Construction and Building Materials 25 (2011) 150–155

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

A study to determine the degree of partial blending of reclaimed asphalt

pavement (RAP) binder for high RAP hot mix asphalt
Prashant Shirodkar a, Yusuf Mehta a,*, Aaron Nolan a, Khyati Sonpal a, Alan Norton a, Chris Tomlinson a,
Eric Dubois a, Patrick Sullivan b, Robert Sauber c
a
Department of Civil Engineering, Rowan University, 201, Mullica Hill, Glassboro, NJ 08028, USA
b
Department of Chemical Engineering, Rowan University, 201, Mullica Hill, Glassboro, NJ 08028, USA
c
New Jersey Department of Transportation (NJDOT), NJ, USA

a r t i c l e i n f o a b s t r a c t

Article history: The objective of this paper is to provide a methodology for determining the degree of partial blending in
Received 5 March 2010 high RAP mixtures. When RAP is mixed with virgin aggregates and virgin binder, partial blending of RAP
Received in revised form 3 May 2010 binder occurs in the hot mix asphalt. Agencies limit the amount of RAP because the degree of blending
Accepted 19 June 2010
between the RAP and the virgin materials is not known. The methodology provides a systematic approach
Available online 13 July 2010
for determining the degree of partial blending in the RAP mixture. The ability to accurately determine the
degree of partial blending will help in determining the virgin binder content to be added in the mixture.
Keywords:
The degree of partial blending measured from this procedure for 25% RAP by weight of aggregates with
Reclaimed asphalt pavement
Recycling
PG 70-28 and 35% RAP by weight of aggregates with PG 58-28 are 70% and 96% respectively.
Partial blending Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction need to determine a methodology that quantifies the degree of par-

The demand to utilize higher percentages of reclaimed asphalt
pavement (RAP) in the construction of hot mix asphalt (HMA)
pavements continues to rise [1]. Several states, including New Jer-
sey, conduct Superpave mixture designs for the high RAP mixtures.
This procedure consists of burning the asphalt from the RAP and
considering the RAP as one of the aggregates (without any RAP bin-
der). After the design binder content is determined, the virgin bin-
der content is established by giving full credit to the RAP binder,
assuming 100% blending of RAP binder. However, researchers have
consistently shown that when RAP is mixed with virgin binder and
aggregates, partial blending occurs [2], that is, something between
100% blending and the ‘‘black rock” effect occurs. The assumption
of 100% working binder may produce an under-asphalted mixture,
resulting in poor pavement performance.
Several state agencies limit the amount of RAP that can be
placed in the different layers of pavement [3]. One of the primary
reasons the RAP usage is limited is because the degree of blending
between the RAP and the virgin materials has not been quantified.
The question that needs to be answered is how much of the RAP
binder blends with the virgin binder. Is there 100% blending or
0% blending or somewhere in between? The answer is probably
somewhere in between the two extremes. Therefore, there is a
* Corresponding author. Tel.: +1 856 256 5327; fax: +1 856 256 53242.
E-mail address: mehta@rowan.edu (Y. Mehta).
tial blending occurring in high RAP mixtures.
The ability to quantify the amount of blending will be critical in
determining the actual virgin binder content and grade that needs
to be added to the mixture after the design binder content is ob-
tained. Determining the amount of RAP binder that is available
for blending with virgin binder is the primary focus of this paper.
2. Objective
The objectives of the study are:
(a) To provide a framework/methodology on how to determine
the degree of partial blending in the high RAP mixtures.
(b) To determine the degree of partial blending occurring in the
HMA consisting of 25% and 35% RAP by weight of aggregates.
3. Recent studies regarding determination of partial blending
McDaniel and Anderson [4] have recommended the use of a
three-tier system based on the assumption of full-blending be-
tween virgin and RAP binder which was later modified to incorpo-
rate the low temperature grade. The new three-tier system allows
a maximum of 20% RAP without a change in the binder selection
and up to 30% RAP by lowering one grade softer. For mixes using
more than 30% RAP, a blending chart is recommended in order to
adjust the binder grade accordingly.

0950-0618/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2010.06.045
 P. Shirodkar et al. / Construction and Building Materials 25 (2011) 150–155
Kandhal and Foo [5] at NCAT confirmed the use of the three-tier
system and also developed a ‘‘sweep blending chart” to determine
the percentage of RAP if a three-tier system was not used. The
‘‘sweep blending chart” required the determination of G/sin(d)
and Gsin(d) for different percentages of virgin binder for high
and intermediate temperatures. The percentage of RAP obtained
by the intermediate temperature sweep blending chart (average
37%) was higher than the typical average practice of around 15–
20%. To rectify the discrepancy between the calculated percentage
of RAP and actual practice, Kandhal and Foo [5] recommended a
‘‘specific grade” blending chart which has reduced the effort of
developing three sweep blending charts.
Huang et al. [2] blended RAP with virgin aggregates without any
new virgin asphalt binder being introduced. The mixture was
mechanically blended (dry blended). The purpose was to find out
the extent at which the aged asphalt from RAP particles would
blend with virgin aggregate.
In addition to the coating study, Huang et al. [2], attempted to
determine the viscosity of RAP mixed with virgin binder by per-
forming staged extractions. After the RAP was mixed with virgin
binder, the RAP was dipped in trichloroethylene in stages and the
solution was recovered. It was clear that the asphalt viscosity in-
creased as it went from outside layers to the inside layers. Based
on the staged extraction described in the study, it was observed
that about 60% of the inner portion of the binder (nearest to the
RAP aggregate) had asphalt properties close to pure RAP aged bin-
der, whereas the outside 40% of the binders were blended with vir-
gin binder.
Al-Qadi et al. [6] conducted performance related tests for four
blending cases, black rock effect (BR), total blending (TB), partial
blending (PB) and actual practice (AP). The results confirmed the
study carried out by McDaniel and Anderson [4] who had studied
blending of 10% and 40% RAP content. They found that at low
RAP contents, the BR, TB, and AP cases were statistically similar
but at high RAP contents, the BR case was different (TB and AP
stayed statistically similar). Al-Qadi et al. [6] also concluded that
the higher complex modulus of AP samples was either due to the
selective absorption of lighter fractions in the aggregate surface
over time or from the change in gradation caused by partial blend-
ing (whose extent is unknown). Gradation change is caused either
by formation of mastic layer or release of fine particles in the RAP
binder.
Al Qadi et al. [6] made three mixes consisting of 0%, 20% and
40% RAP. In all the three cases, the overall gradation was kept
the same. The Superpave mixture design of the above three
mixes indicated that the binder content was the same. The sur-
face area of the aggregates was similar for all the three mixes
due to similar gradations. Due to similar surface area and binder
content of all three mixtures, Al Qadi et al. concluded that 100%
RAP binder was mobilized in all the three cases. However, the
authors of this study believe that the binder content being the
same does not necessarily conclude that the 100% RAP blending
is occurring.
Other studies to evaluate interaction between virgin and RAP
binder is carried out by Bennert et al. [7]. Bennert et al. [7] devel-
oped the analytical procedure, using back calculation methodology
along with analytical methods developed by Bonaquist [8] and
Rowe [9] to determine ‘‘effective” asphalt properties of HMA con-
taining RAP. The term effective asphalt properties is used to de-
scribed degree of interaction between virgin and RAP binder in
RAP HMA. The concept of this procedure is if the back calculated
asphalt binder properties of RAP HMA differ from extracted and
recovered binder which completely blends during the extraction
and recovery process then the degree of blending is less than
100%. Results show that the degree of blending for the 15% and
20% RAP HMA were not as thorough as 25% RAP HMA.
151
The above studies made significant headway in enhancing the
understanding of the degree of partial blending, but none of them
actually quantify the degree of blending. Based on the lessons
learned from the above studies, the authors have proposed a new
methodology to determine the degree of partial blending.
4. Experimental procedure
4.1. Materials
In this study, to demonstrate the proof of concept, two virgin binder grades of
PG 70-28 and PG 58-28 from NuStar refineries and two percentage of RAP – 25% and
35% by weight of aggregates from a single plant were analyzed. The virgin aggre-
gates, the RAP, and the binder were selected from an approved Job Mix Formula
(JMF) from the same plant. The plant selected for the mixture design also provided
the RAP for the laboratory study.
4.2. Estimation of the degree of partial blending
Based on Huang et al. [2], RAP was sieved below the #8 (2.36 mm) sieve and the
virgin aggregates were all above the #4 (4.75 mm) sieve. The gap gradation was cre-
ated to separate the RAP and virgin aggregates at the #4 sieve so that the degree of
blending could be determined. Every effort was made to keep the gradation close to
the JMF provided by the plant. Fig. 1 shows the gradation of the JMF and the gap
gradation of 25% and 35% RAP by weight of aggregates, used in the study.
To obtain a realistic estimate or ‘‘starting point” of the degree of blending, the
virgin aggregates were mixed with RAP without the virgin binder. This was per-
formed to evaluate how much of the RAP binder mobilized and coated the virgin
aggregates during the mixing process without the introduction of virgin binder.
The mixing process was kept as close as possible to the practice followed by the
plant in terms of the mixing time, the mixing process, and temperatures of the vir-
gin aggregates and the RAP. The detailed procedure used for the coating study with-
out virgin binder is as follows:
 Sieve the virgin aggregate above the #4 (4.75 mm) sieve as per the procedure
described above.
 Wash the aggregate to remove any fines that would pass the #4.
 Dry the aggregate in the oven.
 Sieve the RAP to be less than #8 sieve (2.36 mm).
 Measure a total of 2000 grams of aggregate and RAP according to the gradation
determined in the above paragraph.
 Heat the aggregates, bucket and mixing arm to 350 °F in the oven.
 RAP is added to the hot virgin aggregates 30 min prior to the mixing.
 Mix the RAP and virgin aggregates for 10 min in the oven with a mechanical
mixer.
 Put the aggregate and the RAP mix in the oven for 2 and ½ h at 350°F.
 Remove the aggregate and RAP mix from the oven; allow the aggregate mix to
cool until it is ready to be handled.
 Separate the aggregates and the RAP from the aggregate/RAP mix through siev-
ing (be sure to remove the entire RAP from the aggregate as some of it will be
attached).
 Weigh the aggregates and the RAP that has been separated from the mix.
Fig. 1. Gradation of the JMF and the gap gradation for 25% and 35% RAP by weight
of aggregates.
152 P. Shirodkar et al. / Construction and Building Materials 25 (2011) 150–155

Table 1
Material used in coating study (without binder).

25% RAP 35% RAP

Weight (%) Batch 1 (g) Batch 2 (g) Weight (%) Batch 1 (g) Batch 2 (g)
Aggregate bin 3 35.4 708.8 709 26.5 529.1 529.4
Aggregate bin 3 12.6 253.1 253 11.6 232.5 232.4
Aggregate bin 3 26.9 539.9 539 26.9 539.1 538.4
Initial weight of RAP 25 500.1 500.4 35 700.7 700.1
Total 100 2001.9 2001.4 100 2001.4 2000.3
Initial weight of virgin aggregates 1501.8 1501 1300.7 1300.2
Final weight of virgin aggregates 1508.9 1507.6 1305.7 1307
Approximate binder transfer (%) 25.2 23.4 12.7 17.3

4.4. Blending study

The binder content from the JMF was used in the design. A full Superpave mix
design was not deemed necessary because the gradation was modified with the sole
intent to determine the degree of partial blending. The optimum binder content
from the JMF supplied by the plant was 4.8%. The approximate RAP binder transfer
from the above coating study was used to determine the amount of virgin binder
content.

4.5. Virgin binder content to be added based on the approximate RAP working binder

Fig. 2. Schematic representation of procedure of coating study.
The mixing time of 10 min was selected by trial and error. After 10 min, no addi-
tional change in visual coating of virgin aggregates was observed. Table 1 shows the
batch weight of different aggregate bins and results of the coating study. As shown
in Fig. 2 the increase in weight of the virgin aggregates is due to the coating by the
RAP binder, however, the reduction in weight of the RAP aggregates may be due to
four things (a) loss of moisture content, (b) RAP binder lost to bucket and arm, (c)
loss of fine particles of RAP during mixing, and (d) transfer of RAP binder to virgin
aggregates. Therefore, the loss of RAP weight will be greater than the increase in the
weight of virgin aggregates. This coating study without the virgin binder only pro-
vides an estimate of the partial blending because some of the RAP working binder
will also coat the RAP aggregates. Additionally, this cannot be measured in this pro-
cess and the impact of the presence of hot virgin binder on the degree of partial
blending cannot be captured.
Table 2 shows the weights of different aggregate bins by weight of mix (con-
verted from weight of aggregates to weight of mix). The weight of total mix was se-
lected such that sufficient binder can be extracted for determining the binder
properties.
The detailed procedure used for the blending study with virgin binder is similar
to the one outlined above except that the weights changed (shown in Table 2) and
the virgin binder was also heated to mixing temperature. A schematic representa-
tion of the procedure is shown in Fig. 3.
4.6. Binder properties
After mixing, the virgin aggregates were separated from the RAP aggregates
by slight heating and manually separation. To minimize the heating duration,
three batches of 5000 g mix were prepared. Also, while separating the mix, only
a small portion of mix is heated in the oven. The binder from a separate mix was

4.3. Approximate RAP binder transfer

The percentage of RAP binder in the RAP was obtained from the extraction and
recovery process [10]. The percentage of binder in the RAP is 5.63%. The mass of bin-
der in the RAP was determined from weight of RAP aggregates. The RAP binder coat-
ing the coarse aggregates was determined from increase in weight of coarse virgin
aggregates. The approximate RAP binder transferred was calculated based on the
following equation. Table 1 shows the RAP binder transfer for 25% and 35% RAP
averaged 24% and 15%, respectively. These values of binder transfer seem much
lower than what is expected. Hence, it was approximated to higher end. Binder
transfer was considered as 30% and 20% for 25% and 35% RAP.

Approximate RAP binder transfer ð%Þ

¼ 100  ðweight of RAP binder coating the virgin aggregates after mixing
=weight of binder around the RAP before mixingÞ ð1Þ Fig. 3. Schematic representation of procedure of blending study.

Table 2
Materials used in blending study (with virgin binder).

Material 25% RAP by weight of 25% RAP by weight of 35% RAP by weight of 35% RAP by weight of
aggregates (%) aggregates (g) aggregates (%) aggregates (g)
Aggregate bin 3 33.7 1685.0 25.2 1261.4
Aggregate bin 3 12.0 599.8 11.0 552.2
Aggregate bin 3 25.6 1280.4 25.6 1280.4
RAP 24.2 1210.40 33.7 1685.0
Virgin binder 4.4 219.6 4.4 221.0
Total batch weight 99.9 4995.2 99.9 5000.0
Total binder 4.8 240.0 4.8 240
 P. Shirodkar et al. / Construction and Building Materials 25 (2011) 150–155 153

extracted and recovered using the AASHTO T319 procedure [10]. The RTFO G/
sin(d) of extracted binder was conducted at 76 °C and 70 °C. This temperature
selection for the mix prepared with 25% RAP by weight of aggregates with PG
70-28 virgin binder was the same as the high PG-grade of the virgin binder.
From the testing results it was found that binder testing temperature did not af-
fect the determination of degree of partial blending. Hence, further all testing
were carried out at the same temperature (76 °C and 70 °C). The G/sin(d) of
RTFO binder was selected for two reasons: (a) The amount of binder required
for a RTFOT sample can be obtained with one single extraction and recovery
using the AASHTO T319 procedure [10] and (b) the binder properties at high
temperatures are generally more sensitive to blending than low temperature test
results [5].
The concept behind this procedure is that if there is 100% blending, the proper-
ties of the blended binder on the virgin and RAP aggregates would be similar. If
there is 0% blending, the properties of the blended binder around the RAP aggre-
gates after mixing will be a blend of the RAP binder and the proportion of the virgin
binder that coats the RAP aggregates. The properties of the binder at zero blending
condition around the virgin aggregates will be the properties of the virgin binder
only. The following section will explain how the zero percent blending condition
is simulated.
4.7. Determination of the properties of the binder around the RAP aggregate under zero
blending condition
In zero blending condition, RAP binder acts like a rock, in that it does not
coat other virgin aggregates. While preparing mix, virgin binder coats the RAP.
When coarse virgin aggregates are extracted and recovered, theoretically only
virgin binder would be expected. When fine RAP aggregates are extracted and
recovered, mixed RAP and virgin binder in certain proportion would be obtained.
This proportion of RAP binder and virgin binder can be calculated by determin-
ing film thickness of RAP binder and virgin binder. The film thickness of RAP bin-
der calculated by Bailey’s method was found to be 10 micron using the following
equation:
Weight of effective asphalt binder around the aggregate ¼ AC=ð100  ACÞ
Film thickness ¼ weight of effective asphalt binder around the aggregate=ð1000
 specific gravity of asphalt  total surface area
Here,
a. Weight of effective asphalt binder around the aggregate is calculated in kg/
kg of aggregates.
b. AC is asphalt content determined by extraction recovery method AASTHO
T319 [10] and expressed in percentage (%).
c. Specific gravity of asphalt is assumed as 1.02.
d. Total surface area is determined as per Bailey’s method [11], it is sum of
product of surface area factor and gradation (percent passing) of extracted
aggregates and expressed in m2/kg (calculated in Table 3).
Average film thickness of virgin binder around the RAP was assumed as 8 mi-
cron. This assumption was also cross checked using Bailey’s Method. Film thickness
of virgin binder for 4.4% asphalt content for the given RAP aggregate gradation was
found to be approximately 8 micron. Hence, the ratio of the RAP binder and the vir-
gin binder coating the RAP is 56:44 (10:8).
The RAP binder and the virgin binder were mixed according to the above ratio.
This proportioned binder was subjected to short term conditioning in RTFO test and
the Superpave PG properties (RTFO G/sin(d)) were calculated. Table 4 shows the
Superpave PG properties for the binder coating the RAP aggregate under the zero
blending condition.
5. Results and discussion
5.1. Determination of the degree of partial blending
If there is full-blending, the properties of the binder around the
virgin aggregates will be similar to that of the binder around the
RAP. As the degree of blending decreases, the difference in
the properties between the blended binder around the virgin
aggregates and the RAP will approach that of the difference under
ð2Þ
the zero blending condition. After the properties of binder coating
the virgin aggregates and the RAP aggregates under ‘‘black rock
effect” or 0% blending and full (100%) blending are identified, the
degree of partial blending was determined from the following
ð3Þ
equations below:

Blending ratio ¼ ðjðG = sinðdÞÞblend binder virgin agg

Table 3
Surface area using Bailey’s method.  ðG = sinðdÞÞblend binder RAP agg jÞ=ðjðG = sinðdÞÞvirgin binder
Sieve size Surface area factor Percent passing Surface area  ðG = sinðdÞÞRAP virgin binder 0 blend jÞ
(mm) (m2/kg) (%) (m2/kg) ð4Þ
37.5 100
25 0.41 100 0.41
Degree of partial blending ð%Þ ¼ 100j1  blending ratioj ð5Þ
19 0 100 0.00
12.5 100 0.00
where(G/sin(d))blend binder virgin aggRTFO G/sin(d) of blended binder
9.5 100 0.00
4.75 0.41 100.0 0.41 coating the coarse virgin aggregates.(G/sin(d))blend binder RAP aggRTFO
2.36 0.82 100.0 0.82 G/sin(d) of blended binder coating the fine RAP aggregates.
1.18 1.64 37.6 0.62 (G/sin(d))virgin binderRTFO G/sin(d) of the virgin binder.(G/sin(d))RAP
0.6 2.87 22.2 0.64 
virgin binder 0 blendRTFO G /sin(d) of the RAP and virgin binder that is
0.3 6.14 14.0 0.86
0.15 12.29 8.5 1.04
coating the RAP aggregate assuming 0% blending.
0.075 32.77 3.2 1.05 Numerator of this equation is difference of RTFO G/sin(d) of
0.00 blended binder around virgin aggregates and RAP aggregates. As
Total 5.84 the blended binder around the virgin and RAP aggregates is
subjected to aging during mixing and heating (of around 4–4.5 h)

Table 4
Comparison of RTFO G/sin(d) values and degree of partial blending for 25% and 35% RAP by weight of aggregates.

Percentage of Batch Testing G/sin(d) (kPa) Degree of partial

RAP (%) temperature blending (%)
Virgin RAP + virgin (zero RAP Blended binder Blended binder around coarse
(°C)
binder blending condition) binder around fine RAP virgin aggregates
25 (PG 70-28) 1 70 5.3 12.0 29.8 3.2 1.3 69
76 3.2 6.5 14.8 1.8 0.8 72

35 (PG 58- 1 70 1.2 6.0 29.8 1.4 1.0 92

28)
76 0.6 3.1 14.8 0.7 0.5 92
2 70 1.2 6.0 29.8 1.2 1.2 100
76 0.6 3.1 14.8 0.7 0.7 100
154 P. Shirodkar et al. / Construction and Building Materials 25 (2011) 150–155
carried out during separation process. This blended binder is tested
for RTFO DSR without subjecting to RTFO simulation. Denominator
is difference between virgin binder and proportioned virgin and
RAP binder (for zero blending). Proportioning of virgin binder
and RAP binder recovered from RAP below #8 sieve is carried out
as per the proportioned determined in Section 4.7 of this paper.
This unaged binder is subjected to RTFO simulation and tested
for RTFO DSR. In Eq. (4), extend of aging for binders in numerator
and denominator would not be similar, as one is carried out while
preparation of mix and other is carried out by RTFO simulation. But
it is not going to affect the determination of degree of partial
blending as ratio is taken. Table 4 shows a comparison of binder
stiffness (RTFO Gsin(d)) values and degree of partial blending for
25% and 35% RAP by weight of aggregates.
5.2. Discussion
The degree of partial blending of RAP is dependent upon many
factors like aggregate temperature during mixing, grade of binder,
RAP binder property, percentage of RAP, and virgin binder prop-
erty. From the above results, it can be seen that the degree of par-
tial blending is more for softer binder. Also, degrees of partial
blending determined from different binder testing temperatures
are similar. Hence, the degree of partial blending is independent
of binder testing temperature.
5.3. Methodology
The methodology to determine the degree of partial blending is:
Step 1: Determine the binder content of the RAP and the grada-
tion of the extracted aggregates.
Step 2: Determine the Superpave PG properties (from AASHTO
M320) of the RAP binder and the virgin binder.
Step 3: Create a Superpave gradation for a given percentage of
RAP (i.e. 25% and 35%), such that all the fine aggregates (minus
#8–2.36 mm) are RAP and all coarse aggregates (greater than
#4–4.75 mm) are virgin aggregates. The Superpave gradation
created in the lab will be similar to the JMF gradation for a given
percentage of RAP.
Step 4: The design binder content from the JMF could be consid-
ered for the study. If the design binder content is not known,
determine the design binder content (DBC) based on the Super-
pave mixture design.
Step 5: Coating study – mix the RAP and the virgin aggregates.
The mixing process was kept as close as possible to the practice
followed by the plant in terms of the mixing time, the mixing
process, and the temperatures of the virgin aggregates and
the RAP. Calculate the increase in mass of virgin aggregates
before and after mixing to determine the ‘‘the approximate”
amount of RAP binder that coated the aggregates. This will help
in determining the virgin binder content. Or assume initial bin-
der transfer of around 50%.
Step 6: Create the mixture at the virgin binder content (VBC)
determined from the following equation:
Virgin binder content ¼ design binder content from JMF
 estimated RAP working binder
where the RAP working binder depends on the degree of partial
blending is determined from step 5 above.
Step 7: Separate the coated virgin and RAP aggregates after mix-
ing by conducting sieve analysis at the #4 and #8 sieves
Step 8: Extract and recover the binder separately from the coarse
virgin aggregates (plus #4) and fine RAP aggregates (minus #8).
Step 9: Determine the Superpave PG properties (from AASHTO
M320 like RTFO G/sin(d)) of the blended binder on the RAP
and the virgin aggregates.
Step 10: Determine the proportion of RAP and virgin binder that
would coat the fine RAP aggregates under zero blending condi-
tion by estimating surface area of fine RAP aggregates using Bai-
ley’s method.
Step 11: Blend the RAP binder with the proportion of the virgin
binder determined from step 10 above. Subject this mixture of
binder to short term aging by RTFO simulation. Determine the
Superpave PG properties (from AASHTO M320), such as RTFO
G/sin(d).
Step 12: Calculate the degree of partial blending from the fol-
lowing equation:
If property A of the virgin binder and RAP binder is similar
Blending ratio ¼ ðjAðvirgin aggÞblend binder
 AðRAP aggÞblend binder j=jðAvirgin binder
 ARAP virgin binder 0 blend Þj
Degree of partial blending ð%Þ ¼ 100j1  blending ratioj
where A(Virgin_agg)blend binder = binder property ‘‘A”, such as
‘‘G/sin(d)”, of blended binder coating the virgin aggregates
(determined from step 9), A(RAP_agg)blend binder = binder prop-
erty ‘‘A”, such as ‘‘G/sin(d)”, of blended binder coating the RAP
aggregates (determined from step 9), Avirgin binder = binder
property ‘‘A”, such as ‘‘G/sin(d)”, of virgin binder (determined
from step 2), ARAP virgin binder 0 blend = Binder property ‘‘A”, such
as ‘‘G/sin(d)”, of the RAP and virgin binder that is coating the
RAP aggregate assuming 0% blending (determined from step
11).
Step 13: Iteration – if the degree of partial blending (determined
from step 12) is similar to the calculated value in step 5 then the
degree of partial blending has been determined. However, if
considerable difference exists between the two, the process will
be repeated with the revised value of the RAP working binder
that is obtained from step 12 and the steps will be repeated
from step 6 onwards.
6. Summary of findings
The findings based on the study are:
a. The degree of partial blending for 25% RAP by weight of
aggregates of chosen gradation and PG 70-28 virgin binder
is 70%.
b. The degree of partial blending for 35% RAP by weight of
aggregates of chosen gradation and PG 58-28 virgin binder
is 96%.
c. Degree of blending determined by blending study is much
higher than that determined by coating study. Hence step
of determination of approximate binder could be skipped
by assuming first approximate value of 50% for preparing
mix.
d. Also, from the comparison of results due to blending study
and coating study it could be seen than influence of presence
of virgin binder is much higher (46% (70%  24%) and 81%
(96%  15%) for 25% and 35% RAP by weight of aggregates)
and hence change in degree of blending due to change in
gradation will be of smaller magnitude which could be
neglected.
e. Degree of partial blending is independent of binder testing
temperature.
 P. Shirodkar et al. / Construction and Building Materials 25 (2011) 150–155
f. Degree of partial blending is higher for PG 58-28 as compare
to PG 70-28.
g. The recommended methodology was successfully adopted
to determine the degree of blending of RAP binder.
7. Significance of study
The methodology proposed in this paper provides a systematic
approach of determining the degree of partial blending in RAP. The
ability to accurately determine the degree of partial blending will
help in precisely determining the virgin binder content to be added
in a mixture. It will also help in developing a blending chart to
determine the properties of the final binder grade, the required vir-
gin binder grade and the percentage of RAP based on the degree of
partial blending measured from this procedure.
8. Limitations
The limitations of the procedure:
a. The process requires at least three extractions and recover-
ies of the binder; these include the RAP before the blending
study, the virgin aggregates and the RAP after the blending
study.
b. To minimize the heating during separation, prepared sam-
ples should be large enough; heating should be done on por-
tions of sample at a time which can be separated without
cooling the sample. The sieving of the coated aggregates is
very time consuming.
9. Conclusions and recommendations
The study successfully demonstrated a proof of concept of accu-
rately determining the degree of partial blending of RAP binder.
The degree of partial blending for 25% RAP by weight of aggregates
of chosen gradation with PG 70-28 virgin binder was found to be
on an average of 70%. The degree of partial blending for 35% RAP
by weight of aggregates of chosen gradation with PG 58-28 virgin
binder was found to be on an average 96%. However, these results
are depended upon various factors like RAP binder property, grada-
tion, mixing temperature, virgin binder property, hence should not
155
be generalized. From the above results, methodology to determine
degree of partial blending is verified and this should be used to
determine degree of partial blending of desired combination of
materials.
Acknowledgements
This research was funded by the New Jersey Department of
Transportation. 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 re-
flect the official views of the New Jersey Department of
Transportation.
References
[1] Nation Asphalt Pavement Association. Hot mix asphalt is the best buy all of the
time; 1998. <http://www.flexiblepavements.org/documents/BB11.pdf>
[accessed 01.03.09].
[2] Huang BZ, Li G, et al. Laboratory investigation of mixing HMA with RAP. In:
Transportation Research Board: Journal of the Transportation Research Board,
no. 1929. Washington (DC): Transportation Research Board of the National
Academies; 2005. p. 37–45.
[3] Tam K, Marks P, Cautillo C. Results of survey on RAP usage. Material Engineer
and research office, Ministry of Ontario; 2008.
[4] McDaniel RS, Soleymani H, Anderson RM, Turner P, Peterson R. Recommended
use of reclaimed asphalt pavement in the SuperPave™ mixture design method.
NCHRP Final Report (9–12), TRB, National Research Council, Washington, DC;
2000.
[5] Kandhal P, Foo KY. Designing recycled hot mixture asphalt mixtures using
superpave technology. NCAT Report no. 96-5; 1997. p. 7–22.
[6] Al-Qadi IL, Carpentor SH et al. Determination of usable residual asphalt binder
in RAP. Research Report ICT-09-031, Illinois Center for Transportation; 2009.
[7] Bennert T, Dongre R. A back calculation method to determine ‘‘Effective”
asphalt binder properties of RAP mixtures. Transportation Research Board,
Washington, DC; 2010.
[8] Bonaquist R. New Approach for the Design of High RAP HMA. In: Presented at
the 2005 Northeast Asphalt User’s Producer’s Group Meeting, Burlington, VT,
October 19th to 20th, 2005; 2005.
[9] Rowe G. Phase angle determination and interrelationships within bituminous
materials. In: 7th international RILEM symposium on advanced testing and
characterization of bituminous materials, Rhodes, Greece, May 27th to 29th,
2009; 2009.
[10] AASHTO Standard T 319. Standard method of test for quantitative extraction
and recovery of asphalt binder from asphalt mixtures; 2008.
www.transportation.org.
[11] Sengoz B, Topal A. Minimum voids in mineral aggregate in hot-mix asphalt
based on asphalt film thickness. Sci Dir 2007;42(10):3629–35.