Effects of Rejuvenating Agents On Superpave Mixtures Containing Reclaimed Asphalt Pavement
Effects of Rejuvenating Agents On Superpave Mixtures Containing Reclaimed Asphalt Pavement
Effects of Rejuvenating Agents On Superpave Mixtures Containing Reclaimed Asphalt Pavement
Abstract: Rejuvenator is not a commonly used softening agent to be used in recycling of reclaimed asphalt pavement 共RAP兲. In this
study, Superpave mixtures containing RAP were designed using rejuvenating agents, including a rejuvenator and a softer binder, and
subsequently evaluated in terms of the volumetric results, obtained the indirect tensile strength 共ITS兲 of samples as well as evaluating the
mixtures for rutting using the asphalt pavement analyzer 共APA兲. The content of the rejuvenator used for those mixtures containing the
rejuvenator was determined from the blending charts of RAP binders containing the rejuvenator. A total of 12 Superpave mixtures
including 10 containing RAP and two virgin were designed. The results indicated, for the mixtures tested for this project, that: 共1兲
properties of the recycled mixtures using the rejuvenator, such as ITS and APA, were better than those containing the softer binder; 共2兲
10% more RAP could be incorporated in the Superpave mixtures by using the rejuvenator than using the softer binder; and 共3兲 the
blending charts established under the Superpave binder specifications can be used to determine the content of the rejuvenator for the
recycling.
DOI: 10.1061/共ASCE兲0899-1561共2007兲19:5共376兲
CE Database subject headings: Recycling; Asphalt pavements; Asphalt mixes; Flexible pavements; Aging; Shear deformation.
Rejuvenator
One oil type of rejuvenator available commercially in the United
States was selected for this study. The properties of the rejuvena- in this study兲; TRAP⫽critical temperature of extracted RAP binder
tor are listed in Table 4. The content of the rejuvenator 共%兲 共Table 3兲; and %RAP⫽percentage of RAP expressed as a decimal
needed for rejuvenating the RAP binder to PG 64-22 was deter- 共i.e., 0.30 for 30%兲.
mined on the blending charts established by DSR and BBR. When the softer binder was used as a rejuvenating agent, the
When using RAP, there has long been a question regarding to maximum percentages of 38 and 30% were determined to be
what extent the RAP binder blends with the virgin binder/ satisfying the volumetric requirements and the combined grada-
rejuvenator to achieve a target PG grade. To answer this question, tion requirements for the two RAP sources containing Aggregate
three situations of the RAP, i.e., black rock, total blending and Sources C and L, respectively. This process of determining the
real world were proposed. Black rock is the idea that there is RAP percentage was completed by adjusting the percentages of
absolutely no mixing between the RAP binder and virgin binder/ the two fractions of the RAPs into designated aggregate structure
rejuvenator, whereas total blending is assumed that the two bind- and satisfying the volumetric parameters. When the rejuvenator
ers completely and uniformly mix. The research indicated that the was used, 10% more 共48 and 40%兲 of RAPs with Aggregate
real-world results more closely match the total-blending, depend- Sources C and L were incorporated, respectively 共Table 5兲.
ing upon the amount of RAP. Based on this finding, the total
amount of the rejuvenator needed in the mixtures was obtained by
Test Program
multiplying the amount of the RAP binder contained in RAP with
the %rejuvenator. Fig. 1 shows the experimental design used in this study. Table 5
shows the combination of all Superpave mixture designs.
Percentage of RAP
The possible percentage of RAP incorporated was carefully de- Critical Temperatures and Blending Charts–DSR and BBR
termined so that both Superpave binder and mixture specifications Tests
could be met. The percentage of RAPs, 17–43%, was initially Extraction and recovery tests of the RAP binders were performed
calculated using Eq. 共1兲 below 共McDaniel and Anderson 2001兲 according to ASTM D 2172-81 共centrifuge extraction with ro-
共Table 3兲: tavapor recovery兲 and ASTM D 1856-95a 关recovery of asphalt
%RAP = 共Tblend − Tvirgin兲/共TRAP − Tvirgin兲 共1兲 cement 共Abson method兲兴 procedures. Samples were prepared in
accordance with ASTM Method D 979 共sampling bituminous
where Tvirgin⫽critical temperature of virgin binder 共PG58-22, in paving mixtrure兲. The properties of the RAP binders were ob-
this study兲; Tblend⫽critical temperature of blend binder 共PG64-22, tained by DSR and BBR on samples of the extracted RAP binders
in three stages: aged binder 共no further aging兲, RTFO residual,
and RTFO+ PAV residual.
Table 4. Properties of the Rejuvenator Used
Conventional blending chart is established with the viscosity/
Properties Ranges penetration as ordinate and the percentage of a rejuvenator as
Specific gravity 共15.6/15.6C兲 0.98–1.02 abscissa for selecting the type and the content of the rejuvenating
Viscosity, 60C CST 200–500 agent. This chart uses only two percentages of the rejuvenating
Flash Point, COC 共°C兲 204 minimum agent: 0 and 100% of the rejuvenating agent with the relationship
RTFO-C 163° C 关weight loss 共%兲兴 4.0 maximum
showing the viscosity/penetration and the percentage being as-
sumed to be linear 共Kandhal and Foo 1997兲. In the research
RTF-C 共viscosity ratio兲 2.5 maximum
project presented in this paper, blending charts were established
Compatibility 共PC/S ratio兲 0.5 minimum
with performance properties at high temperature 共64° C兲, interme-
Saturates 共w%兲 28 maximum
diate temperature 共25° C兲 and low temperature 共−12° C兲 as ordi-
Chemical compatibility 0.2–1.2
nates and the percentage of the rejuvenator as abscises. Three
percentages of the rejuvenator were added into the RAP binders test, two sets of three samples each after being cured under dif-
so that linear regression equations could be obtained 共Shen and ferent specified conditions were tested by following the SC-T-70
Ohne 2002; Kennedy et al. 1998兲. testing procedures.
tents with which the G* / sin共␦兲 reaches 1.0 and 2.2 kPa for the Volumetric Results
blends in the original state 共no aging兲 and after RTFO, respec-
Results of the 12 Superpave mix designs are shown in Table 7.
tively. These contents of 12.7 and 13.5% were obtained by ex-
The average optimum binder content of Superpave mixtures con-
trapolating the lines or algebraically using the equations for no
taining RAPs using the rejuvenator was lower than that of those
aging and RTFO aging cases; respectively.
using the softer binder regardless of the sources of the RAP
Second, a minimum allowable content was determined by sat-
sources 共i.e., C or L兲. Generally, the optimum binder content of
isfying the fatigue and the shrinkage parameters, i.e., G*sin共␦兲,
the Superpave mixtures containing RAPs decreased slightly with
stiffness and m-value, of the blends after RTFO+ PAV aging.
the increased content of the RAP, but no clear relationship be-
G*sin共␦兲 that was obtained at the intermediate temperature of
tween the optimum binder content and the content of the RAP
25° C was also linearly correlated with the content of the rejuve-
was found, Fig. 7. In addition, optimum binder content of virgin
nator 共Shen and Ohne 2002兲. A minimum rejuvenator content of
mixtures was higher than that of all recycled mixtures.
12.3% was obtained by extrapolating the regression line with the
The change in the optimum binder content may be caused by
G*sin共␦兲 value being less than 5.0 MPa 共Fig. 4兲. Similarly, the
the difference in the viscosity of the binders in the mixtures. The
m-value and stiffness obtained at low temperature of −12° C were
amount of the rejuvenating agents 共rejuvenator or softer binder兲
correlated linearly with the contents 共Figs. 5 and 6兲. A minimum
was obtained based on the estimated weight of the RAP binders
rejuvenator content of 10.8 and 2.9% were obtained by extrapo-
lating the regression lines with the stiffness value being less than
300 MPa and the m value being larger than 0.3, respectively. As a
result, the minimum rejuvenator content satisfying the fatigue and
the shrinkage parameters was found to be 12.3%.
Finally, a mean value of the common contents, 12.5%, with
which all requirements were satisfied 共AI 2001兲, was used as a
design value in this study 共Table 6兲. This content is approximately
2-3% of the weight of the mixtures containing 30–50% of RAP.
This is a reasonable and practical value for this rejuvenator ac-
cording to the manufacturer.
Test results
%Max density at Ndes 96 95.5 95.9 96 95.8 96 96 94.9 95.9 96 95.8 96 96
%VMA 15.5–17.5 16.2 15.8 15.62 15.52 15.55 15.52 16.6 15.8 15.85 15.52 15.58 15.65
%VFA 70–80 72.5 76.1 75 71.9 72 74.5 76.2 76.1 74.5 71.9 70 72
%Max density at Nini ⇐89 88 88.4 88 88.5 88.3 88.2 87 88.4 89 88.5 88.3 87.8
%Max density at Nmax ⇐98 96.5 97.1 96.5 96.9 97.2 97.3 96.1 97.1 97 96.9 97.4 97
Dust-to-asphalt ratio 0.6–1.2 0.95 0.83 0.89 0.94 0.92 1.05 0.91 0.83 0.9 0.94 1.13 1.18
Optimum binder content 共%兲 5.02 5.3 4.93 5.05 5.01 4.87 5.5 5.4 5.09 4.9 4.8 4.85
Note: C⫽Aggregate Source C; V⫽virgin binder; R⫽rejuvenator; L⫽Aggregate Source L; and 0, 15, 38, and 48⫽percent of RAP incorporated in each
mixture.
using the binder content in each RAP source. Considering the 共65 psi兲 required by the SCDOT, regardless of the aggregate
Black Stone effect of some RAPs in the mixtures, the amount of source, the rejuvenating agent, and the curing state. The mixtures
the rejuvenating agents added may overdose, leading to a smaller containing the rejuvenator agent produced higher strengths than
viscosity than expected. The volumetric properties of HMA is those containing the softer binder. Furthermore, the mixtures con-
closely related with the content and viscosity of the binder for a taining RAP, in general, had higher ITS values than the corre-
given compaction condition. In general, a lower viscosity and sponding virgin mixture of the same aggregate source. The mix-
higher binder content will make the compaction easier to accom- ture made with 10% more RAP Source L and containing the
plish. Therefore, to get the same air void content 共i.e., 4%兲, less rejuvenator still had higher strength than required by SCDOT’s
asphalt content is needed for a lower viscosity binder. specifications. This trend was also true for mixtures containing
Percent voids in the mineral aggregate 共%VMA兲 of the mix- Aggregate Source C. This indicated that the mixtures containing
tures, as shown in Fig. 8, were greater than 15.5%, a value re- the rejuvenator produced ITS results as good as or even better
quired by the Superpave specifications. The values of %VMA of than the mixes made with the softer binder. There was no appar-
the Superpave mixtures containing RAPs with the rejuvenator
ent relationship between the percentage of the RAP incorporated
were less than those using the softer binder regardless of the
in the mixtures and the ITS values. In addition, some mixtures
sources of the RAP. The virgin mixtures had a higher %VMA
produced higher ITS values in wet state than in dry state, espe-
than the recycled mixtures using either rejuvenator or the softer
cially for Aggregate Source C. This phenomenon was occasion-
binder. In general, the %VMA decreased with the increase of the
rejuvenator content in mixes made with aggregate C and contain- ally observed in experiments for the mixtures using hydrophobic
ing RAP Source C. The percent voids filled with asphalt 共%VFA兲 aggregates. However, the curing of the recycled mixtures under
and dust asphalt ratio of all the mixtures satisfy the requirements hot water may improve the interaction of the RAP aggregate with
of Superpave mix specifications 共Table 7兲. the binders.
The percent tensile strength ratio 共%TSR兲, defined by the ITS
strength in the wet state divided by that in the dry state, are higher
ITS and TSR Results than the standard of 85% required by the SCDOT 共Fig. 10兲 for all
Fig. 9 shows the ITS average values of three samples of the mixtures. For Aggregate Source C, the ratios ranged from 95.7%,
mixtures incorporating Aggregate Sources C and L in wet and dry the lowest for the virgin mixture and as high as 141% for the
states. Generally, it is shown that all of the Superpave mixtures mixture containing 15% RAP. For Aggregate Source L, the range
have higher wet ITS values than the standard of 0.46 MPa of TSR ratio was from 83.3%, for the mixture containing 30%