Effects of concrete and asphalt surface treatment on bond strength of thin-bonded overlays

Material Information

Effects of concrete and asphalt surface treatment on bond strength of thin-bonded overlays
Bindel, Mary Katherine
Publication Date:
Physical Description:
xii, 107 leaves : ; 28 cm


Subjects / Keywords:
Pavements -- Overlays ( lcsh )
Pavements, Concrete ( lcsh )
Pavements, Asphalt ( lcsh )
Bridges -- Snow and ice control ( lcsh )
Bridges -- Floors ( lcsh )
Bridges -- Floors ( fast )
Bridges -- Snow and ice control ( fast )
Pavements, Asphalt ( fast )
Pavements, Concrete ( fast )
Pavements -- Overlays ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 105-107).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Mary Katherine Bindel.

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:
710983816 ( OCLC )
LD1193.E53 2010m B56 ( lcc )

Full Text

Mary Katherine Bindel
B.S., Civil Engineering, University of Colorado at Denver, 2009
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering

This thesis for the Master of Science
degree by
Mary K. Bindel
has been approved

Bindel, Mary Katherine (MS, Structural, Civil Engineering Department)
Effects of Concrete and Asphalt Surface Treatment on Bond Strength
Thesis directed by Assistant Professor Stephan A. Durham
Departments of Transportation are faced with maintaining and protecting their bridge
decks in an effort to prevent infiltration of water and aggressive chemicals into
bridge decks and supporting structural elements. Bonded overlays are designed to
provide a durable, cost effective wearing surface that reduces or even prevents water
and chlorides migration into the deck. The Colorado Department of Transportation is
evaluating a new thin-bonded overlay, SafeLane, for potential use on their bridge
decks. Though this product is costly, it is marketed as an anti-icing/anti-skid overlay
that can be used on both bare concrete decks and asphalt wearing surfaces.
This thesis investigated the bond strength of the SafeLane overlay to concrete
and asphalt surfaces. Specifically, this study examined the surface treatment applied
to the two pavement types to determine the effects of troweling, tining, sand-
blasting, and roughening on the bond strength of the overlay. A modified version of
the American Concrete Institute 503R, pull-off test, was utilized during the testing
phase of this study. This is an important test to help predict the effectiveness and
durability of the overlay. Failure of the overlay to adhere to the pavement surface
would result in a pathway for aggressive chemicals (i.e. deicing salts) to penetrate
and decrease the life of the bridge deck. In addition, delaminations of the thin-

bonded overlay have the potential to decrease safety of the motoring public as skid
resistance and anti-icing capabilities would be decreased. In this study, a laboratory
investigation was conducted by testing the bond strength of the SafeLane overlay on
concrete and asphalt block samples. Various surface treatments were performed on
the blocks prior to testing.
No clear surface preparation performed the best, none the less the data
suggests that the tined epoxy specimens and the roughened epoxy and aggregate
specimens produced the best adhesion. The ultimate failure stresses had a large
scatter ranging from 73 to 277 psi (0.503 to 1.91 MPa). The ultimate failure stresses
were on average lower than the required failure stresses. Most all of the samples
failed in the concrete thus the failure mechanism was more a concrete material
failure than an epoxy/overlay failure. The recommended value may have been
reached if the concrete had not failed.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Stephan A. Durham


I dedicate this thesis to my family for their patience and support. I never would have
pursued my masters degree without their help.

I genuinely thank my academic advisor, Dr. Stephan Durham, for his patience, and
continued help and guidance. In addition, I would like to thank Dr. Kevin Rens and
Dr. Fred Rutz for their participation on my thesis committee.
I would like to extend my sincere thanks to Mr. Adam Kardos, Dr, Rui Liu,
Mr. Randy Ray, Mr. Tom Thuis, Mr. Logan Young, and the many others who have
contributed to this work. In addition, I would like to thank Dr. Angela Hager, Mr.
William Pat Kennedy and the City and County of Denver for providing assistance
in obtaining the asphalt samples used in this research.
I would like to thank the faculty and staff of the University of Colorado at
Denver Civil Engineering Department, for their support and guidance throughout my
educational career at UCD.

1. Introduction........................................................1
1.1 Research Objectives.................................................3
2. Literature Review...................................................5
2.1 Literature Introduction.............................................5
2.2 Asphalt Overlays....................................................8
2.3 Concrete Overlays...................................................9
2.4 Surface Membranes..................................................12
2.4.1 Bond Strength Testing..............................................19
2.4.2 Pull-off Test......................................................21
2.4.3 SafeLane...........................................................23
3. DOT Survey.........................................................32
3.1 Department of Transportation Survey Findings.......................32
3.1.1 Survey Response....................................................32
3.1.2 Anti-Icing/Anti Skid Use...........................................34 Question #2 Written Response.......................................35
3.1.3 Winter Accidents...................................................37 Question #3 Written Response.......................................37
3.1.4 Skid Resistance....................................................38
3.1.5 Deicing Chemicals..................................................39 Question #5 Written Response.......................................40
3.1.6 Failures...........................................................41 Question #6 Written Response.......................................41
3.1.7 Specifications.....................................................42 Question #7 Written Response.......................................42
3.1.8 Anti-Icing Products................................................43
3.1.9 Deck Surfaces......................................................44
3.1.10 Comments...........................................................44 Question #10 Written Response......................................44

CO >
Response Summary................................................46
DOT Survey Conclusions..........................................49
Problem Statement...............................................51
Statement..................................................... 51
Experimental Design..............................................53
SafeLane Overlay Material........................................53
Experimental Design and Testing Matrix...........................58
Testing Configuration and Procedure..............................60
Test Configuration Design and Construction for Concrete Specimens...61
Test Configuration Design and Construction for Asphalt Specimens.64
Concrete Mixture Design and Sample Fabrication...................66
Mixture Design...................................................66
Form Construction................................................67
Concrete Batching................................................69
Specimen Fabrication.............................................70
Experimental Results.............................................78
Concrete Properties..............................................78
Data Results.....................................................82
Analysis of Data.................................................84
Surface Type Comparisons.........................................85
Epoxy vs. Aggregate Comparisons..................................86
Stress vs. Strain Graphs.........................................87
Manufacturers Adhesion Requirement...............................90
Problems Faced...................................................91
Problems with Test Setup.........................................92
DOT Survey......................................................97
Concrete Design Mixtures.......................................102

2.1 Effects of Magnesium Chloride and Salt Brine on Concrete.......11
2.2 Permeability of Chloride Ion vs. Age of Polymer................12
2.3 Tensile Rupture Strength (VTM 92) vs. Age of Polymer Concrete
2.4 Bald Tire Skid Number (ASTM E 524) vs. Age of Polymer Concrete
2.5 Adhesion Tests (mm)............................................20
2.6 Mechanical Testing Device for ACI 503R.........................23
3.1 DOT Respondents Map............................................33
3.2 Use Anti-Icing/Anti Skid.......................................35
3.3 Winter Accidents...............................................37
3.4 Skid Resistance................................................39
3.5 Amount of Deicing Chemicals....................................40
5.1 Components for Concrete Test Setup.............................62
5.2 Concrete Test Apparatus........................................63
5.3 Components for Asphalt Test Setup..............................64
5.4 Asphalt Test Apparatus.........................................65
5.5 Concrete Form Assemblage.......................................69
5.6 Sample Surface Types...........................................71
5.7 Application of Overlay.........................................73
5.8 Attaching Pipe Cap.............................................75
5.9 Failure Types..................................................77
6.1 Compressive Strengths vs. Time.................................79
6.2 Concrete Epoxy Stress vs. Strain Graph.........................87
6.3 Epoxy and Aggregate Stress vs. Strain Graph....................89
Appendix A
A. 1 DOT Survey, Question #1.........................................97
A.2 DOT Survey, Questions #2, #3, and #4............................98

A.3 DOT Survey, Questions #5, #6, #7, and #8.......................99
A.4 DOT Survey, Questions #9 and #10..............................100
A.5 DOT Survey, Respondent Contact Information....................101

2.1 Overlay type costs..................................................6
2.2 Chloride permeability based on charge passed.......................14
2.3 Friction test......................................................24
2.4 Test sites data....................................................26
2.5 Superior, Wisconsin data...........................................27
2.6 Ironwood Bridge, Indiana data......................................28
2.7 Hibbing, Minnesota data............................................29
3.1 Respondent contact information.....................................34
3.2 Anti-Icing products................................................43
3.3 Overlays on different surfaces.....................................44
5.1 Epoxy properties...................................................54
5.2 Aggregate gradation................................................54
5.3 Application rates..................................................57
5.4 Curing times.......................................................58
5.5 Test matrix........................................................59
5.6 Mixture proportions (SSD)..........................................67
5.7 Fresh and hardened concrete properties tests.......................70
6.1 Fresh concrete properties..........................................78
6.2 Compressive strengths..............................................78
6.3 Batch comparisons..................................................80
6.4 Mixture requirement comparisons....................................81
6.5 Test results.......................................................82
6.6 Manufacture requirement summary....................................91
Appendix B
B. 1 Mixture design # 1................................................102
B. 2 Mixture design #2.................................................103
B.3 Mixture design #3.................................................104

The Federal Highway Administrations (FHWA) National Bridge Inventory reports
that of the 599,766 bridges in the national highway system, approximately 152,316
(25.4%) are structurally deficient or functionally obsolete (NBI, 2009). Furthermore,
Colorado has 1,404 of its 8,366 bridges classified as structurally deficient or
functionally obsolete. Approximately 580 Colorado bridges are structurally
deficient. Though a structurally deficient bridge is not necessarily unsafe, it does
identify the bridge as having a significant level of deterioration and/or damage to
load carrying elements or adequacy of the waterway provided by the bridge. Much
of the deterioration that occurs may be prevented in part by protecting the concrete
(and the steel reinforcement) from water and aggressive chemicals.
Many state Departments of Transportation (DOTs) began incorporating
future wearing surfaces to protect the concrete bridge decks from deterioration and
prevent the infiltration of water and aggressive chemicals into the bridge decks and
supporting structural elements. Bonded overlays are designed to provide a durable,
cost effective wearing surface with low permeability to water and chlorides
[Meggers and Hobson, 2007]. In addition, it is expected that the overlay exhibit an
appropriate level of skid resistance. Types of overlays that have been incorporated

into highway bridge systems included: asphalt, portland cement concrete, non-
reinforced polymer-modified concrete and thin-bonded overlays.
It has become regular practice for state DOTs and county/city road authorities
to take a proactive approach in preventing the formation of snow and ice on
roadways. One method that has been particularly effective is the use of anti-icing
agents on roadway and bridge surfaces prior to the development of severe winter
conditions. These anti-icing agents are liquid forms of salt compounds used to
prevent the formation of ice on the pavement surface, ensuring proper snow/ice
removal [CDOT, 2009]. These chemicals work by decreasing the freezing point of
An innovative surface overlay product called SafeLane has won grants
through the FHWAs IBRC program in four different state DOTs. This overlay
combination of epoxy and aggregate is able to store liquid anti-icing chemicals
within its structure and release them as conditions develop for the formation of ice
and snow (Cargill, 2007). The overlay product allows for a reduction in usage of
deicing chemicals by reducing the amount that is ineffective and wasted, as well as
providing a safer roadway by allowing the chemicals to be more fully effective when
they are most needed. In addition, the product can extend the life of roads and
bridges by acting as an impermeable layer that reduces the effects of chloride and
water intrusion. The overlay can be installed over asphalt, concrete, steel, or wood
and has good traction. Because of these characteristics of SafeLane, this product

may benefit CDOT who is interested in this product by saving them money, making
roads safer, giving them the ability to apply chemicals in advance, and providing
their snow removal crews more time prior to roads needing to be plowed.
Because the effectiveness of the overlay is dependent upon the bond to the
underlying material, this research evaluates the effectiveness of SafeLanes ability to
adhere to concrete and asphalt surfaces, with different surface temperatures and
conditions. The surface conditions included troweled, tined, sand blasted, and
roughened. Increased adhesion will increased the effectiveness and long-term
durability of the overlay. If the overlay has poor adhesion it can peel off due to
stressing being applied to the road. If this occurs the overlay can no longer provide
an impermeable and smooth wearing surface.
1.1 Research Objectives
SafeLanes ability to adhere to different surfaces was evaluated in the lab. Chapter 1
provides background information. Chapter 2 includes an extensive literature review
detailing other research that has been conducted with adhesion testing and/or the use
of SafeLane. The literature review researches reports, journal articles, and other
documents that are beneficial to this study. Chapter 3 includes a national survey of
state DOTs that obtains knowledge regarding the use of SafeLane and other thin
bonded overlays. In addition, the following questions were answered: do state DOTs
included SafeLane on their Approved Products List and have states researched or
incorporated SafeLane in their highway bridge system? Chapter 4 includes the

research problem statement. Chapter 5 discusses the experimental plan for this
study. In this chapter, the fabrication and acquisition of the test samples,
experimental set-up, and test procedure are outlined. A summary of the test results
are included in Chapter 6. Lastly, Chapter 7 summarizes the research program and
provides recommendations for use of the Safelane overlay on asphalt and concrete
surfaces as well as future work in regards to bond strength testing and thin-bonded

Literature Review
2.1 Literature Introduction
The primary goal for any DOT in regards to bridge construction and maintenance is
to prolong the service life of the bridge. The National Association of County
Engineers (NACE) reports the cost to repair or modernize the countries bridges is
$140 billion [Barkin, 2008], This is a substantial cost for federal, state, and local
highway authorities, particularly during this economic down turn and an ever aging
bridge infrastructure. Highway authorities can take a proactive approach in
researching innovative materials that may prolong the life of bridges by preventing
the infiltration of water and deicing chemicals into the bridge structure and
ultimately deteriorating the steel reinforcement. Damage to concrete bridge decks
included, but not limited to, chloride or carbonation initiated steel reinforcement
corrosion, freeze-thaw and deicing chemical deterioration, alkali-silica reaction of
the aggregate, and fatigue due to traffic loading [Paulson and Silfwerbrand, 1998].
Paulson and Silfwerbrand state that in order to provide adequate service life to
bonded overlays (or wearing surfaces), proper bond between the overlay and the
bridge surface must be made, sufficient wear resistance to vehicular traffic, adequate
freeze-thaw resistance, and protection of the reinforcement from chlorides and other
aggressive chemicals.
There is a number of wearing surface overlays available [Carter, et. al., 2007]:

Concrete overlays high performance steel fiber reinforcing, high density,
high performance without fibers, latex modified, etc...
Asphalt (with membranes) multi-layer polymer, hot applied rubberized, and
singly layer polymer,
Surface membranes thin broom and seed epoxy overlays, polymer modified
asphalt, latex-modified asphalt, etc...
Each wearing surface has a different life expectancy and cost which is
summarized in Table 2.1.
Table 2.1 Overlay Type Costs [Knight, et. al., 2004]
Overlay Type Expected Life (years) Average Cost ($/yd2)
Asphalt 15-20 30-40
Reinforced portland cement concrete (PCC) 30+ 70-80
Non-reinforced polymer- modified concrete (PMC) 25-30 55-65
Thin Bonded 20-30 70-110
Some problems that have been experienced with these overlays include
shortened life expectancy, inadequate permeability, cracking, and damage to the
underlying surface during removal and replacement. The corrosion of steel in
concrete will occur when water and oxygen reach the reinforcement and creating a
galvanic cell. An electrochemical reaction begins causing the expansion of the

reinforcing bars resulting in tensile stresses in the concrete and eventually cracking.
Once the cracking process begins, the process of corrosion of the steel reinforcement
is accelerated. Corrosion of the reinforcement will limit the bond between the steel
and concrete, thus reducing the load carrying capacity of the section. The existence
of chlorides accelerates the corrosion process. Chloride ions attack the protective
oxide film formed on the steel by the high alkaline chemical environment present in
concrete. Deicer and salts for snow and ice removal can cause and aggravate surface
scaling. The formation of salt crystals in concrete may contribute to concrete scaling
and deterioration similar to the crumbling of rocks by salt weathering. Scaling and
deterioration is more severe in poorly drained areas because the deicer solution is
retained on the concrete surface during freezing and thawing. [Knight, et. al., 2004]
The life expectancy is a critical factor in overlay design and placement. An
overlay may be considered to be ineffective. Some asphalt overlays have
experienced less than ten years of service life. At the end of the service life, the
overlay must be milled and replaced. During this operation, there is the potential for
damage to the bridge deck surface and the membrane. [Knight, et. al., 2004]
2.2 Asphalt Overlays
Many bridge decks contain an asphalt wearing surface. Engineers at both the
American Concrete Institute (ACI) and the Portland Cement Association have
documented findings which state that asphalt wearing surfaces, when placed over

concrete, cause water to temporarily pond at the interface between the two materials.
The reasoning behind this being that water permeates faster through asphalt than
concrete. In addition, the asphalt layer acts as a protective layer preventing the water
between the two materials to evaporate. This allows the water to remain on the
surface of the concrete and permeate through the concrete if other measures
(impermeable membrane) are not present. [Knight, et. al., 2004]
Asphalt overlays use a combination of asphalt and bridge deck sealant. The
total thickness of this type of overlay is typically 3.25 inch (8.26 cm). Asphalt
overlays do not work well if the bridge deck is noticeably curved, super elevated, or
susceptible to washboard effect in areas of frequent stops. They do not work well
because they have a tendency to move after installation. However, this type of
overlay is the most economical option. [Knight, et al., 2004]
2.3 Concrete Overlays
There are two main types of concrete overlays, reinforced portland cement concrete
(PCC) and non reinforced polymer-modified concrete (PMC). Reinforce PCC is
mainly used when additional construction is being performed on the bridge structure,
a large portion (>50%) of the deck is deemed to be in need of repair, and on older
structurally deficient bridges. The minimum thickness is about 4.5 in. (11.43cm).
(Special consideration needs to be taken for the additional dead load. Non-reinforced
PMC overlays are a typical concrete mixture with a polymer admixture. This type

may be used when additional dead loads needs to be avoided (verses reinforced PCC
and asphalt overlays) and the need of a quick turnaround. The thickness of this type
is typically 1.5 in. (3.81 cm). [Knight, et al., 2004]
Meggers and Hobson tested the four different types of concrete overlays:
[Meggers and Hobson, 2007]
Type IP cement
Type IP cement with 3% silica fume concrete
Type I/II cement with 5% silica fume and polypropylene fibers concrete
Type II cement with 5% silica fume and steel fibers concrete
Type IP is a blended cement that incorporates a pozzolan blended with the
cement at between 15-40% by mass. The overlays were tested for compressive
strength, permeability, chloride concentration, overly adhesion, and cracking
resistance. They found that the least permeable overlays were the ones containing IP
cement, while the steel and polypropylene fiber overlays were the most permeable.
The fiber overlays did not meet the permeability design specification of passing less
than 1,000 coulombs [1.5 in. (3.81 cm) thickness]. After five years, all overlays
appear to be functioning equally unless cracking is present based on the chloride ion
contamination Bond strength was higher when a bonding grout was not used.
Ongoing research conducted with collaboration with the City and County of
Denver (CCD) is examining the effects of various types of deicers (magnesium
chloride, potassium acetate, and salt bring) on concrete deterioration. This

investigation includes the testing of three classes of concrete specified by the CDOT
and utilized by the CCD for use in general concrete construction and pavements. In
addition, pervious (porous) concrete was examined in this study. The four mixtures
and the design requirements are shown below: [Solis, et. al, 2010]
Class B- 3,000 psi 28 day compressive strength, up to 20% class C fly
ash (30% class F), used typically for sidewalks and onsite residential
Class P- 4,200 psi 28 day compressive strength, up to 20% class C fly
ash (30% class F), used typically for pavements and curbs.
Class D or DT- 4,000 psi 28 day compressive strength, up to 20%
class C fly ash (30% class F), used typically on bridges or other
structural elements.
Porous concrete- 2,500 psi 28 day compressive strength, used
primarily in parking lots to lessen the need for onsite drainage
retention facilities.
The testing of the concrete for scaling resistance will be performed using ASTM C
672 Standard Test Method for Scaling Resistance of Concrete Surfaces Exposed to
Deicing Chemicals. Results show significant deterioration of the concrete surface
due to the application of the magnesium chloride. Figure 2.1 shows the effects of (a)
magnesium chloride and (b) salt brine on the concrete. There is a noticeable
dissolution of the cement paste from the aggregate surface when (a) magnesium

chloride is utilized as a deicer (note: the samples shown Figure 1 are pervious
concrete). [Solis, et. al, 2010]
Figure 2.1 Effects of (a) Magnesium Ch
[Solis, et. al, 2010]
oride and (b) Salt Brine on Concrete
2.4 Surface Membranes
Surface membranes are another overlay type. There are two basic material types:
polymer-modified cementitious or polymer-modified epoxy. Thin bonded overlays
consist of a layer of polymer-modified material applied to bridge deck and usually a
rough fine aggregate applied on top of the compound, which, may be repeated
depending on what product is being used. The approximate thickness of this type of
overlay is 0.25 in. (0.635 cm). These types of overlays are extremely expensive.
[Knight, et al., 2004]
Nelsen analyzed the performance of urethane, silicon-based, and epoxy based
overlays. The overlay that performed the best at protecting the concrete, remaining
not cracked, and acceptable levels of skid resistance were epoxy- based overlay.
Silicon based products did not crack, protect the concrete from the wearing effects of

traffic nor improved skid resistance because they seep into the concretes pores.
Urethane- based overlays were consistently out performed by epoxy overlays
without any significant increase in cost or health risk. [Nelsen, 2005]
To determine polymer concrete overlays permeability to chloride ions, tensile
rupture strength and skid resistance Sprinkel analyzed fourteen different bridges with
different overlays that have been in the field of an average of 7 to 12 years.
Sprinkels investigation included three bridges with multiple-layer epoxy (MLE),
three with multiple-layer epoxy urethane (MLEU), three with premixed polyester
(PP), three with metacrylate slurry (MS), and two with multiple layer polyester
(MLP). [Sprinkel, 2001]
Figure 2.2 Permeability of Chloride Ion (AASHTO T 277) vs. Age of Polymer
Concrete Overlays [Sprinkel, 2001]

Figure 2.2 shows that MLE, MLEU, PP and MS provide for very low permeability
for 25 years. MLP overlays were found to only provide good protection for 10 years.
[Sprinkel, 2001]
The two tests most often used in determining a concretes chloride
permeability are the Ponding Test AASHTO T259-80 (ASTM Cl543) and The
Rapid Chloride Permeability Test AASHTO T277 (ASTM Cl202-91). Prior to the
development of the Rapid Chloride Permeability test, chloride permeability of
concrete was measured by the ponding test. Ponding tests take 90 days or longer and
involve taking samples of the concrete at various depths to determine the chloride
profile. The Rapid Chloride Permeability Test was developed after as an alternate
procedure and could be performed in about 24 hours. This test consists of monitoring
the amount of electrical current passing through 51mm (2 in.) thick cores with a
102mm (4 in.) nominal diameter during a period of 6 hours. A potential difference of
60 V dc is maintained across the ends of the specimen, one of which is immersed in
a sodium chloride (3.0% by mass in distilled water) solution, the other in a sodium
hydroxide (0.3N in distilled water) solution. Resistance is than calculated as volts
divided by current. This test method does not measure concrete permeability. What it
does measure is concrete resistivity, which has a fair correlation with concrete
permeability. Table 2.2 shows the ranges of concretes ability to resist chloride
penetration. [Understanding, 2006]

Table 2.2 Chloride Permeability Based on Charge Passed [Understanding,
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
Figure 2.3 Tensile Rupture Strength (VTM 92) vs. Age of Polymer Concrete
Overlays [Sprinkel, 2001]

The tensile rupture strength of MLEU, MLE, and PP stay consistent throughout their
life. MLP should fail in about 10 years from lose of strength with time. This is
shown in Figure 2.3. [Sprinkel, 2001]
Figure 2.4 Bald Tire Skid Number (ASTM E 524) vs. Age of Polymer Concrete
Overlays [Sprinkel, 2001].
In Figure 2.4 the overlays overall, maintain acceptable skid numbers throughout
the life of the overlays with the exception of MS [Sprinkel, 2001]. The principle
measurement of the safety of a pavement is its skid number. This data is collected to
identify the pavements ability to reduce skid related accidents. The higher the skid
number (SK) the better the pavement. The basic formulation of the SK is
SK=100*L/N, where L = the frictional force between the road and tire, N = the force
perpendicular to the two surfaces. The most common test method is the locked-wheel

test or ASTM E 274. This test involves wetting the pavement, pulling a two-wheel
trailer whose tires are standard, locked, and the car is driving 40 mph. The SK is
calculated by measuring the locking force [Garber and Hoel, 2002], SK equal to 35
and above are acceptable for heavily traveled roads. SK typically range from 40-60
but can go as low as the teens and as high as 80.
Conclusions from this study indicate that multiple-layer epoxy, multiple-layer
epoxy-urethane, and premixed polyester polymer concrete overlays can expect to
provide skid resistant wearing and protective surface to bridge decks for 25 years.
[Sprinkel, 2001]
Several failures of polymer overlays have been documented in recent times.
Some of the epoxy polymer overlays lasted only 2 years while others lasted more
than a decade. One paper, Investigations of Failures of Epoxy Polymer Overlays in
Missouri reported that overlays provided lasting protection only when the bridge
deck was still in acceptable condition with 5% or less surface area damage needing
patching. In addition, some of the failed epoxy overlays had bubbles in them
resulting from the use of an improper paddle. The proper paddle that should be used
is a jiffy paddle or a Sika paddle [Harper, 2007]. Scarpinato discussed why
proper evaluation of the polymer overlay and bridge structure needs to be performed
prior to the construction of the overlay to insure quality control. To recognize what is
needed to be analyzed, Scarpinato discusses past failure. Scarpinato found the
following factors were causes of overlay failures: [Scarpinato, 1996]

Inadequate surface preparation
o Bridge needs to have a complete removal of unsound concrete and
surface contaminates
Perform Sand or Shot Blasting
Concrete deterioration too far advanced
o Repair is not feasible
Dissimilar physical properties of the overlay material
o thermal coefficient compared to structure
o shrinkage of polymer compared to concrete
Chemical incompatibility of polymers and concrete
o moisture sensitive polymers
o alkalis in concrete attract the polymers
Improper proportioning, mixing, or application procedure
To prevent failures due to the binder, Scarpinato lists the needed properties a thin
bonded overly binder must have. These include:
Modulus of elasticity between 90,000 and 150,000 psi
Tensile elongation of at least 30 percent
Minimum tensile strength of 2,500 psi
Compressive strength between 5,000 and 8,000 psi
According to the Field Performance of Polymer Bridge Deck Overlays in
Michigan Final Report materials used today exhibit bond strength that is higher

than the tensile strength of the concrete decks and the elastic properties of the
binders allow for major expansion and contraction without failure and bond
interface. In addition, in Michigan, the application of double coat overlays has
become standard practice and appears to ensure a good seal and surface.
Anti-icing characteristics are a desirable property for overlays. Therefore, the
report Field Performance of Polymer Bridge Deck Overlays in Michigan observed
whether the overlays reduced the amount of snow and ice on the bridge. However,
the report found that there was no measurable benefit of anti-icing from the standard
overlays observed. Further investigation of the anti-icing characteristics of
overlays is underway. [Alger, Gruenberg, Wegleitner, 2003]
2.4.1 Bond strength testing
Adhesion testing is necessary to ensure that the overlay has properly adhered to the
surface it is applied to. The bond strength is a combination of the adhesion of the
material and mechanical interlocking.
Austin et. al. proposed these tests: slant -shear, twist-off shear test, flexure
tests, patch test, uniaxial tension test, friction-grip tensile test, a dog-bone test and
pull-off tests. The criteria they proposed were to simulate in-situ conditions, induce
stress states typical of service loads, be sensitive to variations in bond strength, and
the ability to reproduce test results. The simplicity, severe stress state and potential
for in situ measurement are why tensile tests are generally favored. However, these

types of tests can give misleading results if there is a mismatch in the material
properties, such as, shrinkage, thermal movement or modulus mismatch. In addition,
the surface condition of the substrate is important to develop an adequate bond. The
surface needs to be clean, rough, and sound. The introduction of defects due to the
use of mechanical splitting and roughening techniques can significantly reduce the
tensile bond capacity. The best method found to produce the best surface for a sound
bond was the use of water-jetting and sand/grit blasting. They found that the pull-off
test as being the closest test to meeting the requirements of a good bond test. [Austin,
et. al 1995]
The second paper by Momayez, et. al., investigated the pull-off test, splitting
prism, direct shear and slant shear tests as illustrated in Figure 2.5.
a) Pull-ofT b) Splitting Prism c) Direct Shear d) Slant Shear
Figure 2.5 Adhesion Tests (mm)
The results had a large variation between the four tests. Because of the large
variation it was proposed that the test that best simulates the stresses in the field

should be used. The pull-off test gave consistently the lowest bond strength values
followed by splitting and direct shear. Slant shear gave the highest bond strengths.
[Momayez, et. al., 2005]
2.4.2 Pull-off test
The results taken from the pull-off test include average tensile strength, fracture
depth, and mode of failure. The mode of failure includes: cohesive concrete,
adhesive, and cohesive resin. The pull-off test assesses the minimum tensile stress
needed to detach or rupture the overlay perpendicular to the surface. The test is
performed be securing a dolly perpendicular to the surface. The testing apparatus is
attached to the loading fixture and is aligned to apply tension to the surface. The
force that is applied gradually increases and is monitored until a plug of coating is
detached, or a previously specified value is reached [Wikipedia, 2009]. One critical
aspect of the pull-off test is how deep to core drill into the substrate. Austin, et. al.,
claim that shallow drilling depths will underestimate the real bond strength and can
cause replication and comparison of data a problem. Their tests show that at a
drilling depth of 0.0787 in. (2 mm) compared to a 0.5906 in (15 mm) had a lower
failure stress of about 15%. The proposed European Standard sets the drilling depth
into the substrate at 15 + 5 mm. [Austin, et. al., 1995]
ACI 503R pull off test is a field test. The method to perform the test in the
field is as such. Clean the areas were the epoxy compound will be applied according

to prescribed cleaning methods. The area selected should be representative of the
worst areas. [ACI, 1993]
Using the epoxy compound, mix and apply the material to a test patch. Once
the test patch has hardened, core drill through the coating and down barely into the
subsurface by the use of an electric drill fitted with a carbide-tipped or diamond core
bit. Core a 2in. (5 cm) diameter disc with the appropriate bit. It will appear to be a
small island of coated material. Bond a standard 1.5 in. (3.7 cm) diameter pipe cap to
the cored disc using nearly any commercially available room temperature rapid
curing epoxy compound adhesive. The bottom surface of the pipe cap should be
machined smooth and shoulder-cut to provide a 2 in. (5 cm) diameter surface. Mix
the epoxy components according to the suppliers recommendations. Using a spatula
apply a small amount of the mixed adhesive the cored disc and to the bonding face of
the pipe cap. To facilitate the spreading of the adhesive the bonding face may be
heated. Place the pipe cap on the cored disc and direct a flame into the interior of he
pipe cap in such a way that no direct heat reaches the cored disc or the pavement
bond line and heat the pipe cap to about 160 F (70 C) with a small gasoline blow-
torch (an electric heat lamp or a portable gas radiant heater may be used as
alternatives). A surface pyrometer can check the temperature. In less than 1 minute
under these conditions the adhesive should harden. Once the bonded cap has cooled
to air temperature it will be ready for testing. [ACI, 1993]

Using a testing device similar to the one shown in Figure 2.5 the pipe cap and
core can be tested by applying tension. To test, screw the lower hook into the
threaded pipe cap and attach to the loop on the lower portion of a Dillon
dynamometer. Screw the upper hook into the loading arm at the top and attach to the
loop on the upper portion of the dynamometer. The axis of the dynamometer must
coincide with the axis of the pipe cap extended when the force is applied. To apply
tension, rotate the loading arm so that the threaded shaft and its connections are
lifted. An approximate rate of 100 lb (45kg) every 5 seconds should be applied. The
tensile load is indicated on the dynamometer gage. Once the pipe cap and connected
core is separated from the concrete surface the load needs to be recorded. Take the
load at failure and convert to unit stress. Note the type of failure. There are three
possibilities or combinations thereof: failure in the concrete (cohesive concrete
failure), separation of the epoxy compound from the concrete surface (adhesive
failure), and failure in the epoxy compound (cohesive resin failure). Record the load
required to bring about the failure along with the percent of each type of failure. A
concrete failure should result from a properly formulated epoxy compound applied to
a properly prepared surface. The hole created by the test can easily be repaired using
either an epoxy resin compound or the remaining epoxy adhesive. [ACI, 1993]

2.4.3 SafeLane
In recent years, advancements in bonded overlays have produced a product that
promotes anti-wearing/anti-icing. A product called SafeLane Surface Overlay is
marketed as a wearing surface that acts as a rigid sponge [Cargill, 2007]. When
DOTs apply their standard anti-icing chemicals prior to winter weather, the overlay
stores the chemicals and releases deicing chemicals as snow and ice develop on the
road surface. This method prevents the bond of snow and ice to the road surface,
providing for a safer drivable surface. In addition, Cargill mentions that their
product improves the friction for all weather conditions throughout the year due to

the aggregate in the overlay. Since 2004, the SafeLane Surface Overlay product has
been applied to over 50 sites in the United States, Canada, and Great Britain.
The SafeLane HDX Overlay consists of an epoxy-aggregate mixture that is
applied to the surface of the bridge deck. The product requires an epoxy distribution
system, aggregate spreader, application squeegee, and moisture and oil-free
compressed air. Surface preparation may include repair of surface delaminations
such as potholes, cracks, and breakouts. [Cargill Technical Specification, 2007]
Thomas Martinelli from the Wisconsin DOT preformed a friction test on the
bridge on grade approach ramp to the Blatnik Bridge, which was covered with
SafeLane in one direction on June 2005 [Martinelli, 2007]. The results from the test
are as follows:
Table 2.3 Friction Test [Martinelli, 2007]
Date WisDot ASTM Friction Trailer
4/20/2005 47.2 (initial pavement)
7/20/2005 40.4 (adjacent PCC pavement)
7/20/2005 55.7 (anti-icing overlay)
11/8/2006 44.2 (adjacent PCC pavement)
11/8/2006 55.0 (anti-icing overlay)
The higher the skid numbers the better friction the road surface provides. The
results show that the overlay friction is equivalent or better than that of the concrete

Nixon reported on the performance of the SafeLane Surface Overlay during
the winter 2006-2007 [Nixon, 2007]. His report documented the performance of the
overlay at over 31 sites within the United States. Below is a summary of the data he
collected and analyzed of how SafeLane preformed in winter conditions.

Tfeble 2.4 Test Sites Data [Nixcm, 2007]
Site Location Type Date of Installation Snow/lce Accidents ADT Wearing
with without before after
Junction City, Kansas 1-70 bridge over Washington Street Four lane bridge, westbound side treated, eastbound is control Sep-06 snow and ice covered snow and ice covered 25,000
Superior, Wisconsin US 53 and US 2 Interchange Interchange with relatively high accident rate Sep-C6 Table 5 Table 5 37 injury and 46 property damage accidents 1 accident (driving too fast for condition) <10,000
Mibbing, Minnesota Mitchell Bridge on Highway 169 Four lane bridge, slopes and curves, concern over accidents Jul-06 Table 7 Table 7 14 accidents over a 4 year period 14,400
Asheville, North Carolina Buncombe County, US 19/23 Bridge crossing over a creek Oct/Nov 2006 clear and wet becoming slushy 55,000-60,000
Mansfield, Pennsylvania Route 660 near Mansfield A rural bridge installation Nov-06 bare snow covered No accidents
Prosser, Washington I-82, about 1 mile west of Prosser Overpass across the Sunnyside Canal with WB lanes treated and EB as control. Oct-06 36 accidents related to snow and ice from 1999 2004 1 accident (maybe) 18,000
McLean Bridge, Texas 1-40, mile marker 144,70 miles east of Amarillo Two lane elevated bridge, westbound side only, 12,000 square feet Nov-05 kept the ice on the structure from bonding to the pavement Several accidents each winter
Brecksville, Ohio I-80, exit 173 on Ohio Turnpike Exit ramp, incline with curve, 8.0CO square feet Nov-05 49 accidents No accidents
Ironwood Bridge, Indiana Ironwood Overpass, South Bend, Indiana Eastbound lane of the overpass, 11,790 square feet May-05 slushy and wet (Table 6) iced over, snow packed (Table 6) No accidents 30,000
Blatnik Bridge On-ramp, Wisconsin Blatnik bndge on Rt. 53 between Superior and Duluth On-ramp for the bridge, 15,000 square feet Jurt-05 Salted less often, stayed in a wet condition snow covered and icy 20 accidents over the past 4 years No accidents 15,000
Wolf River Bridge, Wisconsin Rural bridge over the Wolf River, near Crandon Wisconsin Two lane bndge, covered on both directions, 4,800 square feet Summer 2003 3-4 accidents per winter No accidents wear in the wheel tracks (1 layer application)

Table 2.5 Superior, Wisconsin Data [Nixon, 2007].
Date and Time Road and Weather Conditions Applications Condition of Test and Control Sites
12/22/06 @2:00 a.m. Freezing rain, Pavement Temp (PT) = 32 None "It worked well during rain"
12/29/06 @8:00 a.m. Wet heavy snow, PT = 32 300 lbs per lane mile of salt Did not stick to wheel path in test section
12/31/06 @ 1:00 a.m. Freezing rain, PT = 32 None Temps were at freezing, but none of the roads froze
2/23/07 @7:00 a.m. Pre-treatment 10 gallons per lane mile No report
2/24/07 @ no time given. Wet snow, PT = 24 20 gallons per lane mile Wet snow compacts to both test and control
2/25/07 @2:30 a.m. Wet snow, PT = 26 3 applications of 300 lbs Wet snow compacts to both test and control sections
2/26/07 @2:30 a.m. . Snow, PT = 26 2 applications of 300 lbs per lane mile Clear wheel tracks on both test and control
same @ 5:00 p.m. Sleet and snow, PT = 37 None Refreeze occurred on control section, but not on test section
3/5/07 @2:00 a.m. Wet, heavy snow, PT = 28 No data given During heavy snow storm, the test section showed no difference, but did clean up quicker afterwards
3/12/07 @ 6:00 a.m. No snow, PT = 20 No data given Control section iced over, test section was clear
3/16/07 @2:30 a.m. Wet snow, PT = 20 2 applications of 300 lbs per lane mile Compacted during snow storm, but cleaned up well afterwards

Table 2.6 Ironwood Bridge, Indiana Data [Nixon, 2007].
Date and Time Road and Weather Conditions Applications Condition of Test and Control Sites
12/4/OB @ 12 p m. Light snow and Pavement Temperature (PT) not recorded Salt applied but no quantity recorded WB (Control) snow covered, EB (Test) bare and wet
12/7/06 @ 1:00 a m. Light snow, PT not recorded Salt applied WB (Control) snow covered, EB (Test) bare and wet
Same @ 2:10 a m Light Snow, PT = 22 Salt applied WB (Control) snow covered, EB (Test) bare and wet
Same @ 3:00 a.m. Light snow, PT = 22 Salt applied, 200 lb per lane mile WB (Control) snow covered, EB (Test) bare and wet
Same @ 4:00 a.m. Light snow, PT = 20 200 lbs per lane mile on both WB slushy, EB bare and wet
1/16/07 @ 12:30 a m. Snow, PT not recorded 200 lbs on both sections WB (Control) snow covered, EB (Test) bare and wet
1/25/07 @3:40 a.m. Snow, PT = 28 250 lbs per lane mile on both sections Both sections wet
Same @ 4:30 a.m. Snow, PT = 26 250 lbs per lane mile on both WB slushy, EB bare and wet
Same @ 6:20 a.m. Light snow, PT = 26 250 lbs per lane mile of salt on both Both sections wet
Same @ 7:50 a m Light snow, PT = 25 250 lbs per lane mile Both sections wet
1/27/07 @7:00 p.m. Light snow, PT not recorded 250 lbs per lane mile to both Both sections had light snow cover
Same @ 10:00 p.m. Heavy snow, PT not recorded 250 lbs per lane mile to both WB snow covered and slippery, EB slushy
Same @ 11:30 p.m. Snow, PT = 27 250 lbs per lane mile to both WB snow covered and slippery, EB bare and wet
1/28/07 @ 1:20 a.m. Light snow, PT = 270 250 lbs per lane mile to both WB slushy, EB bare and wet
Same @ 3:00 a.m Light snow, PT = 27 260 lbs per lane mile to both WB light snow cover, EB bare and wet
Same @ 3:50 a.m. Light snow, PT = 25 260 lbs per lane mile to both WB snow covered and slippery, EB bare and wet
Same @ 5:10 a m. Light snow, PT = 20 260 lbs per lane mile WB snow covered and slippery, EB very light snow cover
Same @ 6:10 a.m. Light snow, PT = 19 260 lbs per lane mile WB snow covered and slippery, EB bare and wet
Same @ 12:10 p.m. Snow, PT not reported 190 lbs per lane mile on both Both sections snow covered
Same @ 1:30 p.m. Snow, PT not reported 190 lbs per lane mile on both WB snow covered, EB wet and sliqht snow cover
Same @ 4:30 p.m Heavy snow, PT not reported 190 lbs per lane mile on control section only Both sections wet
Same @ 10:00 p.m. Heavy snow, PT = 21 300 lbs per lane mile on both EB slushy after treatment, WB snow packed
1/28/07 @ 12:30 a m. Snow, PT = 21 Salt applied in unknown amount EB slushy, WB snow packed
Same @ 3:00 a.m None, PT= 14 300 lbs per lane mile to both EB slushy, WB snow packed
Same @ 4:15 a.m None. PT = 8 300 lbs per lane mile to both EB slushy and wet, WB wet and clearing up

Table 2.7 Hibbing, Minnesota Data [Nixon, 2007],
Date and Time Road and Weather Conditions Applications Condition of Test and Control Sites
10/23/06 @ 5:30 p.m. Snow and Pavement Temperature (PT) = 31 Salt applied, 400 lbs per lane mile Both SB (test) and NB (control) were clear and wet
10/24/06 @3:00 a.m. Frost, PT = 23 Liquid (Mag Chloride) applied at 15 gal/lane mile Both sides had frost
11/2/06 @4:00 a.m. Flurries, PT = 22 No application SB clear and dry, NB clear and wet. Not slippery
11/5/06 @ 1:30 a.m. None, PT= 19 Liquid applied at 30 gal/lane mile Anti-icing run
11/9/06 @2:00 a.m. None, PT= 31 Liquid applied at 30 gal/lane mile Anti-icing run
11/12/06 @7:00 a.m. Wet, heavy snow, PT = 33 400 lbs per lane mile on both sections Both sections had compact
11/13/D6@ 2:30 a.m. Wet snow, PT = 30 200 lbs per lane mile on both sections Both sections had compact
11/15/06 @ 1:00 a m. None, PT = 32 Liquid applied at 30 gal/lane mile Anti-icing run
11/17/06 @9:00 a.m. Snow, PT = 30 260 lbs per lane mile of salt on both Both sections snow covered
11/28/06 @3:00 a.m. Rain and freezing rain, PT = 33 200 lbs per lane mile on both Both sections ice covered
12/4/06 @3:00 a.m. Snow, PT= 12 No application (sander broken) Both sections 50% snow covered
12/7/06 @9:00 a.m. None, PT = -3 600 lbs per lane mile to both Test section 50% frost covered, control
Major conclusions from Nixons study included:
The overlay did not result in any new problems with winter maintenance.
Test section remained clear of snow and ice during winter weather when
snow and ice had accumulated on nearby control sections.
Bonding of snow and ice did not occur on the test sections.
Two years of consistent evidence that the product improved performance
under winter conditions, thereby improving safety for vehicular traffic.

No special or unusual concern with the application of liquids to a surface that
is good at retaining chemicals. That is the overlay with chemicals being
retained did not cause slippery conditions.
When snowfall is wet and heavy there is no noticeable benefit.
o Cause being dilution of ice control chemicals
The overlay provides more time for snow and ice control measures to be
implemented. It does not however, make those measures unnecessary.
May result in the use of less chemical to achieve satisfactory results, up to
50% less.
From these conclusions Cargill observed these benefits: [Cargill, 2007]
Decrease in winter accidents
Increased bridge life
o From blocking chloride and water intrusion
Possibility of spending less on maintenance chemicals
Reduced pavement maintenance
Reduced environmental impact
o From reduced amount of chemical being used
o Chemicals effect the soil, vegetation, water, highway facilities and
Decreased labor and equipment cost
o Call outs and overtime are required less often

o Equipment use is reduced
Reduces reliance on weather forecasting
o Ability to store anti-icing chemicals
Reduces frequency of treating isolated structures
The procedure for installing the SafeLane Overlay System on concrete and asphalt is
as follows [Cargill, 2007]:
Repair pavement surface if needed
Clean off oil and contaminants
Shot/Sand blast surface
Mix and prepare the epoxy
Manually spread the epoxy using squeegees
Immediately broadcast the aggregate across surface
Apply second application
Leave epoxy to cure for 7-8 hours depending on average deck, epoxy and
aggregate temperatures
Sweeper truck scraps off and vacuums up loose rocks
Leaf blowers follow to remove excess dust
Surface is reopened to traffic

State DOT Survey
3.1 Department of Transportation Survey Findings
A questionnaire was developed to investigate whether other state DOTs have used
thin-bonded overlays with anti-icing properties and their experiences with them. The
survey was intended to discover a vast source of valuable information pertaining to
specifications, performance, and construction practices. Information about these
products will be of value to the CDOT who plans to use these products on bridge
structures throughout the state of Colorado. The web-based survey was designed to
be short, simple, and concise in order to keep it user-friendly to maximize the
potential number of responses generated. A list of the targeted recipients consisting
primarily of DOT Materials Engineers is included in Appendix A. The survey is
shown in Appendix B. The responses to this survey are expected to provide valuable
information for CDOT and are summarized below.
3.1.1 Survey Response
Responses were received from 24 of the 50 State DOTs, for a 48% return rate. See
Figure 3.1. In all, a total of 30 responses were received for the survey. A number of
State DOTs provided more than one response. Most of the two-respondent states
included responses from both the Materials and Bridge Engineers.

$y States that Responded
Figure 3.1 DOT Respondents Map
Table 3.1 provides a list of the Material and Bridge Engineers who respond ended to
the DOT survey. Multiple responses were obtained from six states: Colorado,
Nevada, Tennessee, Mississippi, South Dakota, and North Dakota.

Table 3.1 Respondent Contact Information
Respondent Name Department of Transportation Contact E-Mail Address
Richard A Pratt Alaska DOT&PF richard.Dratt@alaska.aov
Phil Brand Arkansas Highway & Transportation Dept.
Richard Land CADOT richard.
Eric Prieve Colorado DOT
Dennis O'Shea Delaware DOT us
Bouzid Choubane Florida DOT
Matt Farrar Idaho Transportation Department matt.farrar@itd.idaho.aov
Todd Hanson Iowa DOT odd.hanson@dot.iowa.aov
Ross Mills Kentucky Transportation Cabinet ross.mills@ky.qov
Tyson D. Rupnow Louisiana Transportation Research Center Tvson, RuDnow@la.aov
Marc Lipnick MD State Highway Administration
James (Jim) A. Lilly MnDOT
Mike Obrien Ms Dept, of Transportation
Adam Browne MS DOT
Fouad Jaber Nebraska department of Road fouad.iaber@nebraska.aov
Mark Elicegui Nevada Dept Of Transportation
Anita Bush Nevada DOT
David L. Scott NHDOT
Andrew Mastel North Dakota Department of Transportation amastel@nd.qov
Larry Schwartz North Dakota DOT lschwart@nd.qov
Guy Hildreth, PE NYS Department of Transportation
Tim Keller Ohio DOT
Walter Peters Oklahoma Dept of Transportation wDeters@odot.orq
Darin Hodges SDDOT
Kevin Goeden South Dakota DOT
Wayne Seger TN Dept, of Transportation wavne.seqer@tn.qov
Edward P. Wasserman TN DOT ed.wassserman@tn.aov
Jim Wild Vermont Agency of Transportation
David Bohnsack WIDOT david.bohnsack@wi.aov
3.1.2 Anti-Icing/Anti Skid Use
Question #2 of the Survey was to determine what states have used a thin bonded
overlay with anti-icing product. A total of 30 responded to the question, with a total
of 9 indicating that they have used a thin bonded overlay with anti-icing/anti-skid.
Responses are shown in Figure 3.2.

Does Your State DOT Utilize Thin-Bonded Overlays
with Anti-Icing and Anti-Skid Properties Such as
Cargil SafeLane? Question #2 Written Responses
1. Vermont DOT The Cargill SafeLane was actually used on an asphalt road
for a section of it that has been know to have numerous accidents which was
placed fall of 08. We have used an anti-skid/wear treatment on a steel grid
concrete filled panel deck bridge done maybe in 98 or 99. We placed a deck
around the end of June of 09 that cracked extensively, so we used a product
called Sikadure 55 to seal the cracks and the contractor broadcast sand into it
2. Wisconsin DOT 30 applications since 1999
3. Idaho DOT one application about 2 years old; one scheduled for fall 2009
or spring 2010

4. New York DOT We use SafeLane, but not as an "anti-icing" overlay as
claimed. We have used several epoxy systems over the past 3 years. Prior to
that they was used only by Bridge Maintenance for added
5. Mississippi DOT- Only recently used.
6. Minnesota DOT- Four applications of SafeLane; two at 2 plus years (both
removed); two at one year
Two Microslurry seals 9 years
Novachip 10 months
One experimental Methymetracralate chip seal.
7. North Dakota DOT- has two experimental projects that were constructed in
8. Kentucky DOT- has used this type of product on 2-3 bridge decks during
new construction due to less than adequate rebar coverage and questionable
9. North Dakota DOT- I'm not sure if yes is the right answer but we did put
SafeLane on two bridges last year as an experimental project. Our plan is to
study it for a period of time before we use it on a wide scale.

3.1.3 Winter Accidents
Question #3 of the survey provided the respondents with an opportunity to state
whether the product was able to reduce the number of accidents during the winter
season. Responses are shown in Figure 3.3.
Has the number of winter weather related accidents been
reduced on bridge decks using SafeLane (or other Anti-
Icing/Anti-Skid overlays)?
yes no same Data Not
Figure 3.3 Winter Accidents Question #3 Written Responses
1. Wisconsin DOT Best guess would be a 80% decrease in accidents
2. New York DOT- The "new" surface provided by these epoxy TPOs contain
aggregates approved by our Geologist section to ensure acceptable friction of
the deck. SafeLane has not been used to correct icing problems.

3. Minnesota DOT- SafeLane Significant the first years, substantial the
second, increased third year. Microsurfacing seals significant >80%
reduction in accidents over ten years of use. Novachip significant reduction
first winter compared to control bridge. Novachip no data
3.1.4 Skid Resistance
Question #4 was designed to see whether the product produced adequate skid
resistance. Skid resistance is important factor in providing safe road ways. It allows
the motorist to stop easier. Responses to this question are provided in Figure 3.4.
What is the typical skid number (SK) that your DOT has
observed on these types of overlays?
>60 50-60 40-50 30-40 <30 Unknown
Figure 3.4 Skid Resistance

3.1.5 Deicing Chemicals
Question #5 was designed to see whether the Anti-Icing/Anti-Skid overlays reduce
the amount of deicing chemicals verses a typical bridge deck. Responses are
provided in Figure 3.5.
Approximately what percent reduction in the amount of
deicing chemicals applied to these types of bridge decks
has been experienced when compared to decks without
Anti-Icing/Anti-Skid overlays?

0) 7

a) 5
DC 3
o 4
20% 40% 60% 80%
Figure 3.5 Amount of deicing chemicals Question #5 Written Responses
1. Mississippi DOT- Deicing salts are used.
2. North Dakota DOT- The two bridges are in urban areas and they apply the
same amount of chemicals to these structures.

3. Minnesota DOT- Variability of winters makes comparison difficult. Using
anti-icing with these products greatly reduces reactive applications of
material. Reduction for these events may be greater than 40%.
3.1.6 Failures
Question #6 gave the respondents an opportunity to voice any concern and to
describe any major failures with the product. Responses are listed below. Question #6 Written Responses
1. Wisconsin DOT- The only failures we have had are related to the mixing of
the epoxy during construction.
2. Idaho DOT-Not as yet
3. New York DOT- Epoxy overlay systems have been performing well.
4. Mississippi DOT- no
5. Minnesota DOT- SafeLane chip seals have seen significant decrease in
friction numbers and increasing accidents after two years of service.
Slurry seals deterioration of bridge decks via surface spalling. We are trying
to investigate if vapor barrier is increasing freeze-thaw deterioration. Not data
on this yet.
Slurry seals We have had some bonding problems, but these appear to be
due to application difficulties (cold weather) and deck spalling of underlying

Novachip none new application of material.
Methyracralate chip seals none new application
6. North Dakota DOT- The Department has not noticed any major failures.
7. Kentucky DOT- Not that we are aware of.
3.1.7 Specifications
Question #7 was designed to determine whether other DOTs have special
specifications for the SafeLane Overlay system. If the bridge deck overlay projects
that include the SafeLane system currently underway prove to be beneficial, the
CDOT will write specifications for these types of overlays. Responses are provided
below. Question #7 Written Responses
1. Vermont DOT- There might be for the SafeLane product and there probably
was for the bridge done back in 98 or 99 but the most recent bridge was a
repair and not bid with that option.
2. Wisconsin DOT- No.
3. Idaho DOT- Yes.
4. New York DOT- Prior to using Cargill, they were required to adhere to NYS
Friction aggregate requirements. This required that Cargill's "special"

aggregate, a limestone with a very low acid insoluble residue content, be
blended with a design minimum of 25% non carbonate stone. We cannot
verify Cargills claim of anti-icing or that it would provide long-term friction
without the addition of the required noncarbonate material. Our specifications
presently in use for epoxy overlays are 584.50 03 and 584.50 06 Epoxy
Overlay Wearing Surface for Structural Slabs.
5. Mississippi DOT- Not yet
6. Minnesota DOT- All applications have been experimental and special
provisions have been written uniquely for each job.
7. North Dakota DOT- Yes
8. Kentucky DOT- Not at this time.
3.1.8 Anti-Icing Products
Question #8 was written to determine if there are any products, other than SafeLane,
that can provide anti-icing/anti-skid properties. Table 3.2 lists the responses.

Table 3.2 Anti-Icing Products
Has your State DOT observed if any of the overlay products listed below exhibit anti-icing properties? Check all that apply.
Products Response Count
E-Bond 526 0
Sikadur22, LO-MOD 0
Euclid Chemical Flexolith 216 1
Unitex Propoxy Type IIIDOT 2
Poly-Carb Flexogrid Mark- 163 1
Degussa Building Systems 1
Transpo T-48 epoxy 2
3.1.9 Deck Surfaces
Question #9 was written to determine what other products states have used and
weather they have used them on an asphalt wearing surface. Responses are listed in
Table 3.3. The table shows that the only product that has been or can be used on
asphalt wearing surface is Cargills SafeLane.
Table 3.3 Overlays on Different Surfaces
Please indicate (check all that apply) the type of deck surface each of the products listed below (and utilized by your State DOT) are placed on within your state's bridge inventory. For example, is the thin-bonded overlay product applied to only Asphalt Wearing Surfaces, only Bare Concrete Decks, or both.
Products Asphalt Wearing Surface Bare Concrete Deck Surface
E-Bond 526 0 1
Sikadur 22, LO-MOD 0 2
Euclid Chemical Flexolith 216 0 2
Unitex Propoxy Type IIIDOT 0 4
Poly-Carb Flexogrid Mark- 163 0 9
Degussa Building Systems Trafficguard El 0 2
Transpo T-48 epoxy 0 5
Cargill SafeLane 2 5

Question #10 provided the respondents with the opportunity for additional comments
that were not captured in the survey. Question #10 Written Response
1. Louisiana DOT- We have utilized some concrete overlays here in LA, and
they have performed rather well over the years. The skid numbers are high as
a concrete is expected and the ride numbers are still relatively good as well.
2. Nevada DOT- We have only tried a few of the above mentioned products
and have not observed any anti-icing properties. Our typical bridge overlay is
3/4" polymer concrete.
3. New York DOT- We are interested in the research you are proposing to
accomplish and the subsequent results please provide us with your findings.
4. Minnesota DOT- We believe high friction chip seals have great promise as a
safety enhancement process and have clearly demonstrated the ability to
reduce accidents.
- Cost remains a significant consideration, so asphalt emulsion products are
very attractive if they can perform as a deck sealer.
- Deck sealing properties also appear to be favorable for many products, but
unknowns remain about possible deleterious effects of placing a vapor barrier

on the decks and perhaps increasing the rate of freeze-thaw deterioration.
- Soft aggregate of SafeLane was a concern initially and results indicate that
the concern was well founded; it is NOT a satisfactory product.
- Other chip seals have demonstrated ability to reduce accidents in all
weather events and the amount of chemical required in some types of weather
5. North Dakota DOT- The Cargill SafeLane product's skid resistance seams
to be down significantly but we do not have actual skid numbers to prove this
6. Delaware DOT- We do not specify thin bonded overlays for anti icing. -
7. Kentucky DOT- is considering the use of these systems but we do not
currently have any specifications or special notes that pertain to them.
Kentucky has not specified their use to date. We have used them a couple of
times on concrete bridge decks to help with issues during new construction.
3.1.11 Response Summary
Question 1: Questionnaire completed by:
Page lists the contact information for each person that completed the survey,
including name, organization, and e-mail address.

Question 2: Does your State DOT utilize thin-bonded overlays with anti-icing and
anti-skid properties such as Cargill SafeLane?
A total of seven states and eight people responded yes that their state have used thin-
bonded overlay with anti-icing and anti-skid properties.
Question 3: Has the number of winter weather related accidents been reduced on
bridge decks using SafeLane (or other Anti-Icing/Anti-Skid overlays)?
Minnesota and Wisconsin of the eight responding states experienced a decrease in
accidents, at least, early in the overlays life. Accident data was unknown to the
remaining survey participants.
Question 4: What is the typical skid number (SK) that your DOT has observed on
these types of overlays?
Half of the responses documented acceptable skid numbers (>35). The remaining
respondents did not have skid numbers for the overlays placed.
Question 5: Approximately what percent reduction in the amount of deicing
chemicals applied to these types of bridge decks has been experienced when
compared to decks without Anti-Icing/Anti-Skid overlays?
The majority of respondents had no recorded data to determine whether the overlay
decreased the needed amount of deicing chemicals. Some of the responses did not

use the overlay for anti-icing. Only one state responded saying they use deicing salts
which does not utilize the overlays anti-icing characteristic. However, one state
experienced at least a 20% decrease in the amount of chemicals used during winter
Question 6: Has there been any major failures or performance issues when using
this type of overlay. Any concerns? Please specify.
Only two of the eight responses experienced failures with the SafeLane Overlay on
their bridge decks. Wisconsin had problems during the mixing of the epoxy.
Minnesota found a decrease in friction and an increase in accidents after a couple of
Question 7: Are there any special specifications your state uses when applying
overlays with Anti-Icing/Anti-Skid properties?
Two out of the seven states have created specifications for this product. Most other
states are using the product through experimentation and waiting to measure the
overlays performance prior to drafting specifications. One respondent from
Vermont was not sure if there were any in place.

Question 8: Has your State DOT observed if any of the overlay products listed
below exhibit anti-icing properties? Check all that apply.
Two states found that other products have produced anti-icing properties. The
products being Euclid Chemical Flexolith 216, Unitex Propoxy Type IIIDOT, Poly
Carb Flexgrid Mark 163, Degussa Building Systems, and Transpo T-48 epoxy.
However, these products can not be placed on an asphalt wearing surface.
Question 9: Please indicate (check all that apply) the types of deck surface each of
the products listed below (and utilized by your State DOT) are placed on within
your state's bridge inventory. For example, is the thin-bonded overlay product
applied to only Asphalt Wearing Surfaces, only Bare Concrete Decks, or both.
From the responses to this question only two states have placed SafeLane on an
asphalt wearing surface. There are no other products on the market that can be or
have been placed on an asphalt wearing surface.
Question 10: Additional Comments Thank you!
Several states showed interest in the SafeLane Overlay. In fact, the New York DOT
wanted the results from this study. It should be noted that Minnesotas initial concern
about the aggregate in the overlay being sub par was well founded because the
overlays performance decreased over a couple of years.

3.1.12 DOT Survey Conclusions
The results from this survey will be used to help the CDOT design and monitor thin-
bonded overlays with anti-icing and anti-skid characteristics. The survey was
successful in finding a solid foundation of information needed to begin evaluated the
use of the SafeLane Overlay system on Colorado bridge decks.
There is limited experience with thin-bonded overlays with anti-icing and
anti-skid characteristics, particularly on asphalt wearing surfaces. There are no other
products that have these characteristics other than Cargills SafeLane. In addition, it
is the only product on the market that can be placed on an asphalt wearing surface.
The majority of states using this product have measured acceptable skid number
(>35). Several states saw reduction in traffic accidents. However, some states
observed a decrease in the skid numbers and increase in accidents within a few years
of placing the product. Based on the survey results, it is unclear whether the product
can reduce the amount of chemicals needed during winter storm events. The majority
of states have not run into any major failures other than mixing problems and
reduced friction numbers after a few years.

Problem Statement
4.1 Statement
All state Department of Transportations (DOTs) face the problem to find an
overlay for bridge decks that provides a lasting wearing surface and an impermeable
surface to prevent water and salts from penetrating into the structural members.
There is a constant struggle to find a balance between cost, performance, and the
effective life of the overlays. A new product called SafeLane has caught Colorado
DOTs (CDOT) interest, to be used on both their concrete and asphalt bridge decks.
SafeLanes anti-icing abilities and ability to be placed on asphalt has separated itself
from the current products on the market. The CDOT could potentially benefit from
the use of the SafeLane overlay, reducing maintenance costs, increasing road safety,
increasing time prior to deicing chemical application, and more time before the roads
need to be plowed.
The problem with thin-bonded overlays is that they sometimes do not
produce good enough adhesion and peel off of the road surface. The failure of the
overlay to adhere to the pavement surface would result in a pathway for aggressive
chemicals (i.e. deicing salts) to penetrate and decrease the life of the bridge deck.
Currently, SafeLane requires that the surface either to be sand blasted or shot blasted
before application. This thesis with the use of a modified American Concrete

Institute (ACI) 503R known as the pull-off test has investigated how well SafeLane
adheres to different surfaces to determine help CDOT determine the best surface
preparation prior to the application of SafeLane, or if they wish to continue to pursue
this product.
SafeLane was placed on both concrete and asphalt substrates. The different
surface treatment of the substrates consisted of: troweled, tined, grounded using
wired brush, and sandblasted. In addition, this laboratory investigated the effect of
aggregate in the overlay and the effect of surface temperatures during overlay
placement on the bond strength. The test produced failure stresses and mode of
failure for each different sample. Failure modes include: cohesive concrete,
adhesive, and cohesive resin failure.

5. Experimental Design
This research examined the pull off strength of the thin-bonded overlay SafeLane to
concrete and asphalt under different surface treatments and surface temperatures.
This study included: sample fabrication and acquisition, sample preparation, test set
up design and construction, and testing of the various samples to failure. The test
procedure modified ACI 503R for the laboratory. This chapter discusses the
SafeLane overlay materials, experimental design and testing matrix, testing
configuration and procedure, and concrete mixture design and sample fabrication.
5.1 SafeLane Overlay Materials
There are two main materials in SafeLane the epoxy and the aggregate. The epoxys
properties are listed in Table 5.1. SafeLanes epoxy is a modified Type III epoxy
resin base and hardener two component system that meets ASTM-C-881, Grade 1,
Classes B and C requirements. [Cargill Technical Specs, 2007]

Table 5.1 Epoxy properties [Cargill Technical Specs, 2007]
Property Requirement Test Method
Pot Life 15 to 45 mi at 75 F (24 C) ASTM C881 (50 ml sample in paper cup)
Tensile Strength 2,000 to 5,000 psi at 7 days ASTM D638
Tensile Elongation 40% to 80% at 7 days ASTM D638
Viscosity 7 to 25 poises ASTM D2393
Minimum Compressive Strength at 3 hours 1,000 psi at 75 F (24 C) ASTM C579
Minimum Compressive Strength at 24 hours 5,000 psi at 75 F (24 C) ASTM C579
Minimum Adhesive Strength a 24 hours 250 psi at 75 F (24 C) ACI 503R
The important value to observe is the minimum adhesive strength at 24 hours of 250
psi (1.72 MPa) at 75 F (24 C) using ACI 503R. This is the test that was modified
and performed in this experiment. The values taken in the experiment are compared
to this value. [Cargill Technical Specs, 2007]
Table 5.2 Aggregate gradation [Cargill Technical Specs, 2007]
Gradation % Passing
3/8" (9.5 mm) 98-100
#4 (4.75 mm) 50-80
#8 (2.36 mm) 0-15
The aggregates gradation of SafeLane is listed in Table 5.2. SafeLanes
aggregate is lightweight, surface dry and free of dirt, clay, asphalt and other foreign
or organic materials. [Cargill Technical Specs, 2007]
Proper surface preparation must be performed before placement of the
overlay. The surface inspection is performed to identify areas of the pavement that

needs repair. This includes delaminations in the concrete or asphalt, potholes, large
cracks or breakouts. [Cargill Technical Specs, 2007]
For concrete bridge deck or road the surface should be prepared as follows.
First the surface shall be steel shot blasted to thoroughly clean to ensure proper
bonding between the epoxy and concrete substrate. The coarse aggregate is supposed
to be exposed and the surface free of asphalt material, oil, dirt, rubber, curing
compounds, paint carbonation, laitance, weak surface mortar and potentially
detrimental materials once shot blasting is completed. Afterwards all traffic marking
lines shall be removed or protected as directed. All dust and other loose material
shall be removed by moisture and oil free compressed air or high volume leaf
blowers. In the case of rain, mechanical brooms, without water, may be used to
remove any residual dust, followed by moisture and oil free compressed air or high
volume leaf blowers. Before the surface is considered acceptable adhesion bond
strength testing may be required. Unless approved SafeLane shall not be placed on
pavement that is less than 28 days old. [Cargill Technical Specs, 2007]
For asphalt bridge deck or road the surface should be prepared as follows.
Sand blasting or planing and texturing to an approved depth should be performed.
Surfaces should be free of oil, dirt, rubber, curing compounds, paint carbonation,
weak surface mortar and other potentially detrimental materials. All dust and other
loose material shall be removed by moisture and oil free compressed air or high
volume leaf blowers. In the case of a rain event mechanical brooms, without water,

may be used to remove any residual dust, followed by moisture and oil free
compressed air or high volume leaf blowers. Asphalt less than six months old shall
not receive an application of SafeLane. After surface preparation is completed it is
recommended that the overlay should be placed as soon as possible and to keep
traffic closed until the overlay has been placed and allowed to fully cure. [Cargill
Technical Specs, 2007]
The overlay application conditions are as follows. When surface conditions
are such that he material cannot be properly handled, placed and cured within the
specified requirements for project sequencing or traffic control, or when rain is
impending within the curing time SafeLane should not be applied. At the time of
epoxy application the prepared surface must be completely dry. To dry the surface
moisture and oil-free heat sources or torches may be used. There must be a minimum
of 55 F (13 C) temperature of the deck surface and all epoxy and aggregate
components at the time of application. If pavement temperatures exceed 115 F (46
C) or the gel time is less than five minutes the epoxy shall not be applied. [Cargill
Technical Specs, 2007]
The application of SafeLane uses a double pass method. This method
requires that the epoxy and aggregate are applied in two separate layers at
corresponding application rates shown in Table 5.3. The minimum application rate is
10 gallons per 100 square feet. The typical application rate ranges from 10-11
gallons per 100 square feet. [Cargill Technical Specs, 2007]

Table 5.3 Application rates [Cargill Technical Specs, 2007]
Double Pass Method Epoxy Rate Gal/100 sq. ft.* Aggregate lbs/sq ft. **
(liters/sq meter) (kg/sq. meter)
1st course 2.0 to 4.0 (0.8 to 1.6) 1-2(4.9-9.75)
2nd course 6.0 to 8.5 (2.4 to 3.4) 3-4 (14.6-19.5)
* Total epoxy applied must equal no less than 10 gal/100 sq ft. (4.1 liters/sq meter)
"Application of aggregate shall be of sufficient quantity to completely cover the
The epoxy must be mixed at a volume ratio of 1:1 Part A to Part B and mechanically
stirred by a paddle type mixer for three minutes or according to the epoxy
manufacturers recommendations. Immediately after mixing it shall be uniformly
applied to the pavement surface with a 3/16 to Va inch (4.8-6.4 mm) V-notched
squeegee. While the epoxy is still fluid, the aggregate shall be applied in such a
manner as to cover the epoxy mixture. First course applications that do not receive
enough aggregate prior to gelling shall be removed and replaced. [Cargill Technical
Specs, 2007]
To prevent tearing or damaging of the surface each course of epoxy overlay
shall be cured before removing the excess unbonded aggregate with an oil- and
moisture- free compressed air or high volume leaf blowers, vacuum or mechanical
broom. Any remaining dust must be removed as well after all the loose aggregate is
removed using the methods described above. The second course may begin being
applied after all dust is removed. Traffic shall not be permitted on the overlay until it

has been cured sufficiently. Typical curing times are listed in Table 5.4. [Cargill
Technical Specs, 2007]
Table 5.4 Curing times [Cargill Technical Specs, 2007]
Average Temperature of Deck, Epoxy, and Aggregate
Course____________________Components in oF (oC)____________
60-64 65-69 70-74 75-79 80-84 *85+
(16-18) (18-21) (21-23) (24-26) (27-29) (29+)
1 4 hr 3 hr 2.5 hr 2 hr 1.5 hr 1 hr
2 6.5 hr 5 hr 4 hr 3 hr 3 hr 3 hr
Application rates for the second course shall follow rates specified in Table 5.3.
Epoxy will be applied using a flat-bladed squeegee. While the epoxy is still fluid, the
aggregate shall be applied in such a manner as to cover the epoxy mixture. Special
care must be taken to ensure that the wear surface (top) of the aggregate does not
become coated with wet epoxy. Once the epoxy overlay is cured, all excess
unbonded aggregate shall be removed with an oil- and moisture- free compressed air
or high volume leaf blowers, vacuum or mechanical broom. A light shot or sand
blast shall be used on any areas where the top surface of the stone has been coated
with epoxy to remove the excess. [Cargill Technical Specs, 2007]
5.2 Experimental Design and Testing Matrix
A total of 21 samples were tested. The samples including pavement type, surface
treatment, and number of samples tested are listed in the testing matrix shown in
Table 5.5.

Table 5.5 Test matrix
Number of Samples
Material Surface Treatment Surface Temp. (F) Epoxy Epoxy w/ Aggregate
Asphalt Sand blasted 1
Smooth 1
Concrete Troweled 2 2
115 1
100 1
85 1
Sand blasted 2 2
Tined 2 2
Roughened 2 2
Total= 13 8
The asphalt samples tested were taken from an existing pavement in the City and
County of Denver. The concrete samples were designed and created in the lab
according to the CDOT Class H structural concrete specification. The specimens
were fabricated in the lab because specimens from the field could not be obtained.
This class is used for all CDOT bridge decks in the state of Colorado. The samples
are separated into two main categories: epoxy Unitex Pro-Poxy Type III DOT, and

SafeLane overlay which uses epoxy Unitex Pro-Poxy Type III DOT with aggregate.
The samples that were tested with only the epoxy applied included 13 specimens:
1 Asphalt sand blasted
1 Asphalt with smooth surface
2 Concrete troweled surface
1 Concrete troweled finish at 115 F
1 Concrete troweled finish at 100 F
1 Concrete troweled finish at 85 F
2 Concrete sand blasted
2 Concrete roughened with a rotating wire brush
2 Concrete with tined surfaces
The samples that were tested with both the epoxy and aggregate (SafeLane Overlay
System) included 8 specimens:
2 Concrete smooth finish
2 Concrete sand blasted
2 Concrete roughened
2 Concrete with tined surfaces
The pull-off test was conducted to examine the effects of the epoxys
adhesion by different surface treatments, surface temperatures, surface material, and
the use of the aggregate.

5.3 Testing Configuration and Procedure
The testing was designed and constructed for laboratory experiments similar to ACI
503R which is used more for field applications. The test was developed to fit into a
20 kip MTS machine which would apply direct tension to the test sample. The
construction utilizes a system of plates that held the samples in place and transferred
the loads to the sample. The test measured the load applied and the displacement
until the sample failed.
5.3.1 Test Configuration Design and Construction for Concrete Specimens
The materials involved to set a concrete sample to the 20 kip MTS machine include
the following shown in figure 5.1:
12 x 12 x Vi in. (30.48 x 30.48 x 1.27 cm) Aluminum Top Plate (Figure 5.1a)
12 x 12 x 1 in. (30.48 x 30.48 x 2.54 cm) Aluminum Bottom Plate (Figure
12 in. (30.48 cm) piece of 1.5 in. (3.81 cm) pipe with threads on both ends
(Figure 5.1c)
1.5 in. (3.81cm) pipe cap- ( 1 for the top connection and 1 for every test
preformed) (Figure 5. Id)
3/8 in. (.953 cm) NutsAVashers
6 x 6 x 14 in. (15.24 x 15.24 x .635 cm) Steel Plate (Figure 5.1e)
Figure 5.2 is a picture of the complete test setup.

Figure 5.1 Components of Concrete Test Setup (a) Aluminum Top Plate (b)
Aluminum Bottom Plate (c) Pipe (d) Pipe Cap with All Thread
(e)Steel Plate

Steel Plate
All Tread
Pipe Cap
All Thread
with Nut and
Top Aluminum
Figure 5.2 Concrete Test Apparatus
The concrete was bolted to the aluminum plates through the outside holes using all
thread that was embedded into the concrete. The center holes on the bottom 1 in.
(2.54 cm) thick aluminum plate and the corresponding holes on the steel plate was
the bolt design on the MTS machine used to connect the test unit to the machine. The
center V2 in. (1.27 cm) threaded hole on the steel plate was used to bolt the top pipe
cap to the plate. Finally, the 5.5 inch (13.97 cm) center hole on the top (Vt in. or 1.27

cm thick) aluminum plate was to reveal the test area and provide a big enough
opening such that the results are not affected by the surrounding aluminum plate.
5.3.2 Test Configuration Design and Construction for Asphalt Specimens
The testing configuration involved to set the asphalt blocks to the MTS machine
included the same materials as the concrete machine setup except the 12x12x1/2 in.
(30.48 x 30.48 x 1.27 cm) aluminum top plate shown in Figure 5.4. In addition, the
setup included the following materials that were not used in the testing configuration
of the concrete specimen shown in Figure 5.3:
1. 4 slotted channel bars (Figure 5.3a)
2. 4- 3/8 in. (0.953 cm) all tread- length needs to be long enough to attach the
top of the sample to the bottom plate through 2 of the slotted channels
(Figure 5.3b)
Figure 5.3 Components of Asphalt Test Setup (a) Slotted Channel Bar (b) All

The complete asphalt setup is pictured in Figure 5.4.
Steel Plate
All Thread
Pipe Cap
Channel Bar
All Thread
with Nuts and
Figure 5.4 Asphalt Test Apparatus

5.4 Concrete Mixture Design and Sample Fabrication
New concrete blocks had to be fabricated to accommodate test requirements. The
fabrication of these blocks included mixture design, form construction, concrete
batching, and specimen fabrication.
5.4.1 Mixture Design
The mixture design chosen for this experiment was CDOTs class H mixture. CDOT
used this class for all bare concrete bridge decks in the state. This experiment tested
SafeLane overlay for bridge decks this concrete mixture class was chosen and used.
This class requires:
A Water Reducing Admixture
Minimum of 55% AASHTO M 43 size No. 67 coarse aggregate by weight of
total aggregate
450 to 500 lb/yd3 (267 to 297 kg/m3) of Type II portland cement
90 tol25 lb/yd3 (53.4 to 74.2 kg/m3) Fly Ash
20 to 30 lb/yd3 (11.9 to 17.8 kg/m3) Silica Fume
Total Cementitious Material between 580 and 640 lb/yd3 (344 and 380
5 to 8 percent air content
0.38 to 0.42 water to cementitious ratio(w/cm)
4500 psi (31.03 MPa) compressive strength at 56 days

Permeability less than 2000 coulombs at 56 days
No cracking at or before 14 days in the cracking tendency test.
A mixture conforming to the previously stated concrete mixture characteristics was
designed for this study. The mixture proportions are listed in Table 5.6. This design
met all the requirements for the mixture proportions required for CDOT Class H
structural concrete mixture.
Table 5.6 Mixture proportions (SSD)
Materials Mixture
Cement (lb/yd3) 467
Fly Ash (lb/yd3) 93
Silica Fume (lb/yd3) 20
Coarse Aggregate (lb/yd3) 1766
Fine Aggregate (lb/yd3) 1267
Water (lb/yd3) 232
Air (%) 6.5
w/cm 0.40
5.4.2 Form Construction
The concrete samples were designed to have dimensions of 12 x 12 x 3.5 in. (30.48 x
30.48 x 8.89 cm). Figure 5.5 shows finished concrete sample forms. The procedure
used to create the concrete forms was:
1. Cut a piece of wood board to 15 in. (38.1 cm) square.
2. Cut 2 x 4s (5.08 x 10.16 cm) into 4- 13.5 in. (34.29 cm) pieces.
3. Screw together the 2 x 4s to form a square with 3.5 in. (8.89 cm) tall sides

4. Screw the 15 in. (38.1 cm) square board to the bottom of the 2 x 4 square,
screwing only into two adjacent pieces of the 2 x 4s. This is done to make
releasing the concrete from the forms easier.
5. Cut 1 x 1 in. (2.54 x 2.54 cm) into 4- 8 in. (20.32 cm) pieces.
6. Attach them to the center edges of the 15 in. (38.1 cm) square board on the
opposite side of the 2 x 4s.
7. Drill 3/8 in. (0.953 cm) holes into the 15 in. (38.1 cm) square board matching
the pattern of the outside holes on the aluminum plates.
8. Bolt down 6 inch pieces of all thread down to the board leaving a 1.5 in.
(3.81 cm) below the top surface of the board. That way there is enough all
thread to bolt the plates down with.
9. Cover the top inch of all thread with tape. This is to protect the all thread
from the concrete.

Figure 5.5 Concrete Form Assemblage
5.4.3 Concrete Batching
The batching of the concrete went as follows. The testing of the fresh and hardened
concrete properties for each batch were tested in accordance to the ASTM standards.
To reduce the batch size a total of three batches were made. This was critical to
insure that the contents fit inside the concrete mixture and the concrete did not
harden before being placed. The batching of the test samples followed ASTM C 192
Standard Practice for Making and Curing Concrete Test Specimens in the Laboratory
(AASHTO T 126-97). The cylinders were moist cured in a water bath at 73F (23C)
until testing. The concrete samples were cured in humidity and temperature
controlled room. The fresh and hardened concrete properties tests performed are
listed in Table 5.7.

Table 5.7 Fresh and hardened concrete properties tests
Standard Time of Test
Fresh Concrete Tests
Slump C 143 T 119 At Batching
Unit Weight C 138 T 121 At Batching
Air Content C 231 T 152 At Batching
Temperature C 1064 T 309 At Batching
Hardened Concrete Tests
Compressive Tests C 39 T 22 7, 28 Days and time of bond testing
5.4.4 Specimen Fabrication
The method used to fabricate the different surface conditions for the samples is
described as follows. Figure 5.6a is an illustration of a troweled sample. The
troweled sample was created by smoothing the concrete with a trowel during the
finishing process of sample fabrication. All samples were troweled to begin with.
Figure 5.6b is an illustration of a tined sample. The tined samples were fabricated
with a line etched into the concrete approximately every lA inch (0.635 cm) once the
concrete had slightly stiffened using a ruler and a flat head screw driver. Figure 5.6c
is an illustration of a roughened sample. The roughened samples were created by
first tinned, and then a Black & Decker 2 in. wire cup brush attached to a drill was
used to remove the surface of the concrete to the fine aggregate. This process took
about 5-10 min. for each sample. Figure 5.6d is an illustration of sand blasted

sample. The sand blasted samples were first tinned and then taken to Roadrunner
Fabrication where they sand blasted the samples. The samples were sand blasted
until the coarse aggregate was uncovered.
Figure 5.6 Sample Surface Types
Next the overlay was applied once the samples met the required strength and
the surface treatments were performed. The epoxy is measured out at a volume ratio
of 1:1 Part A to Part B illustrated in Figure 5.7a. The two parts are combined and

mixed for 3 minutes illustrated in Figure 5.7b. The epoxy was applied using a depth
!4 in (0.635 cm) using a lA in (0.635 cm) v-notch trowel for the first layer of epoxy
illustrated in Figure 5.7c. The aggregate was applied on top of the still wet epoxy
using a spoon to scoop out and scatter the aggregate to completely cover the epoxy
on the samples that require the aggregate illustrated in Figure 5.7d. Once the epoxy
had cured completely based on the curing times listed in Table 5.4, the unbonded
aggregate was removed using compressed air. Right after the removal of the excess
aggregate the second course was applied. A flat squeegee was used to spread and
apply the second layer of epoxy. Figure 5.7e illustrates the application of the second
layer of epoxy on a sample that is epoxy only. Subsequently a second layer of
aggregate was applied to the appropriate samples in a similar manner as the first

Figure 5.7 Application of Overlay

The samples that test the roads temperature effect on the epoxy were
prepared the following way. The samples were heated overnight to their respect
temperatures. The samples had the epoxy placed on them immediately after they
were removed them from the oven.
Following the placement of the overlay or epoxy, a 2 in. (5.08 cm) core drill
bit was drilled to a depth of 0.5 in. (1.27 cm) from the surface of the concrete or
asphalt specimens. To prevent significant stress non-uniformity 0.5 in (1.27 cm)
depth was chosen. Significant stress non-uniformity may occur at shallow drilling
depths according to (Austin et. al., 1995). The paper suggested using a substrate
coring depth of 15 mm (~ 0.5 inch) when performing the pull-off test.
Next, the pipe caps were applied to the center of the 2 in. (5.08 cm) core with
an epoxy. The epoxy used was 3M Scotch-Weld Epoxy Adhesive 1838 B/A green. It
was chosen because it will bond to both metal and concrete with controlled flow. The
epoxy hardens between 6-10 hours or 15-20 minutes at 150 F (65 C). Do not
exceed a temperature of 200 F (93 C). It has an approximate work life of 60
minutes at 73 F (23 C). Apply adhesive at 60 F (16 C) or above. Use a spatula,
notched trowel or pressure-flow equipment to apply the adhesive and only contact
pressure is required for adhesion. The mixture ratio is 1:1 Part A to Part B by weight.
Mixing the two parts takes 10-15 seconds after uniform color is obtained. The epoxy
has an ultimate tensile strength of 4290 psi (27.6 MPa) and a modulus of elasticity of
344,000 psi (2,372 MPa).

To create a level surface and to match the machine the bottom of the concrete
was propped up so the sample was not resting on the all thread before attaching the
pipe cap. The pipe cap was screwed into the pipe. Using a clamp and level, the pipe
cap was applied to the center of the cored section with the adhesive and using the
clamp to hold the pipe vertically while the epoxy cured. This method is pictured in
Figure 5.8.
Figure 5.8 Attaching Pipe Cap

Once the epoxy cured, the samples were placed into the machine setup. The
testing machine applied a load of 5 pounds per second (22.24 N/s) until the specimen
failed. Although, the ACI 503R load rate is specified at 20 pounds per second (88.96
N/s), the lower load rate was chosen due to the fact that higher loading rates result in
higher failure strengths (Darwin, 2003). Therefore, a slower rate was chosen due to
the desire to obtain a more accurate failure load and supply more data points. The
failure load of these specimens was expected to be approximately 250 to 300 psi
(1.723 to 2.068 MPa). However, with the slower loading rate it is expected that the
failure loads will be lower. Once the sample fails record what type of failure that
occurred. There are three possible failure types or any combination thereof. The
failure types are: failure in the concrete (cohesive concrete failure), separation of
epoxy compound from the concrete surface (adhesive failure), and failure in the
epoxy compound (cohesive resin failure). Figure 5.9a is a picture of a cohesive
concrete failure. Figure 5.9b is a picture of an adhesive failure.

Figure 5.9 Failure Types (a) cohesive Concrete Failure (b) Adhesive Failure at
Concrete Surface

6. Experimental Results
This section includes both the concrete properties and data results. All variables were
held constant during the experiment except for the variables being tested. Therefore,
comparisons between the different samples can be made.
6.1 Concrete Properties
The fresh concrete properties for each batch are listed in the Table 6.1.
Table 6.1 Fresh concrete properties
Fresh Concrete Property Batch 1 Batch 2 Batch 3
Slump 2.75 6 0 in.
Air Content 6.4 5 2.9 %
Ambient Temperature 77 76 62 F
Concrete Temperature 78 77 68 F
Unit Weight 144.4 145.8 154 pcf
Three batches were mixed to reduce to batch size. Batch one was used to batch the 8
surface treated epoxy test samples. Batch two was used to batch the 8 surface treated
epoxy and aggregate test samples. Batch three was used to batch the 3 concrete
temperature test samples. Batch three does not need to have the same properties
because it will not be compared to the other batches samples. It is not compared to
the other two batches because this batch contains the temperature differences.

The compressive strength of the concrete batches made during this study is
summarized in Table 6.2.
Table 6.2 Compressive strengths
Average Compression (psi)
Batch 1 Batch 2 Batch 3
7 Day 4992 4772 62*
28 Day 6032 5582 9007*
At Testing 6776 6734 10812
Average of 3- 4 x 8 in. cylinders
* Average of 2- 4 x 8 in. cylinders
Batch 3 at 7 day and 28 day strength was preformed with two cylinders
instead of three because there was only enough concrete left of the batch to make 7
of the 9 cylinders. Batch 3 had a greater compressive strength because the air
content was lower and had a greater amount of high range water reducer. The greater
amount of high range in effect delayed set but increased the ultimate strength.

Compressive Strength vs. Time
O- 10000
c 8000
w 4000
| 2000
0 50 100 150 200
Time (Days)
Figure 6.1 Compressive Strength vs. Time
Figure 6.1 shows that batch 1 and 2 have almost identical compressive
strength verses time graph, while batch 3 is radically different. Both batch 1 and 2
gained the majority of their strength by 7 days with between 70-74% of their total
strength. Batch 3 gained less than 1% of its total strength by 7 days. Batch 3 gained
83% of its total strength between 7 and 28 days. Batch 1 and 2 gained between 12-
15% of their total strength between 7 and 28 days. Batch 3 gained 17% of its total
strength between 28 and 155 days. Batch 1 gained 11% of its total strength between
28 and 79 days. Batch 2 gained 17% of its total strength between 28 and 81 days.

Table 6.3 Batch comparisons
Concrete Property % Difference for Batch 1 and 2
Slump 74.3
Air Content 24.6
Unit Weight 1.0
Compressive strength at testing 0.6
Table 6.3 summarizes the percent difference between batches 1 and 2
different concrete properties. The percent difference between batch 1 and 2 slump is
74.3%. This is a large difference. It should not have an effect on the bond test
comparisons because slump is only an indicator of how workable the fresh concrete
is. The percent difference in the air content of 24.6% could have an effect on the
bond test because the overlay could potentially penetrate the surface more easily and
produce better adhesion. The percent difference in unit weight is 1.0%. This is a
small difference and is due to the difference in air content. Unit weight should have
no effect in bond strength. The percent difference in compressive strength at testing
is the smallest difference with 0.6%. This should play zero role in effecting the bond
strengths. Table 6.4 lists the requirements for CDOTs Class H mixture design and
compares them to the concrete mixtures.

Table 6.4 Mixture requirement comparison
Category Required Mixture Met Requirement?
Water reducing admixture Yes Yes Yes
AASHTO M 43 size No. 67 coarse aggregate by weight of total aggregate >55% 58.20% Yes
Fly Ash 90 to 125 Ib/yd3 93 Ib/yd3 Yes
Silica Fume 20 to 30 Ib/yd3 20 Ib/yd3 Yes
Cementitious Material 80 to 640 Ib/yd3 580 Ib/yd3 Yes
w/cm 0.38 to 0.42 0.4 Yes
Compressive Strength at 28 Days 4500psi at 56 days 6700 10800 psi at 28 days Yes
The concrete failed to meet the 5 to 8 percent air requirement for batch 3. This batch
of concrete had 2.9 percent air content. In addition, the required cracking tendency
and permeability tests were not preformed.
6.2 Data Results
The results from the tests performed are provided in Table 6.5. This table lists the
type of sample, the failure type, and the stress at which the sample failed. The failure
type is an indicator of the bond. If the sample had a cohesive concrete failure it
indicates that a good bond was formed and the concrete had the smallest tensile
strength. Out of the 21 samples, only 16 produced data. Five samples did not produce
data for a couple of reasons. Four of the samples broke during preloading. The
preload is the small initial load applied rapidly to begin the test. The last samples

pipe cap was placed to far from the center that the sample did not fit into the test
setup. Out of the 16 samples, 14 had cohesive concrete failures.
Table 6.5 Test results
1 2
Surface Type Surface Treatment Ultimate Failure (psi) Failure Type Ultimate Failure (psi) Failure Type
Concrete T roweled 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
Sand Blasted
144 Adhesive
90 Cohesive Ashpalt
Epoxy and Aggregate
1 2
Surface Type Surface Treatment Ultimate Failure (psi) Failure Type Ultimate Failure (psi) Failure Type
Concrete Troweled 214 Cohesive Concrete 263 Adhesive
Tined 192 Cohesive Concrete 105 Cohesive Concrete
Roughened 277 Cohesive Concrete 224 Cohesive Concrete
Sand Blasted 146 Cohesive Concrete - Cohesive Concrete
,,, . Force at Failure
UltimateFailure(psi) =----------------
x*D2/4 n*22!4
The troweled epoxy samples failed within the concrete at stresses of 73 and 208 psi
(0.503 and 1.44 MPa). The troweled epoxy and aggregate samples failed at a stress
of 214 psi (1.48 MPa) within the concrete and 263 psi (1.81 MPa) between the

overlay and the concrete. The tined epoxy samples failed within the concrete at
stresses of 219 and 182 psi (1.51 and 1.25 MPa). The tined epoxy and aggregate
samples failed within the concrete at stresses of 192 and 105 psi (1.32 and 0.723
MPa). The roughened epoxy samples failed within the concrete at stresses of 162 and
87 psi (1.12 and 0.60 MPa). The roughened epoxy and aggregate samples failed
within the concrete at stresses of 277 and 224 psi (1.91 and 1.54 MPa). The sand
blasted epoxy sample failed within the concrete at a stress of 159 (1.10 MPa). The
sand blasted epoxy and aggregate failed within the concrete at a stress of 146 psi
(1.01 MPa). The 85 degree Fahrenheit sample was not tested because core was
drilled too far from the center of the sample to attach to the MTS. The 100 degree
Fahrenheit sample failed within the concrete at a stress of 144 psi (0.99 MPa). The
115 degree Fahrenheit sample failed during preloading between the epoxy and the
concrete. Preloading is at the beginning when the sample is. The asphalt no treatment
sample failed within the asphalt at 90 psi (0.62 MPa). The asphalt sand blasted
sample failed during preloading.
6.2.1 Analysis of Data
The small number of samples and the scatter in the data made it difficult to draw
many concrete conclusions. The scatter could be from not having pure tension being
applied to the samples, samples becoming damaged or slight changes in drilling
83 Surface Type Comparisons
When the epoxy was applied to the samples where the surfaces were treated
differently the results varied. The troweled specimens failed at 73 and 208 psi (0.503
and 1.44 MPa) with an average of 140.5 psi (0.969 MPa). While the tined specimens
failed at 219 and 182 psi (1.51 and 1.25 MPa) with an average of 200.5 psi (1.38
MPa. The roughened specimens failed at 162 and 87 psi (1.12 and 0.60 MPa) with an
average of 124.5 psi (0.858 MPa). The sand blasted sample failed at 159 psi (1.10
MPa). Comparing the average of each type the specimens with the best adhesion to
the worst was tined, sand blasted, troweled and roughened.
When the epoxy and aggregate was applied to the specimens where the
surfaces were treated differently the results varied. The troweled specimens failed at
214 and 263 psi (1.48 and 1.81 MPa) with an average of 238.5 psi (1.64 MPa). The
tined specimens failed at 192 and 105 psi (1.32 and 0.724 MPa) with an average of
148.5 psi (1.02 MPa). The roughened specimens failed at 277 and 224 psi (1.91 and
1.54 MPa) with an average of 250.5 psi (1.73 MPa). The sand blasted specimens
failed at 146 psi (1.01 MPa). Comparing the average of each type the specimens with
the best adhesion to the worst was roughened, troweled, tined, and sand blasted.
The epoxy surface treated specimens do not correlate well with the epoxy and
aggregate surface treated specimens. In this case it appears the results of the best to
the worst adhesion are opposite when comparing the epoxy surface treated
specimens verses comparing the epoxy and aggregate surface treated specimens. The

epoxy roughened specimens performed the worst and the epoxy and aggregate
roughened specimens performed the best. The troweled epoxy specimens performed
the third best and the troweled epoxy and aggregate specimens performed the second
best. The tined and the sandblasted in both cases the adhesion was close, with the
epoxy samplings performing the worst and the epoxy and aggregate performing the
No analysis could be performed on the temperature samples since only one
data point was collected. The 100 F (37.8 C) temperature at placement of epoxy
sample failed within the concrete at a stress of 144 psi (0.99 MPa).The 85 F (29.4
C) temperature of the sample at placement of the epoxy could not be tested because
sample was not cored at the center of the sample. The 115 F (46.1 C) sample did
not produce data because it failed at small loads during the preloading of the sample
The asphalt sample with no treatment produced a failure stress of 90 psi (0.62
MPa). No comparison can be made with the sand blasted asphalt sample because the
sample failed at small loads during the preloading of the sample. Epoxy vs. Aggregate Comparisons
To determine whether the aggregate that is applied with the epoxy to form the
SafeLane Overlay System affects the adhesion, the epoxy alone was compared to the
complete system of epoxy and aggregate. The troweled specimens failed at values of
73 and 208 psi (0.503 and 1.44 MPa) with an average of 140.5 psi (0.969 MPa) for

the epoxy alone at 214 and 263 psi (1.48 and 1.81 MPa) with an average of 238.5 psi
(1.64 MPa) for the epoxy and aggregate. The epoxy and aggregate for the troweled
specimens clearly had higher average adhesion. The values of 219 and 182 psi (1.51
and 1.25 MPa) with an average of 200.5 psi (1.38 MPa) for the tined epoxy only
specimens verses 192 and 105 psi (1.32 and 0.724 MPa) with an average of 148.5 psi
(1.02 MPa) for the tined epoxy and aggregate specimens, shows that the epoxy had
on average a higher bond strength for the tined specimens. Comparable results are
shown between the sandblasted specimen with epoxy only, 158 psi (1.09 MPa), and
epoxy and aggregate specimen, 146 psi (1.01 MPa). The roughened specimens had
higher average values for the epoxy and aggregate of 277 and 224 psi (1.91 and 1.54
MPa) with an average of 250.5 psi (1.73 MPa) compared with the epoxy of 162 and
87 psi (1.12 and 0.60 MPa) with an average of 124.5 psi (0.858 MPa).
The samples had split results. The epoxy alone produced higher average
failure stresses for the tined and sand blasted samples. While, the epoxy and
aggregate produced higher average failure stresses for the troweled and roughened
samples. Stress vs. Strain Graphs
Stress verses strain graphs were created to analyze the data further. The slope (E-
Modulus of Elasticity) of the graphs or the stiffness is a function of bond strength.
All variables were held constant during the experiment except for the surface

preparation. The graphs were made taking the load and displacement data and
calculating the stress and strain at each given time. The stress was calculated by
taking the load divided by the area of the drilled circle. The strain was calculated by
taking the displacement divided by the starting length. The starting length was taken
as the approximate distance from the bottom of the steel plate to the bottom of the
sample. A length of 22 inches (55.88 cm) was used for the concrete specimens and
25 inches (63.5 cm) for the asphalt specimens. Figure 6.2 has a stress verses strain
graph for the concrete epoxy specimens. Figure 6.3 has a stress verses strain graph of
the epoxy and aggregate specimens.
Figure 6.2 Concrete Epoxy Stress vs. Strain Graph

The percent difference in the slopes of the graph in Figure 6.2 is discussed
below. The tined 1 had a slope of 87.2 ksi (601 MPa). The trowel 1 had a slope of
135.6 ksi (938 MPa). The 100 F (38 C) surface temperature had a slope of 141.3
ksi (9,720 MPa). The roughened 2 had a slope of 266.6 ksi (1,830 MPa). The trowel
2 had a slope of 489.1 ksi (3,370 MPa). The tined 2 had a slope of 503.6 ksi (3,470
MPa). The roughened 1 had a slope of 509.8 ksi (3,520 MPa). The sand blasted 1
had a slope of 601.3 ksi (4,140 MPa). The percent difference between the two tined
specimens is 70.5%. The percent difference between the two troweled specimens is
57.2%. The smallest percent difference of the same type of sample was between the
two roughened specimens with 31.3%. The largest percent difference in the slope
was between sandblasted 1 and tined 1 with 74.7%. The graphs show that the higher
the bond strength the higher the stiffness.