GENERIC EXPERIMENTAL DESIGN FOR
PRODUCT/ STRATEGY EVALUATION -
CRUMB RUBBER MODIFIED MATERIALS
State of California Department of Transportation
Materials Engineering and Testing Services
Office of Flexible Pavement Materials
5900 Folsom Blvd
Sacramento, California 95819
February 8, 2005
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
i
EXECUTIVE SUMMARY
This report presents the framework for a generic process to evaluate new products and/ or strategies for
possible use within Caltrans. The framework is the result of a collaborative effort among Caltrans, the
University of California Partnered Pavement Research Center ( UC PPRC) and MACTEC.
The generic experimental design process is applicable to all pavement types as well as the component
materials of pavements. The framework includes various types of studies that may be used in the
evaluation process ― laboratory, accelerated pavement testing ( APT) and field pilot studies.
Additionally, it identifies other factors that must be considered in the evaluation of any new
product/ strategy: economic viability and environmental impact. The generic experimental design process
outlined herein is considered appropriate for Caltrans operating units such as METS, Research,
Maintenance and others who may be involved with evaluating paving materials.
Specific ideas/ hypotheses that may broaden and expand the use of crumb rubber modifier ( CRM) in
pavement applications were identified as follows:
• Construction/ Rehabilitation and Maintenance Applications
o New construction
o Thick overlays
o Open graded- high binder mixes
o Recycling
• Materials Studies
o Type 1 vs. Type 2 binders
o Binder testing
• Structural Design Studies
o Gravel factor for RAC- G and MB mixes
o Modification of the deflection based overlay design procedure
A collaborative effort of Caltrans, UC PPRC, and MACTEC staff, two structural design- related studies
were expanded into detailed work plans which address the following:
• development of a gravel factor for RAC- G and MB mixes for use in new construction; and
• update/ modification of Caltrans deflection based overlay design procedure to accommodate
RAC- G and MB mixes.
It is recommended that the generic process presented in this report be reviewed by and discussed with the
affected operating units and enhanced to include detailed information on data collection and testing
requirements associated with each study type.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
ii
TABLE OF CONTENTS
EXECUTIVE SUMMARY.................................................................................................... i
1.0 INTRODUCTION ......................................................................................... 1
1.1 BACKGROUND......................................................................................................... 1
1.2 SCOPE AND OBJECTIVE .......................................................................................... 1
1.3 ORGANIZATION OF REPORT................................................................................... 1
2.0 GENERIC EXPERIMENTAL DESIGN PROCESS................................. 2
2.1 INTRODUCTION ....................................................................................................... 2
2.2 GENERAL PROCESS OVERVIEW............................................................................. 2
2.3 TYPES OF INVESTIGATIONS.................................................................................... 6
2.3.1 Laboratory Studies ...................................................................................... 6
2.3.2 Accelerated Pavement Testing Study .......................................................... 8
2.3.3 Field Pilot Studies ..................................................................................... 10
2.4 ECONOMIC ANALYSIS AND ENVIRONMENTAL IMPACT....................................... 17
2.5 IMPLEMENTATION ................................................................................................ 17
2.6 SUMMARY ............................................................................................................. 17
3.0 PROPOSED CRUMB RUBBER MODIFIED MATERIAL STUDIES 18
3.1 CONSTRUCTION/ REHABILITATION AND MAINTENANCE APPLICATIONS ........... 18
3.1.1 New Construction...................................................................................... 18
3.1.2 Thick Overlays .......................................................................................... 18
3.1.3 Open Graded- High Binder Mixes ............................................................. 18
3.1.4 Recycling ................................................................................................... 19
3.2 MATERIALS STUDIES............................................................................................ 19
3.2.1 Type 1 vs. Type 2 Binders ......................................................................... 19
3.2.2 Binder Testing ........................................................................................... 20
3.3 STRUCTURAL DESIGN STUDIES ............................................................................ 20
3.3.1 Gravel Factor( s) for RAC- G, MB- D, and MB- G ...................................... 20
3.3.2 Modification of Overlay Design Procedure.............................................. 21
3.4 SUMMARY ............................................................................................................. 21
4.0 SUMMARY AND RECOMMENDATIONS ............................................ 23
4.1 SUMMARY ............................................................................................................. 23
4.2 RECOMMENDATIONS ............................................................................................ 23
5.0 REFERENCES ............................................................................................ 24
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
iii
LIST OF APPENDICES
Appendix A Proposed Deflection Test Plan for Performance Evaluation of Modified
Binder Study for All Field Projects
Appendix B Proposed Condition Survey Method for All Flexible Pavement Field Studies
Appendix C Example Experimental Design for Evaluation of Rubber Modified Asphalt
Mixes
Appendix D Work Plan for RAC- G Gravel Factor for Use in Structural Section Design
Appendix E Proposed Work Plan for Validation, Calibration and Improvement of the AC
Overlay Design Procedure in the Caltrans Flexible Pavement Rehabilitation
Manual
LIST OF FIGURES
Figure 2.1 General Process for Evaluating a Product .................................................................................. 3
Figure 2.2 Time and Resource Requirements Associated with Various Types of Studies.......................... 4
Figure 2.3 Proposed Laboratory Study Process for Flexible Pavement/ Materials Evaluation .................... 7
Figure 2.4 Proposed Process for APT Study................................................................................................ 9
Figure 2.5 Proposed Process for Field Pilot Study .................................................................................... 11
LIST OF TABLES
Table 2.1 Proposed Project Initiation Form................................................................................................. 5
Table 2.2 Possible Tests for Inclusion in Laboratory Studies - Flexible Pavements ................................... 8
Table 2.3 Example Matrix for Laboratory Study of Mixes ......................................................................... 8
Table 2.4 Example Matrix for Study of Various Pavement Types and Overlay Thicknesses ................... 10
Table 2.5 Example of Important Variables for Flexible Pavement Studies............................................... 13
Table 2.6 Generic Experimental Design for New Construction ................................................................ 14
Table 2.7 Generic Experimental Design for Rehabilitation and Maintenance Projects............................. 14
Table 2.8 Suggestions for Project Selection and Data Collection – Field Pilot Studies ............................ 15
Table 2.9 Proposed Data Collection Checklist for Various Types of Study.............................................. 16
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
1
GENERIC EXPERIMENTAL DESIGN FOR PRODUCT/ STRATEGY
EVALUATION – CRUMB RUBBER MODIFIED MATERIALS
1.0 INTRODUCTION
1.1 BACKGROUND
The California Department of Transportation ( Caltrans) annually expends tremendous resources to
evaluate new products and/ or develop new strategies that would improve the performance of flexible
pavements. Although there is overlapping interest and need in this evaluation process among Caltrans
operating units ( Materials Engineering Testing Services, Research and Maintenance), it is often
undertaken independently of one another. Not surprisingly, this results in different approaches to data
collection sometimes limiting the use of these data to the “ sponsoring unit.” Accordingly, this broad-based,
generic approach to the evaluation of materials/ strategies is an attempt to weave a more
coordinated approach within Caltrans, one that might help to ensure that a study initiated by “ materials”
considers the needs of and potential effects on its “ sister” operating units, i. e., design, construction,
maintenance, research, etc. Since there is no standardized, consistent approach to product/ strategy
evaluation within Caltrans this document is offered as a first step in that direction, recognizing that this
will be an iterative process.
1.2 SCOPE AND OBJECTIVE
Although the impetus for this report was the evaluation of crumb rubber modified ( CRM) materials, this
approach is not material- specific; i. e., it is applicable to the evaluation of new products and/ or strategies
regardless of pavement type and component materials. The ultimate object of developing a standardized
approach is to ensure that all materials are evaluated in a uniform manner.
1.3 ORGANIZATION OF REPORT
The report is organized as follows:
• Chapter 2 presents the generic experimental design process developed in collaboration with
Caltrans, the University of California Partnered Pavement Research Center ( UC PPRC) and
MACTEC. It is applicable to studies ranging from laboratory to full- scale field studies.
• Chapter 3 presents candidate studies of CRM materials that evolved from the literature review and
discussions with Caltrans, industry, and the UC PPRC. These studies are intended to broaden and
expand Caltrans usage of CRM.
• Chapter 4 presents summary and recommendations resulting from this report.
Appendices are included to provide support information as follows:
• Appendix A includes a proposed FWD deflection test scheme for all field studies.
• Appendix B contains a proposed condition survey method for all flexible pavement field studies.
• Appendix C is an example experimental design for evaluation of rubber modified asphalt mixes.
• Appendices D and E are work plans for studies that are recommended to develop gravel factor( s)
for rubberized asphalt concrete ( RAC- G) and modified binder ( MB) mixes within the Caltrans
new pavement and overlay design procedures.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
2
2.0 GENERIC EXPERIMENTAL DESIGN PROCESS
2.1 INTRODUCTION
The generic experimental design, shown in Figure 2.1, outlines a uniform, consistent approach to evaluate
a new product and/ or design strategy. Each step shown in Figure 2.1 is described in this chapter.
2.2 GENERAL PROCESS OVERVIEW
The first step in the process is to develop an idea or hypothesis for the study. The idea should be
developed by a project “ champion” ( or manager) who would be involved in and oversee the study from
start to finish. This champion ensures ownership and, ultimately, responsibility for implementation. The
study hypothesis also needs to be clearly defined. The hypothesis is a stated premise arising from the idea
that can either be confirmed or rejected as a result of observation and testing. Once the idea and
associated hypothesis are defined, an advisory committee should be established to provide technical
oversight for the duration of the study and to help with implementation. The committee should include
both Caltrans and industry personnel.
A review of existing information should be required as a starting point for all studies. Computer analyses
or simulation studies with existing data may also be carried out if appropriate. Some products and/ or
strategies can be evaluated and recommended for implementation based on an evaluation of existing
information or the experience of other agencies ( assuming there are no economic or environmental
concerns). As an example, consider Caltrans use of high binder open- graded asphalt rubber mixes that
are used routinely in Arizona. These types of projects were constructed in several locations in California
without extensive preliminary studies.
If the hypothesis is not confirmed as a result of the initial evaluation of available information, the project
idea is usually considered invalid and discarded. However, there may be some cases where additional
testing and analyses are required to confirm or reject the hypothesis. If the idea warrants further study, it
may be necessary to consider various types of studies. Current technology within Caltrans allows the
evaluation of new products/ strategies by any or all of the following:
• Laboratory studies
• Accelerated pavement testing studies using the heavy vehicle simulator ( HVS)
• Full scale field test sections ( pilot studies)
Experience, research and technology development suggest that the relationship between time/ resources to
implementation is that shown in Figure 2.2. Among the three alternatives, laboratory studies require the
least amount of time and resources to complete an evaluation. Accelerated pavement testing requires
special facilities and trained personnel, thus adding more time and resources. Field pilot studies require
test sections to be constructed and monitored periodically over time. These require considerably more
time and money. Supplementing laboratory studies with full- scale field studies allows one to characterize
the behavior of the materials/ structure as a function of actual loading and environmental conditions.
Logically, these studies yield data that allow one to draw definitive conclusions. A detailed discussion on
three types of studies ( or investigations) is presented in Section 2.3.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
3
Figure 2.1 General Process for Evaluating a Product
Project Idea/
Hypothesis
• Review Existing
Information & Analysis
• Computer Simulation
Confirm
Hypothesis?
Require Further
Investigation?
Field Pilot
Study
Confirm
Hypothesis?
Implementation
• Reports
• Guidelines or Specs or Test Procedures
• Training
Laboratory
Study
APT/ HVS
Study
Yes
No
Yes
Yes
No
No
Are there concerns with?
• Economic/ Cost
• Environmental Impact
No
Reject
Hypothesis?
Discontinue
Project
No
Yes
Yes
• Champion
• Technical Advisory
Committee
Additional
Analysis
Are there concerns with?
• Economic/ Cost
• Environmental Impact
Yes
No
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
4
Figure 2.2 Time and Resource Requirements Associated with Various Types of Studies
The objective of conducting a study is to confirm or reject the study hypothesis. If the study results do
not confirm the hypothesis, consideration should be given to rejecting the hypothesis, conducting
additional analysis or using another approach for evaluation purposes. The project idea/ hypothesis should
be discarded if it cannot be confirmed.
The economic viability ( benefit/ cost ratio) and environmental impact must be assessed prior to
implementation, and should be conducted as early as possible in the study.
An implementation plan should be developed after confirming the hypothesis and assessing the cost-effectiveness
and environmental impact. As appropriate, the implementation plan may include the
following: a report, guidelines, specifications, test procedures, and/ or training materials.
Table 2.1 is a proposed project initiation form that should be completed for all studies. Items to be
included are as follows:
• Proposed title
• Project champion ( Member of operational unit within Caltrans who will provide oversight and
lead the implementation effort)
• Background/ problem statement ( The background provides information on the extent and
importance of the problem as well as efforts by others to solve the problem. The problem
statement should provide a brief description of the problem and a clear scope of work.)
• Project objective( s)/ hypotheses to be tested ( This should be a concise statement of the critical
issues, and if possible, the appropriate types of studies required to satisfy the project objectives.)
Analysis
With Existing
Data
Idea
Lab Testing
With
Analysis
Accelerated
Pavement
Testing
Long- Term
Performance
Monitoring
1- 6 Months,
$ 10- 50 K
3- 24 Months,
$ 50- 750 K
Typical Time and Resource Requirements
6- 36 Months,
$ 0.5- 5.0 M
5- 30 Years,
$ 1- 20 M
Relation of
results to real
pavements
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
5
• Expected benefits ( This should provide a clear indication of the expected monetary and
operational benefits expected from the study.)
• Implementation plan ( This should identify the expected deliverables, who will lead the
implementation process and anticipated timetable and cost.)
• Potential partners ( This should identify those who have a vested interest in the results ( within
Caltrans as well as industry) and could make a contribution, technical, financial or “ in- kind”.)
Table 2.1 Proposed Project Initiation Form
Item Description
• Title/ Idea
• Project Champion
• Background/ Problem Statement
• Objective/ Hypothesis to be Studied
• Expected Benefits
• Implementation Plan
• Potential Partners
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
6
2.3 TYPES OF INVESTIGATIONS
Within Caltrans, specifically when evaluating products/ strategies related to flexible pavements, there are
three types of studies or investigation alternatives commonly used: laboratory studies, accelerated
pavement testing using the HVS ( heavy vehicle simulator), and field pilot studies. These alternatives may
be carried out sequentially or concurrently depending upon the study hypothesis.
2.3.1 Laboratory Studies
The laboratory study may be conducted alone to evaluate a study hypothesis or in association with an
accelerated pavement testing and/ or field pilot study. Figure 2.3 proposes a general process for
conducting the laboratory study.
The laboratory study should begin with the development of a study plan, which includes identification of
important variables to be investigated, development of a laboratory test plan ( e. g., experimental design),
and establishment of a database framework. The identified variables dictate the laboratory test plan and
the structure of the database framework. The database should include information such as types of test
performed, dates of sample preparation, testing conditions, test results, and comments related to
conducting the test and test results. Laboratory tests typically include standard indicator tests and/ or
performance tests. The final list of tests should be determined based on the study hypothesis and
recommended laboratory test plan.
Table 2.2 summarizes the types of tests that should be considered in the development of a laboratory test
plan. Where feasible, performance tests should always be included to quantify the relative performance
of the new product or strategy. The standard indicator tests used for a given study would vary from
project to project.
The laboratory evaluation should always include a “ control” product for comparison purposes. Table 2.3
shows an example experimental design matrix for the laboratory study of three mixes with two binders
and two types of aggregates. The identified variables include material characteristics related to aggregate,
binder and mix. A full factorial evaluation of these variables would require a minimum of 12 possible
combinations. Replicate specimens could easily double or triple this number. For each combination of
binder, aggregate and mix to be evaluated, a recommended set of tests should be considered. Specific
testing requirements would need to be determined based on the study hypothesis. When conducting the
laboratory tests, initial testing on key indicators may be carried out first for the purpose of fine- tuning the
testing parameters and testing matrix defined in the original experimental design and then followed by
full scale testing. If the aggregate or binder is same for all mixes, the number of specimens and tests
would be proportionally reduced. More variables mean more tests.
Laboratory testing, analysis and report preparation can be accomplished as illustrated in Figure 2.3. To
ensure that the laboratory study addresses the idea/ hypothesis, a work schedule ( including milestones for
testing and reporting) should be developed. It is critical that report and “ deliverable” expectations be
clearly defined.
Ideally, the study should be “ statistically valid and robust.” Time and budget constraints will clearly
affect this. Sensitivity studies of critical variables as well as design/ analysis simulations are also integral
parts of the laboratory study. These types of analyses, including selection of proper testing method( s) and
method( s) of analysis, need to be identified in the problem statement and scope of work. Finally, a report
documenting the findings of the study and including recommendations for implementation should be
developed.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
7
Figure 2.3 Proposed Laboratory Study Process for Flexible Pavement/ Materials Evaluation
Conduct Laboratory Testing
Conduct Data Analysis and Computer Studies
Develop Study Plan
• Identify important variables
• Develop laboratory test plan
• Establish data base framework
Laboratory Study
Prepare Reports
Aggregate Tests*
• Standard
Indicator Tests
• Performance
Tests
Binder Tests*
• Standard Tests
• Performance
Tests
Mixture Tests*
• Standard
Indicator Tests
• Performance
Tests
* Specific tests should be determined based on the study hypothesis
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
8
Table 2.2 Possible Tests for Inclusion in Laboratory Studies - Flexible Pavements
Indicator Tests Performance Tests
a) Aggregate
• Size and Shape • Wear Resistance – studded tires
• Soundness • Polish Resistance
• Specific Gravity • Durability
• Cleanness • Adhesion
• Absorption
b) Binder
• Specific Gravity • Rheological Properties ( e. g., DSR, BBR)
• Consistency ( e. g., Viscosity, Penetration) • Temperature Susceptibility
• Safety ( Flash Point)
c) Mixes
• Air Voids • Fatigue
• Stability • Rutting
• Binder Content • Thermal Cracking
• Gradation
• VMA, VFA, Dust to Binder Ratio
• Moisture Sensitivity
• Aging Resistance
Table 2.3 Example Matrix for Laboratory Study of Mixes
Binder Aggregate Mix 1 Mix 2 Mix 3
Control AB
New Product AB
2.3.2 Accelerated Pavement Testing Study
Accelerated pavement testing ( APT) using the heavy vehicle simulator ( HVS) is particularly useful to
evaluate factors/ variables related to pavement structural properties. These types of studies typically
involve construction of test sections, application of repeated loads to full scale pavement ( including
measuring in- place air void content with time), and monitoring the performance for rutting, cracking, and
other distress modes. These studies typically include a companion laboratory testing component. Figure
2.4 presents a general process for conducting such a study.
The APT study should start with a study plan, which includes identification of important variables,
development of a data collection plan, and establishment of a database framework. The identified
variables dictate the data collection plan and the structure of the database framework. The database
should include basic test section information such as pavement- layer geometry, materials, instrumentation
( if any), types of tests and monitoring performed, testing/ monitoring dates, testing conditions,
testing/ monitoring results, and comments related to the actual testing/ monitoring as well as
testing/ monitoring results.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
9
Figure 2.4 Proposed Process for APT Study
If laboratory testing is required, a specific test plan should also be included as a part of the study plan.
Laboratory tests typically include standard indicator or characterization tests and/ or performance tests.
These tests are selected based on the study hypothesis.
Typical variables that have been considered in APT studies include the following:
• Pavement structure
o Cross section and layer thickness
o Base type
• Surface Materials
o Conventional AC
o Modified AC
Table 2.4 shows an example experimental design matrix for a study of flexible pavement overlays.
Specifically, this experiment attempts to quantify the effects of the following variables: existing pavement
condition, mix type, and overlay thickness. It is assumed that the same base asphalt is used. There are 16
Accelerated Pavement
Testing/ HVS Studies
Develop Study Plan
• Identify important variables
• Develop data collection plan
• Establish data base framework
Construct Test Sections QC/ QA Tests
Performance
Monitoring
and Data
Laboratory
Testing
Construction
Report
Conduct Data Analysis and Computer Studies
Prepare Reports
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
10
possible combinations in this example. The overlays constructed with modified binder materials would
fall into appropriate cells in the matrix. If another variable is added ( e. g., roadbed soil), the number of
test sections needed would double.
Table 2.4 Example Matrix for Study of Various Pavement Types and Overlay Thicknesses
Mix Type
Existing Pavement
Condition
Overlay
Thickness Control MB RAC- G RUMAC
Good HFualllf
Poor HFualllf
Prior to construction, the test section plans, specifications, and a construction quality control plan must be
developed. As with other Caltrans construction projects, all equipment used to construct the test sections
must be calibrated and in good working order. Quality control and assurance ( QC/ QA) tests should be
performed during the construction of the test sections to ensure that sections are built as designed.
A construction report should be developed within 60 days after the construction. This report should
address all construction- related activities: project layout including cross sections, mix design,
instrumentation ( if there is any), QC/ QA results, etc. The construction report defines the “ as- built”
product serving as a baseline for evaluation purposes.
The laboratory testing should be conducted as soon as the test sections have been constructed. This may
help to identify if additional tests are necessary. Performance monitoring and data collection activities
should be carried out according to the schedule and requirements identified in the study plan.
The deliverable from the APT study should be a report documenting the entire effort related to the study
hypothesis: construction of the test sections, analysis of the laboratory testing, performance monitoring
results, and conclusions and recommendations.
2.3.3 Field Pilot Studies
Field pilot studies should be conducted only if the laboratory and accelerated pavements studies cannot
accomplish the intended result or if a field study is deemed to be the best option to evaluate a
product/ strategy. Figure 2.5 provides a proposed process for performing field pilot studies.
The field pilot study should begin with a study plan, which includes identification of important variables,
selection of candidate project ( including new construction) and test section locations, development of a
data collection plan, and establishment of a database framework. The identified variables dictate the
selection of candidate projects, data collection plan and the structure of the database framework. The
database should include basic test section information such as pavement structure and materials,
pavement instrumentation ( if any), test section layout, construction data, traffic data, types of tests and
monitoring to be performed, testing/ monitoring schedule, testing conditions, and testing/ monitoring
results.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
11
Figure 2.5 Proposed Process for Field Pilot Study
Develop Study Plan
• Identify important variables
• Project selection
• Develop data collection plan
• Establish data base framework
Field Pilot Studies
Pre- construction
Evaluation/ Testing/ Establishing
Performance Evaluation Sections ( PES)
Field Sampling Construct Test Sections QC/ QA Tests
Laboratory
Testing
Post-
Construction
Evaluation
Construction
Report
Periodic Performance Monitoring
• Pavement Condition Survey
• Deflection Testing
• Ride Quality
• Skid Testing
Conduct Data Analysis and Computer Studies
Prepare Reports
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
12
If laboratory testing is required, a specific test plan should also be included in the study plan. Laboratory
tests typically include standard indicator and characterization tests and/ or performance tests. These tests
should be determined based on the study hypothesis and recommended and discussed in the laboratory
test plan.
For flexible pavements, examples of important test variables are summarized in Table 2.5. Typical
experimental design matrices for field studies are included in Table 2.6 ( new construction) and Table 2.7
( rehabilitation/ maintenance) projects, respectively. These designs clearly show that the number of
possible sections in any field study increases greatly as the number of variables to be evaluated increases.
This emphasizes the importance of careful planning when doing field studies.
The field pilot study must also consider the geometry of the roadway, cross section, and soil support
condition associated with the project. Ideally, all test sections should be constructed on a uniform
supporting foundation to preclude bias in the results. A list of suggestions for project selection and data
collection is provided in Table 2.8. These should be similar regardless of the operational unit overseeing
the project ( e. g., METS, Research, Maintenance).
The field pilot study typically involves a pre- construction evaluation/ testing and establishment of
performance evaluation sections ( PES). Specific pre- construction evaluation testing requirements are also
presented in Table 2.8. The establishment of PESs should be based on existing roadway condition and
support characteristics. For a rehabilitation project, FWD deflection testing is normally conducted prior
to design to establish these PESs. Appendix A includes a recommended deflection testing scheme for all
field studies. A consistent methodology for pavement condition survey should also be adapted. This
assures all distress measurements are collected consistently and evaluated against the same criteria.
Appendix B presents a proposed method for all field studies and a comparison of several methods
available. Caltrans is encouraged to adopt one method for use by all operating units.
Prior to constructing test sections, plans, specifications, and a construction quality control plan must be
developed. As with other Caltrans construction projects, all equipment used to construct the test sections
must be calibrated and in good working order. Quality assurance and quality control ( QA/ QC) tests
should be performed during the construction. If the study plan or the laboratory test plan requires the use
of sample materials from the field, a specific sampling plan must be developed. The sampling plan
should include the type of materials, quantity, sample size, sample location, and other requirements
associated with the sampling activities.
A construction report should be developed within 60 days of construction. This report should address all
construction- related activities: project layout including cross sections, mix design, instrumentation ( if
there is any), QC/ QA results, etc. The construction report defines the “ as- built” product serving as a
baseline for evaluation purposes. It is important that the resident engineer for the project be aware of the
scope and importance of the study and cooperates with the study team by providing needed information in
a timely manner.
The laboratory testing should be conducted as soon as the test sections have been constructed. This may
help to identify if additional tests are necessary. Post- construction evaluation should be conducted
between one to six months after completion. Periodic performance monitoring should begin one year
later. Periodic performance monitoring and data collection activities should be carried out according to
the schedule and requirements identified in the study plan.
Data collection plans may vary with the field pilot study type, i. e., new construction, rehabilitation, and
maintenance. A suggested data collection checklist for various types of field pilot studies is provided in
Table 2.9. The deliverable from the field pilot study should be a report documenting the entire effort
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
13
related to the study hypothesis, the construction of the test sections, the laboratory testing, performance
monitoring results, conclusions and recommendations.
Table 2.5 Example of Important Variables for Flexible Pavement Studies
Variables Considerations New Pavement Rehabilitation/
Maintenance
Climate • Temperature
• Rainfall X X
Traffic
• Low
• Moderate
• High
X X
Roadbed Soil
• Good
• Fair
• Poor
X X
Base Type
• Aggregate
• ATPB
• CTB
X X
Existing Pavement
Condition
• Good
• Fair
• Poor
X
Overlay Thickness • Full
• Half X
Overlay Materials/
Surface Treatment
• Control Mix
• Other Mixes
• Other Materials
X
ATPB = Asphalt treated permeable base. CTB – Cement treated base
X indicates variables to be considered in field pilot studies
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
14
Table 2.6 Generic Experimental Design for New Construction
Pavement Types
Climate Traffic Roadbed Soil
AC/ AB AC/ ATPB AC/ CTB Full Depth AC
Good
Low
Poor
Good
Coastal
High
Poor
Good
Low
Poor
Good
Valley/ Desert
High
Poor
Good
Low
Poor
Good
High
Desert/ Mountain
High
Poor
Table 2.7 Generic Experimental Design for Rehabilitation and Maintenance Projects
Overlay Thickness/ Surface Treatment
Climate Traffic Existing Pavement
Condition Control Mix Mix 1 Mix 2 Mix 3
Good
Low
Poor
Good
Coastal
High
Poor
Good
Low
Poor
Good
Valley/ Desert
High
Poor
Good
Low
Poor
Good
High
Desert/ Mountain
High
Poor
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
15
Table 2.8 Suggestions for Project Selection and Data Collection – Field Pilot Studies
Data Collection Item Considerations ( MACTEC, 2002a, 2002b, 2003)
Project selection and layout considerations
Project site and test sites
• Project length sufficient to establish performance evaluation
section ( PES)
• Design AC layer thick enough to evaluate effect of full/ half
thicknesses
• Uniform cross section for PES
• Relatively uniform deflection profile for PES
• Relatively uniform pavement condition, including drainage
• Free from serious structural defects, such as pumping or base
failure
• Similar geometrics of roadway
Standard layout 1000 ft is preferred – 500 ft usually is not adequate to stabilize AC
plant mix production or field compaction operations
Sampling location and instrumentation package 30 m or 100 ft away from each end of the test section
Pre- construction evaluation/ testing
Existing pavement profile and material properties AC, base, subbase layer thickness, stiffness, R- value, gradation
Deflection testing data & core sampling information For determination of location of PESs
Pavement condition survey maps and photos Include distress type, severity, extent
Rut measurement of selected PES Measured as part of the pavement condition survey
Traffic information ADT, % trucks, ESALs, TI, growth rate
QC testing plan Contractor QC plan
Mix design data Contractor mix design and Caltrans verification
Plans and special provisions Project specific document
Climate and environmental information Project site information may be obtained from UCB
Maintenance history of the existing pavement Previous repair/ maintenance data
Construction monitoring/ testing
Sources of materials ( binders, aggregates, modifiers) Certificates of compliance from material suppliers
Plant type and condition ( T- 109 data) Calibration data of plant and equipment
Paving dates Start and end dates, including delays in paving operations, the
reasons for them
Paving equipment used Model, make, year
Haul distances and time From plant to paving site
Site weather conditions During paving, include ambient air and pavement temperatures
Mix temperatures at various locations Upon discharge, in windrow, immediately behind screed, during
breakdown, finish rolling
Compaction equipment and methods Type, make, model, weight, vibration or static
In- place air voids Based on core density and maximum theoretical density ( CT309)
RE diaries/ Inspectors notes Copies of resident engineers and inspectors notes
Results of QC/ QA tests AC Pay data – more important for characterizing products actually
supplied for study purposes
Sample requirements for laboratory tests ( binders,
aggregates and mixes)
Number of cores, beams, amount of aggregate, binders, respective
loose mixes
Date opened to traffic
Laboratory testing
Refer Table 2.2 Table 2.2
Post- construction evaluation
Pavement condition data including ride On PESs and the entire project
Deflection data On PESs
Additional sampling needs If needed, 30 m or 100 ft away from each end of the test section
Periodic monitoring
Detailed distress mapping ( PES only) On PESs
Overall condition- photo logs and distress survey For entire project
Deflection testing On PESs
Reporting
Construction report Activities associated with construction, layout, mix design
Initial performance report Data collected during the field performance evaluation
Final report Summary of the construction, performance, and findings
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
16
Table 2.9 Proposed Data Collection Checklist for Various Types of Study
Data Collection Item New
Construction Rehabilitation Maintenance
Project selection and layout considerations
Project location and length X X X
Design AC layer thicknesses X X X
Cross section for PES X X X
Deflection profile for PES X X
Pavement condition, including drainage X X
Description of any structural defects, such as pumping or base failure X X
Geometrics of roadway X X X
Pre- construction evaluation/ testing
Existing pavement profile and material properties X X
Deflection testing data and core sampling information X X
Pavement condition survey maps and photos X X
Rut measurement of selected performance evaluation section X X
Traffic information X X X
QC testing plan X X X
Mix design data X X X
Plans and special provisions X X X
Climate and environmental information X X X
Maintenance history of the existing pavement X X
Construction monitoring/ testing
Sources of materials ( binders, aggregates, modifiers) X X X
Plant type and condition ( T- 109 data) X X X
Paving dates including delays in paving operations, the reasons for them X X X
Paving equipment used X X X
Haul distances and times X X X
Site weather conditions X X X
Mix temperatures at various locations ( truck, paver hopper, behind
paver, etc) X X X
Compaction equipment and methods X X X
In- place air voids X X X
RE diaries/ Inspectors notes X X X
Results of QC/ QA tests X X X
Sample requirements for laboratory tests ( binders, aggregates and
mixes) X X X
Date opened to traffic X X X
Post- construction evaluation
Pavement condition data including ride and skid X X X
Deflection data X X X
Additional sampling needs X X X
Periodic monitoring
Detailed distress mapping ( PES only) X X X
Overall condition- photo logs and distress maps X X X
Deflection testing X X X
Reporting
Construction report X X X
Initial performance report X X X
Final report X X X
X indicates data element should be collected.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
17
2.4 ECONOMIC ANALYSIS AND ENVIRONMENTAL IMPACT
As early in the project as possible, a determination should be made as to the economic viability and
environmental impact of the product and/ or new strategy. Environmental impact encompasses not only
air and water quality, but also worker health and safety. Both initial costs and life cycle costs of
implementing the product and/ or strategy should also be addressed.
2.5 IMPLEMENTATION
The implementation plan that should address the following:
• Expected products and dissemination format.
• Responsible unit for dissemination.
• Timetable and cost for dissemination or “ technology transfer.”
2.6 SUMMARY
This chapter outlined a generic process to evaluate new products and/ or strategies regardless of the
Caltrans operating unit ( METS, Maintenance and Research). Additional work is required to evaluate the
process and to refine the details for data collection and materials testing. An example of this process for
the evaluation of CRM materials is presented in Appendix C.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
18
3.0 PROPOSED CRUMB RUBBER MODIFIED MATERIAL STUDIES
This chapter outlines potential CRM studies generated from discussions with Caltrans, UC PPRC,
industry, the RAC technology transfer centers, and the Caltrans- industry RAC Task Group.
3.1 CONSTRUCTION/ REHABILITATION AND MAINTENANCE APPLICATIONS
3.1.1 New Construction
Caltrans use of CRM mixes has traditionally been limited to pavement rehabilitation and maintenance
applications. However, this does not preclude the possibility of use in new construction as a
comprehensive literature review revealed that some agencies have used CRM successfully for shoulder
widening of an existing roadway or as the wearing course of a new pavement structure. For Caltrans to
use CRM in new construction within its current design framework, a gravel factor ( Gf) for CRM would be
required.
.
Hypothesis: Asphalt rubber products can be used in new construction.
3.1.2 Thick Overlays
In California, RAC- G overlays are typically designed and constructed to 60 mm or less because of the
cost differential with DGAC. Although there are concerns with the potential for shear flow and rutting of
thicker sections, the 60- mm maximum is being re- evaluated in the Firebaugh project and the MB study at
UC Berkeley where 90- mm overlays have been constructed. The Firebaugh and MB studies may provide
some much- needed insight as to the technical and economic benefits of thicker sections, i. e., overlays
thicker than 60 mm. That said, agency experience and research indicate that CRM asphalt tends to be
most effective and economical when placed as a relatively thin wearing course.
Cost is a critical factor in selecting wet process binders in thick layers. RAC- G made with high viscosity
binders is deemed cost effective in its current application because of the reduced thickness associated
with the prevention of reflection cracking. Reduced thickness, however, may be inappropriate in other
applications such as new construction.
Hypothesis: Asphalt rubber overlays can be placed at thickness greater than 60 mm without adverse
effects on construction, cost and performance.
3.1.3 Open Graded- High Binder Mixes
A well- documented benefit of an open- graded asphalt concrete ( OGAC) surface layer is noise reduction
( ATRC, 1996; Sacramento County, 1999; Roschen, 2000; Donavan and Rymer, 2003; Carlson, 2003;
MACTEC, 2004a). The tire- pavement noise reduction is attributed to the open texture and increased
binder film thickness of the OGAC mix, not necessarily to the presence of CRM. Still, current studies
conducted by Arizona Department of Transportation ( ADOT) suggest that mixes with high viscosity
binders and corresponding high binder contents (≥ 9% by total weight of mix) may yield greater noise
reduction. Arizona has found that open- graded mixes ( high viscosity binder) placed as thin ( 25 mm thick)
overlays on portland cement concrete not only retard reflection cracking but also reduce noise ( Way, 2000;
Scofield and Donovan, 2003).
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
19
The Arizona Department of Transportation is a pioneer in the use of high viscosity CRM binders in
paving projects for noise control ( Kuennen, 2004), although that was not the original purpose for
development and use of these materials. Generally, Arizona DOT uses an asphalt rubber asphalt concrete
friction course ( AR- ACFC) with aggregate passing the 9.5 mm sieve combined with 9 to 9.5% asphalt
rubber binder for noise attenuation. A 12.5 mm lift is used on flexible pavements whereas a 25 mm lift is
used on rigid pavements with high traffic volumes. Minimum noise attenuation of 4 dBA has been
consistently attained using these thin open- graded surfaces, and in many cases greater attenuation has
been achieved ( Kuennen, 2004; Scofield and Donovan, 2003).
Similarly, a local agency study ( Sacramento County, 1999) of RAC- G on Alta Arden Expressway
recorded an average 4 dBA noise reduction compared to that measured on conventional asphalt concrete
on Bond Road. This noise reduction continued for six years after the paving with rubberized asphalt
( Sacramento County, 1999).
The use of high binder open- graded mixes has definite potential for use as a wearing course on portland
cement concrete pavements or on flexible pavements. There are numerous opportunities in California to
place high- binder open- graded mixes as a wearing course to achieve the twofold benefits: reducing noise
and retarding reflective cracking. Extending CRM use for overlays of PCC pavements would likely
improve ride quality and reduce noise.
Hypothesis: Arizona DOT’s strategy of open- graded high binder mixes can be effective for Caltrans in
reducing noise and retarding reflective cracking.
3.1.4 Recycling
Recycling of conventional mixes with CRM materials represents another avenue to expand the use of
CRM. Currently, however, the primary concern is with recycling of RAC, an issue addressed in a
companion report titled “ Feasibility of Recycling Rubber Modified Paving Materials” ( MACTEC, 2004b).
Hypothesis: A RAC mix can be recycled.
3.2 MATERIALS STUDIES
3.2.1 Type 1 vs. Type 2 Binders
Two types of high viscosity CRM binders have been used in California: Type 1 includes only scrap tire
CRM in the asphalt cement; Type 2 includes a blend of CRM consisting of 75% scrap tire and 25% high
natural CRM ( typically truck tires), and an extender oil ( Hicks, 2002).
Caltrans currently requires the use of Type 2 high viscosity binders and most cities and counties in
California also use Type 2 binders. However, only Caltrans requires Type 2 binder. Some California
cities and counties and the states of Arizona, Florida and Texas use Type 1 binders ( Hicks, 2002). A
study to evaluate the relative effectiveness of the Type 1 and Type 2 binders is recommended to clarify
the most appropriate applications for the different binders.
The use of the extender oils in asphalt concrete mixes may help reduce the rate of age- hardening and the
development of such surface distresses as thermal cracking and raveling. However, air quality complaints
are reportedly related the presence of extender oils. Also, extender oil may lead to early bleeding when
used in chip seals over a newly constructed asphalt pavement. Thus, the use of extender oils is an area
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
20
which warrants further study. Arizona DOT does not allow extender oil in high viscosity binders; Texas
and Florida DOTs allow it, but do not normally use it.
Hypothesis: Type 1 and Type 2 binders affect performance of mixes. The use of extender oil in asphalt
rubber binders is unnecessary.
3.2.2 Binder Testing
Many have recognized that asphalt binder properties are critical to the control of cracking in asphalt
pavements. This is particularly true for thermal cracking and, to a lesser extent, fatigue cracking.
Modified binders are generally more crack resistant than “ neat” binders, though performance varies
considerably with modifier type and content. For high viscosity binders in California, the cracking
performance is not well understood or documented. This is largely the result of the use of the aged-residue
( AR) grading system that classifies asphalt cements after they have been artificially aged using
the Rolling Thin Film Oven procedure and that provides little information on the properties of the unaged
( or original) asphalt cement. Research has shown cold temperatures properties of the base asphalt cement
govern the properties of CRM binders. Extender oil, used by refiners to soften asphalt cement and
required in Type 2 binders, may enhance low temperature performance. Crumb rubber modification
would provide significant increases in stiffness and elasticity at high temperatures and enhance resistance
to permanent deformation or rutting and low- temperature cracking ( Navarro, 2000).
To develop a better understanding of high viscosity and no agitation CRM binders, a laboratory test
program should be undertaken to quantify the effect of binder type on pavement performance. This can
best be achieved for some materials using the Superpave performance graded binder and mix tests.
However, there are concerns as to the use of these tests with high viscosity binders. The discrete swollen
rubber particles may produce highly variable dynamic shear rheometer ( DSR) test results. The particulate
matter may provide premature failure planes in beams for bending beam rheometer ( BBR) and in direct
tension specimens which are used to determine critical cracking temperature ( low temperature
performance tests). The outcome of this laboratory test program should provide some data for assessing
the relative benefits of the different CRM ( no agitation and high viscosity) binders, but due to the issues
of testing two- phase materials, results may not be definitive.
Hypothesis: The properties of neat asphalt cements and CRM binders affect mix and/ or field pavement
performance.
3.3 STRUCTURAL DESIGN STUDIES
3.3.1 Gravel Factor( s) for RAC- G, MB- D, and MB- G
Caltrans current pavement structural design procedure ( Caltrans, 2004) requires the use of R- value and
Traffic Index ( TI) to develop flexible pavement layer thicknesses for new and reconstruction projects.
The procedure is based upon a layer equivalency approach in which the relative load- carrying capacity of
individual pavement layers is related through a gravel equivalence value. The gravel factor ( Gf) refers to
the relative strength of a given material compared to a standard gravel subbase material. Gravel factors
for dense- graded asphalt concrete and various types of base and subbase materials have been developed
over the years. However, no Gf has been established for rubberized asphalt concrete ( RAC) materials for
use in new pavement as well as in rehabilitation designs.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
21
Recently, Caltrans initiated an effort to develop gravel factors for RAC- G, MB- D and G mixes. Based on
a meeting held with Caltrans and UC PPRC staff, a general framework for the development of the gravel
factor( s) for RAC- G was developed. This framework calls for a review of the related work, especially of
the laboratory test data on dense- graded AC and RAC- G materials. Interim gravel factor( s) for RAC- G
may be developed based on the review of these laboratory test data and refined through mechanistic-empirical
and/ or finite element analysis methods. The interim gravel factor( s) may then be validated
using data gathered from studies already underway: the UC PPRC HVS MB study and the Firebaugh
project. Additional test sections with different thickness of RAC- G may need to be constructed and tested
using the HVS to validate the interim Gf value( s). A draft work plan for this work is presented in
Appendix D.
Hypothesis: The gravel equivalence approach for determining the DGAC thickness in new pavement
design is valid for the thickness design of RAC- G and modified binder ( MB) mixes.
3.3.2 Modification of Overlay Design Procedure
Caltrans current deflection- based flexible pavement rehabilitation design procedure ( Caltrans, 2001)
employs the percent reduction in deflection approach to determine the overlay thickness required for
future traffic. The procedure was originally developed for the design of conventional DGAC overlays
and was adapted in recent years to accommodate RAC- G overlays in lieu of ( or in combination with) a
standard DGAC overlay. Since its development, there has been no field validation of this design
approach.
Accordingly, it is recommended that a study be undertaken to validate and calibrate the design procedure
with specific emphasis on RAC- G. This may be accomplished through a combination of field and
laboratory studies. Some of the needed data will be obtained from the RAC Warranty and Firebaugh
projects. However, it is anticipated that more projects will be needed to provide a statistically valid data
set. The experiments will require different types of RAC- G sections, including RAC- G on conventional
DGAC, RAC- G on RAC- G, and conventional DGAC on RAC- G, as well as DGAC control sections.
Any new sections identified will require deflection testing ( before and after construction), core sampling
and testing ( to determine layer resilient moduli), traffic counts, and performance monitoring. A draft
work plan for this work is provided in Appendix E.
Hypothesis: The deflection based approach used for determining the DGAC overlay thickness in
pavement rehabilitation design is valid for the overlay thickness design of RAC- G and modified binder
( MB) mixes.
3.4 SUMMARY
This chapter presented several studies that could generate data to confirm/ refute the effectiveness of
Caltrans currents strategies for use of CRM. Additionally, studies were suggested that might broaden or
extend Caltrans current use of CRM. Also, a suggested hypothesis for each study was presented.
Caltrans is encouraged to evaluate the use of RAC and MB mixes in new construction, as thick overlays,
as a wearing course for reducing noise and retarding reflective cracking, and in recycling applications.
These ideas can be implemented immediately if the individual hypothesis can be confirmed and the
approach is cost effective and has no adverse environmental impact. Similarly, a study on binder types
( Type 1 versus Type 2) may be implemented if appropriate projects can be solicited after the confirmation
of related hypotheses. Binder testing may be initiated at a later time.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
22
The development of gravel factors for RAC- G and MB mixes requires further investigation. The work
plan presented in Appendix D is the result of several meetings with Caltrans, UC PPRC and MACTEC
staff. It is envisioned that the work will be performed in two phases. Phase 1 is to develop interim gravel
factor and phase 2 is to validate the interim gravel factor. Phase 1 is expected to be completed by
September 2005; phase 2 will require three to five years to complete, after the performance monitoring
data from the various studies are collected. A draft work plan for modification of the rehabilitation design
procedure, which is also the result of several meetings with Caltrans, UC PPRC and MACTEC staff, is
presented in Appendix E. The work will be performed in two phases with an expected completion date
for Phase 1 by September 2005.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
23
4.0 SUMMARY AND RECOMMENDATIONS
4.1 SUMMARY
This report presents the framework for a generic process to evaluate new products and/ or strategies for
possible use within Caltrans. The framework is the result of a collaborative effort among Caltrans, the
University of California Partnered Pavement Research Center ( UC PPRC) and MACTEC.
The generic experimental design process is applicable to all pavement types as well as the component
materials of pavements. The framework includes various types of studies that may be used in the
evaluation process ― laboratory, accelerated pavement testing ( APT) and field pilot studies.
Additionally, it identifies other factors that must be considered in the evaluation of any new
product/ strategy: economic viability and environmental impact.
The process was used to develop numerous project ideas/ hypothesis that may help to broaden and expand
Caltrans use of CRM in paving applications. They are listed below:
• For Construction/ Rehabilitation and Maintenance Applications
o New construction
o Thick overlays
o Open graded- high binder mixes
o Recycling
• For Materials Studies
o Type 1 vs. Type 2 binders
o Binder testing
• For Structural Design Studies
o Gravel factor for RAC- G and MB mixes
o Modification of the deflection based overlay design procedure
In addition, two structural design- related studies were expanded into detailed work plans which address
the following:
• development of a gravel factor for RAC- G and MB mixes for use in new construction; and
• update/ modification of Caltrans deflection based overlay design procedure to accommodate
RAC- G and MB mixes.
4.2 RECOMMENDATIONS
It is recommended that the generic experimental design process outlined herein be considered as a
foundation for Caltrans evaluation of any new product and/ or strategy, as is the case for CRM in paving
applications.
Also, it is recommended that the generic process presented in this report be refined to include detailed
information on data collection and testing requirements associated with each study type.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
24
5.0 REFERENCES
AASHTO, 1993. “ Guide for Design of Pavement Structures,” AASHTO, Washington, D. C., 1993.
ATRC, 1996. “ Asphalt Rubber Friction Course Reduces Traffic Noise,” Arizona Department of
Transportation Research Notes, August 1996.
Caltrans, 1979. “ Asphalt Concrete Overlay Design Manual,” Caltrans, Sacramento, CA, January 1979.
Caltrans, 2000. “ Pavement Evaluation Manual,” Caltrans, Sacramento, CA, January 2000.
Caltrans, 2001. “ Flexible Pavement Rehabilitation Manual,” Revised June 6, 2001. Caltrans, Sacramento,
CA, June 2001.
Caltrans, 2003. “ Guide to the Investigation and Remediation of Distress in Flexible Pavements,” Draft
Field Manual, Caltrans, Sacramento, CA, July 2003.
Caltrans, 2004. “ Highway Design Manual,” Caltrans, Sacramento, CA, 2004.
Carlson, D. D., H. Zhu, and C. Xiao, 2003. “ Analysis of Traffic Noise Before and After Paving with
Asphalt Rubber,” Asphalt Rubber 2003 Conference, pp 413- 428, December 2003
Donavan, P. R. and B. C. Rymer, 2003. “ Measurement of Tire/ Pavement Noise Sound Intensity
Methodology,” Asphalt Rubber 2003 Conference, pp 399- 412, December 2003.
FHWA, 2003. “ Distress Identification Manual for the Long- Term Pavement Performance Program
( Fourth Revised Edition), Report No. FHWA- RD- 03- 031, McLean, VA, June, 2003.
Hicks, R. G., 2002. “ Asphalt Rubber Design and Construction Guidelines: Volume 1 – Design
Guidelines,” RACT2 Center, Sacramento, California, January 2002.
Highway Research Board, 1962. “ The AASHO Road Test, Report 5, Pavement Research,” Special Report
61E, , Washington, D. C., 1962.
Kuennen, T. 2004. “ Asphalt Rubber Makes a Quite Comeback’” Better Road, May 2004.
MACTEC, 2002a. “ Rubberized Asphalt Concrete ( RAC) Warranty Pilot Projects: Information Packet for
Resident Engineers Data Collection Guidelines,” Prepared for Materials Engineering and Testing
Services, Caltrans, Sacramento, CA, November 2002.
MACTEC, 2002b. “ Rubberized Asphalt Concrete ( RAC) Warranty Pilot Projects: Data Collection
Guidelines,” Prepared for Materials Engineering and Testing Services, Caltrans, Sacramento, CA,
November 2002.
MACTEC, 2003. “ Proposed Field Evaluation Plan for Asphalt Concrete Pavements with Asphalt Rubber
and Rubber Modified Binders,” Draft Proposed Field Evaluation Plan to CIWMB and Caltrans
Partnership, September 2003.
MACTEC, 2004a. “ Use of Scrap Tire Rubber – State of the Technology and Best Practices,” Prepared for
METS, Caltrans, Sacramento, September 2004.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
25
MACTEC, 2004b. “ Feasibility of Recycling Rubber- Modified Paving Materials”, Prepared for METS,
Caltrans, Sacramento, September 2004.
Navarro, J., P. Partal, F. Martinez- Boza, and C. Gallegos, 2000. “ Linear Viscoelastic Properties of
Ground Tire Rubber- Modified Bitumen,” Asphalt Rubber 2000 Conference, pp 411- 420, November
2000.
Roschen, T., 2000. “ Report on the Status of Rubberized Asphalt: Traffic Noise Reduction in Sacramento
County,” Asphalt Rubber 2003 Conference, pp 517- 539, November 2000.
Sacramento County, 1999. “ Report on the Status of Rubberized Asphalt Traffic Noise Reduction in
Sacramento County,” Sacramento County Department of Environmental Review and Assessment,
November 1999.
Scofield, L. and P. R. Donovan, 2003. “ Development of Arizona’s Quiet Pavement Research Program,”
Asphalt Rubber 2003 Conference, pp 429- 452, December 2003.
University of California, Berkeley, 2003. “ Comparison of MB, RAC- G and DGAC Mixes under HVS
and Laboratory Testing,” Draft Prepared for California Department of Transportation, University
of Berkeley, July 2003.
Way, G. B., 2000. “ OGFC Meets CRM Where the Rubber Meets the Rubber 12 Years of Durable
Sucess,” Asphalt Rubber 2000 Conference, pp 15- 31, November 2000.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
Appendix A
PROPOSED DEFLECTION TEST PLAN FOR PERFORMANCE
EVALUATION OF MODIFIED BINDER STUDY FOR ALL FIELD
PROJECTS
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
A- 1
PROPOSED DEFLECTION TEST PLAN FOR PERFORMANCE EVALUATION OF
MODIFIED BINDER STUDY FOR ALL FIELD PROJECTS
OVERVIEW
Pavement surface deflections are a structural response of the pavement system to an applied load and
provide the basis for:
• evaluating pavement structural capacity,
• assessing the variability of existing support,
• characterizing the in- situ material properties of the layers, and
• developing rational rehabilitation designs.
For the purpose of performance monitoring of a test section, which has a length of 152.4 m ( 500 ft), the
deflections should be measured at specific, pre- determined locations under certain load levels with
specific sensor configurations, for a desired monitoring period. These conditions must be followed for
the data to be meaningful and useful. For this study, activities associated with the FWD testing are
described below. A comparison of two commonly used deflection test schemes is presented in Table A. 1.
FWD TEST REQUIREMENTS
Based on an evaluation of the schemes, a proposed FWD test plan was developed following extensive
discussions with Caltrans and UC PPRC. The recommended FWD test requirements are described below:
Item Description Comments
Sensor
Configuration
1. - 305, 0, 203, 305, 457, 610, 914,
1219, 1524
2. 0, 203, 305, 610, 914, 1219, 1524
Distance in mm from the center of the
load plate. Plan 1 uses 9 sensors and is
preferred. If seven sensors are used, then
use plan 2.
Load Package
• 26.7 ( Seating load, once)
• 26.7 ( Once, range 24.0- 29.4)
• 40.0 ( Once, range 36.0- 44.0)
• 53.4 ( Once, range 48.1- 58.7)
Load in kN. A seating drop at 26.7 kN
should be applied but not recorded. At
each load level, the load should be applied
once and deflections be recorded
electronically.
Air & Pavement
Temperatures At each test location Temperatures may be measured using
device mounted on the FWD or manually.
Test Location
• Mid- lane, 11 deflections
• Outer wheel path, 11 deflections
• ~ 15.2 m intervals
See Figure A. 1
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
A- 2
Table A. 1 Comparison of Deflection Test Schemes
Item CT 356
( June 2004 Version)
LTPP
( GPS- 1, 2, 6, 7, 500 ft Section)
Proposed Plan
( Revised October 2004)
Sensor
Configuration No description - 305, 0, 203, 305, 457, 610, 914,
1219, 1524
1. - 305, 0, 203, 305, 457, 610,
914, 1219, 1524
2. 0, 203, 305, 610, 914, 1219,
1524
Load Package
• One seating load ( 26.7 kN)
• Three drops with an applied load of 40 kN
± 10%
• Use average of three readings and normalized
to 40 kN
• 53.4 ( Seating load, 3 times)
• 26.7 ( 4 times, 24.0- 29.4)
• 40.0 ( 4 times, 36.0- 44.0)
• 53.4 ( 4 times, 48.1- 58.7)
• 71.2 ( 4 times, 64.1- 78.3)
• 26.7 ( Seating load, once)
• 26.7 ( Once, 24.0- 29.4)
• 40.0 ( Once, 36.0- 44.0)
• 53.4 ( Once, 48.1- 58.7)
Air & Pavement
Temperatures Record the ambient air and pavement surface temperatures
Test Location
and Frequency
Method A:
o Length ≥ 1.6 km: 21 deflections/ 1.6 km; 80- m
intervals outside wheel path
o Length < 1.6 km: determine the size of testing
interval to obtain 21 deflections
Method B:
o Select one 300- m long test section
“ representative” of every 1.6 lane- km. ~ 15 m
intervals to obtain 21 deflections
o Length < 300 m: determine the size of testing
interval to obtain 21 deflections
• Mid- lane ( 21 measurements)
• Outer wheel path ( 21
measurements)
• ~ 7.6 m intervals
• Mid- lane ( 11 measurements)
• Outer wheel path ( 11
measurements)
• ~ 15.2 m intervals
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
A- 3
ONSITE CORING
This activity is performed once during post- construction to verify the thickness of pavement structure
after overlay. Two full depth cores should be collected at each performance evaluation section ( PES).
Each core should be at least 100 mm ( 4- inches) in diameter and be located 30.5 m ( 100 ft) before and
after the PES ( see Figure A. 1). Cores should be inspected to determine its condition ( e. g., stripping) and
pavement layer thickness. Cores should be packaged and retained for future evaluation and/ or testing.
Figure A. 1 FWD Test Pattern
Outer Wheel Path
Performance Evaluation Section ( PES)
100- feet
FWD TESTING LOCATION
Test at 50- foot interval in the
outer wheel path and the mid
lane. See testing requirements
for load drop levels.
100- feet
CORE- A CORE- B
Shoulder
Edge Pavement
0+ 00 1+ 00 2+ 00 3+ 00 4+ 00 5+ 00
Mid Lane
CORING LOCATIONS
At least 4- inch diameter
100- feet from end of each
performance evaluation section
Direction of Traffic
Centerline
Note: Distance is in feet. Not to Scale
50- ft
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
Appendix B
PROPOSED CONDITION SURVEY METHOD FOR ALL FLEXIBLE
PAVEMENT FIELD STUDIES
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
B- 1
Table B. 1 presents a comparison of several distress survey methods for flexible pavements. The method
developed by Caltrans in 2003 ( Caltrans, 2003) contains the majority of the distress types found on
flexible pavements in California, is compatible with the procedures developed by Caltrans in 2000
( Caltrans, 2000), by LTPP ( FHWA, 2003), and by AASHTO ( AASHTO, 1993) and therefore is
recommended for use in all field studies for condition survey on flexible pavements.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
B- 2
Table B. 1 Comparison of Distress Survey Methods for Flexible Pavements
Caltrans ( 2000) Caltrans ( 2003) LTPP AASHTO
Distress Type Severity Measurement Severity Measurement Severity Measurement Severity Measurement
Cracking
Longitudinal Cracking ( Non PCC Slab Joint Reflective) 1/ 4" ft L, M, H m L, M, H m L, M, H linear ft or m
Alligator or Fatigue Cracking A, B, C WP/ NWP L, M, H m or sq. m L, M, H sq. m L, M, H sq. ft or sq. m
Transverse Cracking ( Non PCC Slab Joint Reflective) 1/ 4" number L, M, H number, m L, M, H number, m L, M, H linear ft or m
Joint Reflection Cracking from PCC Slab L, M, H number, m L, M, H number, m L, M, H linear ft or m
Block Cracking L, M, H sq. m L, M, H sq. m L, M, H sq. ft or sq. m
Edge Cracking L, M, H m L, M, H m
Deformation
Rutting NA via PCS profiler L, M, H mm NA mm L, M, H sq. ft/ m & mm
Corrugation NA sq. m L, M, H sq. ft or sq. m
Shoving NA yes/ no NA number, sq. m NA number, sq. m
Depression NA sq. m L, M, H sq. ft or sq. m
Overlay Bumps NA number, m
Deterioration
Delamination/ Slippage Cracking NA sq. m
Slippage Cracking NA sq. ft or sq. m
Potholes fill/ unfill yes/ no L, M, H number, sq. m L, M, H number, sq. m L, M, H number
Patching NA ft L, M, H number, sq. m L, M, H number, sq. m L, M, H sq. ft or sq. m
Raveling and Weathering coarse/ fine 25% or more L, M, H sq. m NA sq. m L, M, H sq. ft or sq. m
Stripping NA yes/ no/ unknown
Polished Aggregate NA sq. m NA sq. m NA sq. ft or sq. m
Pumping and Water Bleeding NA yes/ no NA number, m NA number, m L, M, H yes/ no
Mat Problems
Segregation NA sq. m
Checking NA sq. m
Bleeding NA 25% or more L, M, H sq. m NA sq. m NA sq. ft or sq. m
Other
Lane/ Shoulder Dropoff or Heave L, M, H inches/ 100 ft
Lane Shoulder Joint Separation L, M, H inches/ 50 ft
Swell L, M, H sq. ft or sq. m
Re- opened Cracks Re- open %
Sealed Cracks > 6 mm %
Settlement NA yes/ no NA yes/ no
A = A single or two longtitudinal cracks in the wheel path, cracks are not spalled or sealed Not Included L, M, H = Low, Moderate, High. NA = Not Applicable.
B = An area of interconnected crackes in the wheel path forming a complete pattern WP = Wheel Path. NWP = Non Wheel Path
C = An area of moderately or severely spalled interconnected cracks outside of the wheel path forming a complete pattern
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
Appendix C
EXAMPLE EXPERIMENTAL DESIGN FOR EVALUATION OF
MODIFIED ASPHALT MIXES
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
C- 1
EXAMPLE EXPERIMENTAL DESIGN FOR EVALUATION OF
RUBBER MODIFIED ASPHALT MIXES
Project Idea/ Hypothesis
A generic experimental design specific to the evaluation of wet process CRM- modified binders and mixes
is presented in this section. The project idea is to evaluate the performance of rubberized asphalt concrete
( RAC) made with high viscosity binder, with modified binder ( MB, no agitation), and a dry process
CRM- modified mix ( RUMAC) that Caltrans currently uses for pavement rehabilitation projects. METS
is the proposed champion of this project with the RAC Task Group ( RACTG) as technical advisory
committee.
Review of past performance information indicates that wet process CRM mixes can result in good
pavements if they are designed and constructed properly. The hypothesis for this project is that wet
process mixes perform better than conventional dense- graded asphalt concrete ( DGAC) and are cost
effective.
Available data indicate that the initial cost per ton for using wet process mixes is significantly higher than
the conventional AC. Based on overall performance to date, these products deserve further investigation.
The three types of investigation alternatives described in Section 2.3 are recommended for the evaluation
of RAC and MB materials. Detailed descriptions for each are provided below.
Laboratory Study
The important variables related to aggregate, binders, and mixtures were identified and described in
Section 2.3.1. A dense- graded asphalt concrete ( DGAC) mix should be included to serve as the control
mix for comparison with other identified mixes.
The laboratory study may be conducted by itself or as a part of ATP studies and/ or of field pilot studies.
The laboratory testing program should include proposed tests identified in Table C. 1 for aggregate and
mixtures. For binders, the tests shown in Table C. 1 should be run on base asphalt cements ( AR- 4000 and
AR- 8000), the high viscosity asphalt rubber binder, and the MB binders. It is suggested that the
laboratory testing focus primarily on binder and mixture performance, including evaluation of volumetric
property requirements for mixes made with the two different families of wet process binder ( high
viscosity and no agitation) and appropriate ranges of binder contents for each. ( Current information from
Texas and Arizona indicates that optimum binder content ( OBC) for high viscosity and no agitation
binder types may differ by 2%.) Performance testing of high viscosity binders may be limited due to the
size of the swollen CRM particles relative to the DSR gap. Mixes should be tested for fatigue and
repeated shear. The proposed mixture test program is based on work done by UC PPRC ( University of
California Berkeley, 2003).
Accelerated Pavement Testing Study
The important variables to be considered under APT study were identified and described in Section 2.3.2.
A control AC section should be constructed along with other test sections for study. Table C. 2 shows an
experimental design matrix for studies of various mix types and overlay thicknesses.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
C- 2
Table C. 1 List of Proposed Tests for Each Material
Test Purpose Comments
a) Aggregate
CT 202 Combined Gradation Indicator test
CT 205 % of crushed particles Indicator test
CT 206 SG & Absorption - Coarse Indicator test
CT 211 Abrasion - Coarse Performance test
CT 214 SSoouunnddnneessss -- CFionaer se IInnddiiccaattoorr tteesstt
CT 217 Sand Equivalent - Fine Indicator test
CT 226 MMooiissttuurree CCoonntteenntt -- CFionaer se IInnddiiccaattoorr tteesstt
CT 227 Cleanness Indicator test
CT 105 GGrraaddiinnggss && SSGG -- CFionaer se IInnddiiccaattoorr tteesstt
AASHTO T304 Uncompacted Voids Indicator test. Index of fine aggregate
angularity and texture
ASTM D4791 Flat & Elongated Particles Indicator test
b) Binder
AASHTO T48 Flash and Fire Points Indicator test
AASHTO T49 Penetration Indicator test
ASTM D217 Cone Penetration Indicator test
AASHTO T201 Kinematic Viscosity Indicator test
AASHTO T202 Viscosity Indicator test
Caltrans Special Hand- held Haake Viscosity Indicator test
AASHTO T240 Rolling Thin- Film Oven Indicator test
ASTM D3407 Resilience Indicator test
ASTM D36 Softening Point Indicator test
CT 381 and
AASHTO T315
Dynamic Shear Rheometer
( DSR)
Performance test. High viscosity binders
may not be suitable for DSR testing due to
the size of swollen CRM particles relative
to DSR gap size ( nominal 1 mm opening)
c) Mixture
CT 308 Bulk Specific Gravity Core, and lab compacted Hveem specimens
CT 309 Rice Gravity Loose mix
CT 366 Stability Value Lab compacted specimens
Volumetric Analysis Air Voids Content, VMA,
VFA, dust to binder ratio
Volumetric analysis of lab and field
compacted mixture specimens
CT 371 Moisture Sensitivity Lab mixed lab compacted ( LMLC) mix
CT 382/ CT 202 Binder Content/ Gradation Loose mix or Core
AASHTO T321 * FFaretiqguueen Acys sSewssemepe, n Bt, eBaema m 2T esmtrapisn 5s °@ a n2d0 ° 2C5° C
AASHTO T320 * RTeumttipn gF rAeqss oenss Smtiefnfnt, e Csso, r Ce ore TTeemmppss (( 4200,, 5400,, 6600°° CC))
Long- Term Oven Effect of Aging, Beam 3, 6 days and 2 strains
AASHTO T324 Hamburg Wheel Track Core or lab compacted specimens
AASHTO TP10- 93 Temperature Cracking Eval. Field or lab compacted specimens
* The tests may also be run on lab mixed lab compacted mixes.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
C- 3
Table C. 2 Matrix for Various Overlay Mixes and Thicknesses
Mix Type
Existing
Pavement
Overlay
Thickness Control AC RAC- G MB- D MB- G RUMAC
A HFualllf
Field Pilot Studies
The important variables to be considered under field pilot studies were identified and described in Section
2.3.3. Since the idea is for pavement rehabilitation projects, the primary study variables are climate,
traffic, roadbed soil, and overlay thickness and materials.
Table C. 3 shows an experimental design matrix for evaluation of RAC- G and MB mixes under various
study variables. For each climate, traffic, and existing pavement condition, the field pilot study includes a
control mix of full thickness, RAC- G, MB- G, MB- D, and RUMAC mixes with a full thickness and/ or a
reduced thickness for a total of 60 possible combinations ( not including the full and ½ thickness
possibilities for the new products). This illustrates the importance of clearly defining the important
variables so that the study is a manageable size and can be sold to upper management. For selection of
projects, activities associated with various stages of construction and evaluation, guidelines ( Table 2.8),
sampling requirements ( Table C. 4), and data collection checklist ( Table 2.9) should be followed.
Table C. 3 Experimental Matrix for RAC- G and MB Study
Overlay Thickness
Climate Traffic Existing Pavement
Condition Control Mix RAC- G MB- G MB- D RUMAC
Good Full Full or Half Full or Half Full or Half Full or Half
Low
Poor Full Full or Half Full or Half Full or Half Full or Half
Good Full Full or Half Full or Half Full or Half Full or Half
Coastal
High
Poor Full Full or Half Full or Half Full or Half Full or Half
Good Full Full or Half Full or Half Full or Half Full or Half
Low
Poor Full Full or Half Full or Half Full or Half Full or Half
Good Full Full or Half Full or Half Full or Half Full or Half
Valley/ Desert
High
Poor Full Full or Half Full or Half Full or Half Full or Half
Good Full Full or Half Full or Half Full or Half Full or Half
Low
Poor Full Full or Half Full or Half Full or Half Full or Half
Good Full Full or Half Full or Half Full or Half Full or Half
High
Desert/ Mountain
High
Poor Full Full or Half Full or Half Full or Half Full or Half
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
C- 4
Table C. 4 Proposed Sampling Requirements for Each Mix
Type Quantity Dimension Sample Location Remarks
Aggregate 1000 kg Plant Combined gradation
Binder 40 kg/ type Plant The quantity is for each type
of binder used in project, e. g.,
AR- 4000, wet process high
viscosity, wet process no
agitation ( including MB), etc.
Modifier 10 kg/ type Supplier The quantity is for each type
of modifier used in project,
e. g., CRM. Extender oil is
also a modifier
Loose mix 250 kg Behind paver For making lab compacted
specimens
Core 20 150 mm 10 from each end of
monitoring section
Sample location should be 30
m away from each end of
monitoring section
Slab 6 550 x 350 mm 3 from one end of
monitoring section and
3 from the other end
Each slab is for making 4
beams
Implementation
The various studies should result in a confirmation of the project idea/ hypothesis. If the hypothesis is
confirmed that the RAC- G and/ or MB mixes perform better than conventional AC, are cost effective, and
have little or no environmental impact, an implementation plan needs to be developed. The plan should
include the types of reports to be delivered as well as updated guidelines or specifications, and/ or training
materials. If the hypothesis is not confirmed, additional studies may be necessary or the further
exploration of the idea should be discarded.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
Appendix D
WORK PLAN FOR RAC- G GRAVEL FACTOR FOR USE IN
STRUCTURAL SECTION DESIGN
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 1
WORK PLAN FOR RAC- G GRAVEL FACTOR FOR
USE IN STRUCTURAL SECTION DESIGN
BACKGROUND
Caltrans current pavement structural design procedure ( Caltrans, 2004) involves the use of R- value and
Traffic Index ( TI) to develop flexible pavement layer thicknesses for new and reconstruction projects.
The procedure is based upon a layer equivalency approach in which the relative load- carrying capacity of
individual pavement layers is related through a gravel equivalence value. The gravel factor ( Gf) refers to
the relative strength of a given material compared to a standard gravel subbase material. Gravel factors
for dense- graded asphalt concrete and various types of base and subbase materials have been developed
over the years based primarily on the cohesiometer tests. However, no Gf has been established for
rubberized asphalt concrete ( RAC) or modified binder ( MB) materials for use in new pavement as well as
in rehabilitation designs.
Caltrans employs several types of structural mixes in the design of asphalt concrete overlays for flexible
pavements: dense- graded asphalt concrete ( DGAC), MB mixes, and a gap- graded, rubberized asphalt
concrete ( RAC- G) mix. The thickness of the DGAC overlays needed to limit fatigue cracking is
determined using an empirical relationship relating measured pavement surface deflection, TI, and the
thickness of the existing pavement. The design procedure also uses an empirical relationship to determine
the thickness of the overlay needed to retard reflective cracking. For a RAC- G mix, Caltrans uses
equivalence ratios of 1.5 to 2.0 to reduce the thickness when fatigue cracking is the expected distress
mode and 1.5 to 2.33 when reflection cracking is the expected distress mode. In cases where a greater
additional structure is required, DGAC mix might be placed on the existing surface and then overlaid with
RAC- G of reduced thickness. The equivalencies are based on a 10- year design life for overlays,
assuming that the existing pavement is structurally adequate. Caltrans also specifies minimum and
maximum RAC- G thickness of 30 mm and 60 mm, respectively ( Caltrans 2001).
Other states that have significant experience with the asphalt rubber products generally treat RAC
mixtures as having the same structural value as conventional DGAC. Arizona’s structural design
methodologies for rubber- modified asphalt concrete are the same as that for dense- graded mixture
regardless of application. Texas treats gap- graded rubber- modified mixes the same as the conventional
dense- graded mixes in terms of structural credit. Florida DOT does too by using AASHTO Design Guide
layer coefficients of 0.44 for dense- graded friction course with or without crumb rubber modifier
( MACTEC, 2004a).
In 2004, a literature review was conducted as a part of a study ( MACTEC, 2004a) funded by California
Integrated Waste Management Board ( CIWMB). The results of the literature review indicate that there is
no universal consensus on structural design with rubber- modified asphalt concrete mixes. However, it
appears that treating RAC as a structural equivalent of DGAC has yielded reasonable results.
OBJECTIVE
The objective of this study is to develop gravel factor( s) for both RAC- G and MB mixtures for use in new
pavement structural section design in accordance with Caltrans Highway Design Manual ( Caltrans, 2004).
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 2
HYPOTHESIS
This study will be conducted as outlined in the generic experiment design described previously. The
hypothesis for the experiment is that the gravel equivalence approach used for determining the DGAC
thickness in new pavement design is valid for the thickness design of RAC- G and MB mixtures. This
hypothesis will be tested/ confirmed if analyses of existing data ( and data to be collected) result in valid
criteria and analytical models that accurately characterize the difference in performance among the
various AC mixtures used for new pavement design.
STUDY APPROACH
Based upon the outcome of several meetings held with Caltrans and UC PPRC staff, a general framework
for the development of the gravel factor( s) for RAC was outlined. This framework ( graphically depicted
in Figure D. 1) calls for a review of the related work, especially of laboratory test data on dense- graded
AC and RAC materials and supplemental testing ( i. e., cohesiometer test, indirect tensile and resilient
modulus tests) of DGAC, RAC- G, and MB materials. Interim gravel factor( s) for RAC- G/ MB will be
developed based on these test results, a review of related work, and the mechanistic- empirical ( M- E)
analyses. The interim gravel factor( s) will then be validated using data gathered from the current UC
PPRC Heavy Vehicle Simulator ( HVS) modified binder study and relevant data from other new pavement
or reconstruction projects involving the use of RAC or MB mixtures. Additional test sections with
different thicknesses of RAC may also need to be constructed and tested using the HVS to validate the
interim Gf values. This study will be closely coordinated with a companion study to develop Gf values
and associated design criteria for RAC mixtures used for overlays ( see Appendix E). This study will be a
joint effort of Caltrans, the UC PPRC, and MACTEC.
WORK PLAN
The study will be conducted in two phases. Phase 1 is a short- term plan, which involves the development
of interim gravel factor( s) and design criteria for RAC and MB materials for use in structural design.
Phase 2 is a long- term plan, which involves field validation of the interim gravel factor( s) and the
collection of materials, construction, and performance data for use in the development of a future M- E
design method. The short- term plan should be completed by September 2005. On the other hand, the
long- term plan is anticipated to require an additional three to five years to complete, including the
collection of routine field performance data from the various study sections.
Phase 1 - Short- Term Plan
The proposed short- term plan involves the following tasks:
• Review related work on gravel factors ( Gf).
• Develop interim Gf based on laboratory testing of DGAC, RAC- G, and MB mixes, available
laboratory test data and field performance. These data will be analyzed using M- E principles.
• Implement interim Gf values for RAC- G and MB mixtures.
The bulk of this effort will be carried out by MACTEC in consultation with Caltrans and UC PPRC staff.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 3
Task 1 – Planning Meeting
A one- day planning meeting will be held at the beginning of this phase to review ( and finalize) the
technical approach and work plan and to coordinate the data collection activities. This meeting should
include key Caltrans, UC PPRC, and MACTEC staff.
Task 2 – Review of Related Work on Gravel Factors
This task entails a review of material related to establishing gravel equivalency factors ( Caltrans method)
and layer coefficients using the AASHTO method. Following is a summary from a handout on gravel
factors provided by UC PPRC at the September 7, 2004 brainstorming meeting:
Historically, Caltrans developed gravel factors for aggregate base and subbase materials and
emulsion and cement treated materials based on the Brighton Test Road and laboratory
cohesiometer measurements in the 1950s and 1960s. Gravel factors for other materials were
established relative to that of aggregate subbase, which was assigned a gravel factor of 1.0.
Development of Gravel Factors for RAC and MB Materials
For Structural Section Design Using Caltrans Method for New
Construction and Rehabilitation
Short- Term Plan
• Review related work on gravel factors ( Gf).
• Develop interim Gf based on laboratory tests ( i. e., cohesiometer
test, indirect tensile strength test, and resilient modulus test) on
DGAC, RAC- G, and MB mixes, available laboratory test data, and
field performance data using a mechanistic- empirical approach.
• Implement interim Gf.
Long- Term Plan
• Validate interim Gf using data gathered from the UC PPRC MB
study, the Firebaugh project, the RAC Warranty projects and other
RAC projects.
• Construct additional experimental test sections with varying layer
thickness of RAC/ MB/ DGAC materials and test with HVS.
• Validate interim Gf using data from the experimental test sections.
• Develop data for use in M- E design procedure.
• Finalize Gf for implementation.
Figure D. 1. Study Approach for the Development of Gf for RAC and MB Materials
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 4
Gravel factors for conventional asphalt concrete were developed based on data from the AASHO
Road Test.
Examination of the AASHO Road Test data ( HRB Special Report 61E, 1962) and the “ layer
coefficients” approach included in the original AASHO Interim Design Guide indicates that there
is a great deal of scatter in the data. The ratios of the structural contribution factors for aggregate
base, asphalt concrete and cement treated base developed from the AASHO Road Test for both
the AASHO Interim Design Guide and the Caltrans design method are similar, but not the same.
Conceptually similar, both AASHO layer coefficients and Caltrans gravel factors are used to
describe the structural capacity of various materials.
The most recent gravel factor development was for asphalt treated permeable base ( ATPB)
materials in the 1980s by Caltrans. Test sections with and without ATPB were constructed. The
gravel factor ( 1.4) was estimated based strictly on surface deflections measured with the
Dynaflect. The performance under traffic was not considered in the development of the gravel
factor. Furthermore, there was a great deal of scatter in the data.
Task 3 – Development of Interim Gravel Factor
The review of the previous work indicates that there was a significant scatter in the data and that there
was no consistent method used to develop the gravel factors currently used by Caltrans. Other agencies
have used either laboratory tests or accelerated pavement performance tests to develop structural layer
coefficients for use in the AASHTO design procedure ( MACTEC, 2004a).
In the meeting held on October 7, 2004, personnel from Caltrans, UC PPRC, and MACTEC discussed the
use of a cohesiometer for determining the gravel factor Gf, for RAC and MB materials. The group
acknowledged that the historical data used to develop various Gf might not be easily found in the Caltrans
archives. Additionally, the group acknowledged that cohesiometer results may not relate to pavement
performance. Nevertheless, the group agreed to conduct cohesiometer testing to establish baseline values
for the RAC- G and MB materials. Since the utility of the cohesiometer data is unknown, it would be wise
to begin with a “ limited” test matrix. Should the initial test data yield meaningful results, a more
comprehensive testing program can be undertaken.
Laboratory Test Plan
To assist in the development of gravel factor for rubberized asphalt concrete ( RAC) materials, it was
agreed that cohesiometer tests be run on dense- graded asphalt concrete ( DGAC), MB, and RAC- G
materials. The tests will be performed through a cooperative effort between the UC PPRC laboratory and
Translab. Additionally, indirect tensile and resilient modulus testing are proposed to provide fundamental
engineering properties for use in M- E analyses. The UC PPRC or Caltrans lab will conduct the indirect
tensile and resilient modulus testing.
The objective of the lab test plan is to characterize the mix properties under laboratory and/ or in- situ
conditions. This requires both laboratory prepared samples and field cores. It should be noted that
Caltrans has yet to use RAC of MB mixtures in the construction of a new pavement. Thus, development
of improved gravel factors for new pavement design must rely on field data from pavement sections that
have been overlaid with a RAC- G or MB mixture. These data will also be used in the development of
gravel factors for overlay applications, as discussed in Appendix E.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 5
Materials
The materials for lab prepared samples will come from the Firebaugh and/ or the RAC Warranty projects.
Field cores ( 100 mm or 4” in diameter) may be collected as necessary. The benefit of testing the field
cores is obvious: their properties will reflect the “ as- built” pavement condition. The proposed sampling
program will provide for a valid characterization of the key mixture properties for each project. Also it
helps isolate the effect of material type and environment on pavement performance.
Loose Mixes
The first part of the lab test plan involves the use of field- mixed and lab- compacted mixes. The primary
benefit of testing lab compacted specimens is to capture properties of un- aged mixes.
The Firebaugh project includes four mix types: DGAC, RAC- G, MB- G, and MB- D. For each mix, six
specimens will be compacted by Translab to the “ as- built” air void content. These values are shown in
Table D. 1. Three specimens will be used for the cohesiometer test and three specimens will be used for
resilient modulus and indirect tensile strength tests.
Table D. 1. Target air void contents ( as- built) for laboratory preparation of the different mixtures.
Mix Type Air Void Content, %
DGAC To be provided
RAC- G To be provided
MB- G To be provided
MB- D To be provided
The air void content for each mix type is being determined by Translab based on pavement cores obtained
from the project site. As a reference, the mix design information used on the Firebaugh project is
summarized in Table D. 2 below.
Table D. 2. Mixture design information for the mixes used in the Firebaugh project.
Sieve Size ( mm) RAC- G MB- G MB- D Type A
25 100 100 100 ---
19 97 97 96 97
12.5 85 --- --- ---
9.5 68 68 68 70
4.75 36 36 48 50
2.36 21 21 33 36
0.6 11 11 17 18
0.075 2.9 2.9 4.2 5.4
OBC, ( by dry weight of aggregate), % 7.90 6.30 5.30 4.80
Maximum Density @ OBC 2.35 2.40 2.44 2.46
Asphalt Absorption @ OBC, % 0.70 0.67 0.93 0.98
Stability 37 29 39 46
VMA, % 18.31 15.43 12.80 12.09
VFA, % 83.02 82.38 79.90 76.18
Crushed Coarse, min 90% 99 99 98 100
LA Rattler: @ 100 rev max 10%
@ 500 rev max 45%
4.7
24.1
5.0
23.6
4.5
22.0
4.4
24.6
Crushed Fine, min 70% 98 97 94 98
Sand Equivalent, min 47% 56 56 70 58
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 6
Cores
Lab testing of field cores is contingent upon the success of the testing and analyses of the loose mix.
Table D. 3 shows the number of cores needed for the proposed tests. The cores should be obtained from
two different locations to account for material variability. At each location, six cores should be collected.
Three cores will be used for each: the cohesiometer test and the resilient modulus and indirect tensile
strength tests.
Table D. 3. Core samples needed for proposed testing program.
Project/ Mix Type
Number of
Cores/ Location
Number of
Locations
Total Cores
Firebaugh/
DGAC 6 2 12
RAC- G 6 2 12
MB- G 6 2 12
MB- D 6 2 12
Ventura/ RAC- G 6 2 12
Fresno/ RAC- G 6 2 12
Merced/ RAC- G 6 2 12
San Diego/ RAC- G 6 2 12
Lassen/ MB- D 6 2 12
Total 54 18 108
Test Procedures
The cohesiometer test will be performed in accordance with ASTM D 1560- 92 ( Reapproved 2000)
“ Standard Test Methods for Resistance to Deformation and Cohesion of Bituminous Mixtures by Means
of Hveem Apparatus.”
The indirect tensile strength test will be performed in accordance with ASTM D4123- 82 ( Reapproved
1995) “ Standard Test Method for Indirect Tension Test for Resilient Modulus of Bituminous Mixtures.”
The recommended testing temperature for the indirect tension strength test is 25° C. The resilient
modulus test will be performed in accordance with ASTM D4123. The recommended temperatures are 5,
25, and 40° C. A destructive test, the indirect tensile strength test is conducted after the determination of
resilient modulus.
Expected Outcome
The results of the cohesiometer tests from the above mixes will be reviewed, analyzed, and used to
develop gravel factors. The following relationship will be used to determine a gravel factor ( Gf) for each
mix:
0.2
20
  
   Gf = C
where: C = Cohesion value for material.
20 = Cohesion value for aggregate subbase.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 7
It is noted that the type of aggregate subbase associated with C= 20 is not documented in the literature.
Since unbound aggregate performs poorly in tension, the cohesiometer test does not seem ideally suited to
capture its strength properties. Since a cohesiometer value of 20 for aggregate subbase is the basis for
Caltrans gravel factor design concept, it will be used as a baseline measurement. By extension, aggregate
subbase should be used for calibration purposes. MACTEC will work with Translab to make specimens
for this test.
The results from the DGAC mix will also be used as a baseline to allow comparison with the results from
other types of mixes.
The results from the indirect tensile and resilient modulus tests have a dual purpose: 1) to support the
development of gravel factors; and 2) to modify/ improve the Caltrans overlay design procedure. The
results from the indirect tensile strength test may be used to compute strength ratios between DGAC and
RAC- G/ MB mixes for comparison with the ratios determined from the cohesiometer test. Similarly,
laboratory resilient modulus data will be compared to modulus generated from the analysis of deflection
data.
Mechanisitic- Empirical Approach
The M- E ( layered elastic) and/ or the finite element approaches are rational methods for the development
of gravel factors proposed for RAC and MB materials. In using the mechanistic- empirical method to
develop the gravel factor, the following parameters are needed:
• Resilient modulus for conventional AC, RAC, and MB materials – These data may be obtained
from deflection data through backcalculation and/ or from laboratory tests. As available, UC
PPRC data will be compiled for evaluation.
• Performance data for conventional AC, RAC, and MB materials – Ideally, these data should
encompass conventional AC, RAC- G, and MB materials placed in similar environments and
having sustained similar loading. The Firebaugh project is ideal though the final data from this
project will not be available until 2009. Alternatively, the laboratory data from the fatigue and
rutting tests for the conventional AC, RAC, and materials may be used. Again, UC PPRC data
will be evaluated for developing preliminary performance models.
To develop interim gravel factors, pavement sections of varying thicknesses will be considered.
Stresses/ strains at critical locations in the pavement section will be computed and used in conjunction
with performance models to quantify structural capacity. The analyses will be performed for pavements
with conventional AC, RAC- G, and MB materials. Mix performance ( as characterized by fatigue or
rutting) may be used to develop an interim gravel factor.
The before/ after deflection data from typical overlay projects will be used to estimate changes in stiffness
that could, in turn, be used in the development of Gf. These data may also help develop/ refine the
Caltrans deflection- based overlay design procedure.
To best meet Caltrans needs on the Gf issue, it was agreed that following activities be pursued:
• Identify best use of asphalt rubber products.
• Meet with the pavement design group to identify what else is needed besides Gf for RAC
projects.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 8
• Conduct cohesiometer tests and determine Gf values.
• Obtain available materials test data ( including fatigue, rutting, and stiffness/ modulus) from UC
PPRC Goal- 3 study and other relevant information for developing interim Gf using the M- E
approach.
• Develop interim guides for use and/ or incorporate changes in the Caltrans Highway Design
Manual.
To carry out these activities, on- going consultation with Caltrans, particularly from the Office of
Pavement Rehabilitation ( OPR), is crucial. It is recommended that key personnel from Caltrans and
MACTEC participate in a one- day meeting to discuss the following:
• A critical review of existing Highway Design Manual ( Caltrans, 2004) Chapter 600 on Pavement
Structural Section and Flexible Pavement Rehabilitation Manual ( Caltrans, 2001), specifically on
the gravel factor for RAC and MB materials;
• Modifications to the aforementioned manuals to accommodate RAC and MB mixtures.
• Clarify responsibility and schedule for conducting the cohesiometer test on RAC, MB, and
DGAC materials, assuming the cohesiometer and these materials are available. It may be
necessary to develop an experimental matrix for this purpose.
Task 4 – Implementation of Interim Gravel Factor
It is anticipated that interim gravel factors ( and other relevant design criteria) for RAC- G and MB
mixtures will be developed by August 1, 2005. The implementation plan includes three primary activities.
• Report – The results of the Phase 1 ( short term plan) will be thoroughly documented including a
discussion of the data, analysis, and findings, and recommendations as to modifications to the
Caltrans Highway Design Manual.
• Training – MACTEC will prepare a two- hour training module for Caltrans engineers on the
application of the new gravel factors and design criteria. The training module will include a
hands- on workshop to allow participants to apply the revised methodology using sample data.
Phase 2 – Long- Term Plan
The proposed long- term plan involves the following tasks:
• Construct additional experimental test sections with varying layer thickness of RAC/ MB/ DGAC
materials and test them with HVS.
• Validate the interim Gf value using data gathered from the UC PPRC HVS MB study and any
other relevant new pavement or reconstruction project involving the use of RAC or MB mixtures.
• Develop data and criteria for new Caltrans M- E design procedure.
• Implement the findings.
MACTEC will play a key role in getting this phase laid out. However, work proposed for years 2 through
5 will have to be conducted as a coordinated effort between Caltrans and UC PPRC.
Task 1 – Planning Meeting
A one- day planning meeting will be held near the end of first phase of work, probably near the
completion of Task 3, to review the phase 1 results and to finalize the work plan for the long- term study.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
D- 9
A critical topic of this meeting will be the approach to validation of the RAC and MB gravel factors
through future data collection and analyses. Another key topic will be the data needs and collection
protocol for incorporating RAC and MB mixtures within an M- E design procedure.
Task 2 – Construction of Additional Experimental HVS Test Sections
Time and budget permitting additional construction of additional experimental test sections ( including
conventional DGAC, RAC, and MB mixtures) will be constructed for HVS testing. These will generate
data for the M- E design procedure and to refine the performance models which can be used to
update/ refine the interim gravel factors. It is not recommended that additional full- scale experimental
field projects be constructed. However, it may be possible to use information from routine construction
projects.
Task 3 – Validation of Interim Gravel Factor
The ongoing studies, i. e., the UCB HVS MB study, the Firebaugh project, and the Warranty projects, will
be completed in 3 to 5 years. The results from these studies should address construction, materials
characterization, and laboratory and field performance for various RAC and MB materials. The data from
these studies may provide valuable information related to the validation of the interim gravel factor.
Task 4 – Development of Data and Criteria for an M- E Design Procedure
Data from various projects will be gathered and analyzed. Performance data along with laboratory test
results will be used to develop criteria for use in an M- E procedure.
Task 5 – Implementation
Implementation will include a report that thoroughly documents the entire effort. Conclusions and
recommendations will be included. Technology transfer will be initiated at Caltrans direction.
EXPECTED OUTCOMES
The expected outcomes from this study include the following:
• From the short- term plan – establishing Gf:
o Report documenting the development of the interim gravel factor based on lab test data and
an M- E approach; and
o Interim gravel factor for RAC/ MB materials for use in Caltrans pavement design procedure.
• From the long- term plan – validating Gf:
o Findings from the ongoing projects;
o Construction and evaluation of additional experimental test sections;
o Analysis and validation of the interim gravel factors using data from ongoing studies and
additional experimental test sections; and
o Data and criteria for use of RAC and MB mixtures in Caltrans M- E design procedure.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
Appendix E
PROPOSED WORK PLAN FOR VALIDATION, CALIBRATION, AND
IMPROVEMENT OF THE AC OVERLAY DESIGN PROCEDURE IN THE
CALTRANS FLEXIBLE PAVEMENT REHABILITATION MANUAL
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
E- 1
PROPOSED WORK PLAN FOR
VALIDATION, CALIBRATION, AND IMPROVEMENT OF THE
AC OVERLAY DESIGN PROCEDURE IN THE
CALTRANS FLEXIBLE PAVEMENT REHABILITATION MANUAL
BACKGROUND
The Caltrans Flexible Pavement Rehabilitation Design Manual ( Caltrans 2001) provides technical
guidance for the design of a number of different rehabilitation treatments for flexible pavements. Among
these are the following:
• asphalt concrete ( AC) overlay ( directly on the existing pavement and pre- treated with either a
stress- absorbing membrane interlayer, cushion course, or a drainage layer);
• remove and replace ( including mill and fill);
• cold recycled asphalt concrete ( CRAC); and
• hot recycled asphalt concrete.
The primary component of the design procedure is the overlay design model. It relies on measured
surface deflections to characterize the structural load- carrying capacity of the existing pavement, layer
equivalence factors to define the relative strength of individual pavement layers, and the deflection
reduction approach for thickness design. The overlay method was originally developed in 1979 ( Caltrans,
1979) and has been refined over the years, including the recent adaptation to consider gap- graded,
rubberized AC ( RAC- G) mixtures ( Caltrans, 2001). The method serves as the standard for Caltrans
rehabilitation design, and is widely used by many local agencies.
OBJECTIVE
Despite its general acceptance and widespread use in the state, the Caltrans overlay design method does
not readily accommodate new materials, e. g., rubber modified binder mixes. As indicated above, the
design procedure was adapted to consider RAC- G mixtures based on the field performance of asphalt
rubber mixtures. For AC mixtures containing crumb rubber modified binders, basic design and
performance criteria need to be re- evaluated, modified as necessary and incorporated in the procedure.
The basic objective of this study is to improve to the overlay design component of Caltrans rehabilitation
design procedure. This includes the following:
• validation/ calibration of the RAC- G design criteria currently used by Caltrans;
• development of design and performance criteria for AC mixtures containing modified binders;
• revision of the tolerable deflection criteria for rubber modified mixtures; and
• evaluation of thickness requirements the surface layer combinations including RAC on DGAC,
DGAC on RAC, and RAC on RAC.
In the case of RAC- G, special attention must be given to addressing the RAC- G maximum overlay
thickness limitation and the potential problem of placing a conventional dense- graded AC overlay over an
existing RAC- G surface. Also, one of the internal models for determining pavement deflection reduction
as a function of overlay thickness should also be re- evaluated to account for the effect of overlay material
type.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
E- 2
HYPOTHESIS
The objective of this study will be carried out under the general framework established in the generic
experiment design described previously. The hypothesis for the experiment is that the deflection based
approach used for determining the DGAC thickness overlay design is valid for the overlay thickness
design of RAC- G and modified binder ( MB) mixtures. This hypothesis will be confirmed if analyses of
existing data ( and data to be collected) result in valid criteria and analytical models that accurately
characterize the difference in performance among the different AC mixtures used for overlay design.
WORK PLAN
This study will be carried out in two phases and will involve a joint effort among Caltrans, the UC PPRC,
and MACTEC. The scope of the first phase will be limited primarily to the analysis of existing data from
recent ( and on- going) field and laboratory experiments. Its goal is to make immediate improvements to
the overlay design model.
The second phase, or long- term program, will be carried out over an additional 3- 5 year period and
involves analysis of data from laboratory and field experiments conducted during that period, as well as
data that are already available. Depending on the results of the data analyses and Caltrans transition to
mechanistic- based design, the product of this program may be a completely new design procedure.
For the sake of consistency, this study should be closely coordinated with the companion study to develop
RAC and MB gravel factors for use in new pavements ( see Appendix D). Following is a discussion of
the work tasks currently envisioned for the short- term and long- term programs.
Phase 1 – Short- Term Program
The target completion date for phase 1 is July 31, 2005.
Task 1 – Planning Meeting
The purpose of this meeting is to finalize the work plan for revising the current Caltrans AC overlay
design methodology to accommodate the materials and performance characteristics of rubber- modified
binders. This will be a cooperative effort among Caltrans, the UC PPRC, and the MACTEC project team.
It is important that key staff from each organization participate in the meeting. The meeting was held in
December 2004 and included the following agenda items.
• Overview of current overlay design methodology and areas of needed improvement
• Review of rubber- and polymer- modified materials characteristics and performance properties
• Approach for improvements and validation of design methodology
• Data sources for model calibration/ development
• Data collection plan ( associated with field sampling, field testing, and laboratory testing)
• Staff assignments
• Schedule
• Report outline
• Work plan for long- term program
The last agenda item ( work plan for the long- term program) was included to take advantage of the
consensus agreement reached on the short- term program and to develop consistency with the long- term
program.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
E- 3
Task 2 – Collection and Assembly of Project Data
The short- term program will rely upon a combination of existing data and new data to improve the
overlay design methodology. The bulk of the data will be gathered from existing records available from
the following:
• recent laboratory and HVS studies carried out by the UC PPRC;
• Caltrans Translab field and laboratory experiments;
• RAC- G Warranty and Firebaugh projects; and
• the 1000 plus test sections that make up the Caltrans performance monitoring database, a study
currently being performed for Caltrans Research by Stantec.
Some ( new) field and laboratory testing will also be conducted over the next 6 to 8 months, primarily to
estimate key material properties and/ or laboratory properties of select rubberized- and modified- binder
mixtures and to characterize the in- situ condition of certain field- test sections. The laboratory test
program will include tests to determine the cohesion, stiffness, fatigue cracking, and permanent
deformation properties of select AC mixtures. Because of the deflection reduction basis in the overlay
design methodology, before and after overlay deflection testing is a necessary requirement for the field
test program.
The data will be compiled into one or more databases that will support the development of the improved
analytical models and design criteria.
Work on this task should begin in spring 2005 and be completed by July 2005.
Task 3 – Develop Effective Gravel Equivalence Factors for RAC- G and MB Mixtures
In this task, effective gravel equivalence factors will be established for RAC- G and MB mixtures based
upon their “ performance” relative to the standard gravel or a conventional DGAC mixture. The detailed
field sampling and laboratory test plan presented in Appendix D, under section Task 3 – Development of
Interim Gravel Factor - Laboratory Test Plan, addresses the sampling and testing needs for developing
improved gravel factors for both new pavement and rehabilitation design. Testing included in the
laboratory program are described below:.
• Cohesion Test ( ASTM D1560) – This refers to the laboratory test method that was originally used
to develop gravel equivalence factors for subbase and base courses, as well as the original
standard DGAC mixture.
• Mix Stiffness and Strength ( ASTM D4123) – Laboratory indirect tensile strength and resilient
modulus offer two more alternatives for the determining the gravel equivalence factor. As
fundamental engineering properties, indirect tensile strength and resilient modulus tests provide a
rational basis for deriving gravel factors for asphalt concrete mixtures.
• Fatigue Resistance ( AASHTO T 321) – Both laboratory and field test results for crack resistance
can be used to develop gravel equivalence factors that relate RAC- G and MB mixtures to the
standard DGAC mixture.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
E- 4
• Rut Resistance ( AASHTO T320) – Both laboratory and field test results for rut resistance can
also be used to develop gravel equivalence factors that relate RAC- G and MB mixtures to the
standard DGAC mixture.
The fatigue and rutting tests are likely to provide more rational gravel equivalence factors if enough data
can be extracted from available records or generated from field/ lab testing experiments. They are the
foundation for the M- E based design procedures in the second phase, long- term program. The project
will rely on the staff and equipment available at METS and the UC PPRC for any lab testing.
Work on this task will be coordinated with that outlined in Appendix D to ensure consistency in the
gravel factors used for new pavement and overlay design.
Task 4 – Evaluate Various Structural Layer Combinations of Overlay Types on RAC- G Surface
The purpose of this task is to evaluate the state of stress in two structural layer combinations that have not
yet been tested in a field situation. This evaluation will be conducted using mechanistic tools based on
elastic layer theory and/ or finite element analysis. The alternate layer combinations include DGAC
overlay on existing RAC- G surface and RAC- G overlay on existing RAC- G surface. By comparing the
state of stress in these two layer combinations ( as measured by critical shear and tensile strains) with the
typical RAC- G on DGAC structural combination, it will be possible to identify conditions that could lead
to rapid deterioration or premature failure.
Task 5 – Evaluate Tolerable Surface Deflection and Maximum Thickness Criteria
The current Caltrans overlay design procedure provides criteria for the conversion of a specific design
thickness of DGAC overlay into an equivalent thickness of RAC- G overlay. The conversion does not
directly consider the larger tolerable surface deflection associated with a RAC- G overlay, although it is
probably inherent in the thickness conversion. The thickness reduction was based on field performance
and confirmed in the HVS testing in South Africa and later at UC Berkeley.
The overlay design procedure establishes criteria for maximum and minimum thickness of the RAC- G
overlay. The maximum ( 60- mm) thickness limitation was established to help guard against the potential
for shear flow and rutting in a thicker layer. The validity of this limitation will be evaluated using data
from the Firebaugh project that has test sections of 90 mm thick, in terms of both the field performance
and cost effectiveness.
In this task, the Caltrans overlay design procedure will be revised to allow the direct evaluation and
thickness design for RAC- G overlays. This will be accomplished by treating RAC- G overlay design as a
separate, but similar process, as that for the DGAC overlay thickness design. This will require the
development of a unique tolerable surface deflection relationship that may be appropriate for each type of
combination of overlay and surface materials, e. g., RAC- G over DGAC, RAC- G over RAC- G, and
DGAC over RAC- G. The tolerable deflection relationship can be developed through a two- pronged
evaluation involving: 1) an analysis of laboratory fatigue and permanent deformation test results, and 2)
an investigation of before and after overlay deflection measurements on RAC- G projects. It will also
require an analysis of the data generated in Task 4 to determine if there is a valid maximum thickness
limitation for RAC- G overlays. Performance data from the variable thickness RAC- G overlay sections in
the Firebaugh project should also be helpful in this assessment.
Generic Experimental Design for Product/ Strategy Evaluation - Crumb Rubber Modified Materials February 8, 2005
Caltrans/ CIWMB Partnered Research
E- 5
Task 6 – Develop and Incorporate New Deflection at Milled Depth Relationship
To account for the effect of removing part of an AC surface as part of a cold recycling or mill and fill
operation, the current Caltrans overlay design methodology incorporates a relatively simple model to
estimate the maximum deflection at the milled depth. Because of its simplicity, the accuracy of the model
is questionable.
In this task, a more accurate, yet deterministic, deflection prediction model will be developed as a
replacement for the current model. Development of the new model will be based upon elastic layer
theory. Potential computer programs used for the analyses include the following: LEAP2, ELSYM5,
WESLEA, and EVERSTRESS. The new prediction model should be valid for the design of both DGAC
and RAC- G resurfacings on milled pavements. Caltrans suggested that this approach be verified ( if
possible) with FWD measurements performed incrementally on milled RAC- G surface. This issue