2007-25

Resilient Modulus Development of Aggregate Base

and Subbase Containing Recycled Bituminous and
Concrete for 2002 Design Guide and Mn/Pave
Pavement Design

ve Solutions!
vati
ch . ... Inno
r .. Knowledge
sea
Take the steps... Re

Transportation Research
 Technical Report Documentation Page
1. Report No. 2. 3. Recipients Accession No.
MN/RC-2007-25
4. Title and Subtitle 5. Report Date
Resilient Modulus Development of Aggregate Base and June 2007
Subbase Containing Recycled Bituminous and Concrete for 6.
2002 Design Guide and Mn/Pave Pavement Design
7. Author(s) 8. Performing Organization Report No.
Thomas M. Westover, Joseph F. Labuz, Bojan B. Guzina
9. Performing Organization Name and Address 10. Project/Task/Work Unit No.
Department of Civil Engineering
University of Minnesota 11. Contract (C) or Grant (G) No.
500 Pillsbury Dr. SE (c) 81655 (wo) 156
Minneapolis, MN 55455
12. Sponsoring Organization Name and Address 13. Type of Report and Period Covered
Minnesota Department of Transportation Final Report
395 John Ireland Boulevard Mail Stop 330 14. Sponsoring Agency Code
St. Paul, Minnesota 55155
15. Supplementary Notes
http://www.lrrb.org/PDF/200725.pdf
16. Abstract (Limit: 200 words)
The primary objective of this study was to quantify stiffness (resilient modulus) of aggregate base containing
recycled asphalt and concrete pavements. After a survey of other state’s specifications and implementation
guidelines, Minnesota recycling projects were selected based on the availability of laboratory resilient modulus
(MR) tests and field measurements from FWD. The projects were County State Aid Highway 3, Trunk Highway 23
and Trunk Highway 200. Based on the results of a parametric study, it was found that traditional peak-based
analysis of FWD data can lead to significant errors in elastostatic backcalculation. A procedure for extracting the
static response of the pavement was formulated and implemented in a software package called GopherCalc.
Laboratory resilient modulus measurements were compared with moduli backcalculated from the FWD data. The
FWD data was analyzed using conventional (peak-based) and modified (FRF-based) elastostatic backcalculation
(Evercalc) as well as a simplified mechanistic empirical model called Yonapave. Laboratory values from
sequences in the MR protocol that produced a similar state-of-stress were used. Additionally, a seasonal analysis of
FWD test data revealed a significant increase in stiffness when the pavement is in the frozen state.

17. Document Analysis/Descriptors 18. Availability Statement

Resilient modulus, No restrictions. Document available from:
Falling weight deflectometer, National Technical Information Services,
Recycled asphalt pavement, Springfield, Virginia 22161
Full depth reclamation
19. Security Class (this report) 20. Security Class (this page) 21. No. of Pages 22. Price
Unclassified Unclassified 55
Resilient Modulus Development of Aggregate Base and
Subbase Containing Recycled Bituminous and Concrete
for 2002 Design Guide and Mn/Pave Pavement Design

Final Report

Prepared by
Thomas M. Westover, Joseph F. Labuz, & Bojan B. Guzina,
Department of Civil Engineering
University of Minnesota

June 2007

Prepared for
Minnesota Department of Transportation
Office of Materials and Road Research
1400 Gervais Avenue
Maplewood, MN 55109

This report represents the results of research conducted by the authors and does not
necessarily represent the views or policies of the Minnesota Department of
Transportation and/or the Center for Transportation Studies. This report does not contain
a standard or specified technique.
 TABLE OF CONTENTS

Chapter 1 Introduction ........................................................................................................ 1

1.1 Background ............................................................................................................... 1
1.1.1 Recycled Concrete Aggregate (RCA)................................................................ 1
1.1.2 Reclaimed Asphalt Pavement (RAP)................................................................. 2
1.2 Objectives ................................................................................................................. 2
1.3 Organization.............................................................................................................. 3
Chapter 2 Project Information ............................................................................................ 4
Chapter 3 Data Collection................................................................................................... 6
3.1 Minnesota Reclaiming Projects ................................................................................ 6
3.2 Recycled Materials Survey ....................................................................................... 7
Chapter 4 Data Analysis ..................................................................................................... 9
4.1 FWD Data Analysis .................................................................................................. 9
4.1.1 GopherCalc ...................................................................................................... 11
4.1.2 Yonapave ......................................................................................................... 14
Chapter 5 Results and Discussion..................................................................................... 16
5.1 Wright County State Aid Highway 3...................................................................... 16
5.2 TH-23...................................................................................................................... 19
5.3 TH-200.................................................................................................................... 20
5.4 Seasonal Effects ...................................................................................................... 22
Chapter 6 Summary and Conclusions............................................................................... 24
Appendix A DOT Survey Responses
Appendix B SDOF Initial Conditions
Appendix C Additional Project Selection
Appendix D FWD Parametric Study
Appendix E Seasonal Effects
 LIST OF FIGURES
Figure 4-1 Error in base modulus using peak-based elastostatic backcalculation
(175 test cases)................................................................................................ 10
Figure 4-2 Error in base modulus using FRF-based backcalculation (175 test cases). .... 10
Figure 4-3(a-d) Effect of baseline correction on frequency response functions,
a) Original time-history b) original FRF c) baseline-corrected time-history
d) baseline corrected FRF. .............................................................................. 12
Figure 4-4 Schematic of a single-degree-of-freedom (SDOF) system. ............................ 14
Figure 5-1 CR-3 base modulus prior to reconstruction. ................................................... 17
Figure 5-2 CR-3 base modulus after reconstruction. ........................................................ 17
Figure 5-3 CR-3 equivalent subgrade modulus before reconstruction. ............................ 18
Figure 5-4 CR-3 equivalent subgrade modulus after reconstruction. ............................... 18
Figure 5-5 TH-23 base modulus. ...................................................................................... 19
Figure 5-6 TH-23 equivalent subgrade modulus. ............................................................. 20
Figure 5-7 TH-200 base modulus. .................................................................................... 21
Figure 5-8 TH-200 Equivalent subgrade modulus............................................................ 21
Figure 5-9 Seasonal effect on Evercalc-generated base modulus.
Jan 1st (Day 0) – Dec 31st (Day 365), 1999..................................................... 22
Figure 5-10 Seasonal effect on Yonapave-generated equivalent subgrade modulus.
Jan 1st (Day 0) – Dec 31st (Day 365), 1999..................................................... 23
Figure B-1 Effect of damping ratio on a SDOF system response................................... B-2
Figure D-1 Comparison of static (FRF) and dynamic (P2P) deflection basins,
Case 4............................................................................................................ D-1
Figure D-2 Results of Evercalc backcalculation for the a) AC layer, b) base layer,
c) subgrade layer, d) stiff layer (if present)................................................... D-3
Figure E-1 Mn/ROAD Mainline Cell #2. Seasonal effect on Evercalc -generated base
modulus Jan 1st (Day 0) – Dec 31st (Day 365), 1999.....................................E-1
Figure E-2 Mn/ROAD Mainline Cell #2. Seasonal effect on Yonapave-generated
equivalent subgrade modulus. Jan 1st (Day 0) – Dec 31st (Day 365), 1999...E-1
Figure E-3 Mn/ROAD Mainline Cell #21. Seasonal effect on Evercalc-generated base
modulus Jan 1st (Day 0) – Dec 31st (Day 365), 1999.....................................E-2
Figure E-4 Mn/ROAD Mainline Cell #21. Seasonal effect on Yonapave-generated
equivalent subgrade modulus. Jan 1st (Day 0) – Dec 31st (Day 365), 1999...E-2
Figure E-5 Mn/ROAD Low-volume road Cell #31. Seasonal effect on Yonapave-
generated equivalent subgrade modulus.
Jan 1st (Day 0) – Dec 31st (Day 365), 1999....................................................E-3
 LIST OF TABLES

Table 2.1 Project selection criteria...................................................................................... 5

Table 3.1 Data collected from recycling projects in Minnesota. ........................................ 6
Table 4.1 Yonapave coefficients for determining characteristic length, l0. ...................... 15
Table 4.2 Yonapave curve fitting constants for determining equivalent subgrade
modulus, Esg.................................................................................................... 15
Table 5.1 CR-3 modulus comparison. .............................................................................. 16
Table 5.2 TH-23 Modulus Comparison. ........................................................................... 19
Table 5.3 TH-200 modulus comparison. .......................................................................... 20
Table C.1 Minnesota recycling projects selected for further review. ............................. C-1
Table D.1 Information on Synthetic Layer Profiles........................................................ D-1
 EXECUTIVE SUMMARY

The primary objective of this study was to quantify stiffness (resilient modulus) of
aggregate base containing recycled asphalt and concrete pavements. This was
accomplished by (1) reviewing other states’ specifications related to implementation of
the recycled materials in pavement design and (2) using the statewide (cities, counties
and Mn/DOT) testing data, such as falling weight deflectometer (FWD), to develop
resilient modulus values that can be used as input for Mn/PAVE and the Design Guide.
A brief survey of neighboring states, as well as states that frequently use recycled
materials, revealed that the use of resilient modulus in design is limited, though
considerable research is being conducted. Thus, only rehabilitation projects in the state
of Minnesota were selected for study. The projects were County State Aid Highway 3,
Trunk Highway 23, and Trunk Highway 200. These projects were selected based on the
availability of laboratory results of resilient modulus and field measurements from
FWD. Information regarding pavement thickness and actual design was also collected.
Based on the results of a parametric study, it was found that traditional peak-based
analysis of FWD data can lead to significant errors in elastostatic backcalculation due to
dynamic effects. A procedure for extracting the static response from the dynamic test
data by using the frequency-response-functions (FRFs) of the pavement was
formulated. The analysis was simplified by the development of a software package
called GopherCalc.

Field measurements in the form of FWD testing on rehabilitated pavements were

analyzed and compared with resilient modulus values measured in the laboratory. By
using conventional (peak-based) and modified (FRF-based) elastostatic backcalculation
with Evercalc, as well as a simplified mechanistic-empirical model called Yonapave,
representative values of base layer moduli were determined. These values were then
compared with laboratory resilient modulus values obtained from the NCHRP 1-28A
protocol. The laboratory values selected were from sequences in the protocol that
produced a state-of-stress similar to that imparted by FWD loading. In addition, FWD
data from the MN/Road facility was analyzed to determine the seasonal behavior of base
material. Three sections (Low-volume Section 31, Mainline Sections 2 & 21) were
examined for the year 1999. It was found that the base material exhibited a considerable
increase in stiffness from the thawed months to frozen months, with modulus values
changing from approximately 120 MPa to 760 MPa (17,500 psi to 110,500psi). Surface
deflections in the frozen state were so small as to prevent accurate elastostatic recovery of
the base moduli. For Cells 2 and 21, Evercalc analysis recovered a modulus of 120 MPa
(17,500psi) for both sections, while Yonapave returned values of 100 MPa and 130 MPa
(14,500 psi and 19,000 psi), respectively. Cell 31, analyzed with Yonapave, exhibited a
base modulus of 110 MPa (16,000 psi).
 Chapter 1
Introduction

Owing to the increasing scarcity of economical virgin aggregate supplies and increasing
fuel costs, attention is now being focused on the use of recycled concrete pavements,
reclaimed bituminous pavements, and other poured concrete as replacement materials in
unbound pavement base applications. The practice of using recycled concrete aggregate
(RCA) and reclaimed asphalt pavement (RAP) in unbound base has been in place for
some time. Research on the mechanical and chemical properties of these recycled
materials following their field removal and processing has been extensively conducted
for use as aggregate in bound applications. In unbound applications, recycled materials
have been shown to perform in a similar manner to virgin aggregates in a number of
laboratory comparison studies [1].

Although the pavement behavior is strongly influenced by the base and subgrade material
characteristics, sufficient sampling and laboratory testing of the base materials is
problematic due to the spatial variability of the materials and large number of tests that
would be required over a typical reclamation project. It is therefore advantageous to
determine an appropriate measure of base layer stiffness, such as resilient modulus, from
a simple, non-destructive field pavement test. One such testing device that is commonly
used in pavement evaluation is the Falling-Weight Deflectometer (FWD). In an FWD
test, a load is applied to a pavement by dropping a weight from a specified height onto a
buffered loading plate directly in contact with the pavement surface. Using an array of
sensors, the pavement response near the load can be extracted and used to infer stiffness
properties of the pavement layers. In this study, the stiffness of unbound aggregate bases
containing recycled materials will be examined using FWD field test data and laboratory
resilient modulus tests.

1.1 Background
In the area of pavement recycling, full-depth reclamation is a technique in which the
existing pavement surface course along with a portion of the aggregate base are
uniformly blended and compacted as a rehabilitated base course. Full-depth reclamation
provides an economical alterative to total reconstruction, owing to the conservation of
existing aggregate resources along with reduced material hauling and waste disposal
costs. Due to the substantial contribution of the base layer to overall pavement quality
and longevity, accurate determination of the material properties of this blended material
is central to proper pavement design.

1.1.1 Recycled Concrete Aggregate (RCA)

The specifications regarding the use of recycled materials vary widely from state to state.
In the Federal Highway Administration’s State of Practice National Review from
September 2004, a survey was conducted to determine the amount and type of recycled

1
material application. The five states with the greatest amount of recycled material
experience were further surveyed in the FHWA report: Texas, Virginia, Michigan,
Minnesota and California. Of these states, all strongly encourage the use of RCA in
unbound pavement base courses. It is generally acknowledged that the large amount of
fines present in RCA can lead to workability and dust-control issues in the field, so it is
commonly recommended that the material be placed and compacted at or near the
saturation point to ensure the proper dispersion of fines and a reduction in dust.
Minnesota and Michigan expressed considerable concern over the excess fines in RCA
clogging drains. With regard to the field compaction of this material, it is also
recommended that steel-drum rollers should be used, as small amounts of metal scrap can
be problematic for rubber-tire type rollers.

1.1.2 Reclaimed Asphalt Pavement (RAP)

The use of RAP in both bound and unbound applications is also quite common. Several
of the states reviewed place a limitation on the percentage of RAP, ranging anywhere
from 0.5 to 25% by weight. This is also accomplished by limiting the amount of asphalt
binder in the mix to below 3.0% by weight. As noted in [1] and [2], unbound aggregate
bases may have a greater potential for significant permanent deformation. A study by
Molenaar [3] stated the most important factor controlling the stiffness and strength
properties in unbound recycled materials is the degree of compaction. From a field study
performed by Garg [4] some difficulty was experienced with field compaction of the
material using both vibratory drum and sheep-foot rollers.

1.2 Objectives
The motivation of this study is to improve the quality of pavement design by examining
the relationship between laboratory measurements made on the base materials and field
measurements conducted on the in-place pavement sections. In this context, laboratory
testing was carried out on base course aggregates containing recycled materials in a
concurrent project entitled Resilient Modulus and Strength of Base Course with Recycled
Bituminous Material, following the National Cooperative Highway Research Program
(NCHRP) 1-28A test protocol to obtain a value of resilient modulus. Resilient modulus
(MR) is similar to a secant Young’s modulus based on recoverable axial strain, Δεar, from
an induced cyclic axial stress, Δσa:
Δσ a
MR = (1.1)
Δε ar
The cyclic axial loading applied to a cylindrical specimen within a conventional triaxial
cell is designed to simulate traffic loading. The protocol specifies that the specimen
should be tested at several different levels of axial (deviator) stress and confining
pressures. The samples were a collection of in-place blended material from reclamation
projects within the state of Minnesota.

Field data, in the form of falling-weight deflectometer (FWD) measurements, were

performed on these sections. The FWD data was interpreted in the framework of

2
elastostatic back-analysis, which produces elastic moduli for each layer of the pavement
system. Comparison of the laboratory and field values was accomplished by selecting
sequences from the 1-28A test protocol for which the confining and deviator stresses
were similar to those imparted by FWD loading.

1.3 Organization
Chapter 2 discusses the project selection procedure and identification of recycling
projects suitable for this investigation within the State of Minnesota. Chapter 3 presents
the data collection procedure and describes the information available from each of the
selected projects. Chapter 4 covers the analysis for the field data. Chapter 5 describes
the results, including a comparison with the laboratory findings. Chapter 6 summarizes
and concludes the findings of the research.

3
 Chapter 2
Project Information

Based on information obtained by a survey conducted in conjunction with the LRRB 808
study, 15 projects were selected from a total of 117 for further review based on the
amount of information that was returned by the survey participants. This group was
further narrowed based on the quality and type of information submitted. Special
consideration was given to projects for which FWD data were available. This was done
in order to relate the projects to those that have the laboratory determination of Mr and
shear strength namely, Wright CSAH 3, TH23, and TH200. Table 2.1 displays the list of
projects, including those selected for further review.

4
 Table 2.1 Project selection criteria.

Project Information Original Design

Project # Year of AASHTO Class Soil

City/County/State Hwy R-Value
Const Type

6928-22 Duluth, MN TH 73 1977 Silt Loam 29

3107-27 TH6 to Effie TH1 1955 clay/silty clay 10
5804-51 Pine County TH 48 loamy fine sand na

0110-29 0.1 mi S of CSAH 2 to TH 65 1939 loamy sand and gravel 20

0.1 mi N of CSAH 4
Pre-Rehab Analysis Rehab Information
CIR/ Depth
Project # PSR Year of Emulsi
Date SR PQI FDR/ of Mill/ Oil Type
(Ride) Rehab on %
M&O/ Reclai
6928-22 1993 2.3 2.6 2.4 1995 CIR 4 N/A
3107-27 1995 1.9 1.5 1.7 1997 FDR 6.5 58-28
5804-51 1998 2.1 3 2.5 1999 FDR 12
0110-29 2003 2.2 2 2.1 2005 FDR 14

5
 Chapter 3
Data Collection

To determine a proper correlation between laboratory and field values of resilient

modulus suitable for this climate, projects within Minnesota (and neighboring states)
were selected that could provide data for analysis. Data were collected on three projects
that were recently reconstructed using recycled materials in Minnesota: County State Aid
Highway (CSAH) 3 in Wright County, Trunk Highway (TH) 23 in District 1, and TH 200
in District 4.

3.1 Minnesota Reclaiming Projects

Table 3.1 outlines the data collected on the projects listed above.

Table 3.1 Data collected from recycling projects in Minnesota.

Project Data Type Description Date
CSAH-3
Values of Mr were obtained from a previous project for
the in-situ blend, and the following blends of
Resilient Modulus January, 2007
RAP/Virgin Aggregate: 0/100, 25/75, 50/50, and
75/25.

FWD Data FWD peak values prior to rehabilitation. May 10, 2004
FWD Data FWD peak values after rehabilitation June 10, 2005
FWD Data FWD truncated time-histories after rehabilitation. June 10, 2005
Layer thickness information from Ground Penetrating
GPR Data May 15, 2004
Radar prior to rehabilitation

Original Construction Data Layer thicknesses and material types from GPR report May 15, 2004

TH-23 Values of Mr for the in-situ blend of recycled material

Resilient Modulus January, 2007
from a previous project.

Reclaim depths by station, traffic forecast, existing

Design Recommendations April 30, 2002
pavement, soils and R-value

Obtained via fax from the district materials engineer

Typical Section February 22, 2007
(Rod Garver)

FWD Data FWD peak values prior to rehabilitation October 19, 2000
FWD Data FWD peak values after rehabilitation. July 2, 2002
808 Database SR Data (2 years) 2002, 2004
TH-200 Values of Mr for the in-situ blend of recycled material
Resilient Modulus January, 2007
from a previous project.

Obtained via email from the district materials engineer

Pavement Thickness February 26, 2007
(Perry Collins)

September 1, 2004
FWD Data FWD peak values after rehabilitation. (2 sets)
July 28, 2005
808 Database SR Data (10 years) 1989-2004

6
3.2 Recycled Materials Survey
To determine the proper implementation of resilient modulus for recycled materials used
in pavement design for Minnesota, a short survey was submitted to the DOT’s of
neighboring states (North Dakota, South Dakota, Iowa, Wisconsin, and Michigan) to
determine the state-of-practice for this climate region. In addition, California and
Virginia were contacted due to their widespread use of recycled materials, as outlined in
the FHWA State-of-Practice Review, September 2004.

The survey requested the following information:

1. Resilient moduli and shear strength values used in design for virgin aggregate
bases, with seasonal adjustments;
2. Resilient moduli and shear strength values used in design for aggregate bases
containing recycled materials, with seasonal adjustments;
3. Recycled material implementation guidelines and any additional information
regarding successful/unsuccessful projects using recycled materials.

The responses will be summarized by state. The detailed communications can be found
in Appendix A.

Michigan
Aggregate bases containing recycled materials are treated the same as virgin aggregates.
For dense-graded bases, a resilient modulus of 200 MPa (30,000 psi) is used and assumed
to be seasonally adjusted. Recycled materials are not allowed in open-graded bases, or
where geotextiles or unfiltered underdrains are used. Michigan does not use resilient
modulus for open-graded virgin bases.

Michigan does not have any implementation guidelines. Experience suggests that most
successful recycling projects from the late 1980’s and early 1990’s were recycled
concrete with asphalt stabilization. These pavements outperformed the unbound (virgin)
aggregate projects from the same area.

South Dakota
Virgin aggregates are assigned a resilient modulus value of 145 MPa (21,000 psi).
Alternatively, a Structural Number (SN) is defined for this material as 0.1 per inch of
material. Aggregate bases containing recycled materials have a resilient modulus of 200
MPa (30,000 psi) or an SN of 0.14 per inch of material.

South Dakota has been using recycled materials in aggregate bases for over 20 years.
Reconstruction of any existing pavement involves salvaging all of the in-place asphalt
and granular base for reuse. The target blend is 50% granular/50% recycled asphalt
material, but in the field, a blend of 40% granular/60% recycled asphalt is allowed. They
do not allow Portland Cement Concrete (PCC) pavement as salvaged base below asphalt
pavements. Recycled PCC can be used below concrete pavements, provided it meets the
criteria for “Gravel Cushion”.

7
California
The California design method is currently an empirical method similar to the 1993
AASHTO method using a gravel factor characteristic of the strength of the material.
California is currently developing a mechanistic-empirical method that will utilize
resilient modulus and shear strength.

The California DOT (Caltrans) allows up to 100% substitution of recycled or reclaimed

materials. They are currently developing specifications for cold foam in-place recycling
and pulverization. In general, the target mixture is 30% base material/70% recycled
asphalt for cold-foamed projects.

Wisconsin
Resilient modulus values are not used in pavement design, although research is in
progress.

North Dakota
Resilient modulus values are not used in pavement design.

Iowa
Iowa DOT did not respond to the survey.

Virginia
Virginia DOT did not respond to the survey.

8
 Chapter 4
Data Analysis

4.1 FWD Data Analysis

In the area of nondestructive pavement testing, the FWD test has become a standard way
of characterizing subsurface properties. Interpretation of these results to obtain the
properties of the pavement, base, and subgrade is the focus of considerable research in
areas such as dynamic and static layered elastic and visco-elastic modeling, finite element
analysis, neural networks, experimental or empirical correlations, and laboratory testing.
In this study, the data obtained from the FWD test will be interpreted using an elastostatic
back-calculation method (Evercalc), as well as a simplified mechanistic-empirical model
(Yonapave).

Due in large part to its simplicity, elastostatic back-calculation remains the norm in
estimating the mechanical properties of the pavement layers. Using information on layer
thicknesses, assumed or calculated Poisson’s ratios, and initial or seed moduli values, the
backcalculation procedure mimics the deflection basin obtained from the test by varying
the input to an elastostatic forward model until a proper fit of surface deflection profiles
is achieved.

To improve the accuracy and reliability of FWD data interpretation, a modification of the
existing back-calculation procedure should be executed to remedy a fundamental
inconsistency in the back-calculation input. Traditionally, the peak values of deflection
together with the corresponding peak value of force are used to describe the deflection
basin. These peak values are obtained by dropping a weight from a specified height onto
the buffered loading plate of the FWD. These events, and the peak values that are
generated, are dynamic in nature. The problem arises of performing a dynamic test and
using its dynamic peak values as an input to elastostatic back-calculation. This issue is
especially significant in the case of shallow stiff layer, wherein the contribution of
dynamic effects to surface displacement can be significant. As in Figure 4-1, this
phenomenon can cause substantial errors in recovery of layer moduli through
backcalculation if not properly treated. One such treatment involves using frequency
response functions to extract the static response of the pavement from the dynamic test
data. Shown in Figure 4-2, using the static pavement response greatly reduces the
potential for serious error in backcalculation. A parametric study investigating the effect
of layer thickness, base moduli, and stiff layers on the dynamic (peak) and static (FRF)
backcalculation procedures is discussed in Appendix D.

9
Figure 4-1 Error in base modulus using peak-based elastostatic backcalculation (175 test
cases).

Figure 4-2 Error in base modulus using FRF-based backcalculation (175 test cases).

10
In what follows, a procedure to extract the static response of the pavement from the
dynamic FWD test data using the concept of a frequency-response-function (FRF) will be
presented and discussed as a way to elevate traditional elastostatic backcalculation
procedures. Due to the large amount of data generated by the FWD test, a graphical user
interface, GopherCalc, was developed to expedite the data analysis procedure.

4.1.1 GopherCalc
GopherCalc is a graphical user interface developed using MATLAB. The purpose of the
program is to facilitate both traditional and modified elastostatic back-analysis of
pavement properties and provide a convenient tool to visualize the time- and frequency-
domain signatures of the FWD test. It allows a user to: 1) perform automated averaging
of device-stored data records, 2) apply and visualize a proper baseline correction, 3)
compute and extrapolate frequency-domain response, 4) compare peak-based and
extrapolated-static deflection basins and, 5) export deflection basin data into a form that
is compatible with external elastostatic back-calculation programs.

Using GopherCalc to extract the static basin involves three major steps: 1) baseline
correction, 2) calculation of a frequency-response-function (FRF) and, 3) low-frequency
extrapolation. To perform this analysis, the full time history of the test is required. The
time-history should contain the initial pulse from the loading as well as the free-
vibrations that follow. Performing this procedure using these longer records is essential
to avoid potentially serious errors associated with signal truncation.

Baseline Correction
In an FWD test, geophones record the pavement surface velocity over the duration of the
test. These velocity records are then integrated to obtain the displacements at each
geophone location. Random noise inherent to the transducers and data collection system
is accumulated during this integration resulting in a non-zero displacement at the end of
the record known as baseline offset. While this noise is typically not significant in terms
of peak-based methods, it can lead to significant errors in the frequency-based
interpretation. It is therefore necessary to account for this non-zero displacement with a
proper baseline correction. The effect of baseline correction can be seen in Figure 4-3.
Note the non-zero drift known as baseline offset, Figure 4-3a, and significant reduction in
FRF noise seen in Figure 4-3d.

11
Figure 4-3(a-d) Effect of baseline correction on frequency response functions, a) Original
time-history b) original FRF c) baseline-corrected time-history d) baseline corrected FRF.

For a given geophone displacement record w(t) and corresponding time-history t, the
general form for the baseline correction at time ti is:
(
wbc (ti ) = w ( ti ) −
)
w ( t f ) − w ( t0 )
( ti )
n
(4.1)
( t f − t0 )
n

where n is the power of the baseline correction. In most cases, a linear baseline
correction (n=1) is a proper choice and results in a ”rotation” of the record to achieve the

12
desired zero-offset. The power of the baseline correction is dependent upon the amount
of noise present in the record. Application of the baseline correction is implemented
automatically in the program.

Frequency Response Function

Once the record has been treated with a proper baseline correction, the next step in
extracting the static response is to perform a frequency domain analysis by using a
Fourier transform. Since the signal from an FWD device is of finite duration and
digitized, a discrete Fourier transform
M
⎡⎣G ( f m ) ⎤⎦ = Δt ∑ g ( t j )e
− i ( 2π f m ) t j
, m = 0,1, 2,..., M (4.2)
j =0

is applied to the time-domain record, g(tj), to find its frequency-domain counterpart,

G(fm). This procedure is commonly and efficiently implemented by means of a Fast
Fourier Transform (FFT) in many computer applications. In the context of FWD test
analysis, this FFT procedure is applied to the falling-weight force, q(t), and the deflection
record for each of the N geophones, wk(t) (k=1,2,…,N) to obtain the frequency-domain
representations Q(f) and Wk(f) respectively. Once these records have been transformed, it
is possible to compute the frequency response function (FRF) for the kth geophone as
W (f )
FRFk ( f m ) = k m . (4.3)
Q( f m )
This resulting function represents deflection per unit force at each frequency of excitation,
which is essential to extracting the static response.

When multiple drops are available at a particular test point, the records should be
combined by using the cross-spectral, Sqk and auto-spectral Sqq density functions:
1 T
S qk ( f m ) = ∑ ⎡⎣Q* ( f m ) ⎤⎦ ⎡⎣Wk ( f m ) ⎤⎦ i , m = 0,1, 2,..., M (4.4)
T i =1 i

1 T
S qq ( f m ) = ∑ ⎡⎣Q* ( f m ) ⎤⎦ ⎡⎣Q ( f m ) ⎤⎦ i , m = 0,1, 2,..., M (4.5)
T i =1 i

where ‘*’ denotes the complex conjugate. The FRF is then defined as
S (f )
FRFk ( f m ) = qk m , k = 1, 2,..., N . (4.6)
Sqq ( f m )
It is useful to note that the FRF is defined at each frequency m in the record. The static
response of the system is then given, by definition, as the value of the FRF at a frequency
of zero.

Low-Frequency Extrapolation
Due to the physical construction of a geophone, the data in the lowest frequency range
(<10Hz) is inherently characterized by a poor signal-to-noise ratio and is thus deemed
unreliable for the extraction of the zero-frequency (static) response of the system. It is
therefore necessary to develop an extraction scheme anchored in a frequency range with
better signal-to-noise ratios and extrapolate through the noise polluted region. For FWD
applications, a single-degree-of-freedom (SDOF) model provides a stable, consistent

13
means for dealing with this low frequency noise. Additionally, the SDOF model
provides a proper analog to the physical behavior of the pavements in the low frequency
ranges, manifest in its ability to capture resonant peaks within the fit range and remain
stable when extrapolated towards zero.

f(t)

m x(t)

k c

Figure 4-4 Schematic of a single-degree-of-freedom (SDOF) system.

The governing equation for the motion of a SDOF model consisting of a mass m, a spring
with spring constant k, and a dashpot with damping constant c, can be formulated as
1
k k c
FRFSDOF ( f m ) = , ω0 = ,ξ = (4.7)
⎛ ⎛ω ⎞ ⎞ ⎛2 2 m 2 km
ω⎞
⎜1 − ⎜ ⎟ ⎟ + ⎜ 2ξ
⎜ ⎝ ω0 ⎠ ⎟ ⎝ ω0 ⎟⎠
⎝ ⎠
Using this representation, along with a proper choice of initial values based on the
characteristics of the geophone record being fit, the zero-frequency values can be
recovered despite the limitations imposed by the geophone.

4.1.2 Yonapave
As is often the case in pavement evaluation, accurate layer thicknesses are not always
available. This can have a significant effect on the layer moduli determined from a multi-
layer back-calculation, particularly when the peak values of force and deflection are used.
A way to circumvent this lack of reliable data is to use a methodology that generates an
equivalent subgrade modulus that encompasses both the aggregate base and in-place
subgrade material. One such method, Yonapave [5] is based on the Hogg model of a thin
slab resting on an elastic foundation. By using this representation of the pavement
structure and determining the properties of the elastic layer based on the deflections
obtained, a suitable equivalent modulus can be recovered.

The Yonapave method uses the basic relationships of the Hogg model together with the
MODULUS program to generate curves from which modulus values can be inferred
based on the characteristics of an individual deflection basin. The “area” of a deflection
basin can be determined by using the deflections at 0, 30, 60, and 90 cm from the load
plate by

14
 ⎛ D D D ⎞
Area = 6 ⎜ 1 + 2 30 + 2 60 + 90 ⎟ [inches ] (4.1)
⎝ D0 D0 D0 ⎠
Yonapave suggests that the ratio of the actual depth to the bedrock or stiff layer, h, and
the characteristic length is related to this deflection basin area. The relationship between
the characteristic length, l0, and the area can then be defined as
l0 = A × e B× Area (4.2)
where the values of A and B are determined from curve fitting and are found in Table 4.1
Yonapave coefficients for determining characteristic length, l0.

Table 4.1 Yonapave coefficients for determining characteristic length, l0.

Range of Area
h/l0 A B
Values, inch
Area ≥ 23.0 5 3.275 0.1039
21.0 ≤ Area < 23.0 10 3.691 0.0948
19.0 ≤ Area < 21.0 20 2.800 0.1044
Area < 19.0 40 2.371 0.1096

Based on this characteristic length, the equivalent subgrade modulus Esg in units of MPa
can be determined as a function of the load plate pressure p in units of kPa, center
deflection D0 and an additional set of curve fitting coefficients found in Table 4.2
Yonapave curve fitting constants for determining equivalent subgrade modulus, Esg. as
p
Esg = m × × l0n (4.3)
D0

Table 4.2 Yonapave curve fitting constants for determining equivalent subgrade modulus,
Esg.

h/l0 m n
5 926.9 -0.8595
10 1152.1 -0.8782
20 1277.6 -0.8867
40 1344.2 -0.8945

The Yonapave algorithm presents further procedure for determining the AASHTO
Structural Number (SN), but for the purpose of comparison with other methods in this
report, those methods have been omitted.

15
 Chapter 5
Results and Discussion

The FWD data analyzed following the methodology previously presented will be
discussed in detail and compared with laboratory resilient modulus values.

5.1 Wright County State Aid Highway 3

FWD and Ground-Penetrating-Radar (GPR) testing was performed on the roadway prior
to its reconstruction. During reconstruction, field sampling of the virgin aggregate,
reclaimed asphalt, and in-situ blend created from the recycling process was conducted.
Using these materials, laboratory resilient modulus tests were conducted following
NCHRP protocol 1-28A. Once the rehabilitation was completed, FWD testing was again
performed on the completed roadway. Truncated (60 ms) time-histories were acquired
during this testing procedure. Although full length (120 ms) time-histories are preferable,
a brief investigation from Mn/ROAD data suggests that the value of the error in the fit
range to be less than 5%. While this error is fairly low, its effect on the extrapolation
procedure and subsequent elastostatic backcalculation can not be reliably determined.

Using the methods described in the previous section, the results from data obtained from
CR-3 are presented in Table 5.1. As a basis for comparison, the data obtained from the
field will be compared with similar cycles in the resilient modulus test protocol.

Table 5.1 CR-3 modulus comparison.

Before Reconstruction After Reconstruction

CR-3 Material (Virgin Aggregate) (Unbound RAP)
MPa (psi) MPa (psi)
FWD (Peak) 70 (10,000) 192 (28,000)
FWD (FRF) n/a 177 (25,500)
YONAPAVE 89 (13,000) 152 (22,000)
Mr (σ3 = 21 kPa) 124 (18,000) 154 (22,000)
Mr (σ3 = 41 kPa) 181 (26,500) 230 (33,500)

In the FWD data taken prior to reconstruction, there is no record of the exact station or
mile-post of the test. As a result, the spatial relationship of the data can not be directly
determined. The base moduli generated from Evercalc are presented in Figure 5-1 and
Figure 5-2. The equivalent subgrade modulus calculated using Yonapave is presented in
Figure 5-3 and Figure 5-4.

16
 500

400

300
Eba [MPa]

Peak
200

100

0
0 10 20 30 40 50 60
Test Number

Figure 5-1 CR-3 base modulus prior to reconstruction.

500

400

300
Eba [MPa]

Peak
FRF
200

100

0
0 5 10 15 20 25 30
Test Number

Figure 5-2 CR-3 base modulus after reconstruction.

17
 250

200

150
Esg [MPa]

Before

100

50

0
0 10 20 30 40 50
Test Number

Figure 5-3 CR-3 equivalent subgrade modulus before reconstruction.

250

200

150
Esg [MPa]

After

100

50

0
0 5 10 15 20 25 30
Test Number

Figure 5-4 CR-3 equivalent subgrade modulus after reconstruction.

18
5.2 TH-23
FWD test data (peak values) were available from before and after the reconstruction. A
portion of the in-situ blended material generated by full-depth reclamation was tested in
the laboratory, following the 1-28A protocol. GPR data was unavailable for this project,
and thus asphalt layer thickness could not be accurately determined. Since layer
thickness plays a significant role in elastostatic backcalculation, some anomalous results
were obtained. Based on backcalculation performed with design layer thicknesses, only
stations with a base modulus of less than 500 MPa (72,500 psi) and an asphalt modulus
of between 2-9 GPa (290-1300 ksi) were considered. These criteria were sufficient to
prevent obviously erroneous results produced by backcalculation. For the purpose of
comparison, the average value of base or equivalent modulus is compared with in Table
5.2 with resilient modulus values with a similar stress state to FWD loading. The spatial
variation of the Evercalc and Yonapave analysis can be found in Figure 5-5 and Figure
5-6.

Table 5.2 TH-23 Modulus Comparison.

Before Reconstruction After Reconstruction
TH-23 Material (Virgin Aggregate) (Unbound RAP)
MPa (psi) MPa (psi)
FWD (Peak) 79 (11,500) 97 (14,000)
YONAPAVE 70 (10,000) 93 (13,500)
Mr (σ3 = 21 kPa) n/a 149 (21,500)
Mr (σ3 = 41 kPa) n/a 236 (34,000)

500

400

300
Eba [MPa]

Before
After
200

100

0
315.000 320.000 325.000 330.000 335.000
Station

Figure 5-5 TH-23 base modulus.

19
 250.0

200.0

150.0
Esg [MPa]

Before
After
100.0

50.0

0.0
315.000 320.000 325.000 330.000 335.000
Station

Figure 5-6 TH-23 equivalent subgrade modulus.

5.3 TH-200
FWD test data (peak values) were available from two years after the reconstruction took
place (2004, 2005). The in-situ blended material was tested in the laboratory.
Backcalculation was performed with design layer thicknesses. Though moduli returned
from Evercalc backcalculation (Figure 5-7) are higher than returned by Yonapave (Figure
5-8) or laboratory experiments (Table 5.3), the values are consistent from 2004-2005.
Accurate layer thicknesses at the drop locations would most likely provide moduli values
more similar to those found by other methods.

Table 5.3 TH-200 modulus comparison.

2004 Data 2005 Data

TH-200 Material (Unbound RAP) (Unbound RAP)
MPa (psi) MPa (psi)
FWD (Peak) 325 (47,000) 392 (57,000)
YONAPAVE 106 (15,500) 110 (16,000)
Mr (σ3 = 21 kPa) n/a 109 (16,000)
Mr (σ3 = 41 kPa) n/a 210 (30,500)

20
 2000
1800

1600
1400
1200
Eba [MPa]

2004
1000
2005
800
600
400

200
0
41.000 42.000 43.000 44.000 45.000 46.000 47.000
Station

Figure 5-7 TH-200 base modulus.

250

200

150
Esg [MPa]

2004
2005
100

50

0
41.000 42.000 43.000 44.000 45.000 46.000 47.000
Station

Figure 5-8 TH-200 Equivalent subgrade modulus.

21
5.4 Seasonal Effects
The influence of seasonal changes on FWD-backcalculated resilient moduli was
investigated using data from the Mn/ROAD facility. Low-volume road Section 31,
Mainline Section 2, and Mainline Section 21 were chosen based on the pavement profile
and availability of sufficient FWD data. In this study, seasonal effects were studied for
the year 1999. The FWD data from Sections 2 and 21 was evaluated using both Evercalc
to obtain a base modulus and Yonapave to obtain an equivalent subgrade modulus.
Section 31 was analyzed using Yonapave only. For each test section, the pavement is
tested at 10 locations. At each location, there are two initial drops to seat the loading
plate against the pavement. Following the seating drops, the pavement is subjected to
three drops at each of three different drop heights. A time-history is saved from the last
drop at each drop height, for each location. Variation over all locations in one section is
small compared to the variation between different testing dates. For clarity, the average
value of modulus over all locations is presented for each station in Figure 5-9 (Evercalc)
and Figure 5-10 (Yonapave). It should be noted that in the Evercalc backcalculation
(Figure 5-9), deflections for the beginning of the year (Days 26 & 53) were too small to
produce reasonable backcalculation results and were thus omitted from the plot. These
small deflections suggest a considerably stiffer behavior than the rest of the year, such as
is evident in the Yonapave calculation (Figure 5-10). Plots of the moduli obtained for
individual locations at each station are available in Appendix E.

240

220

200
Eba [MPa]

180
Cell #2
160 Cell #21

140

120

100

80
0 30 60 90 120 150 180 210 240 270 300 330 360
Day

Figure 5-9 Seasonal effect on Evercalc-generated base modulus. Jan 1st (Day 0) – Dec
31st (Day 365), 1999.

22
 900

800

700

600
Esg [MPa]

500 Cell #2
Cell #21
400
Cell #31
300

200

100

0
0 30 60 90 120 150 180 210 240 270 300 330 360
Day

Figure 5-10 Seasonal effect on Yonapave-generated equivalent subgrade modulus. Jan 1st
(Day 0) – Dec 31st (Day 365), 1999.

23
 Chapter 6
Summary and Conclusions

The primary objective of this study was to quantify stiffness (resilient modulus) of
aggregate base containing recycled asphalt and concrete pavements. This was
accomplished by (1) reviewing other state’s specifications related to implementation of
the recycled materials in pavement design and (2) using the statewide (cities, counties
and Mn/DOT) testing data, such as falling weight deflectometer (FWD), to develop
resilient modulus values that can be used as input for Mn/PAVE and the Design Guide.

A brief survey of neighboring states, as well as states that frequently use recycled
materials, revealed that the use of resilient modulus in design is limited, though
considerable research is being conducted. Thus, only rehabilitation projects in the state
of Minnesota were selected for study. The projects were County State Aid Highway 3,
Trunk Highway 23, and Trunk Highway 200. These projects were selected based on the
availability of laboratory results of resilient modulus and field measurements from
FWD. Information regarding pavement thickness and actual design was also collected.

Based on the results of a parametric study, it was found that traditional peak-based
analysis of FWD data can lead to significant errors in elastostatic backcalculation due to
dynamic effects. A procedure for extracting the static response from the dynamic test
data by using the frequency-response-functions (FRFs) of the pavement was
formulated. The analysis was simplified by the development of a software package
called GopherCalc.

Field measurements in the form of FWD testing on rehabilitated pavements were

analyzed and compared with resilient modulus values measured in the laboratory. By
using conventional (peak-based) and modified (FRF-based) elastostatic backcalculation
with Evercalc, as well as a simplified mechanistic-empirical model called Yonapave,
representative values of base layer moduli were determined. These values were then
compared with laboratory resilient modulus values obtained from the NCHRP 1-28A
protocol. The laboratory values selected were from sequences in the protocol that
produced a state-of-stress similar to that imparted by FWD loading.

In addition, FWD data from the MN/Road facility was analyzed to determine the seasonal
behavior of base material. Three sections (Low-volume Section 31, Mainline Sections 2
& 21) were examined for the year 1999. It was found that the base material exhibits a
considerable increase in stiffness from the thawed months (approximately 120 MPa,
17,500 psi) to frozen months (approximately 760 MPa, 110,500 psi). Surface deflections
in the frozen state were so small as to prevent accurate elastostatic recovery of the base
moduli. For Cells 2 and 21, Evercalc analysis recovered a modulus of 120 MPa (17,500
psi) for both sections, while Yonapave returned values of 100 MPa and 130 MPa (14,500

24
psi and 19,000 psi), respectively. Cell 31, analyzed with Yonapave, predicted a base
modulus of 110 MPa (16,000).

25
 REFERENCES
[1] W. Kim, and J.F. Labuz. (2007) Resilient Modulus and Strength of Base Course
with Recycled Bituminous Material. St. Paul, MN (Minnesota Department of
Transportation, Final Report 2007-05.)

[2] T. Bennert, W.J. Papp Jr., A. Maher, and N. Gucunski. (2000). Utilization of
Construction and Demolition Debris Under Traffic-Type Loading in Base and
Subbase Applications. Transportation Research Record, No. 1714,. 33-39.

[3] A.A.A Molenaar, and A.A. van-Niekerk. (2002). Effects of Gradation,

Composition, and Degree of Compaction on the Mechanical Characteristics of
Recycled Unbound Materials. Transportation Research Record, No. 1787, 73-82.

[4] N. Garg, and M.R. Thompson. (1996). Lincoln Avenue Reclaimed Asphalt
Pavement Base Project. Transportation Research Record, No. 1547, 89-95.

[5] M.S. Hoffman. (2003). Direct Method for Evaluating Structural Needs of Flexible
Pavements Based on FWD Deflections. Transportation Research Record, No.
1860 ,41-47.

[6] B.B Guzina, and S. Nintcheu. (2002). A Study of Ground-Structure Interaction in

Dynamic Plate Load Testing. International Journal for Numerical and Analytical
Methods in Geomechanics, Vol. 26, pp 1147-1166.

[7] B.B. Guzina, and R.Y.S. Pak. (2001). On the Analysis of Wave Motions in a
Multi-Layered Solid. Quarterly Journal of Mechanics and Applied Mathematics,
Vol. 40, pp 13-37.

[8] WSDOT Pavement Guide: EVERCALC (1995). Seattle, WA (Washington State

Department of Transportation.)

[9] R.W. Meier, and G.J. Rix. (1995). Backcalculation of Flexible Pavement Moduli
From Dynamic Deflection Basins Using Artificial Neural Networks.
Transportation Research Record, No. 1473, pp 72-81.

26
 Appendix A

DOT Survey Responses

Michigan

Mr. Westover,

Curtis Bleech asked me to respond to your questions:

1. Resilient moduli and shear strength values used in design for virgin

aggregate bases, with seasonal adjustments;

We use a resilient modulus of 30,000 psi for dense-graded aggregate

bases with no seasonal adjustment (assumed to already be seasonally
adjusted). We do not have resilient modulus for our open-graded bases
since AASHTO concrete design does not require it.

2. Resilient moduli and shear strength values used in design for

aggregate bases containing recycled materials, again with seasonal
adjustments;

We do not have separate value for bases utilizing recycled materials.

They are not allowed in open-graded bases, and are restricted in
dense-graded bases to areas where there is no geotextile or unfiltered
underdrains.

3. Recycled material implementation guidelines and any additional

information regarding successful/unsuccessful projects using recycled
materials.

We do not have any implementation guidelines. We do have some success

using recycled concrete as a base for concrete pavements when it is
asphalt stabilized. This is a recent observation of some concrete
pavements built in the late 80's/early 90's. They are performing much
better (almost crack-free) than the unbound projects from the same area
(that utilized virgin aggregate).

Michael Eacker
Pavement Design Engineer
Construction and Technology Support Area
Michigan Department of Transportation
(517) 322-3474
[email protected]

A-1
South Dakota

Responses are shown below after your questions. We do not place a shear
strength value on our granular materials.

We currently do not accept Portland Cement Concrete Pavement as a

portion of our salvaged bases beneath asphalt concrete pavement
sections, but are currently in the middle of a research project that is
studying the use of this material. Any reference in the questions below
will refer to a salvaged base material consisting of a blend of
granular
base & asphalt mix material.

Recycled concrete pavement is allowed beneath new concrete pavement

sections provided it meets the requirements in Section 260 of our
Standard Specifications for the new material specified on the project,
which is usually Gravel Cushion.

Thanks and please respond if you need additional information for your
survey.
Gill L Hedman
Pavement Design Engineer
605-773-5503
[email protected]

-----Original Message-----
From: Thomas Westover [mailto:[email protected]]
Sent: Tuesday, October 17, 2006 2:26 PM
To: Hedman, Gill
Subject: Resilient Modulus of Recycled Base Materials

Dear Mr. Hedman,

As a graduate student at the University of Minnesota, my research is

focused on the development of seasonal resilient moduli values for
aggregate bases containing recycled bituminous and concrete pavements.
In addition to resilient modulus and FWD testing conducted at the
University and the Minnesota Department of Transportation (MnDOT), our
project requires us to ask DOT's from neighboring states how properties
of recycled materials are accounted for in pavement design.

Please provide (any or all of) the following information by the end of
November 2006:

1. Resilient moduli and shear strength values used in design for virgin
aggregate bases, with seasonal adjustments;

We use a value of @ 21,000 psi for the Resilient Modulus

(Structural Number of 0.1 per inch of material) for of a Base
Material in the 1993 version of the Pavement Design Guide.

2. Resilient moduli and shear strength values used in design for

aggregate bases containing recycled materials, again with seasonal
adjustments;

We use a value of 30,000 psi for the Resilient Modulus

A-2
(Structural Number of 0.14 per inch of material) for of a Salvaged
Base Material in the 1993 version of the Pavement Design Guide.

3. Recycled material implementation guidelines and any additional

information regarding successful/unsuccessful projects using recycled
materials.

We have been successfully using salvaged asphalt and granular

base materials in our pavement structure for over 20 years. When we
reconstruct road segments we salvage all in place asphalt and
granular base for reuse as a portion of the total base structure
needed on these roads. We target for a blend of 50% granular &
50%
asphalt material in this base material but will accept a blend
of 40% granular & 60% asphalt material in the finished product.

Section 270 of our Standard Specifications details our salvaging

and stockpiling process and Section 260 details the placement and
compaction for our salvaged bases.

Thank you in advance for your assistance. I would be happy to send you
the final report, which should be completed by April 2007, if you are
interested.

Sincerely,

Tom Westover

--
Thomas Westover
Department of Civil Engineering
University of Minnesota
500 Pillsbury Drive S.E. Rm 350A
Minneapolis, MN 55455
612.626.1538

A-3
California
Forwarded by Imad Basheer/HQ/Caltrans/CAGov on 10/23/2006 12:14 PM

Imad
Basheer/HQ/Caltra
ns/CAGov To

10/23/2006 12:14 cc
PM Bill
Farnbach/D03/Caltrans/CAGov@DOT,
Terrie
Bressette/HQ/Caltrans/CAGov@DOT,
Robert Hogan/HQ/Caltrans/CAGov@DOT
Subject Re:
Resilient Modulus of Recycled Base Materials(Document link: Imad Basheer)

Dear Mr. Westover,

Our response to your questions is below. Let me know if you have other
questions.

Q1+Q2 (Response by Basheer). Currently, we do not use resilient modulus or

shear strength. Instead, California uses an empirical method for design of
new flexible pavements similar to the 1993 AASHTO method in which a gravel
factor reflecting the strength of the material to be used in the pavement
is used. If you would like more information about California design method
check out this web link http://www.dot.ca.gov/hq/oppd/hdm/pdf/chp0630.pdf
California is currently in the process of developing M-E method in which
there will be greater use of resilient modulus for characterizing the
strength of various pavement materials.

Q3 (Response by Hogan). Caltrans currently allows up to 100% substitution

of recycled or reclaimed materials in new aggregate base and subbase.
Additionally, we have experimented with in-place material recycling in the
form of pulverization and cold foam in-place recycling and are currently
developing guidelines and specifications to standardized those processes.
For in-place recycling projects, we typically look for structurally
inadequate roadways with advanced distress, such as severe fatigue
cracking, that may indicate deteriorated base or would otherwise need a
thick overlay or deep milling. We have generally had good success with
pulverization and cold foam, however there have been some problems with
subgrade failures and concerns about the moisture susceptibility of cold
foamed material. As with any in-place recycling process, material control
is limited and field adjustments are required as conditions such as in-situ
moisture content and pavement thickness vary. In general, we strive for a
well graded mixture of 70% asphalt concrete and 30% base material with cold
foam and at least 1 inch of base material for pulverization.

Dear Terrie,

As a graduate student at the University of Minnesota, my research is

focused on the development of seasonal resilient moduli values for
aggregate bases containing recycled bituminous and concrete pavements.
In addition to resilient modulus and FWD testing conducted at the
University and the Minnesota Department of Transportation (MnDOT), our
project requires us to ask DOT’s from neighboring states how properties
of recycled materials are accounted for in pavement design.

A-4
Please provide (any or all of) the following information by the end of
November 2006:

1. Resilient moduli and shear strength values used in design for virgin
aggregate bases, with seasonal adjustments;

2. Resilient moduli and shear strength values used in design for

aggregate bases containing recycled materials, again with seasonal
adjustments;

3. Recycled material implementation guidelines and any additional

information regarding successful/unsuccessful projects using recycled
materials.

Thank you in advance for your assistance. I would be happy to send you
the final report, which should be completed by April 2007, if you are
interested.

Sincerely,

Tom Westover

--
Thomas Westover
Department of Civil Engineering
University of Minnesota
500 Pillsbury Drive S.E. Rm 350A
Minneapolis, MN 55455
612.626.1538

A-5
Wisconsin
I'm sorry I did not respond sooner. We've done testing toward this
effort,
but I don't yet have the information you are looking for.

-----Original Message-----
From: Thomas Westover [mailto:[email protected]]
Sent: Tuesday, October 17, 2006 2:21 PM
To: Fenley, Laura
Subject: Resilient Modulus of Recycled Base Materials

Dear Ms. Fenley,

As a graduate student at the University of Minnesota, my research is

focused on the development of seasonal resilient moduli values for
aggregate bases containing recycled bituminous and concrete pavements.
In addition to resilient modulus and FWD testing conducted at the
University and the Minnesota Department of Transportation (MnDOT), our
project requires us to ask DOT’s from neighboring states how properties
of recycled materials are accounted for in pavement design.

Please provide (any or all of) the following information by the end of
November 2006:

1. Resilient moduli and shear strength values used in design for virgin
aggregate bases, with seasonal adjustments;

2. Resilient moduli and shear strength values used in design for

aggregate bases containing recycled materials, again with seasonal
adjustments;

3. Recycled material implementation guidelines and any additional

information regarding successful/unsuccessful projects using recycled
materials.

Thank you in advance for your assistance. I would be happy to send you
the final report, which should be completed by April 2007, if you are
interested.

Sincerely,

Tom Westover

--
Thomas Westover
Department of Civil Engineering
University of Minnesota
500 Pillsbury Drive S.E. Rm 350A
Minneapolis, MN 55455
612.626.1538

A-6
North Dakota

We do not perform resilient moduli testing on bases in our pavement

designs.

-----Original Message-----
From: Thomas Westover [mailto:[email protected]]
Sent: Tuesday, October 17, 2006 2:24 PM
To: Horner, Ron J.
Subject: Resilient Modulus of Recycled Base Materials

Dear Mr. Horner,

As a graduate student at the University of Minnesota, my research is

focused on the development of seasonal resilient moduli values for
aggregate bases containing recycled bituminous and concrete pavements.
In addition to resilient modulus and FWD testing conducted at the
University and the Minnesota Department of Transportation (MnDOT), our
project requires us to ask DOT's from neighboring states how properties
of recycled materials are accounted for in pavement design.

Please provide (any or all of) the following information by the end of
November 2006:

1. Resilient moduli and shear strength values used in design for virgin
aggregate bases, with seasonal adjustments;

2. Resilient moduli and shear strength values used in design for

aggregate bases containing recycled materials, again with seasonal
adjustments;

3. Recycled material implementation guidelines and any additional

information regarding successful/unsuccessful projects using recycled
materials.

Thank you in advance for your assistance. I would be happy to send you
the final report, which should be completed by April 2007, if you are
interested.

Sincerely,

Tom Westover

--
Thomas Westover
Department of Civil Engineering
University of Minnesota
500 Pillsbury Drive S.E. Rm 350A
Minneapolis, MN 55455
612.626.1538

A-7
 Appendix B

SDOF Initial Conditions

In a typical backcalculation procedure, it is assumed that the layered pavement system is
laterally homogeneous over the length of the FWD sensor array. As a result, the
geophone records from a single FWD test will have similar characteristics in both
frequency- and time-domain signatures. Features of the record, such as resonance peaks
in the frequency record, will be less pronounced for distances further away from the load.
It is therefore necessary to fit each geophone record individually to ensure accurate
extrapolation with the single-degree-of-freedom (SDOF) model. To ensure a reliable and
efficient fit, a procedure is outlined for selecting the initial conditions from each
geophone.

Initial Spring Stiffness

To determine the initial spring stiffness, ki, recall the SDOF fitting equation used:
1
k k c
FRFSDOF ( f m ) = , ω0 = ,ξ = (B.1)
⎛ ⎛ ω ⎞2 ⎞ ⎛
2 m 2 km
ω⎞
⎜1 − ⎜ ⎟ ⎟ + ⎜ 2ξ
⎜ ⎝ ω0 ⎠ ⎟ ⎝ ω0 ⎟⎠
⎝ ⎠

The denominator of this equation approaches one as the circular frequency ω→0 so the
static value approaches 1/k. Selecting the initial value for k using the value of the FRF at
10Hz provides a reasonable approximation while avoiding the noise dominated behavior
typical of the lower frequencies. The expression for the initial spring constant then
becomes
1
ki = (B.2)
FRF (10 Hz )

Initial Damping

Using the results of numerical simulations, it was determined that the ability of a SDOF
system to adequately capture the low-frequency resonance peak, if any, is primarily
dependent upon the proper choice of damping ratio. For records where a resonant peak is
prominent in the fit range, the SDOF system is highly sensitive to the choice of damping
ratio, and the half-power method provides a methodology for determining an initial value.
Let
FRF max
FRF reduced
= (B.3)
2
where FRFmax is the maximum value of the geophone record. Let fL and fR be the
frequencies to the left and right of the maximum whose value is FRFreduced. An
equivalent damping ratio can then be approximated as
f −f
ξi = R L (B.4)
2 f max
This approximation will not yield reasonable values for peaks where fL and fR can not be
readily determined. Based on experience, for peaks within the fit range that are not as
pronounced, a damping ratio of 0.3 was found to provide reliable convergence. In the

B-1
case where there is no peak or the function varies monotonically over the fit range, a
damping ratio of 0.8 is recommended. The effect of damping ratio on the SDOF system
can be seen in Figure B-1

Figure B-1 Effect of damping ratio on a SDOF system response.

Initial Mass

Following the proper selection of spring and damping parameters, an value for the initial
mass can be determined. Using the expression

mi =
(
ki 1 − 2 ( ξ i )
2

ξi < 2
), (B.5)
2π f max
ensures the location of the SDOF peak and field peak are located at the same frequency.
For values of ξi>√2, experience suggests that a value for mi that is sufficiently small (e.g.
1*10^-5) will provide reasonable fits for the suggested fit range around 10-20 Hz.

Implementation in GopherCalc

All of the initial condition selection procedures outlined here are implemented
automatically in GopherCalc. The procedure should be applied to each geophone record
individually to ensure that the SDOF system will fit each geophones particular character.

B-2
 Appendix C

Additional Project Selection

Using information contained in the 808 Database, a short survey of the projects was
performed. The purpose was to locate projects within the database which could provide a
further basis for comparison. The projects listed in this table were chosen based on the
amount of data available.

Table C.1 Minnesota recycling projects selected for further review.

Project Information Original Design

Project # Year of
City/County/State Hwy AASHTO Class Soil Type R-Value
Const

0114-09 Duluth, MN TH 18
0912-23 Duluth, MN TH 73 slpl SL; swamps

6928-22 Duluth, MN TH 73 1977 Silt Loam 29

3107-27 TH6 to Effie TH1 1955 clay/silty clay 10

5803-34 Askov to Nickerson TH 23

0901-72 S.Carlton Co Line to Duluth TH 23 clay

.9 mi N of CSAH 8 to 0.1
0901-66 TH 23 clay
mi N of CSAH

Pine County: CSAH15 to

5804-49 TH 48 1976 loamy fine sand na
state line

5804-51 Pine County TH 48 loamy fine sand na

5804-52 Pine County TH 48

0.1 mi S of CSAH 2 to 0.1
0110-29 TH 65 1939 loamy sand and gravel 20
mi N of CSAH 4

3609-29 TH 65 1939 loamy sand and gravel

W of Moose Lake to Kettle
0903-26 TH 27
River

6927-17 Prairie Lake to Floodwood TH 73 silty loam, sandy loam

6935-85 TH 169

Pre-Rehab Analysis Rehab Information

CIR/ Depth
Project # PSR Year of FDR/ of Emulsio Oil
Date SR PQI
(Ride) Rehab M&O/ Mill/Rec n% Type
Ov laim

0114-09 1995 CIR

0912-23 1994 2 2.3 2.1 CIR 4 3

6928-22 1993 2.3 2.6 2.4 1995 CIR 4 N/A

3107-27 1995 1.9 1.5 1.7 1997 FDR 6.5 58-28

5803-34 1994 2.2-2.4 2.4-2.5 2.3-2.4 1998 FDR 12 to 14

0901-72 2000 1.5-3.4 1.9-4.0 1.7-3.6 2002 FDR 12 to 14

0901-66 2000 1.5-3.4 1.9-4.0 1.7-3.6 2001 FDR 8 to 12

5804-49 1996 2.3 3.2 2.7 1998 FDR 12 52-34

5804-51 1998 2.1 3 2.5 1999 FDR 12

5804-52 2000 FDR

0110-29 2003 2.2 2 2.1 2005 FDR 14

3609-29 2000 FDR 6 52-34

0903-26 1998 1.9 2.2 2 2000 FDR

6927-17 1996 2.4 2.5 2.4 2000 FDR 12 52-34

C-1
 Appendix D

FWD Parametric Study

To investigate the conditions where peak-based (P2P) backcalculated pavement and base
layer elastic moduli differ significantly from their frequency-based (FRF) counterparts, a
parametric study was performed. The study used an elastodynamic FWD forward model
[6] that utilizes the method of propagator matrices, the Fourier transform, and the Hankel
integral transform [7]. A set of six synthetic test profiles were used to examine the
behavior of three and four layer layered systems. See Table D.1 for a description of the
inputs to the forward calculation process. Using noise-polluted time-histories and the
corresponding frequency-domain representations, deflection basins were computed. As
shown in Figure D-1, the contribution of dynamic effects produces a decidedly larger
deflection at most geophone locations, particularly close to the load plate.

Table D.1 Information on Synthetic Layer Profiles.

Case Young's Modulus [MPa] Thickness [m]
Number AC Base Subbase Stiff AC Base Subbase Stiff
1 2700 216 112 N/A 0.1 0.3 N/A ∞
2 2700 54 28 N/A 0.1 0.3 N/A ∞
3 2700 216 112 1160 0.1 0.3 5 ∞
4 2700 54 28 580 0.1 0.3 5 ∞
5 2700 28 N/A 580 0.15 3 N/A ∞
6 2700 270 112 N/A 0.1 0.3 N/A ∞
Case Mass [kg/m3] Poisson's ratio
Number AC Base Subbase Stiff AC Base Subbase Stiff
1 2335 2027 1865 N/A 0.35 0.35 0.4 N/A
2 2335 2027 1865 N/A 0.35 0.35 0.4 N/A
3 2335 2027 1865 2160 0.35 0.35 0.4 0.45
4 2335 2027 1865 2160 0.35 0.35 0.4 0.45
5 2335 1865 N/A 2160 0.35 0.4 N/A 0.45
6 2335 2027 1865 N/A 0.35 0.35 0.4 N/A

SPACING FROM CENTER [cm]

0 50 100 150 200
0
NORMALIZED DEFLECTION [um/kN]

10

15

20
FRF
25
P2P
30

35

40

45

50

Figure D-1 Comparison of static (FRF) and dynamic (P2P) deflection basins, Case 4.

D-1
Using these deflection basins as an input, backcalculation was performed using Evercalc.
In the procedure, the layer thickness, Poisson’s’ ratios, and seed moduli were identical for
both P2P and FRF deflection basins. With the true (input) modulus for reference, the
results of backcalculation are shown in Figure D-2a-d. The dynamic effects of the FWD
test are evident in the backcalculated moduli. In particular, a well-known shortcoming of
P2P data interpretation is faulty results in the presence of a shallow stiff layer. As
demonstrated in Cases 3, 4, and 5, the backcalculation compensates for the dynamic
effects by overestimating the modulus of the AC layer, while leaving the stiff layer
virtually undetected. In contrast, the FRF method consistently provides a more accurate
representation of the pavement moduli. In Case 4, the FRF method fails to fully recover
the stiff layer moduli with the same consistency as other layers in other cases. This may
be a limitation of the FWD testing configuration, as FWD deflection basins are not
significantly affected by stiff layers located at depths greater than 3 meters [9].

While further field study would be necessary to further validate the results from synthetic
data, the FRF analysis method presented produces an input that is more consistent with
the fundamental assumptions underlying elastostatic backcalculation. Using this simple
but effective modification has the potential to elevate the accuracy and reliability of
current backcalculation schemes in a significant way. The frequency domain analysis
requires no modification to existing FWD test procedure, as long as the full 120 ms time-
history of the test is collected.

D-2
Figure D-2 Results of Evercalc backcalculation for the a) AC layer, b) base layer, c)
subgrade layer, d) stiff layer (if present).

D-3
 Appendix E

Seasonal Effects
This section will present the results of the two backcalculation procedures in more detail.
Each value for each location is an average of the backcalculated moduli from each of the
three drop heights at that location.

Loc. 1
260
Loc. 2
Loc. 3
210 Loc. 4
Eba [MPa]

Loc. 5
Loc. 6
160
Loc. 7
Loc. 8
110 Loc. 9
Loc. 10

60
0 30 60 90 120 150 180 210 240 270 300 330 360
Day

Figure E-1 Mn/ROAD Mainline Cell #2. Seasonal effect on Evercalc -generated base
modulus Jan 1st (Day 0) – Dec 31st (Day 365), 1999.

900

800 Loc. 1
700 Loc. 2
Loc. 3
600
Loc. 4
Esg [MPa]

500 Loc. 5
400 Loc. 6
Loc. 7
300
Loc. 8
200 Loc. 9
100 Loc. 10

0
0 30 60 90 120 150 180 210 240 270 300 330 360
Day

Figure E-2 Mn/ROAD Mainline Cell #2. Seasonal effect on Yonapave-generated

equivalent subgrade modulus. Jan 1st (Day 0) – Dec 31st (Day 365), 1999.

E-1
 Loc. 1
260
Loc. 2
Loc. 3
210 Loc. 4
Eba [MPa]

Loc. 5
Loc. 6
160
Loc. 7
Loc. 8
110 Loc. 9
Loc. 10

60
0 30 60 90 120 150 180 210 240 270 300 330 360
Day

Figure E-3 Mn/ROAD Mainline Cell #21. Seasonal effect on Evercalc-generated base
modulus Jan 1st (Day 0) – Dec 31st (Day 365), 1999.

900

800 Loc. 1
700 Loc. 2
Loc. 3
600
Loc. 4
Esg [MPa]

500 Loc. 5
400 Loc. 6
Loc. 7
300
Loc. 8
200 Loc. 9
100 Loc. 10

0
0 30 60 90 120 150 180 210 240 270 300 330 360
Day

Figure E-4 Mn/ROAD Mainline Cell #21. Seasonal effect on Yonapave-generated

equivalent subgrade modulus. Jan 1st (Day 0) – Dec 31st (Day 365), 1999.

E-2
 200
180 Loc. 1
160 Loc. 2
140 Loc. 3
120 Loc. 4
Esg [MPa]

Loc. 5
100
Loc. 6
80 Loc. 7
60 Loc. 8
40 Loc. 9

20 Loc. 10

0
0 30 60 90 120 150 180 210 240 270 300 330 360
Day

Figure E-5 Mn/ROAD Low-volume road Cell #31. Seasonal effect on Yonapave-
generated equivalent subgrade modulus. Jan 1st (Day 0) – Dec 31st (Day 365), 1999.

E-3