (BPN). The
BPN has no applicability to other skid numbers generated by other test systems
(Caltrans, 2008).
3.5.2.3 Dynamic Friction Tester
The Dynamic Friction Tester (DFT) is an additional method of measuring the surface
friction of a roadway. It uses a spinning disk with 3 spring-loaded rubber feet that is
lowered onto the roadway surface (Caltrans, 2008). Water is sprayed on the surface to
reduce the friction. Figure 3.5.2.3 shows a typical DFT.
Figure 3.5.2.3: Dynamic Friction Tester (Caltrans, 2008)
57


As the disk is slowed by the surface friction, the torque is measured and a
corresponding friction number is determined. The disk is usually spun up to 55 mph
and decreases to about 3 mph. DFT's have the advantage of being able to report peak
friction as well as being calibrated to report the friction number relative to the
International Friction Index. However, the DFT cannot account for directional
textures in pavements due to its rotational nature (Caltrans, 2008).
3.5.2.4 Surface Friction Values of Thin Bonded Overlays
Friction values for different road surfaces depend on the type of test used. Since the
majority of DOTs in the United States use the locked-wheel test, the numbers
presented here will be from that test. Normally, for any road surface, a friction
number (FN) greater than 35 is deemed acceptable (Caltrans, 2008). Less than 35
usually mean the roadway needs replacement due to poor texture and skid resistance.
Several states have evaluated SafeLane and have published their data. Since SafeLane
has really only been around since 2001, there is a limited number of available studies,
but several are examined here.
Sprinkel (2003) has published results containing data from the previous 25
years of polymer overlays. Among the result is friction numbers for the majority of
different polymer overlays. Figure 3.5.2.4.1 shows friction numbers corresponding to
overlay types over time.
58


Figure 3.5.2.4.1: Friction Numbers of Polymer Overlays (Sprinkel, 2003)
The data shows that the majority of polymer overlays maintain their skid numbers
over their service life. MMA overlays are shown to have reduced skid numbers over
their lifespan, contrary to what has been found in literature.
VDOT installed several different hydraulic cement based overlay 1994. They
tested these overlays with both with smooth and ribbed tires. The results from this
study are shown in Table 3.5.2.4.1
Table 3.5.2.4.1: Skid results from VDOT trial overlays (Sprinkel, 2000)
Bridge No. Overlay Type 1994 Bald Tire 1994 Treaded Tire 1999 Bald Tire 1999 Treaded Tir e
1 SF 46 45 45 41
2 MMLMC 45 45 45 41
3 LMCHE 42 42 48 45
4 SFHE 51 51 51 51
5 ML 38 41 46 43
6 LMC 33 31 49 46
59


All the overlays tested showed fairly standard numbers for hydraulic concrete based
overlays. Each overlay also shows that it was able to keep it skid numbers over 5
years. Interestingly enough, the LMC overlay actually improved its skid resistance.
However, this could be due to wearing of the surface, which provides better traction,
but ultimately less durability and chloride protection.
WSDOT has experimented with several epoxy and MMA type polymer
overlays throughout the 80's and have a large amount of data including bond, friction,
and permeability test results. The epoxy overlays used were PolyCarb Flexogrid on
Chehalis and Snake roads, and Dural Flexolith on Thrall Rd. The MMA overlay was
Degussa Degadur 330. The friction results are shown below in Figures 3.5.2.4.2 and
3.5.2.4.3.
Epoxy Polymer Overlays]
Figure 3.5.2.4.2: Skid Results from WSDOT Epoxy Overlays (Wilson, 1995)
60


MMA
Polymer Overlays
I
Figure 3.S.2.4.3: Skid Results from WSDOT MMA Overlays (Wilson, 1995)
The WSDOT results show that MMA overlays maintain, and in case of this study
increase their surface friction over time. This is contrary to the Sprinkel (2003)
results, but more in line with what traditional literature says. The epoxy overlays
faired equal to the MMA overlays in the beginning, but quickly lost surface friction
over time. It should be noted that these overlays were installed in the 80's and
WSDOT revised their aggregate gradation specifications in 1991 to provide increased
lifespan of their polymer overlays (Wilson, 1995).
VDOT has two different studies involving SafeLane. The first is a study of
high friction surfaces that compares several different products. Unfortunately, their
test methods in that report do not correspond well to the methods traditionally used,
61


and which are presented here. A second report, solely on SafeLane was published in
2009. For the that study, VDOT used a standard SafeLane mix, a 75% SafeLane /
25% silica aggregate mix, and two VDOT modified EP-5 overlays. Table 3.5.2.4.2
shows the results of the skid resistance tests.
Table 3.5.2.4.2: VDOT skid numbers for SafeLane (Sprinkel, 2009)
Bald Tire Skid Numbers for Travel Lanes of I-8I Before aud After Installation of Overlays
Structure No. Overlay Date Placed Date of Test
6 28 04 10 3105 12 "05
2037 2-laver SafeLane 9:05 28 59 46s
2024 2-layer SafeLane1 1005 27 60 53*
2025 1-layer VDOT modified EP-5 8/05 26 57 -
2036 t-laver \DOT modified EP-5 8:04 22 49 -
Aggregate blends: 75% SafeLane'25% quartz by weight for Spans 1 and 2 and 50% SafeLane'50% quartz
for Spans 3 and 4.
4 After liquid chloride pretreatment.
While the period between tests is fairly short, the VDOT study shows that SafeLane
provides excellent skid numbers after installation. Even during wet conditions, the
skid numbers remain very good.
Minnesota DOT installed SafeLane on four different sites in 2006 and 2007.
Their data has shown mixed results. The skid resistance results from their study are
shown in Figure 3.5.2.4A
62


Skid Test Results or the Hibbing Installation

rib ave b smt a*e
if 'f ^ f of f if ^ cf ^ f if ^
s?
Date
Figure 3.5.Z.4.4: Mn/DOT Skid Numbers for SafeLane (Evans, 2010)
Mn/DOT reports a fairly steep decrease in skid numbers over the course of three
years. While initial numbers were very good, they dropped to barely being
acceptable. In the report, Evans (2010) notes that SafeLane, while not suffering de-
bonding failures, tended to have the aggregate shear off. Figure 3.5.4.2.5 documents
this effect.
63


Figure 3.5.2.4.5: Wearing of SafeLane After 26 Months (Evans, 2010)
One of the largest culprits of this effect was down force from snow plow blades. It
should be noted that Cargill recommends not applying additional down force on plow
blades for this exact reason. It is possible that better traction numbers may have been
reported if less aggressive plowing was used (Evans, 2010).
3.5.3 Bond Strength
Bond testing is conducted to determine the bond strength of the aggregate to the
polymer overlay, as well as the strength of the overlay to the bridge deck. The typical
test used, follows ASTM Cl583 or ACI 503R. Almost all test system used in bond
64


testing have some sort of steel disk adhered to a scored circle on the roadway surface,
and to which a hydraulic or threaded pull device is attached. Force is applied to the
disk until the substrate, substrate/overlay interface, or overlay fail in tension
according to ASTM Cl583. Thin bonded overlays rely on the underlying bridge deck
for structural support, and if that bond fails, the overlay will spall off. As such, bond
testing of thin bonded overlays is highly critical. Figure 3.5.3.1 shows a standard test
setup based upon ASTM Cl583. Figure 3.5.3.2 shows the different failure modes
from ASTM Cl583.
Tensile load axis
coincident with corn axis
and perpendicular to
concrete surface
Tensile loading
device
Swivel joint
Circular cut through
overlay or repair
material to at least 10
mm (0.5 in.) below
interface
Steel disk
' Diameter 50 mm [2.0 in.]
Thickness >25 mm [1 0 in.]
Figure 3.5.3.1: Typical Bond Test Setup According to ASTM C1583
65


Failure in
substrate
Bond failure at
concrete/overlay
interface
Failure in
overiav or repair
material
Bond failure at
epoxy/overlay
interface
Figure 3.S.3.2: Bond Test Failure Modes According to ASTM C1583
AASHTO Task Force 34 found that 250 psi (1.72 MPa) should be the
minimum acceptable bond strength for a polymer overlay. This value reflects a
properly mixed and applied polymer concrete that will develop good adhesion. An
examination of different polymer concrete systems shows that their manufacturers
either specify 250 psi (1.72 MPa) or that the bond failure should occur in the
substrate. SafeLane requires at least 250 psi (1.72 MPa), based upon ACI 503R at 24
hours. The T-48 epoxy overlay system by Transpo, Inc. requires the same. Based
upon this examination, 250 psi (1.72 MPa) seems like a reasonable number to expect
for a polymer based overlay (WSDOT, 1995).
66


3.5.3.1 Bond Strength of Thin Bonded Overlays
Sprinkel's (2003) report on different polymer overlays over the last 25 years included
a tensile rupture test of thin bonded overlays. The bond tests were conducted using
the Virginia Test Method 92, which is equivalent to ACI 503R. Figure 3.5.3.1.1
shows the results from the report.
M-LEU Multiple Liya: Epoxy Urethane
M-L E Multiple Layer Epoxy
? P Pranixei Polyester
M-L P Multiple Layer Polyester
Figure 3.5.3.1.1: Bond Strength of Different Polymer Overlays (Sprinkel, 2003)
Sprinkel's (2003) results show that bond strength for the majority of polymer overlay
types slightly decreases with age. The two epoxy overlays show excellent bond
strength, well above the recommended 250 psi (1.72 MPa). In addition, it is
67


noteworthy that the multilayer epoxy urethanes, similar to what Flexogrid uses, has
higher bond strength than straight epoxies, w'hich SafeLane utilizes.
VDOT's 1994 testing on SF and LMC overlays also included bond testing.
The original data was given in SI units, and for the sake of convenience; these
numbers were converted before being presented below in Table 3.5.3.1.1.
Table 3,5.3.1.1: Bond strength of hydraulic concrete overlays (Sprinkel, 2000)
Property Thic mess Bond Strength
Date 1994 1999 1994 1999
Overlay Type inch cm inch cm psi MPa psi MPa
SF 1.34 3.4 1.22 3.1 145 1.0 145 1.0
MMLMC 1.46 3.7 1.34 3.4 102 0.7 145 1.0
ML 1.3 3.3 1.26 3.2 58 0.4 189 1.3
LMCHE 1.65 4.2 1.57 4.0 87 0.6 189 1.3
SFHE 1.57 4.0 1.54 3.9 116 0.8 160 1.1
ML 1.3 3.3 1.18 3.0 131 0.9 131 0.9
LMC 1.5 3.8 1.38 3.5 116 0.8 203 1.4
Hydraulic cement based overlays seem to have lower adhesive bond strengths than
polymer based overlays. None of the overlays tested by VDOT exceed the 250 psi
(1.72 MPa) recommendation for polymer overlays. Sprinkel (2000) noted that all the
failures were below the bond interface of the overlays. It was hypothesized that the
low failure strengths of the underlying concrete decks was due to damage from the
68


milling machine, and that the results may not be representative of the actual bond
strength of the overlays.
The bond tests from the WSDOT report are shown in table 3.5.3.1.2 for epoxy-
overlays, and table 3.5.3.1.3 for MMA overlays. They show several interesting
results.
Table 3.5.3.1.2: Bond tests for WSDOT epoxy overlays (WSDOT, 1995)
Bridge Number Brand Name Year Applied Initial Ave Bond (psi) {no. of Jests) Latest Ave. Bond (psi) {no. of tests) Overlay Age @ Latest Bond test
161/10 EPI/Flex III 1986 294 (10) not tested
*2/115S Concresivc 3070 I987 392 {8} 276 {3} 3 years
5/316 EPI/Flex IB 1990 363 (15) 266 {5} 4 years
82/10S Flexolith 1985 359 {12} 355 {5} 3 years
900/12W Flexolith 1986 201 {15} 327 {6} 5 years
101/115 Flexognd 1984 399 {6} 191 {5} 4 years
12/915 Flexognd 1986 259 {21) 252 {6} 3 years
167/102 FI exogrid 1987 267 {5} 377 {3} 1 year
167/104 Flexognd 1987 215 {5} 257 {3) 1 year
167/106 Flexognd 1987 342 {5} 287 {3} 1 year
104/5.2 Flexognd 1988 308 {27} 244 {6} 4 years
529/20E Flexognd 1988 267 {5} 187 {6} 3 years
52 9/20W Flexognd 1988 207 {5} not tested
1 Average 297 274
69


Table 3.5.3.1.3: Bond tests for VVSDOT MM A overlays (WSDOT, 1995)
Bridge Number Brand Name Year Applied Initial Ave. Bond (psi) Latest Ave. Bond (psi) Overlay Age @ Latest Bona test
5/523 £ Conxryl 198X 162 not tested
82/114S Concresivc 2020 1987 284 258 3 years
27/3 Silikal R66 1990 229 not tested
101/514 Degadur 330 1985 155 128 3 years
4/106A Degadur 330 1986 113 85 5 years
167/2 IE Degadur 330 1987 290 111 1 year
512/40N Degadur 330 1987 259 135 1 year
16M20 Degadur 3 30 1988 189 not tested
97 '2 Degadur 330 1989 217 not tested
Averaec 211 143
The WSDOT study shows several things. The first is that initial bond strengths are
usually the highest, and that the bond strength of epoxy overlays decreases with time.
Flexogrid in this study performed fairly well; 2 of the 8 averages failed to meet the
250 psi (1.72 MPa) requirement. It should be noted that Flexogrid does not have a
250 psi (1.72 MPa) bond requirement in their specifications, it is only specified that
bond failures should occur in the concrete. However, the report fails to mention the
location of bond failures. Also of interest is that epoxy overlays seem to have better
long term bond characteristics than MMA overlays. The MMA overlays had very low
bond strengths in general, less than the recommended 250 psi (1.72 MPa).
70


The bond strength of SafeLane is expected to be comparative to other thin
bonded overlay systems. Several different DOT reports conducted on SafeLane also
included bond tests, and these results are examined here.
VDOT's 2009 study conducted bond testing on two different SafeLane
overlays. Although they used their own VDOT Test Method 92, it is comparable to
ACI 503R. Table 3.5.3.1.4 show's the results from the VDOT report.
Table 3.5.3.1.4: VDOT bond test results for SafeLane (Sprinkel, 2009)
Bond Strength Results (February 15, 2006)
Structure No. Overlay Tensile Rupture Strength (psi) Failure Location (*-) Thickness (in)
Overlay j Bond Base
2037 2-laver SafeLane 205 o o 100 0.49
2024 2 1avet SafeLane" 218 56 21 23 0.46
2025 Mayer VDOT modified EP-5 274 0 0 100 0 10
2036 1 laver VDOT modified EP-5 230 3 95 0.10
Aggregate blends: 75o SafeLane /25% quartz by weight for Spans 1 and 2 and 50'o SafeLane/50o quartz for
Spans 3 and 4.
VDOTs results show that both SafeLane overlays failed at less than the recommended
250 psi (1.72 MPa). However, the results show that, with the exception of the mixed
aggregate deck, all failures wrere within the concrete. The bond strength was a
limitation of the tensile strength of the concrete, and not the overlay.
71


3.5.4 Permeability
Permeability is the ability of a material to allow a fluid or gas to move through it. It is
an important aspect of bridge structural decks and wearing surfaces since it
determines how well protected embedded reinforcing steel and underlying structural
steel are protected from chloride attack. It is desirable for bridge deck materials to be
as close to impermeable as possible to offer the best protection for structural steel.
Measuring the direct permeability of concrete is difficult since it is determined
by the size, amount, and arrangement of the three-dimensional pore structure within
the concrete. Because of this, several different methods have been developed to
measure the permeability of concrete over time.
3.5.4.1 Ninety-Day Ponding Test
The first method developed was AASHTO T 259 (ASTM Cl543), commonly known
as the 90-day ponding test. As its name implies, the permeability of the concrete is
measured indirectly through the amount of chloride that has penetrated the sample. A
sample slab is obtained and moist cured for a period of either 14 days, with 28 days of
drying (AASHTO T 259) or 14 days of each (ASTM Cl 543). The slab is then
prepared by attaching dikes on the top, and waterproofing the sides. The top dikes are
filled with a 3% saline solution which is maintained at a test level for 90 days. At the
end of 90 days, the sample is cored and the chloride content is determined at different
72


depths using ASTM Cl 152, or a similar method. Based upon the amount of chlorides
measured, and the amount of time that they accumulated, the permeability can be
determined.
The main advantage to this method is that it simulates the same mechanisms
in which chlorides penetrate bridge decks. The disadvantage is that test takes 90 days
at minimum, not including moist curing and drying times. For very low permeable
concretes, the 90 days is often not enough time to allow sufficient ingress of chloride
ions.
3.S.4.2 Rapid Chloride Permeability Test
The rapid chloride permeability test (RCPT) is an electrical method used to determine
the approximate permeability of concrete based upon the flow of electrons through a
sample saturated with sodium chloride on one side and sodium hydroxide on the
other. The standard governing this test is the AASHTO T 277 or the similar ASTM
C1202. Figure 3.5.4.2 shows the layout of the RCPT.
73


Figure 3.5.4.2.1: RCPT Basic Setup (Germann Instruments, 2010)
The RCPT does not directly measure the depth or penetration of the concrete sample
by chloride ions. Instead, it applies a known voltage to the specimen and determines
the current, based upon the resistance of the concrete. The chlorides are driven into
the concrete by the voltage potential, which comes in contact with the NaOH solution
and create an electric circuit. The number of coulombs is a function of the current and
the time the test is run. The specifications call for 6 hours of voltage applied to the
specimen; however similar results have been reported for running the test for 3 hours
and doubling the recorded value. Table 3.5.4.2.1 below shows typical permeability
74


levels with each type of concrete, as well as which coulomb range belongs to each
classification.
Table 3.5.4.2.1: Permeability classification (Grace, 2006)
Charge Passed (Coulombs Chloride Permeability Typical of
>4.000 High High W/C ratio (>0.60) conventional PCC
2.000-4.000 Moderate Moderate W/C ratio (0.40-0.50) conventional PCC
1.000-2.000 Low Low W/C ratio (<0.40) conventional PCC
100-1.000 Very Low Latex-modified concrete or internally-sealed concrete
<100 Negligible Polymer-impregnated concrete. Polymer concrete
The advantage of the RCPT is the speed at which the test can be run.
Including sample preparation time, the test can be run in 24 hours. Some test
instruments allow up to 8 samples to be tested at once.
The biggest disadvantage of the RCPT is the significant variation of results.
The same samples run twice will produce quite different results. According to ASTM
C1202, the coefficient of variation for a single operator is 12.2%, which translates to
up to a 42% difference between the same concrete samples, prepared the same way,
from the same batch. Between different laboratories, the C.O.V. jumps to 18%, which
leads to a difference of 51% between similar samples.
75


Many DOTs are starting to specify the RCPT coulomb value for a given
concrete mix at certain days of age. This can lead to problems with mixes that specify
certain admixtures. Depending on the type of admixture, the permeability of the
concrete can be lower or higher than expected at a given age. Mixes that incorporate
fly ash and/or blast furnace slag in particular, tend to take longer than 28 days to
reach their final permeability level. While mixes that incorporate any type ionic salt
including; calcium nitrite, calcium nitrate, calcium chloride, or sodium thiocyanate,
tend to facilitate the transfer of chlorides and thus show higher coulomb counts than
would otherwise be reported. Because of this, ASTM Cl 202 specifies that
permeability should be based upon the qualitative terms in Table 3.5.4.2.I. (Grace,
2006)
3.5.4.3 Permeability of Thin Bonded Overlays
Many different reports exist on thin bonded overlays used by different states. Similar
to other sections, VDOT and WSDOT have published extensive reports on their use
of thin bonded overlays, and the results of their trials are examined here.
Sprinkel (2003) also tested permeability of polymer concrete overlays. His
results are shown in Figure 3.5.4.3.I.
76


Polymers in Concrete: The First Thirty Years
M-L E U Multiple Layer Epoxy Urethane
M-L E Multiple Layer Epoxy
?? Premixea Polyester
M-L P Multiple Layer Polyester
Figure 3.5.4.3.1: Permeability of Polymer Overlays Over Time (Sprinkel, 2003)
Sprinkel (2003) confirms what has been known; that polymer based overlays provide
very low to negligible permeability. The multi-layer polyester overlays do not fare as
well as the others, but still provide very low permeability for 6-8 years.
VDOT overlay tests from 1994 show that hydraulic concrete overlays can
have low to very low permeability as well. Table 3.5.4.3.1 shows the results and
differences between SF and LMC overlays over 5 years.
77


Table 3.5.4.3.1: RCPT results for hydraulic concrete overlays (Sprinkel, 2000)
Date 1994 1999
Overlay Type Coulombs
SF 1081 911
MMLMC 2533 795
ML 327 670
LMCHE 1665 513
SFHE 815 780
ML 1211 696
LMC 1296 454
Both LMC and SF overlays show excellent permeability results in the low to very low
range. This is likely due to the increased density of both overlay types from the latex
in the LMC's and the silica fume in the SF concretes. The results also validated that
permeability decreases with age for hydraulic concrete overlays.
WSDOT also measured permeability for their overlays. Table 3.5.4.3.2 shows
each overlay type they have installed, as well as measured permeability values.
Table 3.5.4.3.2: WSDOT overlay permeability results (Wilson, 1995)
Overlay Tvt>e Bmi Average
Polymer-Epoxy 0-6 3
Polymer-MMA 0-0 0
Latex Mod. concrete 101-1,117 365
Micros ilica concrete 149-1,410 577
Low Slump concrete 438-2,400 1,443
Standard WSDOT bridge deck cone. 1,400-6,840 2,983
78


Similar to the VDOT results, Polymer overlays exhibit negligible permeability while
the LMC and SF concrete overlays show low to very low permeability (Wilson,
1995).
VDOT's 2009 study conducted RCPT on two different SafeLane Overlays.
Table 3.5.4.3.3 shows the results from the VDOT report.
Table 3.S.4.3.3: VDOT permeability values of SafeLane (Sprinkel, 2009)
Average Permeability Value1:- (2 15'06)
Structure No. Overlay Thickness (in) Permeability (coulombs)
2037 2-layer SafeLane 0.54 23 (negligible)
2024 2 -layer SafeLane 0 0.45 246 (very low)
2025 1 -layer VDOT modified EP-5 0 11 1367 (low)
2036 1 -layer VDOT modified EP-5 0.11 1226 (low)
c Aggregate blends: 75% SafeLane 25% quartz by weight for Spans 1 and 2 and 50% SafeLane'50%
quartz for Spans 3 and 4.
As typically seen for polymer based overlays, and especially epoxy overlays, the
permeability of the SafeLane overlay is negligible. The hybrid SafeLane overlay
tested still recorded very low permeability, even though it uses a mix of aggregates.
3.5.5 Chloride Content
Chloride testing is done to determine the amount of chlorides in a given
sample of concrete. This test differs from the RCPT in that it does not measure the
permeability of a given sample, but rather the total amount of chlorides in the sample.
79


This could be considered a component of the 90-day permeability test. Chloride
testing is important for two reasons. It indirectly measures the permeability of
concrete, and also lets engineers know at which point they can expect corrosion to
occur.
In the absence of protection, steel rusts from exposure to moisture, the
chemical process of corrosion. In normal concrete, the steel develops its own
protection in form of a thin oxidation layer from the high alkalinity of concrete. When
chlorides are present, the corrosion process is accelerated through an autocatalytic
process in which the chlorides attack and destroy this layer (Kosmatka, 2002).
The main method to determine the chloride content of concrete is using
ASTM Cl 152, which has many variations or AASHTO T 260. ASTM Cl 152
references ASTM Cl 14 for actual laboratory procedure, which finds the PPM of
chloride in a concrete sample by titration of the chlorides using Silver Nitrate
(AgN03). In some cases, chloride content can be calculated to pounds per cubic
yard. The test is a laboratory intensive procedure that requires significant preparation.
Some of the variations put out by DOTs use newer methods in which several of the
ASTM Cl 14 steps are condensed by use of specially developed equipment.
Testing is typically performed using 2" (5.08 cm) core samples in which three
sections at different depths are tested for their chloride content. Usually the first depth
is at or just below the surface, w ith additional sections every 1/2" (1.27 cm) thereafter
80


to provide adequate space for saw cutting. Figure 3.5.5.1 shows possible test
locations.
Sample 1 (0 in.)
Sample 2 (1 in.)
Sample 3 (2 in.)
Figure 3.5.5.1: Typical Core Test Locations
Some newer mechanical methods can take much smaller cuts and thus take more
sections per core. The cores are prepared by oven drying, and then ground into
powder. The powder is placed into distilled water, boiled and filtered. This filtered
solution is then titrated to produce the results. ASTM Cl 14 requires silver nitrate to
be added till 60 millivolts is read on a volt meter. Newer methods titrate the sample
based upon color.
At the very least, chloride content is shown as a function of depth of the
sample. The first slice, closest to the wearing surface should have the highest
concentration; with each subsequent sample have less chlorides present. The most
2 in.
------------
81


important sample is the one closest to the reinforcing steel, since this determines if
corrosion will be occurring, termed critical chloride content threshold.
Critical chloride content threshold is the point at which there is sufficient
chloride ions present around reinforcing steel to breach the protective film and cause
corrosion. With some controversy, AC1 318 Build Code limits chlorides by percent
by w eight of concrete depending on the service conditions of the structural member.
These values are shown in Table 3.5.5.1.
Table 3.5.5.1: ACI 318 Chloride Limits of Concrete (Kosmatka, 2002)
Type of Member Maximum Water Soluble Chloride Ion (C1) in Concrete. Percent bv Weight of Concrete
Prestressed concrete 0.06
Reinforced concrete exposed to chloride In service 0.15
Rel.Tioreed concrete that will be dry or protected from'
moisture In service 1.00
Other reinforced concrete construction 030
If the ACI 318 standards are followed, it is ideal for a bridge deck that the chloride
content closest to the rebar does not exceed 0.15% chlorides by weight of concrete.
Past this point it is likely that the chlorides will be in a high enough concentration to
penetrate the protective coating of the steel reinforcement and start the corrosion
process.
82


4. Problem Statement
One of the largest problems facing transportation departments within the United
States is the deterioration of bridge decks and structural members. Specifically in cold
weather climates where the use of harsh deicing chemicals results in rapid corrosion
of steel reinforcement, engineers have spent a considerable amount of time trying to
determine effective and economical ways to combat the deterioration of bridges.
Beginning in the early 1960's engineers began to experiment by applying a thin layer
of concrete or polymers onto bridge decks as a wearing surface. The thin bonded
overlay was bom.
The overall goal of this thesis is twofold. First, is to determine the viability of
using the SafeLane system as an anti-skid/anti-icing and protective bridge deck
overlay for concrete and asphalt bridge decks. Second, is to determine the
effectiveness of Flexogrid as an anti-skid and protective overlay. Several different
factors have been examined to make this determination including the application
process and protection of the bridge deck, as well as durability, cost, and its
effectiveness as an anti-skid/anti-icing surface. Extensive research and testing has
been put into thin bonded bridge deck overlays, with SafeLane and Flexogrid being
evaluated with regards to those lessons.
83


Durability and low weight is a key advantage of thin-bonded overlays. In
general, thin-bonded overlays are effective at stopping chemicals from reaching the
underlying bridge superstructure and reinforcing steel. This study took a closer look
at effectiveness of thin-bonded overlays, and more specifically at how effective
SafeLane and Flexogrid are. Permeability and chloride tests were conducted to
determine if SafeLane and Flexogrid were effective barriers against chemical attack.
Two locations were selected in Colorado for the installation of SafeLane, and
one for Flexogrid. The first SafeLane location is the overpass of Interstate-76 and
Colorado Weld County Road 53. This site was selected because of its asphalt bridge
deck and the proximity of a weather station for instrumentation. The second site is the
Parker Road, Interstate-225 southbound flyover on-ramp in Aurora, CO. This site was
selected for its concrete bridge deck, high traffic volume, and moderate crash rate.
The final site selected was for the Flexogrid installations the flyover ramp from
southbound Interstate-25 to northbound Interstate-225. This ramp was selected
because of its traffic volume, but also because of the automated spray system that was
installed along with Flexogrid.
A large concern for thin-bonded overlays is the high installation costs
associated with them. The installation costs of thin-bonded overlays were examined
and due to their inherent high initial cost, it is important that thin-bonded overlays
have a high preventative maintenance value.
84


SafeLane and Flexogrid have been evaluated in several other states with
mixed results. Since weather, methods, and organizations vary from state to state, this
study looked at both systems in regards to Colorado. If they are found to be
successful at prolonging bridge deck life, as well as increasing the safety of bridges
during winter conditions, it is likely CDOT will recommend the use of thin-bonded
overlays for future bridge decks. This has the two fold effect of decreasing lifecycle
maintenance costs through prolonging the serv ice life of the bridge structures as well
as increasing the safety of a bridges through decreased weather related accidents.
85