Available online at www.sciencedirect.

com
Available online at www.sciencedirect.com
ScienceDirect
ScienceDirect
Available
Available online
online at at www.sciencedirect.com
www.sciencedirect.com
Structural Integrity Procedia 00 (2016) 000–000
www.elsevier.com/locate/procedia
ScienceDirect
Structural Integrity Procedia 00 (2016) 000–000
ScienceDirect www.elsevier.com/locate/procedia

Procedia Structural
Structural IntegrityIntegrity
Procedia2 (2016) 2913–2920
00 (2016) 000–000
www.elsevier.com/locate/procedia

21st European Conference on Fracture, ECF21, 20-24 June 2016, Catania, Italy
21st European Conference on Fracture, ECF21, 20-24 June 2016, Catania, Italy
Stress-strain behaviour of asphalt concrete in compression
Stress-strain
XV Portuguese behaviour
Conference ofPCF
on Fracture, asphalt concrete
2016, 10-12 in2016,
February compression
Paço de Arcos, Portugal
Lee Leon*, Raymond Charles, Nicola Simpson
Thermo-mechanical modeling
Lee Leon*, Raymondof a high
Charles, pressure
Nicola Simpson
University of the West Indies, St. Augustine, Trinidad & Tobago
turbine blade of an
airplane
University gasSt. turbine
of the West Indies, engine
Augustine, Trinidad & Tobago

Abstract
Abstract P. Brandãoa, V. Infanteb, A.M. Deusc*
This paper presents an evaluation of the failure mode of asphalt concrete and describes the stress-strain curve that governs asphalt
a
concrete
This paper Department
beyond
presents oflimit
thean Mechanical
evaluation Engineering,
of elasticity
of the in theInstituto
failure Superior
ascending
mode of Técnico,
branch.
asphalt TheUniversidade
concrete relationship
and describesdeofLisboa, Av. Rovisco
stress-strain
the stress-strain Pais,
curve1,that
is identical 1049-001
to governs
that ofLisboa,
cement
asphalt
Portugal
concrete in compression. Aofgeneral form of the behavioural curve is proposed to of
represent the elastic and plastic – damage of
b
IDMEC, Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001cement
beyond the limit elasticity in the ascending branch. The relationship stress-strain is identical to that of Lisboa,
asphalt
concreteconcrete. The nonlinear
in compression. A generalbehaviour
form of of the
asphalt concrete curve
behavioural beyond is
Portugal the linear
proposed elastic
to limit
represent swelling
the elasticis visible
and to
plastic the
– naked
damage eye.
of
These
asphalt
c parameters
concrete.
CeFEMA, that nonlinear
The
Department define thisbehaviour
formEngineering,
of Mechanical are physically
of asphalt significant
concrete
Instituto and
beyond
Superior arethedetermined
Técnico, linear experimentally.
elastic
Universidade delimit swelling
Lisboa, The parameters
is visible
Av. Rovisco Pais,to ofnaked
1, the the stress-
1049-001 eye.
Lisboa,
strain
These curve are fundamental
parameters that define this characteristics of a solid
form are physically subjected
significant to are
short
Portugal
and term loading.
determined The experiments
experimentally. used shortofterm
The parameters static
the stress-
compression
strain curve are loading on cylindrical
fundamental asphalt concrete
characteristics of a solid specimens,
subjected differing in mix
to short term types (Dense
loading. graded, Stone
The experiments usedMatrix),
short termair voids
static
content,
compressiondensity, specimen
loading size and temperature.
on cylindrical asphalt concrete Therespecimens,
were significant changes
differing in mixintypes
the peak stressgraded,
(Dense and strains
Stonebetween
Matrix),theairasphalt
voids
Abstract
concrete
content, mix types.
density, Different
specimen strengths
size and are attainable
temperature. There werebetween asphalt
significant concrete
changes in thetypes
peakasstress
wellandas strains
the properties
betweenof thethe mix
asphalt
constituents.
concrete mixThe proposed
types. Differentequations fit aare
strengths wide range of between
attainable testing conditions and asphalt
asphalt concrete typesconcrete
as welltype as for
the both the ascending
properties and
of the mix
During their operation,
descending modern aircraft engine components are Thesubjected to increasingly demanding operating andconditions,
constituents.branches
The proposedof the equations
stress-strain
fit adiagram
wide rangein compression.
of testing conditions investigation
and asphalt derives parameters
concrete type for (yield
both thestress
ascending elastic
and
especially the
modulus) high pressure turbine (HPT) blades. Suchcan conditions cause these partssuch to undergo different types ofelastic
time-dependent
descendingforbranches
different ofasphalt concrete
the stress-strain mix types
diagram which
in be used
compression. Thein FE programs
investigation derivesas parameters
Abaqus that(yield
uses the
stress and as well
elastic
degradation, one of which is creep. A model using the finite element method (FEM) was developed, in order to be able to predict
as plastic data
modulus) to modelasphalt
for behaviour
different the behaviour
concrete of different mix types
canof asphalt concrete material used in a pavement structure.
the creep of HPT blades.mix typesdata
Flight whichrecords be(FDR)
used infor FEaprograms
specific such as Abaqus
aircraft, provided thatbyuses the elastic asaviation
a commercial well
© 2016
as plasticThe Authors.
data to model Published
the by Elsevier
behaviour of B.V. mix types of asphalt concrete material used in a pavement structure.
different
company,
Copyright wereThe
© 2016 used to obtain
Authors. thermal
Published by and mechanical
Elsevier B.V. This is data
an foraccess
three article
different
underflight cycles. In order to create the 3D model
Peer-review
© 2016
needed under
the responsibility
TheforAuthors. FEM Published
analysis, ofaElsevier
by the
HPTScientific Committee
B.V. scrap
blade ofopen
ECF21.
was scanned,
the CC BY-NC-ND license
and its chemical composition and material properties were
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review
obtained.under
Peer-review underresponsibility
The responsibility
data of Scientific
that wasofgathered
the the Scientific Committee
was Committee
fed into theof FEM ofmodel
ECF21. ECF21. and different simulations were run, first with a simplified 3D
Keywords:
rectangularelasticity;
blockplastic;
shape,asphalt
in orderconcrete; elastic
to better modulus;
establish thecompression;
model, and stress
then strain . real 3D mesh obtained from the blade scrap. The
with the
Keywords: elasticity; plastic; asphalt concrete; elastic modulus; compression;
overall expected behaviour in terms of displacement was observed, in particular stress strain . at the trailing edge of the blade. Therefore such a
model can be useful in the goal of predicting turbine blade life, given a set of FDR data.
1. Introduction
1. ©
Introduction
2016 The Authors. Published by Elsevier B.V.
Peer-review under responsibility
Asphalt concrete, of the Scientific
hereafter referred to as AC, Committee of PCF 2016.
is a composite material made up of different grades of aggregates
bound together
Asphalt using hereafter
concrete, asphalt/bitumen.
referred The
to ascombination of these components
AC, is a composite material made anduptheir separate grades
of different properties contribute
of aggregates
to Keywords:
bound High properties
the material
together Pressure
using Turbine
of ACBlade;
asphalt/bitumen. Creep;
which The Finite
dictates Element Method; 3D Model;
its performance
combination of these Simulation.
during service.
components and AC
theirhas many properties
separate uses such contribute
as paving
to the material properties of AC which dictates its performance during service. AC has many uses such as paving

* Corresponding author. Tel.: +1-868-719-6590; fax: +1-868-645-7691.

* Corresponding [email protected]
E-mail address:author. Tel.: +1-868-719-6590; fax: +1-868-645-7691.
E-mail address: [email protected]
2452-3216 © 2016 The Authors. Published by Elsevier B.V.
Peer-review underThe
* Corresponding
2452-3216 © 2016 responsibility
author. of218419991.
Tel.: +351
Authors. theby
Published Scientific Committee of ECF21.
Elsevier B.V.
E-mail address:
Peer-review [email protected]
under responsibility of the Scientific Committee of ECF21.

2452-3216 © 2016 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the Scientific Committee of PCF 2016.
Copyright © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer review under responsibility of the Scientific Committee of ECF21.
10.1016/j.prostr.2016.06.364
2914 Lee Leon et al. / Procedia Structural Integrity 2 (2016) 2913–2920
2 Lee Leon, Raymond Charles, Nicola Simpson / Structural Integrity Procedia 00 (2016) 000–000

runways, taxiways and roadways as well as parking areas. The stress-strain relationship of materials is very commonly
used to explain and predict how they will behave during service. Asphalt concrete is widely used, however, the
literature on the stress-strain relationship of asphalt concrete (when compared to that of other widely used materials
such as cement concrete) is limited.
Researchers used models that existed for low to normal strength concrete to compare experimental data and model
the stress strain curve for permeable concrete. This method of comparison as well as the information obtained in the
research done by Hussin et. al (2013) will be used in this investigation by considering the stress-strain relationship of
cement concrete to develop relationships of the stress strain behaviour of asphalt concrete. Hussin et. al (2013),
Almusallam & Alsayed (1995) and Carriera & Chu (1985) concluded in their research that the corresponding strain at
peak stress greatly affects the ascending and descending branches of the stress-strain curves of concrete and results
obtained showed a linear relationship between compressive strength and the corresponding strain under uniaxial
compression. A similar relationship will was explored for AC by using the uniaxial compression test method.
Zheng and Huang (2015) developed a new triaxial method to study the failure criteria of asphalt mixtures. An
increase in confining pressure (giving smaller compressive strength) showed mainly shear failure and a further increase
in confining pressure resulted in rheological failure. In comparison to the uniaxial test, the failure from the triaxial test
would allow for more failure modes to be analyzed. The stress-strain relationship of the specimen tested in Zheng and
Huang (2015) research showed larger values of horizontal stress having a curve with an initially ascending limb that
levels off after failure of the specimen (or after the elastic state is surpassed). However, smaller horizontal stresses
showed a curve with an initially ascending limb that peaks then begins to descend which can be seen in Fig. 1 (b)
below. As the horizontal stresses applied tends to zero the descending limb of the stress strain curve will become
steeper or more apparent.

a b

Fig. 1. (a) Stress-strain relation with high horizontal stress (b) Stress-strain relation with very low horizontal stress (Zheng and Huang 2015)

Starodubsky et, al (1994) conducted a research on the stress strain relationship for asphalt concrete in compression
and used both uniaxial and triaxial compression testing. The failure mode was compared to J. Zheng and T. Huang
(2015) and there were similarities of an initially linearly ascending limb which peaked then descended with a less steep
nonlinear curve. It was also noticed that the yield point (maximum stress) was larger in the compressive triaxial test
than with the same specimen in the uniaxial test. Their research concluded that specimen of different characteristics
(aggregate, compaction effort, bitumen percentage, height, type of loading) all show a common pattern in the stress-
strain relationship which is similar to that of concrete. The stress-strain curve can be described as having a linear part
of the branch with elastic behaviour which is represented by Hooke’s law which transitions smoothly into a nonlinear
part which is characterized by micro cracking and therefore failure mode is described as splitting or shear failure.
Wang et. al (2015) investigated the behaviour of asphalt concrete mixtures under tri-axial compression. They
presented results and discussed of triaxial compression monotonic tests that were conducted on dense and porous
asphalt concrete. They concluded that the compressive failure strength and stiffness of asphalt concrete increase when
the temperature decreases, the loading rate increases, and confinement increases.
Asphalt concrete viscoelastic and plastic behaviour are influenced by binder and aggregate properties respectively.
Under compression AC behaves as an elastoplastic material. When the specimen is first subjected to uniaxial loading,
the idealized graph as shown in Fig. 2 begins with a linearly increasing segment (1). This linear segment occurs
 Lee Leon et al. / Procedia Structural Integrity 2 (2016) 2913–2920 2915
Author name / Structural Integrity Procedia 00 (2016) 000–000 3

because the specimen first reacts to the load as a purely elastic homogenous solid. However, the plastic properties of
the specimen arise when at some point the granular aspects of the specimen bring about friction as the grains make
contact, therefore, a shear plane will begin to develop and the specimen begin to behave as a frictional solid. Stress
increases until the peak friction coefficient is reached and will flatten for a short period (2). This horizontal segment
of the curve represents the period at which the peak friction coefficient, µ equals the friction angle, Φ and this
relationship suggests pure shear. The graph is purely horizontal because the sample does not increase in stress because
of the purely frictional force and there is a constantly increasing strain. After this short period the friction angle will
change causing the coefficient of friction to drop and so the graph descends linearly until it reaches a residual friction
coefficient (3). At this point the graph will again become purely horizontal and continue to increase in strain only (4).
This idealized graph, is seen in figure 2 where the experimental results are also superimposed on the graph. This can
also be seen in figure 4 which shows real results from a 150 mm sample tested under uniaxial unconfined compression.

Fig. 2. Idealized stress-strain curve showing elastoplastic behaviour.

Applying pressure to AC through compression causes unbalanced forces in the system from which results in
deformation, cracks and fatigue failure. These applications are useful in determining the plastic and elastic limits of
AC. Plastic deformation and fatigue failure occur when the yield point of the material has been exceeded and this is a
problem seen in many AC pavements today, hence, a connection between the stress and strain in AC needs to be
explored especially for the purpose of designing. The research aims to investigate the failure mode of AC cylindrical
samples when subjected to unconfined uniaxial compression. It will conclude that the general attributes of the stress-
strain curve and AC failure mode as affected by the physical and mix properties of an AC layer.

2. Materials and testing

2.1. Materials

Commonly used construction materials such as Crushed Blue Limestone and Sharp Sand were the types of
aggregates used in preparing the different mixes. The aggregate blending produced respective gradations as shown in
Table 1. The specimens were dense graded (HMA 2, HMA 3) and stone matrix (SMA 3) mixes, which classification
were governed by the gradation specifications and NMAS (nominal maximum aggregate size). The Trinidad Lake
Asphalt (TLA) with a penetration of (60/75) as specified by ASTM (2015) was the study binder.

Table 1. The properties of the mix gradations

Mix type Coarse to fine ratio NMAS
HMA 2 50% - 50% 12.5 mm
HMA 3 49% - 51% 9.5 mm
SMA 3 71% - 29% 12.5 mm
2916 Lee Leon et al. / Procedia Structural Integrity 2 (2016) 2913–2920
4 Lee Leon, Raymond Charles, Nicola Simpson / Structural Integrity Procedia 00 (2016) 000–000

2.2. Mix design and testing

The mix design was conducted using the current and commonly used method of mix design which is the Marshall
Mix design procedure. The mixes were prepared to meet the requirements of Schedule 20 of the Central Tenders
Board of Trinidad and Tobago (GORTT 2000). The optimum binder content was found to be 5% and using the
gyratory compactor the samples were prepared for each mix type which differed in height (100mm and 150mm) and
air voids (3% – 8%). The original sample size was 110 samples.
The samples were tested under static unconfined uniaxial compression test procedure using an MTS (material
testing machine) and platens placed on the surface of each sample prior to loading. The samples were subjected to
continuous loading at a constant rate to complete failure and values of stress and strain values were recorded. Samples
were tested at two temperatures (270C and 450C) which simulates the high and low temperature of road ways in the
region.

3. Results and analysis

3.1. Mohr coulomb behavior of asphalt concrete mixes

The Mohr-Coulomb law is a linear law relating the normal to the shear stress. This line is a yielding condition for
shearing. Below this yield line, the material response will be rigid and does not suffer any strain. If the shear stress is
increased for a given normal stress such that the stress state of the material is exactly on the yield line, then plastic
strain or yielding will result. The average failure (shear) angles for all mix types were measured to be an average angle
of θ = 700. The Mohr’s Coulomb Circle was used to validate this failure angle to be associated with the mode of failure
as being true shear failure. The failure (shear) angle and the maximum axial (major principal) stress were used to
calculate friction angle. This value could also be obtained by the use of the derived Mohr’s Circle equation which is:

Fig. 3. (a) Failed 100 mm height sample (b) Failed 150 mm height sample.

Fig. 4. Mohr-Coulomb circles in the Stress -Normal stress plane

2  90o   (1)
 Author name / Structural Integrity Procedia 00 (2016) 000–000 5
Lee Leon et al. / Procedia Structural Integrity 2 (2016) 2913–2920 2917

The friction value was average to be 500 which is just about that of crushed rock. The angle is higher because of
the adhesion behaviour of the binder within the mix. This is sufficient since the mixes were made of crushed limestone.
Since this relationship can be represented by the Mohr-Coulomb Failure Criteria, the failure mode is in fact shear
failure. We can also say that the test method used (uniaxial unconfined compression test) does in fact give
representative results for an asphalt concrete layer in compression.

Table 2. The shear failure and friction angles of asphalt concrete mixes.
Mix type Average measured Calculated friction
failure angle angle
HMA 2 69o 50o
HMA 3 70 o
51o
SMA 3 68o 48o

3.2. Failure mode of asphalt concrete

Experimental results will deviate more and more away from this idealized model as more properties are changed
that affect the elastoplastic and frictional characteristics of the specimen. This is proven in Fig. 5 showing the
discretized graph of stress vs. strain of a 100mm sample, where the shape of the graph varies greatly from the
theoretical model in Fig. 2. This variation is due to the change in mode of failure from a pure shear failure to a forced
shear combination failure caused by the relatively small height of those samples. This smaller height changes the
interaction of frictional forces within the macrostructure and so the graph and ultimately the mode of failure will also
change.

a b

Fig. 5. (a) Stress-strain curve for 100 mm height sample (b) Stress-strain curve for 150 mm height sample

a b
100 mm sample 100 mm sample

150 mm sample 150 mm sample

Fig. 6. (a) Stress-strain curve for HMA 2 at 27oC (b) Stress-strain curve for HMA 2 at 45oC
2918 Lee Leon et al. / Procedia Structural Integrity 2 (2016) 2913–2920
6 Lee Leon, Raymond Charles, Nicola Simpson / Structural Integrity Procedia 00 (2016) 000–000

a b
100 mm sample 100 mm sample

150 mm sample 150 mm sample

Fig. 7. (a) Stress-strain curve for HMA 3 at 27oC (b) Stress-strain curve for HMA 3 at 45oC

a 100 mm sample b
100 mm sample
150 mm sample
150 mm sample

Fig. 8. (a) Stress-strain curve for SMA 3 at 27oC (b) Stress-strain curve for SMA 3 at 45oC

The experimental results continues to deviate from the idealized model as the physical and constituent properties
of the mix are changed. The effect is evident in the elastoplastic and frictional behaviour of the specimen as shown in
Fig. 6 to Fig. 8. The general shape of the resulting stress-strain graphs, shows that irrespective of mix type, sample
specimens which were 150mm in height were shaped differently and had less discrete segments than samples of
100mm height. They also failed at a lower stress level than the 100mm samples. This was due to the effect the sample
height had on the mode of failure due to effects such as shearing, failure angles, residual strengths, cap friction and
combinations of the same.
There are shear forces which exist on the surface where the sample is held by the platens during testing. This is
because the platens act as shear caps which create a vector frictional force. This is physically seen on any compressed
sample as it bulges from the centre (away from either surface experiencing cap friction). This vector force is combined
with the internal shearing force that develop during loading of the sample, if the samples height allows these two
vectors to meet during compression. The frictional force at the surface, however, does not interfere with the internal
shear forces which cause failure within the sample. The internal failure angle will have an interface with the friction
at the surfaces of the sample because of the short height of the 100mm samples (short given that the ratio of the sample
height to diameter is < or =1) and this is why they combine to give a higher strength.
In the first region of the graph of the 100mm samples, friction failure begins to develop within the sample as a
crack (linearly rising limb). This limb only extends up to approximately 300kPa where the interface is reached and
there is a combination of the forces which cause the graph to continue to rise (second region) up to ultimate failure.
This occurs because the shear friction failure crack first develops within the sample near the centre then migrates at
an angle to the edges of the sample on both ends. At the edge of the surface, it meets and combines with the frictional
force from cap friction. This is the point of true failure.
 Lee Leon et al. / Procedia Structural Integrity 2 (2016) 2913–2920 2919
Author name / Structural Integrity Procedia 00 (2016) 000–000 7

If the crack is not at 45 degrees, but at a greater angle, it does not meet the surface of the sample at the edge but
rather on the inside of the surface and this results in two vectors, a horizontal vector caused by the cap friction and
confining stress, and an angular vector (at the internal failure angle) caused by the internal shearing forces. These two
vectors combine to form a resultant force which runs from edge to edge on both surface and the sample is forced to
fail along this resultant. This resultant is the forced shearing failure plane and is referred to as τ (tau). Tau is reached
at the point where the graphs ultimately peak and the internal shear forces have dissipated. From this peak point (where
the sample has completely failed), the graph then begins to descend since the ultimate failure stress has been reached
and the sample is not able to withstand any more stress but continues to deform as the force is still being applied.
From this point there is only cap frictional force.
The graph descends until it reaches approximately 300kPa where there is again a combination of forces but instead
the cap friction meets and combines with the residual shear friction and a kink is observed in the descending limb of
the graph similar to that in the same position on the ascending branch. From this point the sample reacts to the uniaxial
compression with more strain per unit stress. The failure mode of the 100mm samples is not one of pure shear, but
rather a bimodal failure where there is a combination of internal shear friction and cap friction which results in forced
shear failure.

3.3. Elastic and plastic behaviour of asphalt concrete in compression

An asphalt concrete specimen under compression behaves as an elastoplastic material. This behaviour is due to the
material properties such as viscoelasticity and plasticity of the binder and the aggregates respectively. By changing
different characteristics of the mix the following experimental relationship conclusions can be made so that if the
specific value of a parameter is needed, then one will know which other parameter needs to be changed accordingly.
The parameters compared from these experiments include failure stress, strain at failure, air voids percentage, sample
height and modulus of elasticity. These parameters were compared in different pairs for each mix type and testing
temperature used.

Table 3. The elastic and plastic properties of 150 mm height samples

Mix type Young’s (Elastic) Failure Stress – kPa Plastic strain at Failure Stress
Modulus – MPa (270C / (270C / 450C) – mm/mm (270C / 450C)
450C)
HMA 2 8.7 – 3.9 250.2 – 79.7 0.051 – 0.034
HMA 3 9.7 – 4.2 286.4 – 68.4 0.053 – 0.032
SMA 3 10.1 – 6.0 323.2 – 152.2 0.059 – 0.041

The elastic properties of the bitumen contribute greatly to the elastic properties (and therefore the elastic modulus)
of the sample as a whole. The elastic modulus decreases as the temperature decreases, and this behaviour is due to the
changes in the properties of the bitumen due to temperature. As sample height increases the elastic modulus increases.
In all cases, testing temperature had a clear effect on the outcome of the results. It is seen that testing samples at a
higher temperature decreases elastic modulus, and stress and strain at failure. The level of correlation of the parameters
being compared only slightly changes (increases) with a higher temperature. This effect is due to the increase in
temperature which softens binder and increases AC flow. It will therefore compress faster which result in samples
reaching the failure stress faster.
Failure stress decreases with an increase in sample height no matter the mix type or temperature. This is the due to
the failure mode previously mentioned. The strain at failure decreases as the sample height increases. However, this
does not affect the difference between the modes of failure between two different sample heights. It just shows that
smaller samples fail at a larger strain since more stress has been endured. This relationship has to be defined differently
for a dense graded mix compared to a stone matrix mix. The difference may be due to the presence of more coarse
aggregates within the SMA as compared to the HMA mixes.
2920 Lee Leon et al. / Procedia Structural Integrity 2 (2016) 2913–2920
8 Lee Leon, Raymond Charles, Nicola Simpson / Structural Integrity Procedia 00 (2016) 000–000

4. Summary and conclusions

If an asphalt mixture deforms (ruts) and exhibits fatigue failure, it is normally because the mixture has insufficient
shear strength to support the stresses to which it is subjected. Aggregates are responsible for minimizing shear failure
within an asphalt concrete mix. By subjecting dense graded and stone matrix asphalt concrete mixes in the form of
cylinders of varying properties to unconfined uniaxial compression testing, the failure mode of asphalt concrete was
determined to be shear failure.
The theories of the stress-strain properties of asphalt concrete which were formulated by different researchers
proved to be true, in that there is a common pattern in the stress-strain graphs obtained from the results. However,
there is a trend which arises from samples of 100 mm height having a stress-strain curve shaped differently from those
of 150mm which was not previously mentioned. Sample height therefore has a strong correlation with failure stress
and strain as taller samples have a lower stress and strain at failure. This relationship between height and failure stress
was the determining factor in the type of shear failure. This is seen in the trends existing in the stress and strain curves
which showed true failure occurring at approximately 300kPa, despite the varied parameters in the samples tested at
270C. Residual shear friction also comes into play at this same point but in the plastic limits of the graphs for samples
of 100mm height. The two types of shear failure experienced were bimodal forced frictional as in the case of the
100mm samples and pure shear as in the case of the 150mm samples. The only effect of mix type on failure stress was
that stone matrix asphalt mix samples achieved a higher failure stress when compared to dense graded mixes.
Other trends include the effects of temperature in the stress strain diagrams where a higher temperature lowers the
failure stress in all mix types and sample heights. All dense graded mix type samples (and most stone matrix samples)
of 150mm followed the idealized stress-strain graph when tested at room temperature. All samples, despite the varied
parameters (therefore both heights), followed the idealized stress-strain graph pattern when tested at a higher
temperature of approximately 450C which was assumed to be field temperature. All samples tested at 450C had pure
shear failure and all samples of 100mm height tested at room temperature had a bimodal forced shear failure and did
not follow the pattern of the idealized stress-strain graph.
The investigation derives parameters (yield stress and elastic modulus) for different asphalt concrete mix types
which can be used in FE programs such as Abaqus that uses the elastic as well as plastic data to model and simulate
the elastic and plastic behaviour of different mix types of asphalt concrete material used in a pavement structures.

References

ASTM, 2015. ASTM D 5710 - Standard Specification for Trinidad Lake Modified Asphalt. American Society of the International Association for
Testing and Materials (ASTM). Designation: D5710/D5710M – 15
ASTM, 2015. ASTM D 6927 - Standard Test Method for Marshall Stability and Flow of Asphalt Mixtures. American Society of the International
Association for Testing and Materials (ASTM). Designation: D6927 − 15
Carreira, D.J., and K.H. Chu. 1985. Stress-Strain Relationship for Plain Concrete in Compression. Journal of the American Concrete Institute 82
(6), 797-804.
Government of the Republic of Trinidad and Tobago (GORTT), 2000. Hot Asphalt Mixtures for Use in Road Paving – Schedule 20. Central
Tenders Board Standard Specifications. No. CTB: HAM-R 2000
Hsu, L.S., C. Hsu. 1994. Complete Stress-Strain Behaviour of High-Strength Concrete under Compression. Magazine of Concrete Research 46
(169), 301-312.
Starodubsky, S., Blechman, I., Livneh, M. 1994. Stress-strain relationship for asphalt concrete in compression. Materials and Structures 27 (8),
474-482.
Wang, P., S. Shah, A. Naaman. 1978. Stress-Strain Curves of Normal and Lightweight Concrete Compression. American Concrete Institute 75
(11), 603-611.
Wang, J., Molenaar, A.A, Van De Ven, M. F. C., Wu, S. 2015. Behaviour of asphalt concrete mixtures under tri-axial compression. Construction
and Building Materials. 105, 269-274.
Zheng, J., Tuo, H., 2015. Study on Triaxial Test Method and Failure Criterion of Asphalt Mixture. Journal of Traffic and Transportation Engineering
(English Edition). 2 (2), 93-106.