ARIZONA DEPARTMENT OF TRANSPORTATION

REPORT NUMBER: FHWA-AZ90-350

LABORATORY EVALUATION
OF TYPICAL ADOT
ASPHALT CONCRETE
Final Report

Prepared by:
Michael S. Marn(ouk
Punya P. Khanal
Center for Advanced Research in Transportailon.
and the Department of Civil E~gineering
College o! Eng~neeringand Applied Sciences
Arizona State University
Tempe, Arizona 85287

August 1990

Prepared for:
Arizona Department of Transportation
206 South 1 7th Avenue
Phoenix. Arizona 85007
in cooperation with
U.S. Department of Transportation
Federal Highway Administration

The contents of this report reflect the views of the authors
who are responsible for the facts and the accuracy of the
data presented herein.
The contents do not necessarily
reflect the official views or policies of the Arizona
Depertment of Transportztion or the Federal Highways
Administration. This report does not constitute a standard,
specification, or regulation. Trade or manufacturer's names
which may appear herein are cited only because they are
considered essential to the objectives of the report. The
U S . Government and the State of Arizona do not endorse
products or manufacturers.

Technical Report Documentation Page

/'

i 2 . Governmen: A C C ~ S SNO.
D~

Rem.l N3.

I

I

I

I
I

1

3 Kecpex's C a : a q ho.

1 5 . Repon Dare

LABORATORY n7ALUATION OF TYPICAL ADOT ASPHALT
CONCRETE

A u g u s t 1990

I

1 6. P e r i o r m ~ qGraan~za~ion
Cooe
I

1

16

7. Adnor ( s )

P e r i o r m q 0gan::aton

R e x ? ti:.

I

M i c h a e l S . Mamlouk a n d P u n y a P . K h a n a l

I

1 10. Work Uni7 Nc.

?. Pedormlng Organlzalnn Name and Address

C e n t e r f o r Advanced R e s e a r c h i n T r a n s p o r t a t i o n
Arizona S t a t e University
Tempe, A r i z o n a
85287-6306

11. Contraa or G:an: fuo.

HPR-PL-1-37

1 13. T y ~ ea! f i e ~ o nL

12. Sp0nso:lng Agency Name and Aodress

I
I

I

I

1
I

1350)
Ferod b v e r e c

ARIZONA DEPARTMENT OF TRANSPORTATlON
206 S. l 7 T H AVENUE
PHOENIX, ARIZONA B50D7

I
I

Prepared i n cooperation with the U.S. Departmen1 o l ?ranspor.at~on, Federal H ~ g h w a yA o m ~ n ~ c l r a t ~ o n

!

i

i

I

1

!'

A t y p i c a l ADOT a s p h a l t c o n c r e t e m i x t u r e w a s e v a i u a t e d b a s e d o n t h e

!

A s p h a l t - A g g r e g a t e M i x t u r e A n a l y s i s S y s t e m ( W A S ) p r o c e d u r e (NCHRP
P r o j e c t 9 - 6 ( 1 ) ) . Two s e t s o f ADOT a s p h a l t c o n c r e t e s p e c i m e n s were
prepared u i n g t h e C a l i f o r n i a kneadinq compactor and t h e Marshall
hammer.
A i i t e s t s recommended by t h e PA-MAS p r o j e c t w e r e p e r f o r m e d .
t e s t r e s u l t s were a n a l y z e d u s i n g t h e MblAS g u i d e l i n e s .

I

I

I
I

i
The i

I t w a s f o u n d t h a t t h e d i a m e t r a l r e s i l i e n t m o d u l i o f t h e ADOT m i x t u r e a r e
within r h e acceptable range.
A t y p i c a l MSHTO s t r u c t u r a l l a y e r
T h e r u t t i n g p o t e n t i a l i s l o w in some c a s e s
c o e f f i c i e n t i s recommended.
R e c o n m e n d a t i o n s f o r t h e e v a l u a r i o r of
and m o d e r a t e i n o t h e r c a s e s .
f a t i g u e c r a c k i n g and t h e r m a l c r a c k i n q a r e p r o v i d e d .
The p o t e n t i a l for
m o i s t u r e damage i s h i g h , w h i l e t h e p o t e n t i a l f o r d i s i n t e g r a t i o n i s
marqinal.

i

j

.

; iT. Key L4<3rcs
i

/

'

I

A s p h a l t C o n c r e t e , IIlIMAS, Laboratory i
E v a l u a t i o n , R e s i l i e n t blodulus,
'
Strength, Creep
I

1

I

:G. jezurny C I ~ S S I ~ (c!
I ~ :n~s
: I ~r ~
exr.)

1

Unz12ssri1ed

i/

LV.

e
c

31s~,:12~113?
S;ze3e?.:
Dzcu~en1
:s aval;a31e:-, :ne

~inc;assi~ec

i

i

k a ~ ~ o nTaei z n ~ ~ z ir,:o:rr,a!~or;,
ai
Servrze. S~rln2ileJc'.V~:cin;a ,
,
2216'1

Sec;l:~iy C;assk:,a:~zr. (o! :n~soa?e;

C)P

.

u.S.
DUD~I: : n r ~ > ?
;r.e

, 2:

k3

o! Cases

1

j
ii sr,~,

I

CHAPTER 1. INTRODUCTION
1.1

PROBLEM STATEMENT
The asphalt concrete mix design in Arizona is based on the Marshall procedure (Ariz.

method 81%). Since the Marshail method of mix design is basically empirical, it results in index-

type values such as Marshall stability and flow. In general, empirical characterization parameters
are useful for comparison of materials under specific conditions. However, empirical correlations

are valid only for conditions similar to those under which they were ongirlally developed. Further,
empirical methods of characterization do not provide material properties needed for fundamental or
theory-based structural analysis of pavements. With the continuous increase in truck weight, tire
pressure and traffic -dolume, and with the fast deterioration of the nation's highway system, more
rational philosophy for asphalt concrete characterization is needed so that the pavement design can
be based on a more optimal manner.

NCHRP Project 9-6 (I), "Development of Asphalt-Aggregate Mixture Analysis System"
(AAMAS) has been recently completed and the final report is being revised(1). One of the
objectives of that project was to develop a more rational mixture characterization procedure based
on perfom.ance-related criteria. Phase II, Volume I of the final report is a procedural manual that
provides a complete evaluation procedure of hot-mixed asphalt concrete. Although the AAMAS
method is not a mixture design pnxedure bjr itself, it provides rational evaluation procedure that is
directly related to the mixture performance in the field. Since the project has been recently
completed, the AAMAS procedure has not been implemented by most states. However, various
highway agencies are planning to implement it in the near future. The implementation of the
AAMAS procedure by ADOT could be a major step forward towards rationalizing the asphalt
concrete mix design process.
According to the AAMAS procedure, six test types should k performed. Since some of
these tests are nondestructive, each specimen could be tested using different test types and at
different test temperatures. Some specimens are to be tested without "conditioning" while others

should be "conditioned" in which they are subjected to some treatments before testing. The
detailed test and analysis procedure are reported in Reference 1. A summary of the test procedure
is presented in Chapter 2.
1.2

OBJECTIVES
The objectives of this study are:
1)

To evaluate typical ADOT hot-mixed asphalt ccncsetes using the AAMAS (NCHRP
Roject 9-6 (1j) recommendations.

2)

To expand the ADOT database by providing typical lab test values for ADOT asphalt
concrete.

3)

To evaluate the amount of effort and equipment cost required for the AAMAS

procedure and discuss its potential use by ADOT engineers.

1.3

SCOPE OF WORK
The study includes laboratory evaluation of two sets of typical ADOT hot-mix asphalt

concrete specimens prepared using the California kneading compactor and the Marshall hammer.
All tests recommended by the AAMAS project (1) were performed at ASU highway materials
laboratory. The test results are analyzed using the AAMAS guidelines. The study is limited to one
asphalt grade (AC-30). one aggregate type and gradation, and one asphalt content.

CHAFTER 2. SUMMARY OF THE AAMAS RECOMMENDED PROCEDURE
2.1

BACKGROUND
The AAhlAS project was performed under NCHRP Project 9-6 (1) by Brent Rauhut

Engineering as the. prime contractor. The objective of the project was to develop an asphalt
aggregate mixture analysis system (AAMAS) for design of optimum paving mixtures based on
performance-related criteria. The AAMAS concept as it currently exists is applicable to hot-mixed
asphalt concrete, and includes mixture variables such as binders, aggregates and fillers used in the
construction of asphalt concrete pavements.
Specific items addressed in the current version of the AAMAS report include compaction of
laboratory mixtures to simulate the characteristics of mixtures placed in the filed, preparation and
mixing of materials in the laboratory to simulate the asphalt concrete plant production process,
simulation of the long-term effects of traffic and the environment (this includes accelerated aging
and densificaticn of the mixes caused by traffic), and the conditioning of laboratory samples to
simulate the effects of moisture induced damage and hzdening of the asphalt.
Recommendations are made for laboratory methods of executing the AAMAS and
evaluating the expected performance of dense graded asphalt concrete mixtures. Suggested
reco~~mendations
for incorporating results of the AAMAS program into a final mixture design
procedure are also provided. However, it should be noted that the AAMAS program developed
and reported in the current version is an evaluation procedure of a selected mixtllre and not a mix
design procedure in itself.
Current version of the AAMAS report(1) is only interim because of additional work being
cor~ductedunder Phase 111 of Project 9-6 (1) and because of the extensive SHRP (Strategic
Highway Research Program) asphalt research program currently underway. Some modifications
to the current procedure may be required after these other multi-million dollar research programs

are completed. The expected year of completion for the SHRP program is 1993.
At the time of this report, only a draft version of the AAMAS report was available. The

final version of the AAMAS report is expected to be released in the near future. According to the
principal author of the repoP, the difference between the current version and the final version is not
significant.

2.2

LABORATORY TESTS
Volume I of the AAMAS report includes a detailed labora;zy 9rogr;m to simulate the

characteristics of mixtures placed in the field. The complete AAMAS procedure q u i r e s six types
of tests as shown in Table 2-1. The fo!lowing paragraphs summarize these tests. These
paragraphs are not intended to provide a step-by-step procedure, but to discuss the significanceand
use of each test. The relation Setween the test results and the pavemen: response is also discussed.

Table 2 - 1. Lab tests recommended

Test
No.

Test
Name

the 1 I\.4MASproject

Loading

Sketch
1

I

Diametral resilient modulus
test

Paulses with 0.1 sec. duration and 0.9 sec.
rest period (ASTM D 4123)

2

Indirect tensile strength test

Loading until failure with a constant rate
of deformation of 0.05 or 2 inimin.

3

Indirect tensile creep test

Static load with a specified magnitude for
60 min. and unloading for 60 rnin.

4

Uniaxial compression
resilient modulus test

Pulses with 0.1 sec. duration and 0.9 sec.
rest period to compute resilient modulus

5

Unconfined compressive
strength test

Axial loading at a rate of 0.6 inlmin. mtil
failure

6

Uniaxial creep test

Static load with a specified magnitude for
60 min. and unloading for 60 min.

TEST 1 - DIAMETRALRESILIENTMODULUSTEST
When traffic moves over a pavement structure, a large number of stress pulses are rapidly
appiied to the different pavement layers. The concept of repeated load tests was developed to
approximate the dynamic loading conditions that actually occur beneath the pavement surface. One
of the common repeated load tests is the resilient modulus test. Briefly, pulse loads are applied to
asphaltic concrete specimens and the corresponding recoverable strains are measured. The resilient
modulus (ER)is defined as the ratio of the applied stress to the recoverable strain when a pulsating

I d is applied. It is used as one of the inputs of the multilayer elastic and the finite element design
methods of the highway pavement. In f x t , the resilient modulus of a visco-elastic material such as
asphait concrete is similar to Young's modulus of a linear elastic material.
The resilient modulus for the asphalt concrete can be determined in the laboratory by using
several different modes of repeated loads. Among these modes are the triaxial compression test,
uniaxial compression test, flexural beam test, direct tension test and diametral indirect tension test.
The diarnetral indirect tension test is prefemd over the other tests because it is simple, rapid and
requires Marshall size specimens (2).
The diarnetral test procedure to determine the resilient modulus of asphaltic concrete was
developed by Schmidt (3) and is standardized by ASTM D4123-82 test procedure. According to
this method, a pulsating compressive load is applied across a vertical diarnetral plane of Marshall
specimens every 1-3 seconds with a 0.1 second duration and the corresponding horizontal
deformation is recorded. This type of load produces a relatively uniform tensile stress acting
perpendicular to the applied load plane. Either the horizontal deformation only or both horizontal
and vertical deformations are measured. If both horizontal and vertical deformations are measured,
both resilient modulus and Poisson's ratio can be determined. If the horizontal deformation only is
measured, Poisson's ratio has to be assumed in order to determine the resilient modulus. The test
is recommended to be performed at 41, 77 and 1040F since the mixture response is very
temperature susceptible. Typical values of 0.3, 0.35 and 0.4 have been commonly assumed for
Poisson's ratio. Figure 2-1 shows a device that can be used to attach horizontal and verticd

LWTs to measure deforrnatims.
Typical Plots for load versus time and horizontal deformation versus time are shown in
Figure 2-2. Two resilient mcduli can be determined, instantanzous and total, depending on
whether the instantaneous or the total deformation is used. The AAMAS procedure requires the
determination of the total resilient modulus only which can bz calculated as follows:

where

ERT = Total indirect tension resilient modulus (psi)

P

= Repeated load (lb)

v

= Total resilient Poisson's ratio

t

= Thickness of specimen (in.)

HRT = Total recoverable horizontal deformation (in.)

The test is performed at two perpendicular positions and the results are averaged. The test
procedure is currently being revised by ASTM iv order to ensure more accurate and consistent
results.

-

TEST 2 INDIRECT TENSILE STRENGTH

When the load is applied on the pavement a tensile stress is developed at the bottom of the
asphalt concrete layer in most cases. Due to the repeated traffic load, these tensile stresses might
result in cracking the asphalt concrete layer. The knowledge of the tensile strength (the tensile
stress at failure) is important in developing a mechanistic method of pavement design.
The test is summarized in applying a compressive load w ~ L ;ii, zzazzint rate of deformation
along a diametrical plane of a Marshall-size specimen until failure. This type of loading produces a
failure due to tensile stresses acting perpendicular to the applied load plane. The load at failure is
recorded from which the indirect tensile strength is computed. The vertical and horizontal
deformations at failure are also recorded. The horizontal deformation at failure can be used to
compute the indirect tensile strain at fai!ure. The AAMAS procedure requires that the diameml

0.1 sec.

0.9 sec.

-It--------------l

Time

-

5

.C

0

E

s

INSTANTANEOUS
HORI ZONTAL
DEFORMATION

TOTAL
HORIZONTAL
DEFORWTION

C
W

5I

"I

C

Time

-

Figure 2-2. Typical plot of load and horizontal deformation during the resilient modulus test

resilient modulus test to be perhmed before the indirect tensile strength test. The indirect tensile
strength test is required to be performed along the axis of lower resilient modulus. I t is also
required to use a deformation rate of 0.05 in./minute at 41°F and 2 in./minute at 77 and !M°F.
The indirect tensile strength and the tensile strain at failure are calculated as follows:

where:

TEST 3

St

= Indirect tensile strength (psi)

Pfail

= Total load at failure (lb)

h

= Thickness of specimen (in.)

~t

= Tensile strain at failure (in.iin.)

Xt

= Total horizontal defamation at failure (in.)

v

= Poisson's ratio

- INDIRECT

TENSILE CREEP TEST

The response of a viscoelastic material such as asphalt concrete can be divided into two
parts, elastic and viscous. The elastic response is instantaneous, while the viscous response is
time dependent. Therefore, when a constant load is applied on an asphalt concrete specimen, it
will deform instantaneously and it will continuously deform (creep) as long as the load is applid.
The longer the load is applied, the larger the deformation. An example of this phenomenon is
when a heavy truck is parked for a long time on an asphalt pavement on a hot day, deformation
under the wheels could be noticed. On the other hand, if the same truck is driven on the same
pavement, no deformation could be noticed since the load is applied for a short period of time.
When the parked truck is removed most of the deformation will eventually be recovered. A small
portion of the deformation, however, may not recover causing permanent deformation (rutting).

In the creep test, a constant load is applied to the specimen and the deformation is

continuously measured. The load is then removed and the deformation is continuously measured.

In the indirect tensile creep test, a constant magnitude compression load is continuously
applied along the vertical diametral plane of a Marshall-size specimen. ?his type of load will result

in a tensile stress perpendicular to the applied load plane. Both vertical and horizontal
deformations are continuously recorded. The horizontal deformation is then used to calculate the
tensile creep srrain and the tensile creep modulus at a particular duration of time. After the load is
released, the rebound vertical and horizontal deformations are recorded over a fixed duntion of
time. The indirect tensile creep modulus is calculated at any l d n g time as follows:
0
1
c,(t) = -

&t(t)

where:
Ct(t)

= Indirect tensile creep modulus at time t (psi)

at

= Tensile stress (psi) = 0.156

P

= Applied load flb)

I-,

= Thickness of specimen (in.)

P

~ ~ ( t )= Tensile strain at time t (in./in.)

A H (t) = Horizontal deformation at time r (in.)

v

= Poisson's ratio

The detailed test procedure is shown in the AAMAS report. Currently, no ASTM or
AASHTO procedure exists for the indirect tension creep test.
TFST 4

- UNIAXIAL

COMPRESSION RESILIENT MODULUS TEST

The concept of the uniaxial compression resilient modulus test is similar to the diametral
indirect tension resilient modulus test except that an axial load is applied. Unlike the diametral
resilient modulus test, the uniaxial test results in uniform axial compressive stresses in the
specimen. In this test a pulsating uniaxial load is applied on a cylindrical specimen every one

second with a 0.1 second duration and the corresponding axial deformation is recorded. Similar to
the diamem! method of loading, two types of resilient moduli can l
x determined, instantaneous
and total, depending on whether the instantaneous or total axial deformation is used. The AAMAS
procedure require the determination of the total resilient modulus only which can be calculated as
follows:

ECT = ~

Repeated stress (psi)
~ tmav al a b l e axial swin

where:

E~ = Total uniaxial compressive resilient modulus (psi)
Repeated stress = Repeated axial load (Ib)/A
A = Cross sectional area (in.2)

Total recoverable axial strain = total recoverable axial deformation (in.)Ki
G = Gage length or specimen height (in.)
A 4 in. diameter and 4 in. high specimen is recommended by the AAMAS report. No

ASTM or AASHTO test procedure is currently available. The axial load versus time and axial
deformation versus time plots are similar to those in Figure 2-2 for the diarned loading.
TEST 5

- UNCONFINED

COMPRESSIVE STRENGTH TEST

In this test a uniaxial compressive laid is applied on a cylindrical specimen with a specified
rate of deformation until failure. The tea is standardized by AASHTO T167lASTM D 1074. In the
AAMAS report, however, a rate of strain of 0.15 inche$inch/minute is required. Therefore, for 4
in. high specimens the rate of deformation is 0.6 inches/minute. Also, the AAMAS procedure
requires a test temperature of 1040F.
At failure, the compressive stress, strain and strength are calculated as follows:

where:
= Unconfined compressive stress at failure (psi)

P

= Unconfined compressive load at failure (ib)

A

= Cross sectional area (in.2)

*

TEST 6

= Compressive strain at failure (inJin.)

A

= Axial deformation at failure (in.)

G

= Axial length or specimen height (in.)

Squ

= Unconfined compressive strength (psi)

- UNIAXIAL

CREEP TEST

In this test, a uniaxial constant magnitude compressive load is applied on a cylindrical
specimen, while the axial deformation is continuously recorded. The compressive stress,
compressive creep strain and the compressive creep modulus can in computed at any loading time

as follows:

E (t)

A 0)
=-

G

Oc
& (t) = -

E (t)

where:

oc

= Compressive stress (psi)

P

= Compressive load (:b)

A

= Cross sectional area (in.2)

E (t)

= Compressive creep strain at time t (in.fin.)

A (1)

= Vertical deformation at time t (in.)

G

= Gage length or specimen height (in.)

C,(t)

= Compressive creep modulus (psij

Currently, a tentative ASTM procedure for the uniaxial creep rest is being developed. No
AASHTO procedure is available at that time. The AAMAS procedure requires 4" x4" specimen
size and a test temperature of 1040F. A loading time of 60 minutes and an unloading time of 60
minutes are required. A typical plot of vertical deformation versus time is shown in Figure 2-3.
The slope and intercept of the creep curve can be obtained from this plot

2.3

SPECIMEN CONDITIONING
A rational evaluation of asphalt concrete specimens should simulate the field condition in

the lab as much as possible. Of c o r n , it would be impractical, and may be impossible, to exactly
duplicate the field condition in the lab because of the many factors involved. However, a
reasonable and practical laboratory conditioning is needed.
Conditioning the specimen in the lab to simulate field conditions should not be confused
with load conditioning used in some lab tests. Conditioning to simulate field conditions include
moisture conditioning, aging, traffic densification, etc. On the other hand, load conditioning is
used for other purposes in order to ensure proper deformation measurements. For example,
during the resilient modulus test the load has to be applied to the specimen a few hundred times
before the results are recorded in order to exclude the permanent deformation and to ensure full
contact between the specimen and the loading heads. In the AAMAS report, as well as this report,
the term "conditioning" is used for both purposes. Sometimes, the term "preconditioning" is also
used.
The AAMAS study recommends three types of specimen conditioning. The detailed
procedure is shown in the AAMAS report. A summary of the procedure is as follows:
a. Moisture Conditioning
Specimens are vacuum saturated until the water absorption is greater than 80%.
Specimens are then frozen for 16 hours and thawed at 1400F for 24 hours and at 770F for 2
hours.
14

b.

Envinxmental

Aging (Temperature Conditioning)

Specimens are heated at 1400F for 2 days and at 225°F for 5 days. Specimens are
then cooled at 410F for one day.
c . Traffic Densification
Specimens are further compacted to simulate traffic until refusal or until the find air
void level is reached. The gyratory testing machine (ASTM D3387) or the gyratory shear
compactor (ASTM D4031) is recommended. If either of the gyratory machines is not available, the
California kneading compactor can be used (AASHTOT2471ASTM Dl 561).
Table 2-2 shows a summary of specimen conditioning names and procedu~.
Table 2-2. Specimen conditioning

Conditioning

No.
Vacuum saturation + freeze and thaw
b

Environmental aging

Heating for 7 days + cooling for one

&Y
c
2.4

Traf3ic densifiwtion

Heating and further compaction

GROUPING AND TEST SEQUENCE
The complete AAMAS procedure requires 24 specimens; eighteen 4" x 2.5"specimens and

six 4" x 4" specimens. Specimens are grouped into 8 sets of 3 specimens each. The first 6 sets
have 4" x 2.5" specimens, while the last 2 sets have 4" x 4" specimens. The specimens are
grouped in such a way that the average unit weight (and air voids) of the different sets are
approximately equal. Table 2-3 shows the conditioning as well as test type, sequence, temperature
and measurements for each specimen set. Table 2-4 shows the approximate time required to
complete the AAMAS evaluation of one set of a..phalt mixture.

Table 2-3. b-ditioning, test sequence and measurements

Set No.

No. of
Specimens

Size
(in.)

Conditioning*

Test No.**,
Sequence &
Terrq>erature

None

1, 2 (410F)

Diametral resilient modulus, indirect tensile
strength, tensile strain at failure

None

1,2 (770F)

Same as set 1

None

1,2(1400F) Sameasset!

a

1, 2 (77OF)

Same as set 1

3 (41oF)

Slope and intercept of creep c w e , tensile
creep modulus

Measurement

b

b

4, 5 (I WOF) Axid resilient modulus, unconfined
compressive strength, compressive strain at
failure
6 (104OF)

* See Table 2-2

**

See Table 2-1

Compressive creep strain, compressive
creep modulus

CHAPTER 3. EVALUATION OF ADOT MlXTURE PROPERTIES
3.1

SPECIMEN PREPARATION
The complete AAMAS procedure requires 24 specimens; eighteen 4" x 2.5" specimens and

six 4" x 4" specimens. The AAMAS report recommends the use of the gyratory compactor
(ASTM D3387 or D4013) since it closely simulates the mix compaction in the field. The repon
further concludes that the California kneading compactor (AASHTO T247/ASTM Dl5611 is the
next preferred device, while the hlarshall hammer (AASTO T24SjASTM D1559) is the least
desirable comTactor. Since the Marshall hammer is the most commonly used compactor and it is
the compactor currently used by ADOT, there is a need to compare between the responses of
Marshall-compacted specimens and specimens compacted by other devices. Also, since the
gyratory compactor is not currently available at either ADOT or ASU lab, it was decided to evaluate
two sets of 24 specimens; one set compacted by the kneading compactor and the other set
compacted by the Marshall hammer. It was also decided that the bulk specific gravities and air
voids of all specimens should be similar to those of typical ADOT mixes.
The aggregate was provided by ADOT which was originally obtained from United Metro

No. 1 located in Phoenix, Arizona. The aggregate is river deposit produced from the Salt River in
Arizona. Table 3-1 shows the aggregate gradation.
Table 3- 1. Aggregate gradation

Sieve Size

96 Passing

314 in.
1/2 in.
3/8 in
1/4 in
No. 4
No. 8
No. 40
No. 200

100
98
83
64
58
45

19
4.5

An AC-30 asphalt cement was used in this study which was provided by ADOT and
originally obtained from Sahuaro Petroleum and Asphalt Company in Phoenix, Arizona.
The mix ingredient was designed by ADOT according to ARIZ 8i5c procedure (4). The
design asphalt content was 4.7% by total weight of mix. The bulk density according to ARIZ
41 5a test procedure was 145.2 pcf. The maximum theoretical density of the loose mixture (Rice

test) was determined by ADOT according to ARIZ 806c procedure and found to be 152.4 pcf. The
air voids of the compacted specimens is 4.7%.

3 . 1 . 1 KNEADING COMPACTED SPECIMENS
The first set of specimens was compacted at ASU using the California kneading compactor
(Figure 3-1). Although the AASHTO T247 calls for 20 kneading blows at 250 psi and 150 blows
at 500 psi for the 4" x 2.5" specimens, the compacting effort was adjusted on a trial-anderror
basis to achieve a bulk specific gravity and air voids similar to those of typical ADOT mixtures.
After several trials it was found that 20 blows at 250 psi followed by 100 blows at 500 psi provide
the required density. The 4" x 4" specimens were compacted using the same pmedure except for
20 blows at 250 psi and 120 blows at 500 psi. Table 3-2 shows the thicknesses and unit weights
of the kneading compacted specimens. Note that specimens are numbered and p p e d into 8 sets
in such a way that the average unit weights of all sets are similar as recommended in the AAMAS
report (1). The percent air voids and actual diameters are shown in the data sheets in Appendix A.
3.1.2. MARSHALL COMPACTED SPECIMENS
The second set of specimens was compacted by ADOT using the manual Marshall
hammer. Seventy five blows on each side of the specimen were applied on the 4" x 2.5"
specimens, while 100 blows on each side were appiied on the 4" x 4" specimens. Table 3-3
shows the thicknesses and unit weights of the Marshall compacted specimens. As before,
specimens were numbered and grouped into 8 sets in such a way that the avenge unit weights of
all sets are approximately equal as recommended in the AAMAS report. The percent air voids and

actual diarneters are shown in the data sheets in Appendix B.
20

Table 3-2 Thickness and unit weights of kneading compacted specimens

Set No.

Specirnen
No.

Nominal
Actual
Thickness(in.) Thickness(in.)

Unit Weight
( ~ f )

Average Set
Unit Weight

(PO

Table 3-3 Thicknesses and unit weights of Mmhall mmpactcd spccinrns

Set No.

Specimen
No.

Actual
Unit Weight
Nominal
Thickness(in.) Thickness(in.)
(pcf)

Average Set
Unit weight
(pcf)

3.2

SPECIMEN CONDITIONING
Both kneading compacted and Marshall compacted specimens were conditioned according

to the AAMAS procedure. Three types of conditioning were used; moisture conditioning,
environmental aging and traffic densification as discussed in Chapter 2.
The moisture conditioning was performed as recommended by the AAMAS study. When

the watcr absorption was computed after vacuum saturation a value of slightly more than 100%
was obtained for all specimens. This indicates that there was a minor error in either the bulk
specific gravity or in the theoretical maximum specific gravity determination. It was felt,
however, that the 80% minimum saturation requirement has been satisfied.
Since the gyratory compactor was not available, the California kneading compactor was
used for the traffic

densification of both kneading ~0qXictedand Marshall compacted specimens.

Since it was not easy to define "refusal" using the kneading compactor, 500 kneading blows were
applied to each specimen in order to ensure that refusal have been reached. Table 3-4 shows the
bulk unit weights and air voids before and after oaftic densification.

3.3

LABORATORY TESTING
All tests were conducted at the ASU highway materials lab. The two main pieces of

equipment used in the study are two electrohydraulic closed-loop testing machines. The first
machine is manufactured by the Structural Behavior Engineering Laboratory (SBEL) in Phoenix,
Arizona (Figure 3-2), while the second machine is manufactured by Instron Corporation in
Canton, Massachusetts (Figure 3-3). Both machines are connected to microcomputers and are
capable of applying either static or dynamic loads with different loading types and magnitudes.
The reason for using two machines in this study is that the load cell size and the computer pmgram
of the SBEL machines are suitable for applying small loads necessary for resilient mdulus testing,
while the Instron machine is suitable for applying heavy loads required for other tests. It should be
noted, howevrr, that these equipment qualifications are specific to the machine models available to
the ASU highway materials lab and either monufactui-ingcompany is capable of manufact~ring

Figure 3-3. Insmn elmhydraulic closed-loop testing machine during the indirect tensile
strtngth test

machines with a wide range of capabilities. In this s t ~ d ythe
, SBEL machine was used to perform
tests l , 3 , 4 and 6, while the Instron machine was used for tests 2 and 5 as defined in Table 2- 1.
Table 3-4 Average unit weights 2nd air voids before and after traffic densification

Propeny
Unit weight before (pcf)
Unit weight after (pcf)

Kneading

Marshall

Set No.

Compacted

Compacted

Average

7

145.3
148.4

145.2
147.1

145.3
147.8

7

Unit weight before (pcf)
Unit weight after (pcf)
Air voids before (96)
Air void after (%)
Air voids before (8)
Air void after (8)

An environmental chamber manufactured by BEMCO Inc. in Pacoima, California was used
during testing. The chamber can be attached to either testing machines so that the test could be
performed at the required temperature. The chamber is capable of maintaining a wide range of
temperatures above and below mom temperature.
Testing was performed as specified in the AAMAS study. In the unconfied compressive
strength test (test 5 ) , however, an i n c o m t rate of deformation was initially used. Therefore, the
incorrect results were discarded and the test was later repeated using the correct rate of deformation
on the specimens which were tested for uniaxial creep (test 6). It is believed that the creep effect
was fully removed after several days of storing the specimens in an unloading condition.

In the indirect tensile strength test (test 2) the horizontal deformation could not be measured
due to technical difficulties in the elecuonic quipment. Instead, the vertical deformation was
measured from which the horizontal deformation was computed according to Equation 3-1 from

where:

v = Poisson's ratio
AH = horizontal deformation, in.
AV = vertical deformation, in.
3.4

TEST RESULTS
The detailed test results are shown in Appendices A and B for kneading compacted and

Marshall compacted specimens, respectively. Tables 3-5 through 3-8 show the average test results
of unconditioned, moisture conditioned, environmental conditioned and traffic densified
specimens, respectively. Figure 3-4 illustrates the average diameml resilient modulus at different
temperatures obtained from Table 3-5.
Table 3-5 Average test results of unconditioned spec:rnens

Set
No.

Temp
F

Diametrai ER(ksi)
Tensile strength (psi)
Tensile strain (milshn)

1
1
1

41
41
41

2,414
274
3.0

2,37 1
357
4.0

2,393

Diarnetral ER(ksi)
Tensile strength (psi)
Tensile strain (milsfin)

2
2
2

77
77
77

785
283
6.7

943
322
6.5

864
303
6.6

Diametral Ell (ksi)
Tensile strength (psi)
Tensile strain (mils/in)

3
3
3

104
104

i 92

179
107
6.0

186
100

Propeny

104

Kneading

Compacted

93
8 .?

Marshall
Compacted

Average
316
3.5

7.1

Table 3-6 Average test results of rn0istu.e conditioned specimens

Property
Diametral ER(ksi)
Tensile strength (psi)
Tensile saain (milsh)

Set
No.

Tern P
F'

4

77
77
77

4
4

Kneadinn
compact&

Marshall
Compacted

Average

296
81
4.0

399
87
6.4

348
84
5.2

Table 3-7 Average test results of environmental conditioned specimens
Kneading

Property

No.

Marshdl
Compacted Average

Diametral ER(ksi)
Tensile strengrh (psi)
Tensile strain (rniishn)
Slope of creep curve (1@ 5 milsjsec)
Intercept of creep curve (mils)
Tensile creep modulus at 3,600 sec.(ksi)

6
6
6

41
41
41

2.9
0.1
92

0.6
0.1
475

1.8
0.1
284

Table 3-8 Average test results of traffic densified specimens

Property

Set
No.

Temp Kneading Marshall
F
Compacted Compacted

Average

Axial ER(ksi)
Compressive strength (psi)
Compressive strain (mildin)
Slope of creep curve (1@5 mildsec)
Intercept of creep curve (mils)
Compressive creep modulus (ksi)
at
10 sec
100 sec
1,000 sec
3,600 sec

8
8

104
104

25
7.2

31
8 .O

28
7.6

8
8
8
8

101
104

15.5
11.1
9.8
9.2

15.2
10.4
8.0
8.1

15.4
10.8
8.9
8.7

104
104

40

60

80

Temperature, F

Figure 3-4. Avenge total diarnetral resilient moduli of AMYT specimens at varims ternmtures

CHAPTER 4. PREDICTION OF ADOT MIXTURE PERFORMANCE
4.1

BACKGROUND
The concept of relating the mixture properties to the pavement performance is logical and

appropriate in order to optimize mixture and structural designs. Many models have been developed
by previous researchers for this purpose. However, these models are limited in use to some
degree. The main reason is that the available models are not comprehensive enough to cover all
variables involved in pavement performance. The AAMAS study presents some guidelines to
provide a recommended practice for evaluating asphalt concrete mixtures based on performance
related criteria. These guidelines are based on models suggested for use in NCHRP Project 1-26.
The AAMAS procedure consists of a series of steps using results from the test program,
discussed in Chapter 2, as well as interactions with various models predicting the types of distress
more common with asphalt concrete pavements. The final product of the AAMAS are the
structural and material combinations needed to meet the design requirements or assumptions used
hy the pavement design engineer. In this chapter, the properties of typical ADOT mixtures
prewnted in Chapter 3 are compared with the performance-related criteria reported in the AAMAS
study.

4.2

AASHTO STRUCTURAL LAYER COEFFICIENT
The 1986 AASHTO guide(5) recommends the estimation of the smctural layer coefficient

from the resilient modulus measured at 680F in accordance with the ASTM D4123. The AAMAS
study, however, recommends the consideration of the environmental effects on the structural
design by considering the seasonal fatigue damage. In other words, use seasonal resilient moduli
to calculate seasonal fatigue damage and sum the seasonal damages to determine an annual damage.
From the annual damage, an effective asphalt concrete resilient modulus can be calculated which
can be used to estimate the structural layer coefficient.
The following is a step by step prmedure that can be used to ensure that the asphalt
concrete mixture meets or exceeds the layer coefficient assumed during structural design.

Obtain the seasonal average pavement temperature for each season.
Determine the total resilient mdulus at each seasonal temperature. Figure 4-1 shows
the acceptable range of moduli (unconditioned)at various temperanrres.
Obtain the fatigue factor for each seasonal Resilient Modulus from Figure 4-2.
Calculate the effective resilient modulus using Equation 4- 1.

where:

ERE = Effective resilient modulus based on a fatigue damage approach
E ~ ~= ( T~ota)l resilient moduius as measured by ASTM D 4123 at the avenge

pavement temperature for season i
FF

= The fatigue factors obtained from Figure 4-2

This effective resilient modulus should equal or exceed the mdulus value used to estimate
the AASHT'O structural layer coefficient used for design (Figure 4-3 (5)). The GPS (General
Pavement Sections) projects of the SHRP LTPP (Long-Term Pavement P e r f m c e ) program are
to provide the necessary pavement performance data to find the resilient modulus - AASHTO layer
coefficient relationship to be adequate, with or without modificstion, or inappropriate.
..
Exam* for Dcmmnmg Stmctural Lrl-ver Coefficier~
Asphalt concrete; AC- 30
Aggregate type and gradation: as used in this study
Asphalt content: 4.78 by total weight of mix
Seasonal average pavement temperatures:
Fall:

80oF

Winter

70°F

Spring:

85OF

Summer

lWF

Temperature, F
Figure 4-1. Acceptable range of &ametral total resilient moduli at various ternperatures(1)

Struclural Layer Coelliciont, a! , lor
Aspllalt Concrelo Surface Course

Comparing the test results for unconditioned A D O T specimens (Table 3-5) with the
recommendations in Figure 4-1, it can be seen that the average ADOT moduli are within the
appropriate range at all 3 test temperatures: 41,77 and 104OF.
From Figure 3-4 the moduli at 80, 70, 85 and lOOOF are 700, 1050, 560 and 260 ksi.
From Figure 4-2 the corresponding fatigue factors are 0.50, 0.22,0.80 and 3.6. Using Equation
4-1, the effective resilient modulus is 384 ksi. Figure 4-3 shows that the structural coefficient (al j
for this material should be 0.42. The ADOT design manual shows that if the modulus at 700F is
1,050 ksi according to this study, the structural coefficient a, is outside of the normal range and
should be limited to about 0.54 (Reference 4 , Figure 202.02-3).
4.3

RUTTING
Two types of rutting are considered; 1) densification or one dimensional consolidation and

2) the lateral movement or plastic flow of asphalt from wheel loads. The more severe premature
mmng failures and distortion of asphalt concrete materials are related to lateral flow andlor loss of
shear strength of the mix, rather than densification. Currently, there is no mechanistic/empirical
model that adequately considers the lateral flow problem (1).
Rumng from onedimensional consolidation can be estimated using the traffic densification
procedure recommended in the AAMAS repon (1). Limiting the air voids at mixture ~zfusallimits
the amount of additional densification caused by traffic, assuming that the mixture is properly
compacted on the roadway to an air void level between 5 to 7 percent. The air voids at mixture
refusal should be greater than 2 percent when compacted with the gyratory devices (1). Table 3-4
indicates that the air voids after densification of specimens tested in this study are greater than 2%.
Therefore, the possibility of tuning due to one-dimensional consolidation is small.
A few mathematical models are reported in the AAMAS report to estimate the rutting rate of

asphalt concrete layers in the field. Figures 4-4 through 4-7 illustrate graphical solutions of the
range of data that can be generated for different pavements, climates and loading conditions. The
figures can be used as p s s guidelines for mixture evaluation on high-volume roadways.

The average compressive creep moduli reported in Table 3-8 for typical ADOT mixtures are
plotted in Figures 4-4 through 4-7. It can be concluded that the ADOT mixture has a low rutting
potential for the lower layers of full-depth asphalt pavements (Figure 4-4). On the other hand, it
has a moderate rutting ptential (marginal) for the intermediate layers in thick or fulldepth asphalt
pavements (Figure 4-3,for surface layers (Figure 4-6). and for layers placed over rigid pavements
or rigid base maserials (Figure 4-7).
4.4

FATIGUE CRACKING
Fatigue failures are accelerated by high air voids, which in addition to creating ;1weaker

mix,also increases the oxidation rate of the asphalt film. The development of fatigue cracks is
related to the tensile strain at the bottom of the asphalt concrete layer and to the modulus of the
asphalt concrete material as follows:

K1 (et)-'"

N

=

N

= Number of allowable wheel load applications

ct

= Tensile strain at the bottom of the asphalt concrete layer

K1

=

n

= 1.75 -

EA

= Resilient Modulus of the Asphalt concrete at a selected temperature, psi

E

= Reference Modulus (from the AASHO Road Test; ER = 500,000 psi)

KIR

= Reference coefficient for EA = ER (from AASHO Road Test data, KIR =

where:

K 1R ( E A E R ) - ~
0.252 log K 1

7.87 x 10-7)

The tensile strain at the bottom of the asphalt layer can be calculared using the elastic layer
theory (e.g. ELSYM5(6) or Chevron(7) computer program). The modulus or the asphalt concrete
layer can be obtained from the lab test using the ASTM D4123 procedure (total resilient modulus).
To ensure that the asphalt concrete layer has the necessary fatigue resistance for the specific
structure, the following equation can be used as a check for each season:

~f

(i) < &t (i)

where:
€f(i)

= E* (i) + [ A

A

= at [l - (0.9)btl E* (i)

m

= 1.4 (bt)

et (i)

= Iadirect tensile strain at failure (unconditioned)for season i

~f

+ m log N ]

(i) = Accumulated permanent tensile s

E,-~(i) =

~ (fatigue)
n
for season i

T d recovered tensile strain at the average seasonal temperature for season i

bt, at = Slope and intercept of the indirect tensile creep - time curve at the average
seasonal temperature for season i
Either Equation 4-2 or 4-3 can be used to estimate the "fatigue life" or fatigue resistance of
a specific pavement structure. Equation 4-2 uses the resilient modulus data in Figure 3-4 together
with the use of a multilayer elastic program. On the other hand, Equation 4-3 requires additional
laboratory fatigue tests in addition to the test results obtained in this study.
4.5

THERMAL CRACKING
Thermal cracking is a non-traffic associated type of failure. This type of cracking presents

a serious problem during mixture design, because it is difficult to evaluate and predict. The reason
for this difficulty is related to the aging characteristics and kisco-elastic pr0peni.e~of the asphalt.
Low-temperature cracking results when the tensile stresses, caused by temperature drops, exceed
the mixttlres fracture sue~gth.The e t e at which thermal cracks occur is dependent on the asphalt
rheology prciperties, the mixture propemes and environmental factors.
To evaluate thermal cracking, certain critical mixture properties must be determined, as well
as, project specific environmental conditions. The mixture propemes include indirect tensile
strength, low-temperature creepmodulus, failure stxiins and the thermal coefficient of contraction.
These parameters can !x used to calculate the occurrence of thermal crack with time.
The critical temperature at which cncking occurs can be calculated as follows:

- [ St(14)/aAbC]

,
T

= Tg

,T

= Critical Temperature that cracking is expected to occur at, OF

TB

= Base Temperature, normally assumed to be the Ring and Ball Temperature of the

where:

asphalt, O F

n

= Slope of the relationship between indirect tensile strength and total resilient

modulus of the mixture measured at temperatures of 41,77 and 104F
(unconditioned)
bc

= Intercept on the indirect tensile creep modulus axis of the relationship between

creep modulus and indirect tensile strength measured with a loading rate of 0.05
inches/minute
log bc = n log St(41)/logCt(41)
St(41) and Ct(41) = Indirect tensile strength and creep modulus measured at 41°F,
respectively
aA

= Thermal coefficient of contraction of the asphalt concrete (typical values range

from 1.0 x 10-5 to 1.8 x 10-5 inJin.PF
Based on this criteria the cracking potential of ADOT mixture can be evaluated. The
available data, however, are not enough to provide complete evaluation.
4.6

MOISTURE DAMAGE
Moisture damage is a serious problem, particularly on high traffic roadways. It is caused

by a loss of adhesion or bond between the asphalt and aggregate in the presence of moisture.
Currently, the moisture damage evaluation (tensile strength and resilient modulus ratios, TSR and

MRR) of AAMAS is simply used as a means of accepting or rejecting a mixture. Both of these
ratios should exceed a value of 0.80 for dense graded asphalt concrete. If values less than 0.80 are
measured, an asphalt additive or antistripping agent may k required or the aggregate blend may
need moditication.

For the -4DOTmixture, the tensile strength ratio and the resilient modulus ratio at 770F can
be obtained from Table 3-5 (moistureconditioned) and Table 3-5 (unconditioned) as follows:

TSR = 54 = 0.28
303

It can be seen that both ratios are less than 0.80 which indicate high potential for moisture damage.
Therzfore, an asphalt additive or antistripping agent may be required or the aggregate blend may
need modification. According to the AAMAS study, however, additional work is being conducted

in the moisture damage area and these results are not considered final.
4.7

DISINTEGRATION
Disintegration is primarily related to environmental and material factors, but the severity of

the distress is dependent upon the magnitude and number of wheel load applications. Raveling and
reduced skid resistance are the two disintegration distresses consickred in AAMAS. Increasing t!!
asphalt content in the mix will increase film thickness and decrease asphalt aging, reducing the
severity of raveling. Conversely, this increase in asphalt content will also reduce air voids, which
can increase the possibility of flushing (or bleeding) and reduce skid resistance. Thus, both upper
and lower bounds on asphalt content exist and must be considered in mixture design to reduce
disintegration distresses (1).
The following summarizes the AAhlAS criteria that can be used as guidelines in the
interim, 10 evaluate the acceptability of surface mixtures as related to disintegration:
Air voids at refusal > 3%
Indirect tensile smngth ratio, TSR > 0.90
Retained bond > 0.35
Tensile strain at failure (77F) > 10 milgin.
Retained bond = E~
where

q A = Indirect tensile strain at failure measured on specimens that have been temperature
conditioned (accelerated aging).

E,

= Indirect tensile strain a: failure measured on unconditioned specimens

The following information are obtained h m the ADOT mixture.
* Air voids at refusal (Table 3-4) = 3%

(OK)

84

Indirect tensile strength ratio, TSR (from Tables 3-6 and 3-5) = 303 = 0.28
Retained bond at 410F (from Tables 3-7 and 3-5)

=%

= 0.77(OK)

Tensile strain at failure (770F) (Table 3-5) = 6.6 mildin.

It can be seen that two of the conditions are satisfied, while the other two are not satisfied.
This indicates that the ADOT mix is subjected to disintegration to some extent.

CHAPTER 5. SUhlMARY AND CONCLUSIONS
In this study a typical ADOT asphalt concrete mixture was evaluated based on the AsphaltAggregate Mixture Analysis System (AAMAS) procedure (NCHRPProject 9-6(1)). Two sets of

.4DOTasphalt concrete specimens were prepared using the California kneading compactor and the
manual Marshall ha-.

All tests recommended by the AAMAS project were performed. The

test results were analyzed using the AAhlAS guidelines.
It was found that the diametral resilient moduli of the ADOT mixture are within the
acceptable range. A typical AASHTO structural layer coefficient is recommended. The rutting
potential is low in some cases and moderate in other cases. Recommendations for the evaluation of
fatigue cracking and thermal cracking are provided. The potential for moisture damage is high,
while the potential for disintegration is marginal.
Further research is currently being conducted by the Strategic Highway Research Program

(SHRP). Thus, the results obtained in this study should not be considered final. More studies are
needed for comprehensive evaluations of ADOT asphdt concrete performance.

CHAPTER 6. RECOMMENDATIONS FOR FURTHER RESEARCH
This study serves as a preliminary evaluation of a typical ADOT asphalt concrete based on
performance-related procedure. Further work is needed for mare comprehensive evaluations. The
following are some guidelines for future studies.
1. A wide range of asphalt cement grades and aggregate gradations should k used.

2. Mixtures with asphalt modifiers and/or antistripping agents should be evaluated.
3. The use of the gyratory compaction device should be investigated and compared with
the Marshall compactor currently being used by ADOT.
4. Field performance of pavements in Arizona should be evaluated and cor.elated with
laboratory results in order to develop more rational models for pavemen: performance

and a more optimal method of mixture desip.
5. The need for upgrading the capability of ADOT materiais lab should be considered.
Possible equipment that can be obtained include the following.

m
1. Electrohydradic testing machine

2. Environmental chamber to be attached to the
electrohydrdulic machine

3. Gyratory shear compactor (ASTM D4013)
or Corps of Engineers Gyratory
Testing Machine (ASTM D3387)

xlmate Cost

70-90

REFERENCES
1.

Von Quintus, H. L., et. al., "Interim AAMAS-Procedural Mmual Asphalt-Aggregate Mixture
Analysis System," Final Report, NCHRP Project 9 - q l ) , Brent Rauhut Engineering, Austin,
TX, August 1989.

2. Marnlouk, M.S. and Sarofirr., R. T., "Modillus of Asphalt Mixtures - An Unresolved
Dilemna," Record No. 1171, Transportation Research Board, Washington, D.C., 1989.

3. Schmidt, R. J., "A Practical Method for Measuring the Resilient Modulus of Asphalt Treated
Mixes," Record No. 404, Highway Research Board, Washington, D.C., 1972.
4.

"Preliminary Engineering and Design Mmual," Arizona Department of Transportation,
Phoenix, AZ, March 1989.

5.

"AASHTO Guide for Design of Pavement Structures," American Association of State
Highway and Transportation Officials, Washington, D.C. 1986.

6. Kopperman, S., et. al., "ELSYMS Interactive Version User'sManual," IBM-PCversion,
SRA Technologies Inc., Arlington, VA, July 1985.

7. Michelow, J., "Analysis of Stresses and Displacements in an N-LayeredElastic System under
a Load Uniformly Distributed on a Circular .4rea,"California Research Carporation,
Richmond, CA, Sept. 1963.