Structural Evaluation of Asphalt

Pavements with Full-Depth

Reclaimed Base

Joseph F. Labuz, Principal Investigator

Department of Civil Engineering
University of Minnesota

December 2012
Research Project
Final Report 2012-36
To request this document in an alternative format, please contact the Affirmative Action Office
at 651-366-4723 or 1-800-657-3774 (Greater Minnesota); 711 or 1-800-627-3529 (Minnesota
Relay). You may also send an e-mail to ADArequest.dot@state.mn.us.

(Please request at least one week in advance).

 Technical Report Documentation Page
1. Report No. 2. 3. Recipients Accession No.
MN/RC 2012-36
4. Title and Subtitle 5. Report Date
Structural Evaluation of Asphalt Pavements with Full-Depth December 2012
Reclaimed Base 6.

7. Author(s) 8. Performing Organization Report No.

Shuling Tang, Yuejian Cao, and Joseph F. Labuz
9. Performing Organization Name and Address 10. Project/Task/Work Unit No.
Department of Civil Engineering CTS #2009083
University of Minnesota 11. Contract (C) or Grant (G) No.
500 Pillsbury Drive SE
Minneapolis, MN 55455 (C) 89261 (WO) 156
12. Sponsoring Organization Name and Address 13. Type of Report and Period Covered
Minnesota Department of Transportation Final Report
Research Services 14. Sponsoring Agency Code
395 John Ireland Blvd., MS 330
St. Paul, MN 55155
15. Supplementary Notes
http://www.lrrb.org/pdf/201236.pdf
16. Abstract (Limit: 250 words)

Currently, MnDOT pavement design recommends granular equivalency, GE = 1.0 for non-stabilized full-depth
reclamation (FDR) material, which is equivalent to class 5 material. For stabilized full-depth reclamation (SFDR),
there was no guideline for GE at the time this project was initiated (2009). Some local engineers believe that GE of
FDR material should be greater than 1.0 (Class 5), especially for SFDR. In addition, very little information is
available on seasonal effects on FDR base, especially on SFDR base. Because it is known from laboratory studies
that SFDR contains less moisture and has higher stiffness (modulus) than aggregate base, it is assumed that SFDR
should be less susceptible to springtime thawing.
Falling Weight Deflectometer (FWD) tests were performed on seven selected test sections on county roads in
Minnesota over a period of three years. During spring thaw of each year, FWD testing was conducted daily during
the first week of thawing in an attempt to capture spring thaw weakening of the aggregate base. After the spring
thaw period, FWD testing was conducted monthly to study base recovery and stiffness changes through the
seasons.
GE of SFDR was estimated using a method established by MnDOT using FWD deflections, and the GE of SFDR is
about 1.5. The value varies from project to project as construction and material varies from project to project. All
the materials tested showed seasonal effects on stiffness. In general, the stiffness is weaker in spring than that in
summer and fall.
17. Document Analysis/Descriptors 18. Availability Statement
Full-depth reclamation, Stabilized full-depth reclamation, No restrictions. Document available from:
Falling weight deflectometers, Backcalculation, Granular National Technical Information Services,
equivalency Alexandria, Virginia 22312
19. Security Class (this report) 20. Security Class (this page) 21. No. of Pages 22. Price
Unclassified Unclassified 53
 Structural Evaluation of Asphalt Pavements with
Full-Depth Reclaimed Base

Final Report

Prepared by:

Shuling Tang
Yuejian Cao
Joseph F. Labuz

Department of Civil Engineering

University of Minnesota

December 2012

Published by:

Minnesota Department of Transportation

Research Services
395 John Ireland Boulevard, MS 330
St. Paul, Minnesota 55155

This report documents the results of research conducted by the authors and does not necessarily represent the views
or policies of the Minnesota Department of Transportation or the University of Minnesota. This report does not
contain a standard or specified technique.

The authors, the Minnesota Department of Transportation, and the University of Minnesota do not endorse products
or manufacturers. Trade or manufacturers’ names appear herein solely because they are considered essential to this
report.
 Table of Contents
Chapter 1. Literature Review ................................................................................................... 1
1.1 Background ...................................................................................................................... 1
1.2 Recycling Methods........................................................................................................... 1
1.3 Overview of Full Depth Reclamation .............................................................................. 2
1.4 Stabilizing Methods.......................................................................................................... 3
1.4.1 Chemical Stabilizers ................................................................................................. 3
1.4.2 Bituminous Stabilizers .............................................................................................. 4
1.4.3 Nontraditional Stabilizers ......................................................................................... 4
1.5 Case History 1 .................................................................................................................. 5
1.6 Case History 2 .................................................................................................................. 7
1.7 Case History 3 .................................................................................................................. 7
1.8 Summary .......................................................................................................................... 8

Chapter 2. Analysis Methods .................................................................................................... 9

2.1 The AASHTO Method and Structural Number ............................................................... 9
2.2 The Minnesota Method and Granular Equivalency ....................................................... 11
2.3 Backcalculation and Falling Weight Deflectometer ...................................................... 14

Chapter 3. Selected Sections for Field Testing ...................................................................... 15

Chapter 4. Analysis Results..................................................................................................... 22

4.1 MnROAD Cell 21, Class 5 Base Analysis ..................................................................... 22
4.2 LeSueur County Road 2 ................................................................................................. 22
4.3 LeSueur County Road 13 ............................................................................................... 24
4.4 Pope County Road 28 ..................................................................................................... 26
4.5 Pope County Road 29 ..................................................................................................... 29
4.6 Goodhue County Road 30 Eastern Section .................................................................... 30
4.7 Goodhue County Road 30 Western Section................................................................... 33
4.8 Olmsted County Road 13 ............................................................................................... 34
4.9 Modulus Values Summary ............................................................................................. 37
 4.10 Method Based on Hogg Model ...................................................................................... 39

Chapter 5. Summary and Recommendations........................................................................ 42

References ................................................................................................................................. 43

2
 List of Figures
Figure 1: Minnesota Method Design Chart [23] .......................................................................... 12
Figure 2: LeSueur County Road Sections: a. CR 2; b. Road 13 .................................................. 15
Figure 3: Pope County Road Sections: a. CR 28; b. CR 29......................................................... 15
Figure 4: Goodhue East and West County Road 30 Sections: a. West; b. East........................... 16
Figure 5: Olmsted County Road 13 Section ................................................................................ 16
Figure 6: GPR Equipment............................................................................................................ 16
Figure 7: Example of Base Stiffness Changes during the Year ................................................... 18
Figure 8: Goodhue County Road 30 Pavement Layer Thickness from GPR Survey .................. 18
Figure 9: LeSeuer County Road 2 Pavement Layer Thickness from GPR Survey ..................... 19
Figure 10: LeSeuer County Road 13 Pavement Layer Thickness from GPR Survey ................. 19
Figure 11: Olmsted County Road 13 Pavement Layer Thickness from GPR Survey ................. 20
Figure 12: Pope County Road 29 Pavement Layer Thickness from GPR Survey....................... 20
Figure 13: Pope County Road 28 Pavement Layer Thickness from GPR Survey....................... 21
Figure 14: Young’s Modulus Values for LeSueur CR2 in 2011 ................................................. 23
Figure 15: Young’s Modulus values for LeSueur CR2 in 2010 .................................................. 24
Figure 16: Young’s Modulus values for LeSueur CR2 in 2009 .................................................. 24
Figure 17: Young’s Modulus Values for LeSueur CR 13 in 2011 .............................................. 25
Figure 18: Young’s Modulus Values for LeSueur CR 13 in 2010 .............................................. 26
Figure 19: Young’s Modulus Values for LeSueur CR 13 in 2009 .............................................. 26
Figure 20: Young’s Modulus Values for Pope CR28 in 2011 ..................................................... 27
Figure 21: Young’s Modulus Values for Pope CR28 in 2010 ..................................................... 28
Figure 22: Young’s Modulus Values for Pope CR 28 in 2009 .................................................... 28
Figure 23: Young’s Modulus Values for Pope CR 29 in 2011 .................................................... 29
Figure 24: Young’s Modulus Values for Pope CR 29 in 2010 .................................................... 30
Figure 25: Young’s Modulus Values for Pope CR 29 in 2009 .................................................... 30
Figure 26: Young’s Modulus Values for Goodhue CR 30 East in 2011 ..................................... 31
Figure 27: Young’s Modulus Values for Goodhue CR 30 East in 2010 ..................................... 32
Figure 28: Young’s Modulus Values for Goodhue CR 30 East in 2009 ..................................... 32
Figure 29: Young’s Modulus Values for Goodhue CR 30 West in 2011 .................................... 33

3
Figure 30: Young’s Modulus Values for Goodhue CR30 West in 2010 ..................................... 34
Figure 31: Young’s Modulus Values for Goodhue CR 30 West in 2009 .................................... 34
Figure 32: Young’s Modulus Values for Olmsted CR13 in 2011 ............................................... 35
Figure 33: Young’s Modulus Values for Olmsted CR13 in 2010 ............................................... 36
Figure 34: Young’s Modulus Values for Olmsted CR 13 in 2009 .............................................. 36
Figure 35: Modulus Plot Summary with Stiff Layer (2009)........................................................ 37
Figure 36: Modulus Plot Summary without Stiff Layer (2010) .................................................. 37
Figure 37: Modulus Plot Summary with Stiff Layer (2011)........................................................ 38
Figure 38: Modulus Ratios........................................................................................................... 38
Figure 39 2011 GE Summary Plot ............................................................................................... 40
Figure 40: 2010 GE Summary Plot.............................................................................................. 40
Figure 41: 2009 GE Summary Plot.............................................................................................. 41

4
 List of Tables
Table 1: Subgrade Resilient Modulus Results ............................................................................... 5
Table 2: FWD Data ........................................................................................................................ 6
Table 3: Foamed Asphalt Base Layer Coefficients ....................................................................... 7
Table 4: GE Factors ..................................................................................................................... 14
Table 5: FWD Testing Schedule .................................................................................................. 17
Table 6: Young’s Modulus Values in ksi (1000 lb/in.2) for MnROAD Cell 21 .......................... 22
Table 7: Young’s Modulus Values in ksi (1000 lb/in.2) for LeSueur CR 2................................. 23
Table 8: Young’s Modulus Values in ksi (1000 lb/in.2) for LeSueur CR 13............................... 25
Table 9: Young’s Modulus Values in ksi (1000 lb/in.2) for CR 28 ............................................. 27
Table 10: Young’s Modulus Values in ksi (1000 lb/in.2) for Pope CR29 ................................... 29
Table 11: Young’s Modulus Values in ksi (1000 lb/in.2) for CR 30 East ................................... 31
Table 12: Young’s Modulus Values for Goodhue CR 30 West .................................................. 33
Table 13: Young’s Modulus Values in ksi (1000 lb/in.2) for Olmsted County Road 13 ............. 35

5
 Chapter 1. Literature Review
Full-depth reclamation (FDR) is a recycling technique where the existing asphalt pavement and a
predetermined portion of the underlying granular material are blended to produce an improved
base course. FDR is an attractive alternative in road rehabilitation: resources are conserved, and
material and transportation costs are reduced as recycling eliminates the need for purchasing and
hauling new materials and disposing of old materials. An additive is sometimes used, and this
process is referred to as stabilized full-depth reclamation (SFDR).

Two approaches in pavement design involve the Structural Number and Granular Equivalency.
The Structural Number, which is used widely throughout the United States, has been applied in
many FDR projects. The Granular Equivalency (GE), which is popular in Minnesota, has no
known reclaimed asphalt pavement designs. In this work, GE will be evaluated for both
stabilized and standard FDR sections through Falling Weight Deflectometer (FWD) testing.

1.1 Background
It is inevitable that, over time, asphalt pavements degrade due to a variety of reasons including
thermal cracking, traffic loading, or poor construction. In past years, common methods to
rehabilitate failed asphalt pavements were to either apply hot-mix asphalt (HMA) overlays or to
perform complete reconstruction of the pavement section. However, to fully reconstruct a
pavement is expensive and time consuming, and although the overlay method is fast and less
expensive, it does not always provide a long lasting solution. With overlays, previous distresses
and cracks eventually reflect up to the new layer of pavement, thus requiring further
rehabilitation. In the past few decades, in-place asphalt recycling has become cost-effective and
has gained popularity. The Asphalt Recycling and Reclaiming Association has categorized
recycling into five methods: cold planning, hot recycling, hot-in-place recycling, cold recycling,
and full-depth reclamation [1].

1.2 Recycling Methods

Both cold and hot recycling can be performed on-site or off-site at a central facility. There are
advantages and disadvantages for in-situ recycling, which will be briefly discussed in this
section.

For hot-in-place recycling, a vehicle train with four components is used. The first two units in
the vehicle soften the existing asphalt. Then the next unit mills the softened pavement, adds a
rejuvenator, and then “windrows” the mix. Last, a mixing unit combines the rejuvenated blend
with a virgin material and then agitates the combination in a pugmill. After the final mix is
discharged, conventional methods of HMA paving proceed. Because all material is mixed and
blended on-site, no material must be transported. However, the equipment used for hot-in-place
recycling is expensive and specialized [2].

In hot recycling, the pavement is milled and then transported to a central facility where the
material is heated and blended into a recycled mix and then hauled back to be re-used. Although
the equipment is more common and less expensive, higher transportation costs are incurred.

1
An obvious benefit of cold recycling is that energy is conserved by not heating the materials.
There are two methods of cold recycling commonly used: cold planing and cold-in-place
recycling. During cold planing, the existing asphalt course is cold milled to achieve a specified
vertical profile. Any surface distresses and irregularities are removed, leaving a uniform surface.
The benefits of cold planing include an improved pavement cross-section, minimal traffic
interruption, and a low cost [1].

Cold in-place recycling is also performed by using specialized train of equipment. First, while
the existing pavement is milled, an asphalt or chemical is injected to achieve the proper
compaction moisture content. Next, the newly mixed pavement is profiled with a grader, and
then compacted with a vibratory roller. Last, a fresh surface is applied [3].

It is important to determine if the cause of pavement distress is structural. There are many
distresses, such as fatigue in wheel paths, rutting, and reflective cracking that can indicate
structural inadequacy [4]. The first four methods are very effective for fixing minor pavement
distresses, but do not address structural or base problems because only the top layer of
bituminous material is recycled [5]. The fifth method, full-depth reclamation (FDR) eliminates
more serious base and structural pavement issues by recycling the entire asphalt section and a
predetermined amount of granular base material.

1.3 Overview of Full Depth Reclamation

The advantages of FDR include, (a) improving the pavement structure without changing the
geometry of the road, and (b) restoring the pavement to its desired profile while eliminating
rutting, thermal cracking, and reflective cracking. Also, because FDR can include stabilizing the
base, frost susceptibility can be reduced [2]. It has been stated that FDR is 25-50% lower in
costs than conventional pavement rehabilitation efforts [6]. FDR is sustainable by preserving
virgin materials and preventing the disposal of used material, and air quality issues, such as dust
and smoke, are minimized due to the nature of the processes [2]. After the FDR process is
performed, a surface layer is applied.

The process for FDR is very similar to cold-in-place recycling, and often times the same
specialized vehicle is used for the two methods. There are five main steps in FDR: pulverization,
blending of materials, shaping, compaction, and application of the surface course [7]. In
addition, there are four types of operations used to reclaim pavement: multistep sequence, two
step sequence, single machine, and single pass equipment [2].

In a multiple-step sequence, there are different machines used for each step. To mill the
pavement, a modified roadheader can be used, and to blend mixed materials, a rotary mixer can
be used. The advantages of the multiple-step sequence are that the equipment is readily
available and costs are relatively low. The disadvantages are that there is a lack of uniformity in
the depth of the cut, multiple passes are required to achieve the desired size of granular material,
the production rate is low, and the multiple machines can pose traffic control issues [2].

In a two-step sequence, a cold milling machine is used to grind existing pavement. Next, soil
stabilization mixing equipment is used to blend materials. The advantages of using the two-step
sequence are that the machines can accurately control the depth of removal, and they can

2
perform pulverizing and grading in a single pass, which results in less traffic interference. A
disadvantage is the possible production of oversized aggregates [2].

By using a single machine, the breaking, pulverizing, and blending of the pavement and
stabilizing agents can be performed in a single pass. The single machine operation is
advantageous because of the high production rate, but disadvantages include the possibility of
yielding oversized aggregates, a depth limitation for cutting, and the use of specialized
equipment [2]. The single-pass equipment train uses a special vehicle that first mills the existing
pavement, crushes the material to a desired gradation, adds stabilizing agents and blends the
mixed materials, and finally paves. There are many advantages to this operation, including a
high production rate, the ability to accurately remove the desired quantity of asphalt, and the
elimination of oversized particles. However, the equipment necessary is specialized and very
expensive [2].

1.4 Stabilizing Methods

As previously mentioned, a stabilizer may be added to increase the strength of the base. There
are three types of stabilizing additives: chemical, bituminous, and non-traditional, including
enzymes [5]. Compaction densifies the material and is sometimes considered as stabilization,
but in this work it will be called FDR, as the reclaimed material should always be compacted in
appropriate lifts prior to paving [8].

1.4.1 Chemical Stabilizers

Chemical stabilizers include Portland cement, hydrated lime, calcium chloride, and coal fly ash
[5]. Portland cement reacts with moisture in the soil, and the cementitious, hydration reaction
causes the particles to bond. Also, because the cementitious reaction continues with time, long
term strength improvement can be observed. In a study conducted by the Georgia Department of
Transportation using a cement treated base, it was concluded that approximately 6 in. (152 mm)
of cement treated base (CTB) was equivalent in strength and stiffness to 8 in. (203 mm) of a
crushed stone base. In addition, using FDR with a CTB resulted in a 42% reduction in costs
from the conventional rehabilitation methods [9].

Lime products used to stabilize soil include (1) quicklime, which is calcium oxide, (2) hydrated
lime, which is calcium hydroxide, and (3) lime slurry, which is hydrated lime and water [10].
Lime, however, requires a minimum amount of clay to react favorably. Lime stabilization
causes a significant improvement in soil texture and structure by reducing plasticity and by
providing pozzolanic strength gain. The pozzolanic reaction is the formation of calcium silicate
hydrates as the lime reacts with the aluminates and silicates in the clay minerals. The calcium
silicate hydrates produced in the reaction exhibit high strength and are the molecules responsible
for the strength gain in cementitious reactions. Like Portland cement, the pozzolanic reaction is
time dependent, and long-term strength improvement in lime stabilized soils and aggregates are
possible [11].

Both Portland cement and lime stabilized bases increase the shear strength of the base, decrease
permanent deformation, and reduce moisture susceptibility. However, shrinkage can often be a

3
problem. In addition, due to the high calcium contents in cement and lime, the materials are
subject to sulfate attack [8].

Unlike lime, coal fly ash does not require clay to react. However, the quantity required to
produce a similar outcome is three to four times that of lime [12]. Coal fly ash, which can often
be used as a Portland cement substitute, is a by-product of coal manufacturing. During
combustion, the fly ash becomes infused with inorganic particles. There are two classes of fly
ash. Class F fly ash comes from bituminous coals that have lower levels of calcium content.
Class F fly ash shows no self-cementing properties, and is used less often than Class C fly ash.
Class C fly ash is produced from sub-bituminous coal which has higher concentrations of
calcium carbonate and thus becomes self-cementing [13]. Also, because Class C fly ash exhibits
such rapid rates of hydration, retarders are often required in construction. Like cement and lime,
fly ash is also susceptible to sulfate attack due to the calcium concentrations.

Calcium chloride, another chemical stabilizer, is often used to control dust due to its hydrophilic
and deliquescent properties, meaning it absorbs water and then dissolves. This reaction can
produce high strength bonds [14]. Calcium chloride penetrates the aggregate in the base and
coats the particles and binds them together. Calcium chloride is an alkaline earth metal salt and
is most stable in liquid form with six molecules of water in its structure. However, it is also
commercially available in a dry, flake form. In addition, it can reduce the plasticity index (PI) of
a soil and improves workability while maintaining strength. Research has shown that calcium
chloride used along with Class F fly ash leads to high early strength [15].

1.4.2 Bituminous Stabilizers

Bituminous stabilizers commonly used are slow or medium asphalt emulsions, which can be
polymer modified. Asphalt emulsions work by reducing moisture susceptibility while
maintaining strength and providing flexibility. However, using an emulsion in moist soil can
increase the moisture content to detrimental levels [5]. In addition, asphalt emulsions take time to
cure.

Another popular stabilizer is foamed asphalt. In one study, it was determined that foamed
asphalt stabilized recycled asphalt pavement (RAP) outperformed RAP stabilized with asphalt
emulsion. Foamed asphalt is less expensive than asphalt emulsions and it exhibits rapid strength
gain [5, 16]. Foamed asphalt typically requires a higher percentage of fines, and cement is
sometimes added to meet this requirement.

Foamed asphalt is a mixture of air, water, and hot asphalt. Cold water is introduced to hot
asphalt, causing the asphalt to foam and expand by more than ten times its original volume.
During this foaming action, the asphalt has a reduced viscosity making it much easier to mix
with aggregates. Due to the immediate strength gain, traffic can operate on the stabilized base
until a hot mix asphalt base and wearing surface is applied [18].

1.4.3 Nontraditional Stabilizers

Although lime, cement, fly ash, and bituminous materials are sufficient for improving the
stability of granular materials, the cost has increased in recent years. The increased cost has
pushed companies to develop new, enzyme additives [19]. Scholen has categorized

4
nontraditional stabilizers into five categories: ionic, enzymatic, mineral filling, clay filling, and
polymer. Ionic stabilizers catalyze a cation exchange and flocculation of clay and soil particles.
Ionic stabilizers reduce plasticity and swell potential, which increase strength. In an enzymatic
reaction, the enzymes bond and are attracted to the large, negatively charged organic molecules
in the soil minerals [8, 20].

Stabilizers must be chosen with careful consideration of the base material to ensure proper
strength gain and limited sulfate attack. Kearney [5] summarizes general guidelines for selecting
an appropriate stabilizer. Other states have used SFDR, and the following are some examples.

1.5 Case History 1

In 2003, four projects in Maine were selected to determine the structural strength of stabilized
FDR with foamed asphalt. The test plan consisted of performing FWD tests, obtaining samples,
and conducting laboratory tests. The objective of the project was to determine, from test data,
the structural layer coefficients of foamed asphalt layers and recommend appropriate strength of
foamed asphalt mixes to be used in Maine [18].

The work involved pulverizing the existing HMA surface together with approximately 2 in. (50
mm) of the underlying gravel. After initial reclaiming, the material was then graded and
compacted. The roadway was reclaimed with foamed asphalt to a depth of 5.9 in. (150 mm), and
then surfaced with 1.2 in. (30 mm) of asphalt.

The subgrade resilient moduli of the four Maine projects were determined through the
backcalculation of the falling weight deflectometer results using the following equation:

0.24 P
Mr =
dr × r (1)

where,

P = applied load
d = deflection at a distance r from center of the load
r
r = distance from the center of the load
M = backcalculated subgrade resilient modulus
r

Table 1: Subgrade Resilient Modulus Results

Belgrade Orient Farmington Macowahoc
Mr, psi 21564 19383 15842 29342

After computing the resilient modulus of each section, the effective modulus above the subgrade
was determined using the following equation:

5
   1− 1
  2
   D 
1+  
 
0.5
d o = 1.5 Pa 
1
+ a
  1/ 3 2 
 Ep
  D  Ep    
 M r 1 +    
 A  Mr    
    

(2)

where,

d = temperature corrected central deflection, in.

0
a = load plate radius, in.
A = load plate area, sq.in.
p = load pressure, psi
D = total thickness of all pavement layers above subgrade, in.
E = effective modulus of pavement layers above subgrade, psi
p

Table 2: FWD Data

Belgrade Orient Farmington Macowahoc
d0,mils 8.52 16.6 11.37 8.1
a, in 6 6 6 6
A, sq.
113.1 113.1 113.1 113.1
in
P 80.5 80.6 79.8 78.9
D, in 32.2 22 28 22
Ep, psi 135979 67615 109165 152375

Next, the structural coefficients for each layer were determined:

1
SN eff = 0.0045DE p 3
(3)

Because the subgrade resilient modulus was known, the layer coefficient for the subbase could
be determined. Once the effective structural number was calculated, the layer coefficient for the
base could be computed:

SN eff − a1 D1 − a3 D3
a2 =
D2 (4)

6
Table 3 lists the computed structural layer coefficients for the FDR foamed asphalt stabilized
base.

Table 3: Foamed Asphalt Base Layer Coefficients

Layer
Project Section
Coefficient
Belgrade - Rt 8 0.22
Orient Cary - Rt 1 0.23
Farmington - Rt 156 0.22
Macwahoc - Rt 2A 0.35

1.6 Case History 2

Romanosci [16] conducted an experiment to determine the effective structural coefficients of
foamed asphalt SFDR. The experiment consisted of four pavement sections—all with a silty
clay subgrade and three inches of HMA. Out of the four sections, one was constructed with a 9
in. (229 mm) conventional AB-3 granular base and three constructed with a foamed asphalt
SFDR base of 6, 9, and 12 inches. The blended base material was composed of 50% RAP, 37%
AB-3 granular material, 12% A7-6 soil, and 1% Portland cement (to offset the high plasticity of
the clay). The optimum water content was determined to be 3% and asphalt content was
determined to be 3% of PG 64-28 [16].

Following the preparation of the pavement mixes and sections, accelerated testing was
performed. A single axle and dual axle wheel were used to simulate loading, and then in
addition FWD and weight drop tests were performed. Weight drop tests are similar to FWD
while using a smaller load, a larger load plate, and longer spacing between sensors.

The effective structural number was computed using the AASHTO method and the measured
deflections. In this experiment, the layer coefficients needed to be determined for the asphalt
wearing course and the base. Romanosci assumed a structural coefficient for the asphalt surface
was assumed to be 0.42, a typical value. Because the layer coefficient for the asphalt course and
the depths of each layer were known, the coefficient for the base could be determined:

SN eff − a1 D1
a2 =
D2 (5)

By using the AASHTO method of pavement design, the average structural layer coefficient of
the foamed asphalt stabilized base was estimated to be 0.18.

1.7 Case History 3

In another study conducted by Wen et al. [25], Class C fly ash was incorporated into the FDR.
The asphalt layer was milled and blended with 8% fly ash and 5% water in-place to become the
rejuvenated base material. To evaluate the structural contribution of the fly ash, FWD tests were
performed both four days and one year after construction [26].

7
Twenty-three FWD tests were conducted on CTH JK in 2001 and 22 tests in 2002 at an interval
of 100 ft along the roadway. The data were then used in the programs Modulus 5.1 and
Michback to backcalculate the moduli of each pavement layer. It is important to note that the
backcalculated modulus of the fly ash stabilized base increased 49% from 2001 to 2002, which
reinforces the observation that fly ash reactions continue over time.

Using the backcalculated moduli values, the structural number was estimated using:
1
SN = [1.49 × ET 3 ] 3
(6)

log( ET ) 3 = 5.03 − 1.309 log( AUPP) (7)

AUPP =
1
(5D0 − 2 D1 − 2 D3 − D4 )
2 (8)

where,

SN = structural number of pavement (mm),

ET = flexural rigidity of pavement (mm),
AUPP = area under pavement profile (mm), and
Di = surface deflection (mm).

With SN, the structural coefficient of the asphalt layer, a1, was estimated using the AASHTO
method, and then the layer coefficient could be determined for the base layer using equation (5).
The coefficient of the fly ash stabilized base was 0.16 (2001) and 0.23 (2002).

1.8 Summary
Full-depth reclamation (FDR) is a recycling technique gaining popularity as it decreases or
eliminates the need for purchasing and transporting new material. In this literature review, an
overview of the FDR process was presented along with methods and materials used for
stabilization. In addition, the structural layer coefficients were identified in various projects for
SFDR pavement.

8
 Chapter 2. Analysis Methods
2.1 The AASHTO Method and Structural Number
In the American Association of State Highway and Transportation Officials (AASHTO) method
of flexible pavement design, the structural number (SN) is an AASHTO index of pavement
strength based on layer thickness and material properties. SN is commonly used in pavement
design practices and expresses the capacity of pavements to carry loads for a given combination
of soil support, estimated traffic, terminal serviceability, and environment.

On the other hand, the “Minnesota Method” of pavement design incorporates the Granular
Equivalency (GE), which indicates the contribution of a given layer of pavement material
relative to the performance of the entire pavement section. It is dependent upon the properties of
that layer in relation to the properties of the other layers. The relative thickness between the
layers is known as the granular equivalency factor. The layer equivalency can be determined by
laboratory and field tests.

Layer coefficients used in the AASHTO pavement design method are also used to describe
material stiffness, which is similar to the GE factor. Therefore, layer coefficients of the base
materials of the tested project were calculated and used to estimate GE values.

The AASHTO Guide for the Design of Pavement Structures, originally published in 1961, has
been the primary pavement design approach in the United States. The AASHTO Guide is based
on the results of the AASHO (American Association of State Highway Officials) Road Test
conducted in Illinois in the late 1950’s. The first interim design guide was published in 1961,
and subsequent revisions occurred in 1972 and 1981, with the current edition expanded and
revised in 1986 and 1993. This method incorporates several design variables such as traffic
loading, environmental effects, serviceability, pavement layer thickness, and pavement layer
materials. In addition, the AASHTO method incorporates a level of uncertainty in the process to
ensure that the design will last; the level of reliability must increase as the traffic volume
increases [21].

The Guide for Design of Pavement Structures first requires the desired terminal serviceability to
be determined. The serviceability is expressed as an index from 4.2 to 0, where 4.2 is a newly
constructed flexible pavement and 2.0 is a pavement in need of rehabilitation. Next, the known
traffic volumes must be converted to the number of equivalent 18 kip single axle loads (ESAL).
Then the structural number (SN) can be determined by using design charts or a computer
program. After the SN is known, the layer coefficients are evaluated and the required layer
thicknesses are computed.

From the AASHO Road Test, the original equation is as follows:

log[(4.2 − pt ) /(4.2 − 1.5)]

log Wt18 = 9.36 log(SN + 1) − 0.2 +
0.4 + 1094 /( SN + 1) 5.19 (9)

9
where,

Wt18 = Number of 18-kip single axle load applications at time, t

SN = Structural Number of pavement
pt = Serviceability at time, t

In the original test, the Structural Number was determined from the following equation:

SN = a1 D1 + a 2 D2 + a3 D3 (10)

where,

a1, a2, a3 = layer coefficients for the surface, base, and subbase, respectively
D1, D2, D3 = layer thickness, respectively

The layer coefficients are a measure of the relative ability of a unit thickness of a material to
function as a structural component of the pavement. The coefficients can be computed with
regression equations. For the asphalt layer, the coefficient, a1, used in AASHO road tests are
determined from a design chart or equation, but is often taken as 0.44. For a base and subbase
courses, the coefficients are determined from the following equations:

a 2 = 0.249(log E 2 ) − 0.977 (11)

a3 = 0.227(log E3 ) − 0.839 (12)

where E2 and E3 are the resilient moduli for the base and subbase courses can be determined
either by testing or from an AASHTO equation as a function of the stress state and moisture
conditions:

Ei = K 1θ K 2
(13)

The constants K and θ can be determined from AASHTO tables [22].

Equation (9) is only applicable for pavements with an effective subgrade resilient modulus of
3000 psi (20.7 MPa). To apply the equation to other subgrade conditions, and incorporate
reliability, the equation was modified:

log[∆PSI /(4.2 − 1.5)]

log Wt18 = Z R S 0+ 9.36 log(SN + 1) − 0.2 + + 2.32 log M R − 8.07
0.4 + 1094 /( SN + 1) 5.19 (14)

where,

∆PSI = Loss of serviceability (represents the level of serviceability loss the designer is
willing to accept due to traffic loads)
MR = Effective subgrade soil resilient modulus

10
 ZR = Normal deviate for a given reliability, R
S0 = Standard deviation

In the modified equation, the Structural Number includes drainage conditions:

SN = a1 D1 + a 2 D2 m3 + a3 D3 m3 (15)

where,

a1, a2, a3 = layer coefficients for the surface, base, and subbase, respectively
D1, D2, D3 = layer thickness for the surface, base, and subbase, respectively
m1, m2, m3 = drainage coefficients for the surface, base, and subbase, respectively

The following are typical AASHTO structural layer coefficients obtained from a variety of
recycled test sections using several types of recycled materials. Layer coefficients for cold-
recycled mixes can be determined from these values [4]:

• Coefficients for foamed-asphalt recycled layers range from 0.20 - 0.42

• Coefficients for emulsion recycled layers range from 0.17 - 0.41
• Coefficients for cold recycled mixes are between 0.30 - 0.35

2.2 The Minnesota Method and Granular Equivalency

In Minnesota, a different method is used in flexible pavement design, which is based on the
subgrade R-value determined by laboratory testing [23]. The design method includes traffic
loading (ESALs) and material properties (R-value) [27].

The R-value can be determined in a laboratory with the Hveem Stabilometer test, which is a type
of triaxial test performed by measuring a compacted soil’s resistance to deformation [24]. In
addition, in Minnesota Investigation 201 conducted by the Minnesota Department of Highways,
a relationship between the subgrade modulus and R-value was determined:

R-value = (0.41 + 0.873M R )1.28 (16)

After the R-value is computed, and the design number of ESALs is known, the required total
Granular Equivalency can be determined from Figure 1.

11
 Figure 1: Minnesota Method Design Chart [23]

The design chart not only provides the required Granular Equivalent (GE) for the entire
pavement section, but it provides the minimum GE for the asphalt and base courses. Once the
required GE is known, the pavement can be designed by using

G.E. = a1 D1 + a 2 D2 + a3 D3 (17)

where,
GE = Granular Equivalent,
ai = granular equivalent factors for surface, base, and subbase, respectively,
Di = thickness of respective layers.

In equation (17), the constants, a1, a2, and a3 represent the required depth of a given material to
replace a class 5 or 6 base. The granular equivalent factors, ai, were determined in the late
1960’s through extensive testing and data analysis.

In order to determine the GE, the relative effect of the layers, based on deflections, was
established using the Benkelman Beam. The following equations express the relationship
between load deflection and thickness.

log(d s ) = 0.76 + 1.09 log( L1 ) − 3.32 log(0.056 D1 + 0.016 D2 + 0.026 D3 + 1) (18)

log(d f ) = 1.06 + 1.54 log( L1 ) − 4.60 log(0.140 D1 + 0.021D2 + 0.031D3 + 1) (19)

12
where,

ds = spring rebound deflection, in units of 0.001 in.

df = fall rebound deflection, in units of 0.001 in.
L1 = axle load, in units of kips
D = thickness of surface, base, and subbase.

Using equations (18) and (19), the thickness indices can be converted to gravel equivalent
factors.

From elastic theory, pavement deflections can be predicted if the elastic properties of the
materials are determined under the same conditions of test as in the field. The following
equation can be fit to the elastic theory prediction of deflection.

log(d ) = a 0 − a1 D1 − a 2 D2 − a3 log( E ) (20)

where,

d = deflection, 0.001 in.

D1 = surface thickness, in.
D2 = granular base plus subbase thickness, in.
E = Young’s modulus of embankment soil, psi.
a0, a1, a2, a3 = constants determined from regression analysis, which correspond with
pavement layers

The deflections measured were correlated with the thickness of the pavement layers and the
stiffness of the pavement using equation (20). For each test section, the Benkelman Beam
deflection for surface, base, and subbase thicknesses were used in a multiple regression analysis
to determine the values of constants, ai. The following equations are a result of the regression
analysis.

log(d s ) = 3.125 − 0.070 D1 − 0.027 D2 − 0.024 D3 − 0.601 log( R) (21)

log(d s ) = 2.781 − 0.056 D1 − 0.016 D2 − 0.019 D3 − 0.507 log( R) (22)

log(d s ) = 2.733 − 0.056 D1 − 0.019 D2 − 0.025 D3 − 0.416 log( R) (23)

where Log(ds) = Benkelman beam deflection, 0.001 in. using a 9-ton axle load.

To determine the granular equivalency for the top surface, the surface constant, a1, was divided
by the granular base constant, a2. The lower base equivalency was determined by dividing a3 by
a2. For equation (15), the granular equivalency for the asphalt surface is determined by using the
regression analysis constants, a1 = 0.07 and a2 = 0.027. Thus, the G.E. factor for asphalt is 2.59
(=0.07/0.027). The GE factor for Class 3 and 4 Base is 0.89 (=0.024/0.027). Using equations
(21 - 23), the equivalency factors in Table 4 for the surface, base, and subbase were determined.

13
 Table 4: GE Factors
Equivalency Factors
Year Surface Base Subbase
1965 2.59 1 0.89
1966 3.5 1 1.2
1967 2.95 1 1.32

The Minnesota Department of Highways evaluated several GE factors in the Minnesota

Investigation 183 for various pavement layers [23]:

• plant-mix surface layers: 2.25;

• plant-mix base: 2.0;
• road-mix surface: 1.5;
• road-mix base: 1.5;
• bituminous treated bases range for lean and rich mixes, respectively: 1.25-1.5;
• gravel bases: 0.9-1.0
• crushed rock base: 1.0
• sand gravel subbase: 0.75

2.3 Backcalculation and Falling Weight Deflectometer

Another method that can be used to estimate GE is backcalculation to obtain resilient modulus of
the base materials. It is known that GE of Class 5 material is 1. Therefore, the ratio of resilient
modulus between base materials and Cl5 should give an estimate of the GE factors of these
materials. To measure pavement deflection, impulse load tests are commonly performed by using
a Falling Weight Deflectometer (FWD). The FWD is a device capable of applying dynamic
loads to the pavement surface, similar in magnitude and duration to that of a single heavy
moving wheel load. The FWD test measures the pavement response with seismometers and
generates a deflection basin that provides valuable information [24].

In an FWD test, an impulse load is applied by dropping a weight (usually 9000 lb) on the
pavement, and the resulting deflections are measured at specified distances from the point of
load application. The number of load applications can be adjusted. After obtaining the
deflections, Young’s moduli of the different layers are determined by backcalculation, which
involves an estimate of the elastic properties from the measured surface deflections for an
assumed layer profile [25]. In this study, EVERCALC software is used to backcalculate
modulus of base materials.

14
 Chapter 3. Selected Sections for Field Testing
FWD testing was performed on the selected sections (Figs. 2-5) over a three year period. During
spring thaw of each year, FWD was conducted daily in the first week of thawing in an attempt to
capture spring thaw weakening of base. After spring thaw, FWD was conducted monthly to
study base recovery and stiffness changes through seasons. Table 5 contains the FWD testing
schedule. Ground penetrating radar (GPR) was also conducted on the sections to obtain
pavement thickness profile. GPR is a non-destructive testing tool that has wide applications in
pavements. It detects changes in the underground profile due to contrasts in the electromagnetic
conductivity across material interfaces. It can be used at relatively high speeds and gives a
continuous pavement profile. GPR surveys have been successful in determining stripping zones
in asphalt pavements, detecting subsurface voids, detecting subsurface anomalies (bedrock/peat),
bridge deck delamination, tie bar locations, underground utility locates, sub-grade profiling, and
pavement thickness. Figure 6 shows the GPR equipment.

Figure 2: LeSueur County Road Sections: a. CR 2; b. Road 13

Figure 3: Pope County Road Sections: a. CR 28; b. CR 29

15
Figure 4: Goodhue East and West County Road 30 Sections: a. West; b. East

Figure 5: Olmsted County Road 13 Section

Figure 6: GPR Equipment

16
 Table 5: FWD Testing Schedule

FWD data were used to obtain base stiffness through backcalculation. Figure 7 shows examples
of base stiffness change during the year. GPR data were analyzed to obtain pavement layer
thickness profiles. The layer thickness is determined at every 0.1 miles (Figs. 8-13).

17
 Figure 7: Example of Base Stiffness Changes during the Year

Figure 8: Goodhue County Road 30 Pavement Layer Thickness from GPR Survey

18
 Figure 9: LeSeuer County Road 2 Pavement Layer Thickness from GPR Survey

Figure 10: LeSeuer County Road 13 Pavement Layer Thickness from GPR Survey

19
Figure 11: Olmsted County Road 13 Pavement Layer Thickness from GPR Survey

Figure 12: Pope County Road 29 Pavement Layer Thickness from GPR Survey

20
 Pope Co Rd 28 Estimated Layer Depths
Distance From Start
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
4.0
4.1
4.2
4.3
4.4
0.0

2.0
Northbound Pavmnt Depth
4.0
Southbound Pavmnt Depth
6.0
Depths

Northbound Stbl Base Depth

8.0

10.0 Southbound Stbl Base Depth

12.0

14.0

16.0

18.0

20.0

Figure 13: Pope County Road 28 Pavement Layer Thickness from GPR Survey

21
 Chapter 4. Analysis Results
4.1 MnROAD Cell 21, Class 5 Base Analysis
In order to analyze stiffness values of the SFDR projects, moduli from Class 5 projects must be
compared to determine if the SFDR roads yield higher performance.

The Young’s modulus for the base course was determined using an FWD analysis program,
EVERCALC, for one project using Class 5 as a base, and seven other projects using both
stabilized and standard FDR base. Because the geology under the pavement subgrade was not
specified for all projects, it is unknown whether a stiff layer (location of zero deflection) is
present. In EVERCALC, the modulus values were computed twice, once with a stiff layer and
once with no stiff layer (semi-infinite space). The modulus values with the lowest RMS (root
mean square) error were used for the analysis.

MnROAD cell 21 consists of 8 in. (203 mm) of asphalt and 23 in. (584 mm) of Class 5 base.
The stiffness values for the road section constructed with Class 5 base are detailed as follows:
Unless noted otherwise, only one modulus value is shown per testing season. Spring testing was
conducted from January to May. Summer testing was conducted from June to August. Fall
testing was conducted from September to November. No testing was performed in December.
Unless noted, each modulus is an average of several measurements and is measured in ksi (1000
lb/in.2). MnROAD cell 21 consists of 8 in. (203 mm) of asphalt and 23 in. (584 mm) of Class 5
base.

Table 6: Young’s Modulus Values in ksi (1000 lb/in.2) for MnROAD Cell 21
Cell 21
Spring Summer Fall
HMA nostiff 1008.15 311.26 1520.56
stiff 907.65 278.34 1423.04
BASE nostiff 14.76 20.36 18.39
stiff 24.00 26.23 28.29

4.2 LeSueur County Road 2

The 4.8 mile section of LeSueur County Road (CR) 2 is constructed of 6 in. (152 mm) asphalt
followed by 12 in. (305 mm) of 6% Class C fly ash SFDR base. The subgrade material is
classified as “plastic.”

22
 Table 7: Young’s Modulus Values in ksi (1000 lb/in.2) for LeSueur CR 2

Spring Summer Fall

2011 w/r Cell21 w/r Cell21 w/r Cell21
HMA nostiff 942.27 0.93 536.03 1.72 1152.79 0.76
stiff 822.71 0.91 369.82 1.33 917.89 0.65
BASE nostiff 23.78 1.61 46.57 2.29 37.88 2.06
stiff 51.16 2.13 73.61 2.81 62.91 2.22
2010
HMA nostiff 1246.89 1.24 291.12 0.94 1003.56 0.66
stiff 1106.47 1.22 196.54 0.71 788.66 0.55
BASE nostiff 42.09 2.85 48.19 2.37 49.03 2.67
stiff 59.26 2.47 75.5 2.88 73.19 2.59
2009
HMA nostiff 835.89 0.83 313.02 1.01 1094.04 0.72
stiff 752.06 0.83 215.85 0.78 880.48 0.62
BASE nostiff 41.17 2.79 39.89 1.96 46.39 2.52
stiff 71.57 2.98 62.76 2.39 69.05 2.44
The white columns represent the ratio of the modulus of the FDR base over the modulus of the
Cell 21 Class 5 base. The Young’s modulus values were calculated for each individual pavement
layer for SFDR roads and the Class 5 base roads in spring, summer, and fall. The moduli were
backcalculated using EVERCALC, both with a stiff layer and without a stiff layer. A ratio
greater than 1.0 designates that the FDR section is stiffer. Observing the ratios between LeSueur
CR 2 and Class 5 roads, it can be seen that, for the most part, the base modulus values calculated
for LeSueur CR 2 are generally higher than those of Class 5 base roads.

Figure 14: Young’s Modulus Values for LeSueur CR2 in 2011

23
 Figure 15: Young’s Modulus values for LeSueur CR2 in 2010

Figure 16: Young’s Modulus values for LeSueur CR2 in 2009

It can be observed from Figures 14-16 that when the modulus is calculated with a stiff layer, the
modulus values are consecutively higher throughout the year compared to when calculated
without a stiff layer. In the spring months there is a rapid decrease in both sets of moduli due to
the spring thawing effect on the roads.

4.3 LeSueur County Road 13

The 5.5 mile section of CR13 in LeSueur County is constructed of 6 in. (152 mm) of asphalt
course followed by 7 in. (178 mm) of 3.5% emulsion and 2% Class C fly ash SFDR base, and 1-
3 in. (25-76 mm) of non-stabilized FDR base; the stabilized and non-stabilized were treated as
one layer in Evercalc. The subgrade material is classified as “plastic.”

24
 Table 8: Young’s Modulus Values in ksi (1000 lb/in.2) for LeSueur CR 13
Spring Summer Fall
2011 w/r Cell21 w/r Cell21 w/r Cell21
HMA nostiff 1311.82 1.30 250.60 0.81 1028.73 0.68
stiff 1385.39 1.53 297.61 1.07 1130.52 0.79
BASE nostiff 155.45 10.53 135.90 6.67 230.17 12.51
stiff 144.86 6.04 90.50 3.45 193.15 6.83
2010
HMA nostiff 1814.90 1.80 173.52 0.56 776.37 0.51
stiff 1888.96 2.08 226.88 0.82 865.67 0.61
BASE nostiff 154.37 10.46 165.18 8.11 252.49 13.73
stiff 139.26 5.80 108.35 4.13 221.83 7.84
2009
HMA nostiff 1100.75 1.09 160.01 0.51 1253.12 0.82
stiff 1218.97 1.34 217.13 0.78 1380.82 0.97
BASE nostiff 69.29 4.69 93.83 4.61 171.09 9.30
stiff 42.82 1.78 41.07 1.57 119.20 4.21

Figure 17: Young’s Modulus Values for LeSueur CR 13 in 2011

25
 Figure 18: Young’s Modulus Values for LeSueur CR 13 in 2010

Figure 19: Young’s Modulus Values for LeSueur CR 13 in 2009

Looking at the ratios between LeSueur CR 13 and Class 5 roads, it is clear that the base modulus
for CR 13 is much higher than the values calculated for Class 5 base roads (Table 6). The
moduli calculated without a stiff layer are generally higher throughout the year on CR 13. A
spring thawing effect can be observed as well, which decreased the stiffness in late spring.

4.4 Pope County Road 28

The 4.2 mile section of CR 28 is constructed of 3.5 in. (89 mm) of asphalt followed by 4 in. (102
mm) of 0.004 gal/yd2/in., T15 Base One SFDR, followed by 4 in. (102 mm) of FDR base course.
The subgrade soil is classified as Class 4 material.

26
 Table 9: Young’s Modulus Values in ksi (1000 lb/in.2) for CR 28
Spring Summer Fall
2011 w/r Cell21 w/r Cell21 w/r Cell21
HMA nostiff 1674.06 1.66 403.28 1.30 1022.71 0.67
stiff 2102.68 2.32 753.63 2.71 1524.97 1.07
BASE nostiff 51.32 3.48 125.38 6.16 161.51 8.78
stiff 12.23 0.51 31.26 1.19 53.53 1.89
2010
HMA nostiff 2062.22 2.05 667.97 2.15 658.69 0.43
stiff 1963.20 2.16 699.35 2.51 1017.78 0.72
BASE nostiff 11.12 0.75 76.97 3.78 134.63 7.32
stiff 11.64 0.49 28.28 1.08 45.75 1.62
2009
HMA nostiff 1457.01 1.45 369.92 1.19 624.79 0.41
stiff 1641.56 1.81 664.67 2.39 1071.24 0.75
BASE nostiff 61.34 4.16 167.50 8.23 162.56 8.84
stiff 15.03 0.63 39.31 1.50 39.54 1.40

Figure 20: Young’s Modulus Values for Pope CR28 in 2011

27
 Figure 21: Young’s Modulus Values for Pope CR28 in 2010

Figure 22: Young’s Modulus Values for Pope CR 28 in 2009

Looking at the ratios for CR 28, it can be observed that the base modulus values computed for
CR 28 are higher compared to the values computed for Class 5 roads (Table 6). Although there
are some minor variations in which set of data shows a higher modulus, those calculated without
a stiff layer are generally higher in the three years of testing. It can also be seen that in 2011 and
2010 there were several short periods of thawing and then refreezing again in the spring months
that caused fluctuations in stiffness. Furthermore, in 2011 and 2010, there appears to have been a
short freezing period in September that increased the stiffness as well.

28
4.5 Pope County Road 29
The 5.0 mile section of CSAH 29 is constructed of 3.5 (89 mm) in. of asphalt followed by 8 in.
(204 mm) of FDR base course. The subgrade soil is classified as Class 4 material.

Table 10: Young’s Modulus Values in ksi (1000 lb/in.2) for Pope CR29
Spring Summer Fall
2011 w/r Cell21 w/r Cell21 w/r Cell21
HMA nostiff 2293.59 2.28 1784.39 5.73 1311.76 0.86
stiff 2488.87 2.74 988.92 3.55 3319.05 2.33
BASE nostiff 15.22 1.03 36.24 1.78 33.53 1.82
stiff 15.32 0.64 48.53 1.85 53.38 1.89
2010
HMA nostiff 2355.84 2.34 650.83 2.09 1707.17 1.12
stiff 2144.76 2.36 398.51 1.43 1254.01 0.88
BASE nostiff 12.44 0.84 29.05 1.43 25.86 1.41
stiff 16.86 0.70 43.80 1.67 38.97 1.38
2009
HMA nostiff 2345.98 2.33 945.60 3.04 2016.34 1.33
stiff 1339.11 1.48 333.61 1.20 940.41 0.66
BASE nostiff 12.75 0.86 23.21 1.14 20.78 1.13
stiff 44.18 1.84 54.52 2.08 53.05 1.88

Figure 23: Young’s Modulus Values for Pope CR 29 in 2011

29
 Figure 24: Young’s Modulus Values for Pope CR 29 in 2010

Figure 25: Young’s Modulus Values for Pope CR 29 in 2009

Observing the ratios between CR 29 and the Class 5 roads, it can be seen that the calculated base
moduli are higher for CR 29 than those calculated for Class 5 roads. The moduli calculated with
a stiff layer are consecutively higher than those without a stiff layer. A spring thawing effect can
only be observed in 2010 and 2009, while in 2011 the road section has higher stiffness in the
summer compared to spring, and an even greater stiffness in the fall.

4.6 Goodhue County Road 30 Eastern Section

About 3 miles of CR 30 East is constructed of 4 in. (102 mm) of asphalt followed by 6 in. (152
mm) of FDR, and 8 in. (203 mm) of aggregate. The subgrade soil was not classified.

30
 Table 11: Young’s Modulus Values in ksi (1000 lb/in.2) for CR 30 East
Spring Summer Fall
2011 w/r Cell21 w/r Cell21 w/r Cell21
HMA nostiff 1808.91 1.79 550.45 1.77 1327.37 0.87
stiff 1867.66 2.06 699.07 2.51 1521.12 1.07
BASE nostiff 14.58 0.99 58.26 2.86 53.94 2.93
stiff 9.11 0.38 38.03 1.45 28.64 1.01
2010
HMA nostiff 1831.94 1.82 383.83 1.23 1212.20 0.80
stiff 1915.57 2.11 479.69 1.72 1355.86 0.95
BASE nostiff 16.54 1.12 46.11 2.26 59.73 3.25
stiff 7.97 0.33 29.00 1.11 38.68 1.37
2009
HMA nostiff 1875.05 1.86 338.49 1.09 1811.55 1.19
stiff 1777.05 1.96 371.27 1.33 1882.70 1.32
BASE nostiff 17.94 1.22 38.33 1.88 22.08 1.20
stiff 19.85 0.83 30.40 1.16 13.93 0.49

2011 Base Modulus

1000
nostiff stiff
Modulus (ksi)

100

10

1
Feb Mar Apr May Jun Jul Jul Aug Sep Oct Nov Dec

Figure 26: Young’s Modulus Values for Goodhue CR 30 East in 2011

31
 Figure 27: Young’s Modulus Values for Goodhue CR 30 East in 2010

Figure 28: Young’s Modulus Values for Goodhue CR 30 East in 2009

Observing the ratios between Goodhue CR 30 East and Class 5 roads, it can be seen that the base
moduli calculated for CR 30 are, for the most part, higher than those values calculated for Class
5 roads. In 2011 and 2010, it can be clearly seen that the moduli calculated without a stiff layer
are higher than those calculated with a stiff layer. However, in 2009, the results change halfway
through the year. In spring and early summer, the values calculated with a stiff layer are higher,
while in late summer and fall, the opposite is true. A spring thawing effect can be observed each
year, decreasing the stiffness. In 2009, there was a late spring freeze that caused the stiffness to
rise once again before the summer months.

32
4.7 Goodhue County Road 30 Western Section
The 2.8 mile section of CR 30 West is constructed of 2 in. (52 mm) of asphalt followed by 6 in.
(152 mm) of 4.5% Fortress SFDR, and 8 in. (203 mm) of aggregate. The subgrade soil was not
classified.

Table 12: Young’s Modulus Values for Goodhue CR 30 West

Spring Summer Fall
2011 w/r Cell21 w/r Cell21 w/r Cell21
HMA nostiff 2597.23 2.58 850.32 2.73 2181.99 1.43
stiff 4686.98 5.16 2335.22 8.39 4089.24 2.87
BASE nostiff 38.99 2.64 73.17 3.59 134.58 7.32
stiff 13.97 0.58 47.28 1.80 89.56 3.17
2010
HMA nostiff 4108.65 4.08 616.45 1.98 1772.56 1.17
stiff 5357.84 5.90 1607.88 5.78 3138.07 2.21
BASE nostiff 56.68 3.84 72.82 3.58 123.48 6.71
stiff 27.74 1.16 34.53 1.32 76.93 2.72
2009
HMA nostiff 4767.63 4.73 530.81 1.71 2179.49 1.43
stiff 6037.35 6.65 781.27 2.81 3357.11 2.36
BASE nostiff 41.93 2.84 63.93 3.14 141.53 7.69
stiff 29.02 1.21 52.30 1.99 99.31 3.51

2011 Base Modulus

1000 nostiff
stiff

100
Modulus (ksi)

10

1
Feb Mar Apr May Jun Jul Jul Aug Sep Oct Nov Dec

Figure 29: Young’s Modulus Values for Goodhue CR 30 West in 2011

33
 Figure 30: Young’s Modulus Values for Goodhue CR30 West in 2010

Figure 31: Young’s Modulus Values for Goodhue CR 30 West in 2009

Observing the ratios between Goodhue CR 30 West and Class 5 roads, it can be seen that the
base moduli for CR 30 West are higher than those of the Class 5 roads. It can be observed that
the moduli calculated without a stiff layer are higher than those calculated with one during the
late spring, summer, and fall months. However, there are some discrepancies as to which one is
higher in the early spring months. The spring thawing effect causes several spikes all three years
in both sets of data, which makes it difficult to reach a firm conclusion.

4.8 Olmsted County Road 13

When comparing the ratios between Olmsted CR 13 and Class 5 roads, it can be seen that the
base moduli for CR 13 are much higher than those of the Class 5 roads.

34
Table 13: Young’s Modulus Values in ksi (1000 lb/in.2) for Olmsted County Road 13
Spring Summer Fall
2011 w/r Cell21 w/r Cell21 w/r Cell21
HMA nostiff 2902.63 2.88 563.18 1.81 2058.16 1.35
stiff 3038.08 3.35 642.29 2.31 2121.97 1.49
BASE nostiff 188.14 12.74 274.23 13.47 373.84 20.32
stiff 174.74 7.28 219.61 8.37 365.43 12.92
2010
HMA nostiff 3636.55 3.61 594.69 1.91 1740.05 1.14
stiff 3783.48 4.17 799.37 2.87 2510.08 1.76
BASE nostiff 138.77 9.40 182.66 8.97 228.95 12.45
stiff 120.08 5.00 134.32 5.12 123.30 4.36
2009
HMA nostiff 3669.15 3.64 583.04 1.87 2480.74 1.63
stiff 583.04 0.64 675.53 2.43 2743.04 1.93
BASE nostiff 118.09 8.00 182.33 8.96 228.95 12.45
stiff 66.91 2.79 81.42 3.10 123.30 4.36

2011 Base Modulus

1000

100
Modulus (ksi)

nostiff
10 stiff

1
Feb Mar Apr May Jun Jul Jul Aug Sep Oct Nov Dec

Figure 32: Young’s Modulus Values for Olmsted CR13 in 2011

35
 Figure 33: Young’s Modulus Values for Olmsted CR13 in 2010

Figure 34: Young’s Modulus Values for Olmsted CR 13 in 2009

For Olmsted CR 13, it can be seen that the moduli calculated without using a stiff layer are only
slightly higher than those calculated with. However, the values have very small differences from
one another- they are nearly identical across the span of all 3 years. The spring thawing effect
also caused some large fluctuations in the data pattern in 2009. There appear to be multiple
periods of freezing and thawing that year. On the other hand, there was hardly any variation
between spring and summer values in 2011 and 2010.

36
4.9 Modulus Values Summary
The backcalculated modulus values indicate that these materials also have seasonal effects.
Figures 35-37 show the comparison of backcalculated moduli of all base materials for 2009,
2010 and 2011, respectively.

Figure 35: Modulus Plot Summary with Stiff Layer (2009)

Figure 36: Modulus Plot Summary without Stiff Layer (2010)

37
 Modulus Summary with Stiff
(2011)
10000

1000
GoodhueE
Modulus (ksi)

GoodhueW
100 LeCR2
LeCR13
OlCR13
10
Pope28
Pope29

1
12/28 2/16 4/7 5/27 7/16 9/4 10/24 12/13
Date

Figure 37: Modulus Plot Summary with Stiff Layer (2011)

In general, it can be seen that non-stabilized FDR (Pope 29 and Goodhue E) has lower strength
than other materials and also has weaker strength in the spring time. This illustrates that SFDR is
typically less sensitive to spring thaw than FDR materials.

Figure 38 shows modulus ratios between backcalculated modulus of the base materials and
modulus of Class 5 material. GE of Class 5 is 1, so, the ratio is a measure of the GE of the
material.

5.00
4.50
Modulus ratio with cl5

4.00
2009 Summer
3.50
3.00 2009 Fall
2.50 2010 Summer
2.00
1.50 2010 Fall
1.00 2011 Summer
0.50
2011 Fall
0.00
Goodhue Goodhue LeSueur LeSueur Olmsted Pope CR28 Pope CR29
Eastern Western CR2 CR13 CR13

Figure 38: Modulus Ratios

38
Figure 38 shows that some ratios are very high, which indicates that the method used to estimate
GE may not be appropriate.

4.10 Method Based on Hogg Model

As with any field study, discrepancies and inconsistencies are present. Nonetheless, it is
apparent that the stabilizer is exerting an impact on the base, but it cannot be quantified. Many
other factors are different for each county, including the year of construction, AADT, asphalt
concrete, and stabilized depth, which contribute to the effectiveness of the stabilizer.

To provide a summary, the Hogg Model method is used to calculate the Granular Equivalencies
(GE) from the FWD deflection data. The Hogg model is based on a hypothetical two-layer
system consisting of a relatively thin plate on an elastic foundation. The method in effect
simplifies the typical multilayered elastic system with an equivalent two-layer stiff-layer-on-
elastic foundation model. Depending on the choice of values along the deflection basin used to
calculate subgrade stiffness, the tendency exists to either over- or underestimate the subgrade
modulus. The Hogg model uses the deflection at the center of the load and one of the offset
deflections. Hogg showed that the offset distance where the deflection is approximately one-half
of that under the center of the load plate was effective at removing estimation bias. The
calculations consider variations in pavement thickness and the ratio of pavement stiffness to
subgrade stiffness, since the distance to where the deflection is one-half of the deflection under
the load plate is controlled by these factors.

The method also takes temperature, season, time of the day, and thicknesses of the layers into
account. The Effective Granular Equivalencies (EGE) can be obtained using this method, which
is the sum of the GEs for all layers; the seasonal factors are applicable for Jun-Oct. The effective
depth (Hp) is assumed as 2/3 of the distance where 50% percent of the maximum deflection
occurs. This value is interpolated from the locations of two sensors that are closest to 50% of the
maximum deflection. Since the GE of the asphalt layer is assumed to be 2.25 and that of the
subgrade layer is assumed to be Class 5 material, which has GE = 1.0, the GE of the base
material can be calculated with the following equation:
𝐸𝐺𝐸−2.25×𝐴𝑠𝑝ℎ𝑎𝑙𝑡 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠−1.0×𝐻𝑝
𝐺𝐸𝑏𝑎𝑠𝑒 = (24)
𝐵𝑎𝑠𝑒 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠

The equation is adjusted accordingly if the effective depth does not reach the subgrade layer.

Figures 39-41 are a summary of the GE values of each road site test section from 2009-2011. It
can be seen that, similar to the modulus values, spring thawing affects GE values as well. Higher
GE values can be seen in early spring and late fall when temperatures are still low. As soon as
temperatures rise, the GE values decrease.

In general, modulus for the base is higher when a stiff layer is assumed to be present for most
counties, except LeSueur 13, for which a non-stabilized aggregate depth was assumed to be 3 in.,
and it could vary from 1 – 3 in. The values of the modulus for each county are generally
consistent throughout the three year period, varying from 10 – 100 ksi. In terms of modulus
comparison between counties, the ranking obtained from the calculation without a stiff layer
appears reasonable. For example, for counties that have a stabilized base such as the Pope CR

39
28 and Goodhue Western section, the base modulus appears to be higher than their non-
stabilized counterparts.

The calculated GE values appear to be quite high, which might be associated with the method
used to obtain the effective depth. If the GE values are ranked for each year, similar results can
be obtained. Comparing counties that are stabilized and those that are not, it is observed that the
stabilized base exhibits a higher GE value than that of the non-stabilized.

Figure 39 2011 GE Summary Plot

Figure 40: 2010 GE Summary Plot

40
 Figure 41: 2009 GE Summary Plot

Comparing the GE ranking and the modulus ranking between counties, the results are similar,
but not completely consistent. For example, the modulus for Olmsted CR 13 is the highest for all
three years, while the GE value for Olmsted CR 13 is only ranked around the middle. One
possible reason for these inconsistencies is the method used to calculate GE values. Apart from
the fact that there might be some problems with the calculation of the effective depth, another
important parameter used to calculate the GE for the base, the effective GE, is greatly dependent
on the seasonal adjustment factor (SAF), which is also dependent upon the sub-grade soil type,
where plastic, semi-plastic, and non-plastic make a difference. Such information for each county
is incomplete. Only LeSueur County Roads are specifically given as plastic sub-grade soil; Pope
County Roads are class 4 (assumed to be non-plastic) and Goodhue and Olmsted CR 13 are
given as N/A (assumed to be non-plastic as well). The back-calculation routine might also
contribute to the discrepancies, as the results depend significantly on the initial input.

41
 Chapter 5. Summary and Recommendations
Full-depth reclamation (FDR) is a recycling technique where the existing asphalt pavement and a
predetermined portion of the underlying granular material are blended to produce an improved
base course. FDR is an attractive alternative in road rehabilitation: resources are conserved, and
material and transportation costs are reduced as recycling eliminates the need for purchasing and
hauling new materials and disposing of old materials. An additive is sometimes used, and this
process is referred to as stabilized full-depth reclamation (SFDR). Previous research has
demonstrated that the strength of a traditional aggregate base (such as Class 5) normally shows a
weakening during springtime thaw. It is part of the reason that spring load restrictions have been
applied on some local pavements during each year’s spring thaw period. However, not much
research has been conducted on seasonal effects of SFDR base.

Currently, MnDOT pavement design recommends granular equivalency, GE = 1.0 for non-
stabilized FDR material, which is equivalent to Class 5 material. For SFDR, there was no
guideline for the GE value at the time this project was initiated (2009). Some local engineers
believe that GE of FDR material should be greater than 1.0 (Class 5), especially for SFDR.

The objective of this project was to (1) estimate GE values of both non-stabilized and stabilized
full-depth reclamation materials used for pavement base layer, and (2) assess spring thaw effects
on stiffness of both stabilized and non-stabilized FDR. Falling Weight Deflectometer (FWD)
tests were performed on seven selected sections on county roads in Minnesota. FWD tests were
performed over a three year period. During spring thaw of each year, FWD was conducted daily
in the first week of thawing in an attempt to capture spring thaw weakening of the base. After
the spring thaw period, FWD was conducted monthly to study base recovery and stiffness
changes through the seasons.

It is known that the GE factor is an empirical number, which is used by MnDOT to describe
stiffness of asphalt and base materials. There is no well-defined method to determine GE either
through mathematical computation or laboratory test. In this work, three different approaches
were used in an attempt to estimate GE factor from FWD deflections. The first method is the
AASHTO method, the second one is backcalculation using EVERCALC, and the third one is a
MnDOT method developed by Erland Lukanen. It was found that the third method provides
reasonable GE values.

Based on the data collected for this project, the average GE of SFDR is estimated to be 1.5.
Certainly, the value varies from project to project as construction and material varies from
project to project. It appears that all the materials tested showed seasonal effects on stiffness. In
general, the stiffness is lower in spring than that in summer and fall, typical behavior for
unbound base materials. Most of the SFDR materials tested in this project, but not all, showed
improved seasonal stiffness.

42
 References
1. “Industry Segments.” Industry Segments. Asphalt Recycling and Reclaiming Association.
Web.

2. Full Depth Reclamation – Construction Methods and Equipment. FHWA.

3. Alfred Crawley. “Innovative Hot-in-Place Recycling of Hot-Mix Asphalt Pavement in

Mississippi.” Transportation Research Record 1654 ser: 36-42.

4. Gene Skok, Thomas Westover, Joseph Labuz, Shongtao Dai, and Erland Lukanen. Pavement
Rehabilitation Selection. Tech. no. MN/RC 2008-06. St. Paul: Minnesota Department of
Transportation, MN.

5. Edward Kearney. “Full Depth Reclamation Process.” Transportation Research Record 1684
ser: 203-209.

6. W. S. Guthrie. “Cement Stabilization of Aggregate Base Material Blended with Reclaimed

Asphalt.” Transportation Research Record 2026 ser: (2007) 47-53.

7. Rajib Mallick. “Evaluation of Performance of Full-Depth Reclamation Mixes.”

Transportation Research Record 1809 ser: 199-208.

8. Jeb Tingle, Santoni, Rosa. "Stabilization of Clay Soils with Nontraditional Additives"
Transportation Research Record LVR8-1136 1819 (2003): 72-84.

9. Dwane Lewis. “Georgia’s Use of Cement-Stabilized Reclaimed Base in Full Depth

Reclamation.” Transportation Research Record 1952 ser: (2006) 125-133.

10. N. Bandara and M. Grazioli. “Improving subgrade strength and pavement performance by
chemically treating subgrade soils.” Bearing Capacity of Roads, Railways and Airfields 1
(2009): 29-36.

11. Dallas Little. Evaluation of Structural Properties of Lime Stabilized Soil and Aggregates.
Rep. National Lime Organization, 1999.

12. T. McCleary. “Subgrade modification—practitioner’s experience.” Bearing Capacity of

Roads, Railways and Airfields 1 (2009): 97-106.

13. S. M. Mackiewicz, and E. G. Gerguson. Stabilization of Soil with Self-Cementing Coal

Ashes. http://www.flyash.info/2005/108mac.pdf. Accessed July 20, 2009.

14. D. Saylak, C. Estakhri, S. Mishra, and D. Sinn. Base Stabilization and Dust Control Using
Calcium Chloride and Fly Ash. Rep. Texas: American Coal Ash Association, 2003.

43
15. Hyung Choi. Soil Stabilization Using Optimum Quantity of Calcium Chloride. Thesis. Texas
A&M University, 2005.

16. Stefan Romanosci, Mustaque Hossain, Andrew Gisi, and Michael Heitman. “Accelerated
Pavement Testing Evaluation of the Structural Contribution of Full-Depth Reclamation
Material Stabilized with Foamed Asphalt.” Transportation Research Record 1896 ser (2004):
199-207.

17. Brian Marquis, Dale Peabody, Rajib Mallick, and Tim Soucie. Determination of Structural
Layer Coefficient for Roadway Recycling Using Foamed Asphalt. Tech. Durham: Recycled
Materials Resource Center University of New Hampshire, 2003.

18. Y. Yilmaz, A. Gungor, and C. Avsar. “Stabilization of Clays using liquid enzymes.” Bearing
Capacity of Roads, Railways and Airfields 1 (2009): 65-78

19. D. E. Scholen. Nonstandard Stabilizers. Report FHWA-FLP-92-011.FHWA, U.S.

Department of Transportation, Washington, DC, July 1992.

20. American Association of State Highway and Transportation Officials. Guide for Design of
Flexible Pavement Structures. AASHTO, Washington, DC, 1993.

21. Yang Huang. Pavement Analysis and Design. 2nd ed. Upper Saddle River: Pearson-Prentice
Hall, 2004.

22. David E. Newcomb. Measuring In Situ Mechanical Properties of Pavement Subgrade Soils.
Transportation Research, 1999.

23. F. C. Fredrickson, P. J. Diethelm, and D. M. Zwiers. Flexible Pavement Design 1969.

Minnesota Department of Highways, 1969.

24. M. Ahmed, R. Bisht, and R. Tarefder. “Analysis of FWD data and characterization of airfield
pavement materials in Mexico.” Bearing Capacity of Roads, Railways and Airfields 1 (2009):
669-678.

25. Haifang Wen, Mathew P. Tharaniyil, Bruce Ramme, and Steve Krebs. “Field Performance
Evaluation of Class C Fly Ash in Full Depth Reclamation: Case History Study.”
Transportation Research Record 1869th ser. (2004): 41-46.

26. “Pavement Design.” Minnesota Local Road Research Board.

44