You are on page 1of 12

The International Journal of Pavement Engineering, Vol. 5 (3) September 2004, pp.

163–173

A Comparison of the Marshall and Superpave Design


Procedure for Materials Sourced in India
B.L. SWAMIa,*, Y.A. MEHTAb and S. BOSEc

a
Department of Civil Engineering, Malaviya National Institute of Technology, A-161 Malviya Nagar, Jaipur, Rajasthan 302017, India;
b
Department of Civil and Environmental Engineering, Rowan University, NJ, USA; cFlexible Pavement Division, CRRI, New Delhi, India

(Received 15 August 2003; Revised 25 June 2004; In final form 24 July 2004)

Developing countries are striving for the development of infrastructure facilities including
transportation systems. India has a national highway with a length over 52,000 km. A further
14,000 km length of express highway is currently being built under the golden quadrangle and North–
South and East – West corridor projects. These projects have large scope for adopting superpave
technology.
The strategic highway research program (SHRP) conducted a $ 50 million research effort from 1987
to 1993 to develop a new concept for the design of bituminous mixes referred to as superior performing
asphalt pavements (superpave). Superpave mixes have been widely used by the developed countries for
the last few years but the developing countries are still working with the conventional mixes i.e.
marshall mixes. Flexible pavements, with bituminous surfacing as a wearing course, are widely used in
India. The objective of this study was to compare the design of asphaltic concrete by the superpave and
marshall methods of mix design. A detailed laboratory study was carried out using aggregate from the
Delhi region and bitumen from the Mathura refinery. From the analysis of design of asphaltic concrete,
it was observed that superpave mixes fulfill all the criteria for easy and good construction at lesser
binder content than the marshall mixes. Apart from superpave mix design, the effect of angle of
gyration, number of gyrations on mix properties like density, stability, indirect tensile strength were
also determined. Results revealed that percentage theoretical maximum density increases with the
increase in number of gyrations. At any level of gyration, theoretical maximum density percentage
increases with increase in gyratory angle. Further, it was observed that indirect tensile strength
decreases with increase in angle of gyrations.
Comparing both the design methods, it was observed that marshall compactor is unable to answer the
rutting resistance of the designed mixture. The marshall compaction effort in the field laboratory
corresponded to an air void content of slightly less than six percent, and the mixture appeared to
stabilize under traffic at an air void content between three and four percent. In contrast, superpave
gyratory compactor is capable of achieving air void contents much lower than achieved by mechanical
marshall hammer compaction. This prevents additional compaction under traffic, which could result in
rutting in the wheel paths. Hence, marshall mix design should be replaced by superpave mix design for
Indian national highways/express highways, which compact the specimens in the same manner as
compacted under actual pavement climate and loading conditions.

Keywords: Gyratory compactor; Rutting resistance; Golden quadrangle; National highways; Mathura
refinery

INTRODUCTION has a national highway spanning over 52,000 km and, out


of this, 14,000 km is being upgraded to express highway
General
under the golden quadrangle and North – South and East –
Flexible pavements with bituminous surfacing are widely West corridor projects. These projects involve huge scope
used in India. Developing countries have to raise their for advance technology in various stages of construction
transportation system to higher levels, both in terms of of roads, including new generation products, instruments
length and quality, so as to meet the demand which is and methodology for speedy construction, better riding
being generated by the development plans. The country quality and less maintenance requirements.

*Corresponding author. Tel.: þ 91-141-2522898/3101946. Fax: þ91-141-2702107. E-mail: swami_bls2003@yahoo.com

ISSN 1029-8436 print/ISSN 1477-268X online q 2004 Taylor & Francis Ltd
DOI: 10.1080/10298430412331309115
164 B.L. SWAMI et al.

Need Rutting also results from lateral plastic flow (permanent


deformation) of the HMA from the wheel tracks. Use of
The roads in India are performing poorly with pavement life
excessive asphalt cement is the most common cause for
much shorter than the expected life. The high traffic intensity
this phenomenon. Too much asphalt cement in the mix
in terms of commercial vehicles, the serious overloading of
causes the loss of internal friction between aggregate
trucks and significant variation in daily and seasonal
particles and results in the loads being carried by the
temperature of the pavement have been responsible for early
asphalt cement rather than the aggregate structure. Plastic
development of distress like raveling, undulations, rutting,
flow can be minimized by using large size aggregate,
bleeding shoving and potholes on bituminous surfacing.
angular and rough textured coarse and fine aggregates,
The superpave technology was developed in the United
and providing adequate compaction at the time of
States with proven success. The strategic highway
construction.
research program (SHRP) conducted a $ 50 million
The consistency (penetration or viscosity) of asphalt
research effort from 1987 to 1993. McGennis et al. (1994)
cement plays a relatively small role in the rut resistance of
suggested a concept for the design of bituminous mixes
HMA if well-graded, angular and rough textured
referred to as superior performing asphalt pavements
aggregate are used. Some increased resistance to rutting
(superpave). Superpave mixes have been widely used by
can be obtained by using a more stiff (high viscosity or
developed countries over the last few years. It is currently
low penetration) asphalt cement.
being implemented by the European Union, Japan and
South Korea but the developing countries are still working
with the conventional mixes, i.e. marshall mixes. Objective of the Study
Superpave technology is replacing the marshall method,
The marshall method of mix design is being widely used
which was used for asphalt concrete mixture design for
all over the world. Marshall compactor does not simulate
almost half a century. The marshall method was based
the field compaction for mixture stability and compaction,
mostly on experience and statistical analysis. The flexible
which occurs in actual field situations. Marshall stability
pavement sections designed using the marshall method
also does not exactly estimate the shear strength of mixes.
have had mixed success due to poor understanding of
These two situations make it difficult to assure the rutting
mechanism of failure. The partial success has been mainly
resistance of the design mixture. In most cases, the
due to very thick and uneconomical sections. The roads in
requirements for superpave mixes are very similar to the
India are in a highly distressed condition with pavement
requirement for marshall mix design, however use of a
life much shorter than expected. A new design metho-
superpave gyratory compactor and criteria for selecting
dology, that is more thorough and comprehensive, is
design binder content in superpave mix design make a
required. Superpave technology can be rigorously tested
difference between superpave and marshall method of mix
under varying traffic and environmental conditions.
design.
This technology has a tremendous potential to be
Objective of the study:
implemented in India, which will pay for itself with higher
performance and longer lasting roads. Hence, there is need
(a) to compare the design of asphalt concrete by
to have a comprehensive study comparing the design of
superpave and marshall method of mix design for
bituminous mixes using both superpave and the marshall
Indian conditions,
method of mix design.
(b) to study the properties of superpave mixes at different
angles and different numbers of gyrations.
Rutting
LABORATORY STUDIES FOR DESIGN OF AC
There are some specific types of pavement distresses
USING MARSHALL AND SUPERPAVE METHODS
(affecting the pavement performance); the rutting
phenomenon is one of them. Hence, it is important to
General
understand rutting and its causes.
Rutting is caused by the progressive movement of A detailed laboratory study was carried out using
materials under repeated loads either in the asphalt aggregate of sizes 20 and 10 mm, stone dust and lime
pavement layers or the underlying base. This can occur filler obtained from the Delhi region and bitumen of grade
either through consolidation or through plastic flow. 60/70 from Mathura refinery.
Consolidation is the further compaction of hot mix
asphalt (HMA) pavement by traffic, after construction. Job Mix Formula for Asphaltic Concrete by Marshall
When compaction is poor (10 –12% or more air voids for a Method
mix design at 3– 5%), the channelized traffic provides a
Aggregate Gradation and Proportioning
repeated kneading action in the wheel track areas and
completes the consolidation to the designed air voids level Sieve analysis and proportioning of material i.e. 20 and
(usually 3– 5%). A substantial amount of rutting can occur 10 mm, aggregates, stone dust and lime filler are presented
if very thick asphalt layers are consolidated by the traffic. in Table I.
SUPERPAVE—AN INDIAN CASE STUDY 165

TABLE I Resultant grading along with specified limit TABLE III Design parameters for AC and specified limits

Passing % S.No Parameters Results Specified limits MORTH


Sieve
Size 20 mm 10 mm S.D. Lime 1 Optimum binder % 5.31
S.No (mm) (21%) (42%) (34%) (3%) Combined Specified 2 Stability, kg 1140 900 minimum
3 Flow (mm) 3.2 2–4
1. 26.5 21 42 34 3.0 100 100 4 % Air voids 3.25 3–5
2. 19.0 20 42 34 3.0 99 99 –100 5 % VFA 74 65–75
3. 9.5 3.0 25 34 3.0 65 56 –80 6 % Retained stability 80 75
4. 4.75 – 6 34 3.0 43 35 –65
5. 2.36 – 1 32 3.0 36 23 –49 MORTH—Ministry of road transport and highways.
6. 0.300 – – 11.0 3.0 14 5 –19
7. 0.075 – – 1.0 3.0 4 2 –8
Superpave Mix Design for Asphaltic Concrete
Superpave mix design using superpave gyratory compactor
was carried out in laboratory through the following steps:
Design Details of Asphaltic Concrete
(1) Material selection.
The marshall test method, as described in the asphalt (2) Selection of design aggregate structure.
institute (TAI) manual (MS-2), was used to select the (3) Selection of design binder content.
optimum binder content for all the mixes used. The asphalt (4) Evaluation of moisture sensitivity.
concrete was designed by marshall method of mix design
as per procedure included in ASTM D1559. The samples
were compacted using 75 blows per side in a cylindrical Material Selection
mold of 100 mm diameter.
Marshall mixes were prepared with blended mineral For superpave mix design, the same materials i.e. mineral
aggregates with increasing percentages of binder content aggregate of 20 and 10 mm, stone dust and lime from
from 4.5 to 6.0% by weight of mineral aggregate with an Delhi and bitumen from the Mathura refinery of grade
increment of 0.5%. For each binder content, three test 60/70, were used.
specimens were prepared. Marshall test specimens are
tested for bulk density, stability and flow value. The Selection of Design Aggregate Structure
observations and results are shown in Table II.
The Table III shows the specified values as per ministry (i) Establish trial blends.
of road transport & highways (MORTH) and various (ii) Compact specimens.
design parameters for asphaltic concrete. (iii) Evaluate trial blends.
In this study, the optimum binder content is calculated (iv) Select design aggregate structure.
as per asphalt institute method in MS-2 considering 4.0%
as design air void content. Three trial blends were established having proportion of
Optimum binder content is calculated by taking the aggregate as:
average of:

(i) binder content for maximum stability ¼ 5.4%, Proportions of aggregate in percentage
(ii) binder content for maximum density ¼ 5.5%,
Blends 20 mm 10 mm Stone dust Filler
(iii) binder content for 4% air voids ¼ 5.05%.
Trial blend 1 21 42 34 03
Average i.e. optimum binder content ¼ 5.31% Trial blend 2 20 48 29 03
Trial blend 3 25 44 28 03
All the parameter values at optimum binder are within
the specified range hence mix design for asphaltic
concrete can be acceptable and the design binder content Combined gradation of all three trial blends are shown
is taken as 5.31% by weight of aggregate. in Table IV with MORTH specified limits. As all the
gradation passes through the control points, hence can be
accepted as a trial blends for further computation.
TABLE II Observations and results of marshall design for AC
After establishing the trial blends, specimens were
Average Voids in compacted as per Cominsky et al. (1994) to evaluate the
Binder bulk Average compacted trial blends for each trial blend using the superpave
content density Average flow mix VFB
S.No % gm/cc stability (mm) % % Remark
gyratory compactor.

1 4.5 2.370 1006 2.0 5.2 66 Computation of Trial Binder Content for Each
2 5.0 2.383 1092 2.6 4.1 70 Blend
3 5.5 2.394 1150 3.3 2.96 76 After computing effective specific gravity (Gse), volume
4 6.0 2.388 950 4.2 2.53 83
of binder absorbed (Vba) and volume of effective binder
166 B.L. SWAMI et al.

TABLE IV Combined gradation of three trial blends asphalt content and it is referred as:

Passing % dðACÞ ¼ A:C: percent ðsuperpaveÞ


Sieve Specified 2 A:C: percent ðmarshallÞ
size Trial Trial Trial limit
S.No (mm) blend 1 blend 2 blend 3 MORTH The second approach was to determine the equivalent
1 26.5 100 100 100 100 number of gyrations (Neq), for the 75 blows Marshall
2 19.0 99 99 99 90 –100 results, that when used in the superpave procedure will
3 9.5 65 63 63 56 –80
4 4.75 43 41 45 35 –65 yield the same design binder content as determined by the
5 2.36 35 33 34 23 –49 marshall approach. In the present study the first approach
6 0.300 14 14 13 5–19 has been adopted.
7 0.075 4 6 4 2–8
The number of gyrations used for compaction was
MORTH—Ministry of road transport and highways. determined for the temperature range of 43 – 448C and
design traffic range of 30 – 100 msa. For these ranges, the
number of gyrations is:
(Vbe) for each trial blend, the observed initial trial binder
(Pbi) content were as follows: Nini ¼ 10,
Ndes ¼ 153, and,
For blend 1 Pbi ¼ 4.0291%, Nmax ¼ 253.
For blend 2 Pbi ¼ 4.0319%,
For blend 3 Pbi ¼ 4.033%. Each specimen is compacted using maximum number
of gyrations and the height of specimen during
For design of mixes, generally, the increment of binder compaction was continuously noted. Using initial mass
content considered is 0.25 or 0.5%. Hence next higher of the mix, the fixed volume of the mold and the measured
increment (of the observed initial trial binder content for height, the estimated bulk specific gravity (Gmb(est)) is
all the three trial blends) i.e. 4.25% is considered for Pbi. calculated for any number of gyrations. Soysa et al. (1997)
For all the three trial blends the Pbi is taken as 4.25% by proposed the procedure for compacting asphalt aggregate
weight of aggregate. mixtures. After compaction, the specimen’s bulk specific
Specimen Compaction and Evaluation of Trial gravity Gmb(meas) is measured by ASTM-D-2726 taking
Blends the weight of the specimen in air and water.
Two samples were compacted at 4.25% binder content The final Gmb(est) of each compacted specimen is then
by weight of aggregate for each trial blend using super- compared to the Gmb(meas) and a correction factor is
pave gyratory compactor. calculated. The Gmb(est) at other gyration levels were
Mineral aggregates and binder were heated and mixed corrected (Gmb(corr)) using the correction factor. Finally, the
at the appropriate temperature range of 150– 1638C and percent of Gmm for various gyration levels is calculated by
then compacted at 1558C temperature. dividing the Gmb(corr) by calculated value of Gmm i.e.
In this method of mix design, the gyratory compactor is theoretical maximum specific gravity of mix and loose
used to compact the specimen having consolidation pressure samples of mixture were used to determine Gmm or
of 600 kPa (87 psi) with an angle of gyration 1.258, speed of theoretical maximum density (TMD). Tables V,VI and VII
gyration 30 rpm and mold of 150 mm diameter. shows the densification data for trial blends 1, 2 and 3.
Usually asphalt mixtures are designed at a specific level The percent of Gmm at Nini, Ndes and Nmax along percent
of comp active effort. In superpave, this is a function of the of air voids (Va) and VMA were computed and are shown
design number of gyration Ndes. Ndes is a function of climate in Table VIII.
and traffic level. The average design high air temperature The compaction summary of trial blends is shown in
represents the climate and the traffic is represented by Table VIII.
the design ESALs. Two other gyration levels are also of From these data, an estimated binder content to achieve
interest, the initial number of gyration (Nini) and maximum 4% air voids (96% Gmm) at Ndes is determined for each
number of gyrations (Nmax). Test specimens are compacted trial blend using the equation:
using Nmax gyration, an estimation of the compatibility of PbðestÞ ¼ Pbi 2 ½0:4 £ ð4 2 V a Þ
the mixture is determined using Nini.
Nmax and Nini can be calculated using: where
Log N max ¼ 1:10 Log N des ;
Pb(est), estimated percent binder content;
Log N ini ¼ 0:45 Log N des : Pbi, trial percent binder;
Va, Percent air voids @ Ndes.
The comparison of the design asphalt contents by Blend 1: Pb(est) ¼ 4.906,
marshall and superpave methods is presented in two Blend 2: Pb(est) ¼ 4.64,
approaches. The first approach uses the difference in Blend 3: Pb(est) ¼ 4.66.
SUPERPAVE—AN INDIAN CASE STUDY 167

TABLE V Densification data for trial blend 1

Specimen 1 Specimen 2

Gyrations Ht (mm) Gmb (est) Gmb (corr) Gmm % Ht (mm) Gmb (est) Gmb(corr) Gmm % Avg Gmm %
0 75.44 1.869 1.934 77.00 76.22 1.902 1.949 77.58 77.29
5 70.41 2.003 2.072 82.50 71.35 2.032 2.082 82.88 82.69
10 68.13 2.070 2.142 85.26 69.10 2.098 2.150 85.57 85.42
15 66.90 2.108 2.181 86.83 67.79 2.139 2.191 87.23 87.03
30 64.91 2.173 2.248 89.49 65.86 2.202 2.255 89.78 89.64
50 63.70 2.214 2.291 91.19 64.63 2.244 2.298 91.49 91.34
100 62.24 2.266 2.344 93.33 63.24 2.293 2.349 93.50 93.41
120 61.97 2.276 2.355 93.73 62.93 2.304 2.360 93.96 93.85
153 61.56 2.291 2.370 94.36 62.65 2.314 2.371 94.38 94.37
180 61.32 2.300 2.380 94.73 62.41 2.323 2.380 94.75 94.74
200 61.23 2.303 2.383 94.87 62.31 2.327 2.384 94.90 94.88
225 61.18 2.305 2.385 94.94 62.15 2.333 2.390 95.14 95.04
253 61.04 2.310 2.390 95.16 62.01 2.338 2.395 95.36 95.26
Gmb ðmeas:Þ ¼ 2:391; C ¼ 1:035 Gmb ðmeas:Þ ¼ 2:395; C ¼ 1:0244

Gmm ¼ 2.512.

TABLE VI Densification data for trial blend 2

Specimen 1 Specimen 2

Gyrations Ht (mm) Gmb (est) Gmb (corr) Gmm % Ht (mm) Gmb (est) Gmb(corr) Gmm % Avg Gmm %

0 78.19 1.855 1.921 76.53 79.08 1.836 1.914 76.25 76.39


5 73.13 1.984 2.054 81.83 73.68 1.971 2.054 81.84 81.83
10 70.26 2.065 2.138 85.17 70.96 2.047 2.133 84.97 85.07
15 69.20 2.096 2.170 86.47 69.46 2.091 2.179 86.81 86.64
30 66.53 2.181 2.258 89.94 67.21 2.161 2.252 89.72 89.83
50 65.20 2.225 2.304 91.78 65.81 2.207 2.300 91.62 91.70
100 63.65 2.279 2.360 94.01 64.17 2.263 2.359 93.97 93.99
120 63.41 2.288 2.369 94.37 63.85 2.274 2.370 94.44 94.40
153 62.99 2.303 2.384 95.00 63.44 2.289 2.386 95.05 95.02
180 62.78 2.311 2.392 95.32 63.27 2.295 2.392 95.30 95.31
200 62.66 2.315 2.397 95.50 63.09 2.302 2.399 95.57 95.54
225 62.49 2.322 2.404 95.76 62.81 2.312 2.410 96.00 95.88
253 62.40 2.325 2.407 95.90 62.80 2.313 2.410 96.02 95.96
Gmb ðmeas:Þ ¼ 2:407; C ¼ 1:035 Gmb ðmeas:Þ ¼ 2:410; C ¼ 1:042

Gmm (meas.) ¼ 2.510.

TABLE VII Densification data for trial blend 3

Specimen 1 Specimen 2

Gyrations Ht (mm) Gmb (est) Gmb (corr) Gmm % Ht (mm) Gmb (est) Gmb (corr) Gmm % Avg Gmm %

0 77.79 1.819 1.888 75.26 79.02 1.845 1.924 76.66 75.96


5 72.10 1.962 2.037 81.20 73.70 1.978 2.062 82.20 81.70
10 69.36 2.040 2.118 84.40 70.95 2.055 2.142 85.38 84.89
15 67.90 2.083 2.163 86.22 69.59 2.095 2.184 87.05 86.64
30 65.52 2.159 2.242 89.35 67.54 2.159 2.250 89.70 89.52
50 64.20 2.203 2.288 91.19 66.05 2.207 2.301 91.72 91.45
100 62.56 2.261 2.348 93.58 64.35 2.266 2.362 94.14 93.86
120 62.16 2.276 2.363 94.18 63.96 2.280 2.376 94.72 94.45
153 61.81 2.289 2.376 94.71 63.64 2.291 2.388 95.19 94.95
180 61.53 2.299 2.387 95.14 63.34 2.302 2.400 95.64 95.39
200 61.30 2.308 2.396 95.50 63.18 2.308 2.406 95.88 95.69
225 61.23 2.310 2.399 95.61 63.15 2.309 2.407 95.93 95.77
253 61.15 2.313 2.402 95.74 62.86 2.319 2.418 96.37 96.05
Gmb (meas.) ¼ 2.402, C ¼ 1.038 Gmb (meas.) ¼ 2.418, C ¼ 1.043

Gmm ¼ 2.509.
168 B.L. SWAMI et al.

TABLE VIII Compaction summary of trial blends as its value is 1.29, which is higher than the maximum
required value of 1.2. Trial blends 1 and 3 fulfil the design
B.C. Gmm at Gmm at Gmm at criteria; however, trial blend 1 is selected as the design
Blend (%) Nini(%) Ndes(%) Nmax% Va(%) VMA(%)
aggregate structure so that it is easy to make a comparison
1 4.25 85.41 94.36 95.26 5.64 15.76 with marshall mix design criteria (trial blend 1 has also
2 4.25 85.10 95.02 95.96 4.98 15.18
3 4.25 84.90 94.98 96.10 5.02 15.21 been used for marshall method of mix design)

Selection of Design Binder Content


TABLE IX Estimated mixture volumetric properties at Ndesign After selecting the design aggregate structure, two
specimens were compacted at each of 4.25, 4.75, 5.25
B.C.
Dust and 5.75% binder content by weight of mineral aggregate.
Trial Estimated Air VMA VFB proportion Specimens were prepared and tested, and for this the
Blend (%). (%) Voids (%) (%) (%) ratio gradation and proportioning of different materials are
1 4.25 4.906 4 15.43 74.00 0.82 adopted as per Table I.
2 4.25 4.64 4 14.98 73.29 1.29 Tables XII–XV indicate the test results for each
3 4.25 4.66 4 15.00 73.33 1.16
binder content. Figure 1 shows the average densifica-
tion curves for each binder content.
Ndes values from this study are less than the superpave
specified values by approximately 30 gyrations for
TABLE X Estimated mixture density properties example, at 1.0 million ESALs the superpave Ndes value
is 76, whereas Ndes value obtained from this study is 46
B.C.
gyrations. Hence, in this study instead of Ndes, the gyratory
Trial Estimated Gmm at Gmm at compaction characteristics using superpave gyratory
Blend (%) (%) N=10 (%) N=253 (%) compactor were used.
1 4.25 4.906 87.05 96.88 Tables XVI and XVII shows the mixture properties at
2 4.25 4.64 86.01 96.94 Nini, Ndes and Nmax and volumetric properties of selected
3 4.25 4.66 85.92 97.12
blend at different binder contents at Ndes.
Kandhal et al. (1997) discussed in detail design of
bituminous mixes. Design binder content is established at
Estimated volumetric and mixture density properties for 4.0% air voids. From Figure 2 the value of design binder
trial blends at the binder content that will result in 4% air content is 4.4% corresponding to 4.0% air voids.
voids at Ndes, are shown in Tables IX and X. Table XVIII shows the other properties corresponding to
Dust proportion is calculated using: design binder content i.e. at 4.4% binder content along
with mix design criteria.
P0:075
DP ¼
Pbe
Computation of Moisture Sensitivity
where, P0.075, is aggregate content passing 0.075 mm
sieve, and Pbe, is effective asphalt content percent by mass To evaluate moisture sensitivity, specimens were com-
of mixture. pacted to approximately 7% air voids at design binder
An acceptable range of DP is 0.6 – 1.2. content. One subset of three specimens was considered as
Dust proportion ratio for trial blend 1, 2 and 3 were control specimen. The other subsets of three specimens
observed to be 0.82,1.29 and 1.16, respectively. were considered as conditional subsets. All the specimens
Air voids at Ni is more than 11% of required norms and were tested to determine their indirect tensile strengths.
air voids at Nmax is more than 2.0%, thus satisfying the The moisture sensitivity is determined as the ratio of the
design requirements. On comparison of estimated proper- average tensile strengths of the conditioned subset to the
ties with design criteria as shown in Table XI, trial blend 2 average tensile strengths of controlled subset.
cannot be selected because it fails in dust proportion ratio,
Average tensile strength of conditioned samples
TSR ¼
Average tensile strength of controlled samples
TABLE XI Mix design criteria
where TSR, tensile strength ratio (percent). Average
S. No. Properties Specified Limits
tensile strength of conditioned sample, 8.13 kg/cm2.
1 Air Voids (%) 4% Average tensile strength of controlled samples,
2 VMA (%) 13% min 8.86 kg/cm2
3 VFA (%) 65%–75%
4 Gmm at Nini (%) Less than 89% 8:13
5 Gmm at Nmax (%) Less than 98% TSR ¼ £ 100 ¼ 91:75%
8:86
SUPERPAVE—AN INDIAN CASE STUDY 169

TABLE XII Densification data for blend 1, 4.25% binder content

Specimen 1 Specimen 2

Ht Volume Gmb Gmb Ht Volume Gmb Gmb


Gyrations (mm) (cc) (est.) (corr) Gmm (%) (mm) (cc) (est.) (corr) Gmm (%) Avg.
0 71.00 575.62 1.949 2.073 82.56 76.66 621.51 1.843 1.960 78.04 80.3
5 69.62 564.43 1.987 2.114 84.20 71.71 581.37 1.971 2.095 83.43 83.8
10 67.64 548.38 2.045 2.176 86.66 69.25 561.43 2.041 2.169 86.39 86.5
15 66.55 539.54 2.079 2.212 88.08 67.94 550.81 2.080 2.211 88.06 88.1
30 64.65 524.14 2.140 2.277 90.67 65.96 534.76 2.142 2.277 90.70 90.7
50 63.52 514.98 2.178 2.317 92.29 64.73 524.79 2.183 2.321 92.42 92.4
100 62.22 504.44 2.224 2.366 94.21 63.37 513.76 2.230 2.371 94.41 94.3
120 61.87 501.60 2.236 2.379 94.75 63.10 511.57 2.239 2.381 94.81 94.8
153 61.49 498.52 2.250 2.394 95.33 62.82 509.30 2.249 2.391 95.23 95.3
180 61.36 497.46 2.255 2.399 95.53 62.47 506.46 2.262 2.405 95.77 95.7
200 61.15 495.76 2.262 2.407 95.86 62.23 504.52 2.271 2.414 96.14 96.0
225 61.03 494.79 2.267 2.412 96.05 62.19 504.19 2.272 2.416 96.20 96.1
253 60.90 493.73 2.272 2.417 96.26 62.05 503.06 2.277 2.421 96.42 96.3
Gmb ðmeas:Þ ¼ 2:417; C ¼ 1:0638 Gmb ðmeas:Þ ¼ 2:421; C ¼ 1:0632

Gmm ¼ 2.511.

TABLE XIII Densification data for blend 1, 4.75% binder content

Specimen 1 Specimen 2

Gyrations Ht (mm) Volume (cc) Gmb (est.) Gmb (corr) Gmm (%) Ht (mm) Volume (cc) Gmb (est.) Gmb(corr) Gmm (%) Avg.

0 75.54 612.43 1.867 1.987 79.68 75.61 612.99 1.872 1.959 78.59 79.1
5 70.79 573.92 1.992 2.120 85.03 70.27 569.70 2.014 2.108 84.56 84.8
10 68.45 554.94 2.060 2.192 87.94 67.76 549.35 2.089 2.186 87.69 87.8
15 67.25 545.22 2.097 2.231 89.51 66.48 538.97 2.129 2.228 89.38 89.4
30 65.35 529.81 2.158 2.296 92.11 64.57 523.49 2.192 2.294 92.02 92.1
50 64.11 519.76 2.199 2.341 93.89 63.37 513.76 2.233 2.338 93.77 93.8
100 62.66 508.00 2.250 2.395 96.06 62.03 502.90 2.282 2.388 95.79 95.9
120 62.30 505.08 2.263 2.409 96.62 61.78 500.87 2.291 2.398 96.18 96.4
153 62.04 502.98 2.273 2.419 97.02 61.38 497.63 2.306 2.413 96.81 96.9
180 61.79 500.95 2.282 2.429 97.42 61.16 495.84 2.314 2.422 97.15 97.3
200 61.67 499.98 2.286 2.433 97.60 61.05 494.95 2.318 2.426 97.33 97.5
225 61.52 498.76 2.292 2.439 97.84 60.97 494.30 2.321 2.430 97.46 97.6
253 61.45 498.19 2.295 2.442 97.95 60.71 492.19 2.331 2.440 97.87 97.9
Gmb ðmeas:Þ ¼ 2:442; C ¼ 1:064 Gmb ðmeas:Þ ¼ 2:440; C ¼ 1:046

Gmm ¼ 2.493.

TABLE XIV Densification data for blend 1, 5.25% binder content

Specimen 1 Specimen 2

Ht Volume Gmb Ht Volume Gmb Gmb


Gyrations (mm) (cc) (est.) Gmb (corr) Gmm (%) (mm) (cc) (est.) (corr) Gmm (%) Avg.
0 76.43 619.64 1.854 1.929 77.91 75.23 609.91 1.857 1.944 78.5 78.2
5 70.92 574.97 1.998 2.079 83.96 69.76 565.57 2.003 2.096 84.7 84.3
10 68.48 555.19 2.069 2.153 86.95 67.28 545.46 2.077 2.174 87.8 87.4
15 67.06 543.68 2.113 2.198 88.79 65.95 534.68 2.118 2.217 89.6 89.2
30 64.87 525.92 2.184 2.273 91.79 63.93 518.30 2.185 2.287 92.4 92.1
50 63.42 514.16 2.234 2.325 93.89 62.67 508.08 2.229 2.333 94.2 94.1
100 62.06 503.14 2.283 2.376 95.94 61.35 497.38 2.277 2.384 96.3 96.1
120 61.75 500.63 2.294 2.387 96.43 60.93 493.98 2.293 2.400 96.9 96.7
153 61.33 497.22 2.310 2.404 97.09 60.57 491.06 2.307 2.414 97.5 97.3
180 61.16 495.84 2.316 2.411 97.36 60.35 489.28 2.315 2.423 97.9 97.6
200 61.05 494.95 2.321 2.415 97.53 60.25 488.46 2.319 2.427 98.0 97.8
225 60.96 494.22 2.324 2.418 97.68 60.21 488.14 2.320 2.429 98.1 97.9
253 60.77 492.68 2.331 2.426 97.98 60.00 486.44 2.328 2.437 98.4 98.2
Gmb ðmeas:Þ ¼ 2:426; C ¼ 1:0406 Gmb ðmeas:Þ ¼ 2:439; C ¼ 1:047

Gmm ¼ 2.476.
170 B.L. SWAMI et al.

TABLE XV Densification data for blend 1, 5.75% binder content

Specimen 1 Specimen 2

Ht Volume Gmb Ht Volume Gmb


Gyrations (mm) (cc) (est.) Gmb (corr) Gmm (%) (mm) (cc) (est.) Gmb(corr) % Gmm Avg.
0 76.08 616.80 1.849 1.924 78.23 76.05 616.56 1.902 1.976 80.37 79.300
5 70.75 573.59 1.988 2.069 84.13 70.71 573.27 2.045 2.125 86.44 85.282
10 68.20 552.92 2.062 2.146 87.27 68.39 554.46 2.115 2.198 89.37 88.321
15 66.65 540.35 2.110 2.196 89.30 67.07 543.76 2.156 2.241 91.13 90.215
30 64.39 522.03 2.184 2.273 92.44 65.06 527.46 2.223 2.310 93.94 93.190
50 63.05 511.17 2.231 2.321 94.40 63.82 517.41 2.266 2.355 95.77 95.085
100 61.76 500.71 2.277 2.370 96.37 62.72 508.49 2.306 2.396 97.45 96.910
120 61.36 497.46 2.292 2.385 97.00 62.48 506.54 2.315 2.405 97.82 97.412
153 61.05 494.95 2.304 2.397 97.49 62.27 504.84 2.322 2.414 98.15 97.823
180 60.89 493.65 2.310 2.404 97.75 62.03 502.90 2.331 2.423 98.53 98.141
200 60.68 491.95 2.318 2.412 98.09 62.01 502.73 2.332 2.424 98.56 98.326
225 60.62 491.46 2.320 2.414 98.19 61.94 502.17 2.335 2.426 98.67 98.430
253 60.51 490.57 2.324 2.419 98.37 61.95 502.25 2.334 2.426 98.66 98.512
Gmb ðmeas:Þ ¼ 2:427; C ¼ 1:044 Gmb ðmeas:Þ ¼ 2:426; C ¼ 1:039

Gmm ¼ 2.459.

FIGURE 1 Average densification curves at different binder content.

The moisture susceptibility or the deterioration of a actually washes off the aggregate so that bare aggregate
HMA due to the detrimental influences of moistures, remains still. The more typical situation is that there is a
called stripping. Stripping produces a loss of strength gradual loss of strength over a period of years, which
through the weakening of the bond between the asphalt contributes to the development of rutting and shoving in
cement and the aggregate. This loss of strength can be the wheel paths. The tensile strength ratio comes out to
sudden and catastrophic where the asphalt peels off the be 91.75%, which is more than the minimum criteria of
aggregate, the cohesion of the mixture is lost, and 80%. Hence, this mix will be able to resist the rutting in a
distresses develop rapidly. Sometimes the asphalt cement better way.

TABLE XVI Mix compaction properties TABLE XVII Mix volumetric properties at Ndes

Binder (%) Gmm at N10 (%) Gmm at N153 (%) Gmm at N253 (%) Binder (%) Air voids (%) VMA (%) VFB (%)
4.25 86.5 95.3 96.3 4.25 4.7 14.4 67.2
4.75 87.8 96.9 97.9 4.75 3.1 13.9 77.8
5.25 87.4 97.3 98.2 5.25 2.7 14.5 81.4
5.75 88.3 97.8 98.5 5.75 2.2 15.1 85.6
SUPERPAVE—AN INDIAN CASE STUDY 171

FIGURE 2 Binder content versus air voids.

Since all the parameters fulfil the design criteria of Stability value of the superpave mix is 27% higher than
superpave mix design, the mix design for asphaltic the marshall mix, while flow value of superpave mix is
concrete layer can be accepted at a design binder content slightly less than that of marshall mix.
of 4.4% by weight of aggregates.
Indirect Tensile Strength
COMPARISON OF MARSHALL METHOD WITH To compare the indirect tensile strength of two mixes, a set
SUPERPAVE MIX DESIGN METHOD of three specimens were prepared with marshall compactor
at optimum binder content and another set of three
A comparison of the design results and design criteria specimens were prepared with gyratory compactor at
between the superpave mix design and marshall mix optimum binder content. These six specimens were tested
design has been made. for indirect tensile strength as per ASTM procedure and the
results were tabulated in the Table XX. As can be seen from
Optimum Binder Content Table XX, the indirect tensile strength of the marshall mix
In the marshall method of mix design, the optimum binder at 5.31% binder content was 9.067 kg/cm2 and that of
content was 5.31% by weight of aggregate while in the superpave mix at 4.4% binder content was 13.49 kg/cm2.
superpave mix designs it was 4.4% by weight of mineral
aggregate. Retained Stability and Moisture Sensitivity
Retained stability is the stability of the marshall specimen
Stability and Flow
tested after placing it in water bath at 608C for 24 h. This
Stability and flow value is not the design criteria in test is conducted to find out the effect of water on
superpave mix design. However, to make the comparison bituminous mixes. The similar test for the superpave mix
with marshall mix design, specimens were prepared at is called moisture sensitivity test. Here, the sample is
4.4% binder content on gyratory compactor and tested for tested for indirect tensile strength test after 24 h in water
stability and flow values. Comparative results are shown in bath maintained at 608C and 2 h at 258C. In marshall
Table XIX. mixes, the retained stability shall be 75% of the stability as

TABLE XIX Stability and flow values of marshall and superpave mixes
TABLE XVIII Design mixture properties at 4.4% binder content
Method Marshall Superpave
Mix property Result Criteria
Percent B.C. 5.31 4.4
Air voids % 4.0 4.0 Specimen A1 A2 A3 B1 B2 B3
VMA% 14.15 13.0 min Stability (kg) 1100 1090 1110 1505 1520 1505
VFB% 71 65–75 Average 1100 1510
Dust proportion 0.91 0.6–1.2 stability (kg)
Gmm@Nini ¼ 10 (%) 87 Less than 89 Flow (mm) 3.2 3.1 3.0 2.9 3.0 2.8
Gmm@Nmax ¼ 253 (%) 97 Less than 98 Average flow (mm) 3.1 2.9
172 B.L. SWAMI et al.

TABLE XX Indirect tensile strength value for marshall and superpave EFFECT OF ANGLE OF GYRATION ON
mixes PROPERTY OF SUPERPAVE MIXES
Method Marshall Superpave
Effect of angle gyration on:
Percent B.C. 5.31 4.4
Specimen A1 A2 A3 B1 B2 B3
ITS (kg/cm2) 9.2 9.0 9.0 13.45 13.50 13.52 (i) Density
Average ITS 9.067 13.49 (ii) Indirect tensile strength.

To observe the effect of angle variation on mix


per MORTH specifications but for superpave mixes properties, same mineral aggregate and blend was used for
moisture sensitivity shall be 80% as per ASTM standards. superpave mix design at 5% binder content.
The retained stability and moisture sensitivity comes
out 80 and 91%, respectively, for marshall and superpave
Density of Mix
mixes; therefore, superpave mixes are less affected by
water compared to marshall mixes. To observe the effect of angle of gyration on compaction
characteristics of the mixes, a number of specimens were
Asphalt Film Thickness compacted at 5% binder content, with varying gyratory
angle i.e. at 1, 1.25, 1.5 and 28. Average percent TMD of
The asphalt film thicknesses were calculated for marshall
specimens was calculated at each gyratory angle and
and superpave mixes and are mentioned below:
plotted against number of gyrations as shown in Figure 3.
The graph shows that percentage TMD increases with
Asphalt film thickness for marshall mix ¼ 9.5 microns
the increase in number of gyrations. At any level of
Asphalt film thickness for superpave mix ¼ 7.5
gyration, percentage TMD increases with increase in
microns.
gyratory angle.
For this particular mix, number of gyrations required to
Densification of Superpave Mixes achieve 96% TMD at different gyratory angle are as given
To determine the densification trend of superpave mixes in Table XXI.
with variation in gyrations at different binder contents
graphs were plotted between percentage theoretical
Indirect Tensile Strength
density and number of gyrations (log scale) as shown in
Figure 1. To observe the effect of angle on indirect tensile strength,
Graph shows that the percentage theoretical density specimens were compacted at 1, 1.25, 1.5 and 28 at
increases linearly with the increase in number of optimum binder content by giving 153 gyrations (design
gyrations. From the graph it can also be observed that number of gyrations). Figure 4 and Table XXII show the
the percentage theoretical density increases with the variation in indirect tensile strength with angle of
increase in binder content. gyration.

FIGURE 3 Mixture compaction characteristics with variation in angle.


SUPERPAVE—AN INDIAN CASE STUDY 173

FIGURE 4 Indirect tensile strength at different gyratory angle.

TABLE XXI Number of gyration required to obtained 96% TMD . Indirect tensile stress decreases as the angle of gyration
increases, with a dramatic change when angle of
Angle of gyration 18 1.258 1.58 28
Number of Gyrations 130 80 70 65 gyration is greater than 1.258.

RECOMMENDATIONS
TABLE XXII Indirect tensile strength values for different angle of
gyration
. In developing countries like India, where a lot of
Angle 18 1.258 1.58 28
ITS (kg/cm2) 13.48 13.32 12.69 12.48
investment is underway in highway infrastructure,
it is recommended that superpave technology is
adopted. This will not only prove economical, but
From this graph it was observed that indirect tensile will also increase the lifetime of pavements.
strength decreases with increase in angle of gyrations but . To evaluate the performance of superpave mixes in
there is a sudden decrease after an angle 1.258, which is actual field conditions, a test track shall be laid and the
the angle used for superpave mix design. performance shall be studied.
The analysis of results reveal that the gyratory . Further research should be carried out to compare
compaction better simulates the effect of roller compac- more bituminous mixes, i.e. semi dense bituminous
tion and secondary compaction by traffic and therefore concrete, dense bituminous concrete and bituminous
will eventually replace the marshall design procedure macadam using marshall and superpave methods.
in the developing world. However the marshall method The comparison should be done with similar traffic
will, no doubt, be used for some time to come till the levels i.e. equal million esa.
superpave technology is fully developed in developing
countries like India.
References
CONCLUSIONS
Cominsky, R., Leahy, R.B. and Harrigan, E.T. (1994) Level One Mix
Design: Materials Selection SHRP-A-408 (National Research
. Superpave mixes fulfill all the criteria for easy and Council, Washington, D.C), Strategic Highway Research Program.
good construction at lesser binder content than the Kandhal, P.S. and Robert, L. (1997) “Hot mix asphalt materials”, Mixture
marshall mixes. Design and Construction, 1997, Oct. 1997.
McGennis, Anderson, R.M., Kennedy, T.W. and Solaimanian, M. (1994)
. Increase in binder content will not yield for a good mix Background of Superpave Asphalt Mixture Design and Analysis,
for the same number of gyrations. Report FHWA-SA-95-003, FHWA, US Department of Transpor-
. Superpave mixes are least affected by water. tation.
Soysa, J., et al. (1997) Evaluation of Laboratory Procedure for
. Density increases with increase in angle of gyration but Compacting Asphalt Aggregate Mixtures, Report SHRP-A-91-523,
increment is negligible between the angles 1.25 and 28. NRC, September.

You might also like