close

Вход

Забыли?

вход по аккаунту

?

14680629.2017.1389088

код для вставкиСкачать
Road Materials and Pavement Design
ISSN: 1468-0629 (Print) 2164-7402 (Online) Journal homepage: http://www.tandfonline.com/loi/trmp20
Optimising water content in cold recycled foamed
asphalt mixtures
Wangyu Ma, Randy West, Nam Tran & Nathan Moore
To cite this article: Wangyu Ma, Randy West, Nam Tran & Nathan Moore (2017): Optimising water
content in cold recycled foamed asphalt mixtures, Road Materials and Pavement Design, DOI:
10.1080/14680629.2017.1389088
To link to this article: http://dx.doi.org/10.1080/14680629.2017.1389088
Published online: 24 Oct 2017.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=trmp20
Download by: [UAE University]
Date: 25 October 2017, At: 17:37
Road Materials and Pavement Design, 2017
https://doi.org/10.1080/14680629.2017.1389088
Optimising water content in cold recycled foamed asphalt mixtures
Wangyu Ma ∗ , Randy West, Nam Tran and Nathan Moore
National Center for Asphalt Technology, Auburn University, Auburn, AL, USA
Downloaded by [UAE University] at 17:37 25 October 2017
(Received 15 August 2016; accepted 25 October 2016 )
During cold recycling, water is added to facilitate the dispersion of foamed asphalt in the
mixture and to achieve uniform mixing and help compaction by providing sufficient lubrication. Too little water may cause difficulty in workability and compaction of the mixture,
but too much water may extend the curing time and reduce density and strength. Therefore,
the optimum water content (OWC) was considered as one of the most important factors in
mix design procedures for cold recycling. Currently, mix design procedures for cold recycled foamed asphalt mixtures suggest adding water to the mixture at an optimum content to
facilitate mixing and compaction. However, there is no standard method for determining the
optimum total water content (OTWC) for cold recycling mixtures. Several empirical relationships were developed to determine the OTWC based on modified Proctor test results for
Reclaimed Asphalt Pavement (RAP)/aggregate. However, the compaction effort in the modified Proctor test for RAP/aggregate may not match that for mixtures, which is compacted
using the Superpave Gyratory Compactor (SGC) or Marshall hammer. A study is underway to
improve the design method for cold recycled foamed asphalt mixtures with 100% RAP. The
purpose of this paper is to optimise the design procedure by developing a new method to determine OTWC. SGC was used to compact RAP instead of the modified Proctor test to match
the compaction effort recommended for foamed asphalt mixtures. A regression model was
developed to calculate the OTWC for a mixture based on the determined OWC of the RAP,
foamed asphalt content, and binder type as factors. The method for determining OTWC for
a mixture was validated using six different mixtures and was found to correlate well with the
measured OTWC, even though two of six mixtures had underestimated OTWC due to different binder source. Further comparisons with other two OTWC determining methods showed
the mixtures at the proposed OTWC had improvement in indirect tensile strength.
Keywords: cold recycling; foamed asphalt; water; modelling
1. Introduction
Cold recycling is one of the most sustainable technologies in pavement construction and rehabilitation (Chan, Tighe, & Chan, 2010; Stroup-Gardiner, 2011; Thenoux, González, & Dowling,
2007). As it reclaims asphalt pavement up to 98%, cold recycling can conserve large quantities
of natural resources (Diefenderfer, Bowers, Schwartz, Farzaneh, & Zhang, 2016). Also, since
cold mixtures are produced at ambient temperatures, cold recycling can significantly reduce fuel
consumption and greenhouse gases generated during construction.
Foamed asphalt was first proposed for use in pavement recycling in 1957 (Csanyi, 1957). It
is commonly used as a recycling agent for cold recycling. To make foamed asphalt, water and
air are injected into hot asphalt binder. The process creates bubbles of water vapour within the
*Corresponding author. Email: wzm0006@auburn.edu
© 2017 Informa UK Limited, trading as Taylor & Francis Group
Downloaded by [UAE University] at 17:37 25 October 2017
2
W. Ma et al.
binder that help expand the asphalt volume and reduce the binder viscosity. With the increased
volume and lower viscosity, foamed asphalt can easily disperse among and mix with Reclaimed
Asphalt Pavement (RAP) particles to provide non-continuous bonding (Diefenderfer et al., 2016;
Muthen, 1998). During cold recycling, water is also added to facilitate the dispersion of foamed
asphalt in the mixture and to achieve uniform mixing and help compaction by providing sufficient
lubrication (Asphalt Recycling & Reclaiming Association [ARRA], 2015). Too little water may
cause difficulty in workability and compaction of the mixture, but too much water may extend
the curing time and reduce density and strength (Jenkins, 2000). Therefore, the optimum water
content (OWC) was considered as one of the most important factors in the mix design procedure.
Since there is currently no standard mix design procedure for cold recycled mixtures with
foamed asphalt, different methods have been used to determine the optimum total water content (OTWC) in mixture. Most of them are based on the OWC of RAP/aggregate determined
by the modified Proctor test in accordance with American Association of State Highway and
Transportation Officials (AASHTO) T180. Mohammad, Abu-Farsakh, Wu, and Abadie (2003)
recommended using 100% of the OWC determined by the modified Proctor test. Wirtgen GmbH
(2008) has recommended a reduction factor for OWC by the modified Proctor test. Muthen
(1998) recommended using OWC by the modified Proctor test as the optimum total fluid content, including water and foamed asphalt. Sakr and Manke (1985) proposed a regression model
to calculate the OTWC using different factors such as OWC determined by the modified Proctor
test, proportion of fines, and proportion of foamed asphalt.
While these methods for determining the OTWC have been used in the mix design, there are
two key issues need to be addressed. First, the compaction effort in the modified Proctor test for
RAP/aggregate does not match the compaction effort used to design the foamed asphalt mixture,
which is typically done using Marshall Compactor or Superpave Gyratory Compactor (SGC).
Second, the relationship used to determine the OTWC varied from one study to another because
it was developed based on limited types of mixtures. These issues may affect the accuracy of
the determined OTWC, influencing the density and strength of cold recycled foamed asphalt
mixtures. A new method for determining the OTWC for an improved mix design method for
cold recycled foamed asphalt mixture was investigated in a recent study. This paper discusses
the results of that study.
2. Objective and scope
The objective of this study was to develop a new method for determining the OTWC for an
improved cold recycled foamed asphalt mixture design method. The new method uses a SGC at
30 gyrations to determine the OWC of RAP which matches the compaction effort recommended
for mixture (ARRA, 2015). This compaction effort was considered equivalent to the field compaction effort (Cross, 2003). The new method also uses a regression model to determine the
OTWC based on RAP type, binder type, and foamed asphalt content.
In the following sections, material properties and the mixture production method are introduced followed by a discussion of the effect of water on mixtures. Finally, a new method to
determine OTWC in cold recycled foamed asphalt mixture was proposed and validated in the
laboratory.
3. Materials
Materials used in this study are discussed in this section. RAP is discussed first followed by the
asphalt binder used to produce foamed asphalt. Cold recycled foamed asphalt mixtures produced
with these RAP and binder are discussed finally.
Road Materials and Pavement Design
3
Table 1. Summary of residual binder for RAP.
RAP type
DF (A)
DM (B)
DC (C)
VF (D)
VC (E)
Gradation categorya
NMAS (mm)
Residual binder
content (%)
Residual
binder PG
Fine
Medium
Coarse
Fine
Coarse
12.5
19.0
19.0
19.0
25.0
4.92
4.85
4.10
5.19
5.36
100-10
100-4
100-10
100-10
NAb
a Classification is based on relative location of the gradation curve compared to recommended gradation range by Wirtgen
GmbH (2012).
binder PG for Type E RAP was not tested due to availability of materials.
Downloaded by [UAE University] at 17:37 25 October 2017
b Residual
3.1. RAP
Table 1 shows the information of five different RAP types (from A to E) used in this study. They
were sampled from stockpiles in the NCAT Pavement Test Track or East Alabama Paving plant,
which had been milled, crushed, and screened. Gradation was tested following procedures suggested in ARRA guidelines (American Standards for Testing and Materials (ASTM) C136 and
C117). Residual binder content was tested using ignition method in accordance with AASHTO
T308. Residual binder was extracted and recovered to determine performance grade following
AASHTO M320. RAP types are listed in Table 1. The first letter ‘D’ refers to the optimising
design and ‘V’ refers to validation. The second letter (F, M, or C) represents the gradation category. For example, ‘DF’ RAP is design optimising RAP in the fine gradation category. Types
A–E indicate the sequence when they were used in this study. Residual binder content of these
RAP was similar except for VC RAP with less binder content. Available residual binder PG
grades of these RAP sources are also similar.
Figure 1 plots five RAP gradations measured following ASTM C136 and C117. To categorise
these RAP, the gradation range recommended by Wirtgen GmbH (2012) was also plotted as a
reference. Most gradation curves were basically within the range or slightly offset. DF gradation
has more offsets on the large sieves and DM gradation has more on the small sieves. These
gradations could be altered by further processing RAP or blending with virgin aggregates, but
this is out of scope of this study.
3.2. Asphalt binder
Three types of virgin asphalt binder, including two PG 67-22 binders and one PG 58-34 binder,
were used to produce foamed asphalt. The information for these binder types are shown in
Table 2. Binder PG 67-22 is commonly used in state of Alabama for asphalt pavement construction. The two PG 67-22 binders were obtained from different refinery plants with different
sources. Even with the same performance grade, they were still considered different binder types
when producing foamed asphalt because foaming properties such as expansion ratio (ER) and
half-life (HL) may be affected by different sources (Newcomb et al., 2015). The foaming properties of each binder were tested and summarised in Table 2. Binder 58-34 (B) produced a foamed
asphalt with the highest ER and HL. Binder 67-22 (C) could not produce stable foamed asphalt
at 160°C; so the foaming temperature was increased to 170°C. Binder C had marginal foaming
properties in accordance with the criteria in the ARRA guidelines.
3.3. Cold recycled foamed asphalt mixes
Thirteen cold recycled foamed asphalt mixtures were produced with the previously mentioned
RAP and binder. The method to produce the mixtures is discussed later in experimental plan.
4
W. Ma et al.
Table 2. Summary of asphalt binder PG and foaming properties.
Foaming Properties
Binder type
Temp. (°C)
Water Content (%)
ER (times)
HL (s.)
Lake city, FL
Alberta, CA
Birmingham, AL
160
160
170
1.8
2.0
1.3
11
19
9
9
10
6
Portland cement was added to mix 1–10 at 1.0% to improve durability, asphalt–aggregate
bonding, and moisture damage resistance (ARRA, 2015; Asphalt Academy, 2009; Wirtgen
GmbH, 2012). It was assumed that the effect of cement on the mixtures was small and consistent
regardless of water content when equal amounts of cement was added because water evaporates
quickly once curing starts. Mix 11 and 12 have no cement, allowing for the effect of cement on
OTWC to be evaluated. Mix 13 with 1.5% cement was only used for validating the proposed
OTWC determination method (Table 3).
4. Experimental plan
4.1. Producing mixture
Cold recycled foamed asphalt mixtures were produced in the laboratory for testing in this study
following the ARRA guidelines with modified procedures for determining the OWC of RAP
and testing indirect tensile strength (ITS). Figure 2 shows the basic steps for producing foamed
asphalt mixtures. Each step is discussed briefly in the following sections. FAC stands for foamed
asphalt content and TWC stands for total water content. More detailed information about the
procedures can be found in the ARRA cold recycling guideline (ARRA, 2015).
100
90
80
70
60
50
20
10
Figure 1.
Gradation plot of different RAP types.
12.5
9.50
4.75
2.36
1.18
Sieve Size
(mm)
0.075
0.15
0.30
0.60
0
31.0
30
25.0
DF
DM
DC
VF
VC
Wirtgen coarse limit
Wirtgen fine limit
40
19.0
Percent Passing
Downloaded by [UAE University] at 17:37 25 October 2017
67-22 (A)
58-34 (B)
67-22 (C)
Binder source
Road Materials and Pavement Design
Table 3.
Downloaded by [UAE University] at 17:37 25 October 2017
Mix no.
1
2
3
4
5
6
7
8
9
10
11
12
13
5
Summary of mixtures with different binder and RAP types.
Asphalt binder
RAP
Additive (Type
I/II cement)
67-22 (A)
67-22 (A)
67-22 (A)
58-34 (B)
58-34 (B)
58-34 (B)
58-34 (B)
58-34 (B)
67-22 (C)
67-22 (C)
58-34 (B)
58-34 (B)
67-22 (C)
DF (A)
DM (B)
DC (C)
DF (A)
DM (B)
DC (C)
VF (D)
VC (E)
VF (D)
VC (E)
DF (A)
DM (B)
DF (A)
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
1.0%
No cement
No cement
1.5%
4.1.1. Step 1: preparing and testing materials’ properties
RAP materials were sampled from an East Alabama Paving’s plant stockpile. The RAP materials
had been milled, crushed, and screened before stockpiling. After sampling, the RAP was stored
in buckets and dried in front of a fan for a week before mixing. A mechanical splitter was used
to reduce RAP to testing size following AASHTO R47. Gradation, binder content, and binder
performance grade of the RAP were tested. Virgin asphalt was heated and split into 1-gallon cans
for ease of foaming properties’ tests. ER and HL were tested using Wirtgen WLB10S laboratory
plant. Also, the optimum foaming water content and temperature were determined to produce
foamed asphalt. The machine in Figure 3 is the laboratory foaming plant.
4.1.2. Step 2: determining RAP OWC
RAP can be compacted to its maximum dry density after mixing with the optimum amount of
water. This water content is called the OWC and used as guidance for determining the OTWC
in cold recycled asphalt mixture. The current ARRA guideline recommends using the modified
Proctor method to compact RAP in accordance with AASHTO T180 (or ASTM D1557) and
using 100% OWC of RAP as the OTWC for mixture. However, this compaction effort does
not match that used for compacting foamed asphalt mixtures, which is 30 gyrations in a SGC.
Instead of using the modified Proctor method, the compaction effort by SGC is adopted in this
Step 1: Preparing and testing materials’ properties for RAP & binder
Step 2: Determining RAP OWC based on proposed method using SGC
Step 3: Mixing and compacting at desired FAC and TWC
Step 4: Laboratory curing (40 °C, 3 days)
Step 5: Testing compacted density and indirect tensile strength (d/w)
Figure 2.
Basic steps to produce cold recycled foam asphalt mixtures.
Downloaded by [UAE University] at 17:37 25 October 2017
6
W. Ma et al.
Figure 3.
Laboratory pug mill (left) and foaming plant (right).
Figure 4.
Procedure to determine RAP OWC using SGC.
study to compact RAP. Testing procedure for determining the OWC of RAP is similar to ASTM
D1557 but has some changes. The basic steps of determining the OWC of RAP in this study are
summarised in Figure 4. A description of each step follows.
• Dry RAP with a fan for a week;
• Remove particles above ¾-inch sieve for use with 100-mm diameter molds;
• Add water to RAP at different water contents; at each water content, prepare at least 5kg RAP (using mechanical splitter to reduce RAP to test size) in order to compact three
replicates. Add water slowly and mix with a scoop by hand until uniform;
• Load wet RAP into a 100-mm mold and compact for 30 gyrations at an internal angle of
1.16° in a SGC;
• Record the compacted height and wet weight after compaction. Dry the compacted specimens overnight at 110°C to determine the molding water content, which is used to
calculate dry density;
• Plot dry density against water content and determine OWC at the maximum dry density.
All the TWC levels selected for mixtures are based on the OWC of RAP determined from
this procedure. The OWC of RAP was compared with that determined from the modified Proctor
method. Four RAP types were selected for comparison. Tested water contents are listed in the
experimental plan shown in Table 4. The OWC determined by the proposed method and the
modified Proctor method are listed and compared in Table 5.
Table 5 shows that the OWCs of RAP determined by the proposed method are lower than those
determined by the modified Proctor method. However, the maximum dry density from the two
methods is similar, while those determined by proposed method are slightly lower. The difference
Road Materials and Pavement Design
7
Table 4. Experimental plan for determining RAP OWC.
Selected water contents for testing (%)
RAP types
DF (A)
DM (B)
VF (D)
VC (E)
Proposed SGC method
Modified Proctor method (ASTM D1557)
3, 4, 5, 6
1, 2, 3, 4, 5, 6, 7
2, 3, 4, 5, 6, 7
2, 3, 4, 5
5, 6, 7, 8
2, 3, 4, 5, 6
3, 4, 5, 6, 7, 8
4, 5, 6, 7
Table 5. Summary of RAP OWC and maximum dry density determined from proposed
method and modified Proctor method.
Maximum dry density (kg/m3 )
Downloaded by [UAE University] at 17:37 25 October 2017
RAP OWC (%)
RAP types
DF (A)
DM (B)
VF (D)
VC (E)
Proposed SGC
Mod. Proctor
Proposed SGC
Mod. Proctor
4.0
3.0
4.0
3.0
6.5
3.8
6.5
5.3
2046.8
1940.8
1969.3
1899.6
2101.1
1944.6
2018.3
1955.5
between two OWC values of the same RAP sample may be due to different compaction methods.
However, further evaluation is needed in the future.
4.1.3. Step 3: mixing and compacting
RAP was mixed with cement, water, and foamed asphalt in a laboratory pug mill, as shown in
Figure 3. Materials were added following a recommended sequence mentioned in ARRA (2015)
guideline, which is RAP, mineral additive, followed by water and foamed asphalt. Added water
helps spreading foamed asphalt in RAP and aids compaction of mixture (ARRA CR101). The
twin-shaft pug mill has 10–30 kg mixing capacity. At each time, 10–15 kg RAP was added for
mixing. Mixing speed and mixing time are adjustable. This study applied a medium driving speed
(about 72 rpm) and 60-s mixing time for each mixture. Wirtgen GmbH (2008) recommended 30s for aggregate/RAP blending with cement and water. Therefore, 15-s pre-mixing was applied for
both cement and water in this study to prevent non-uniformity of these two non-asphalt materials
in mixture.
The compaction method recommended in ARRA (2015) guideline was applied. Cold recycled
foamed asphalt mixture was compacted for 30 gyrations in SGC. Specimens are compacted to
the geometry of 100-mm diameter and 63.5 ± 2.5 mm height.
4.1.4. Step 4: curing
The curing procedure used in this study followed the ARRA (2015) guideline. Compacted specimens were cured in oven at 40°C right after compaction. Curing time was 3 days for all the
specimens.
4.1.5. Step 5: testing density and ITS
Dry density of each compacted specimen is an indicator of durability and compaction quality. It
was determined based on dry weight of specimen and measured dimension after curing.
8
W. Ma et al.
Table 6. Experimental plan for determining effect of mixing water on density and dry ITS.
Mix no.
2
2
2
Binder
RAP
FAC (%)
MWC (% OWC)
TWC (% OWC)
67-22 (A)
67-22 (A)
67-22 (A)
DM (B)
DM (B)
DM (B)
3.0
3.0
3.0
50, 75
50, 75
50, 75
75
100
125
Downloaded by [UAE University] at 17:37 25 October 2017
ITS test was conducted for each mixture in both dry and wet conditions. Dry specimens were
cured at 25°C 1 h before testing and wet specimens were cured in a 25°C water bath for 24 h
before testing. The wet ITS testing method is simplified from ARRA (2015) guideline because
vacuum saturation was not applied. This method was recommended in Wirtgen GmbH (2012)
cold recycling manual.
4.2. Mechanism and modelling
The effect of mixing and compaction water on mixture performance needs to be investigated to
optimise water content. After this mechanism was fully understood and influencing factors were
identified, modelling was conducted to find a relationship between the OTWC and the influencing
factors.
4.2.1. Mechanism
In cold recycling, water is added to the cold mix to achieve uniform mixing and to facilitate
compaction by providing lubrication in the mix (ARRA, 2015). This section investigates the
role of water during mixing and compaction in a laboratory environment. Understanding this
mechanism is essential to foamed asphalt mix design because it answers the question of why
water is needed. First, the effects of water during mixing and compaction on the compacted
density and ITS were assessed separately. Only one mixture (no. 2) was used in this step. Water
was added in two steps: mixing water was added before foamed asphalt and compaction water
was added afterwards but before compaction. Then, the combined effect was determined for six
different mixtures (nos. 1–6).
4.2.1.1. Effect of mixing water During mixing, water helps distribute foamed asphalt. This
concept was proposed in 1950s by Csanyi (1957). This effect determines how uniform the foamed
asphalt is distributed in the mixture. The difference in distribution may cause changes in ITS
because strength is dependent on the uniformity of bonding between foamed asphalt and RAP
particles. In this study, the effect was evaluated quantitatively by examining dry ITS at different
mixing water contents (MWCs), while keeping total compaction water constant. Wet ITS was
not tested because of a lack of availability of specimens. The experimental plan is summarised in
Table 6. The same mixture was tested for density and dry ITS at two MWCs at each TWC level.
Analysis of variance (ANOVA) was performed in Minitab 17 software to determine the effect of
MWC and compare it with TWC.
4.2.1.2. Effect of compaction water During compaction, water provides lubrication between
particles and facilitates the compaction process. The effect of compaction water content may be
observed from the change of compacted density or ITS at different compaction water contents
while keeping MWC constant. An experiment was designed based on the mix design procedure
proposed by Wirtgen GmbH (2012), in which water is introduced in two steps. Wirtgen GmbH
(2012) suggests adding 75% of the OWC before foamed asphalt to achieve ideal mixing and then
Road Materials and Pavement Design
9
Table 7. Experimental plan for determining effect of compaction water on density and dry ITS.
Mix no.
2
2
2
2
Binder
RAP
TWC (%OWC)
FAC (%)
67-22 (A)
67-22 (A)
67-22 (A)
67-22 (A)
DM (B)
DM (B)
DM (B)
DM (B)
50, 75, 100, 125
50, 75, 100, 125
50, 75, 100, 125
50, 75, 100, 125
1.5
2.0
2.5
3.0
Table 8. Experimental plan for determining combined effect of total water on density, dry and wet ITSs.
TWC (% OWC)
Downloaded by [UAE University] at 17:37 25 October 2017
Mix no.
1
2
3
4
5
6
Binder
RAP
Density
Dry ITS
Wet ITS
FAC (%)
67-22 (A)
67-22 (A)
67-22 (A)
58-34 (B)
58-34 (B)
58-34 (B)
DF (A)
DM (B)
DC (C)
DF (A)
DM (B)
DC (C)
50, 75
75, 100, 125, 150
100, 125, 150, 175
50, 75
75, 100, 125,
100, 125, 150
50, 75
75, 100, 125, 150
100, 125, 150, 175
50, 75
75, 100, 125
100, 125, 150
50, 75
75, 100, 125, 150
100, 125, 150, 175
50, 75
75, 100, 125
100, 125, 150
2, 3
2, 3
2, 3
2, 3
2, 3
2, 3
adding the remaining water after the foamed asphalt to facilitate compaction. In this study, 50%
of the OWC was added first and remaining water was added after foamed asphalt. Mixing was
performed by hand to ensure that the water was distributed uniformly. The effect of compaction
water on dry density and ITS is examined at different FACs. Table 7 shows the experimental plan
for determining effect of compaction water on density and dry ITS.
4.2.1.3. Effect of TWC for each mixture The above experimental plan in Tables 6 and 7 investigated the effect of water during mixing and compaction separately. However, water was added
only once either in the laboratory mix design or during field production. Same amount of water
not only facilitates mixing but also compaction. Therefore, this water has a combined effect on
compacted mixture. This study added water only once to investigate the combined effect on
density and strengths. The combined effects of water on density as well as dry and wet ITSs
were evaluated for six mixtures. Wet ITS was included because more specimens were prepared
for strength testing. Table 8 summarises experimental plan for determining the effect of TWC
on density and strength (dry and wet ITSs). Each mixture was tested at two FAC levels. This
allowed an ANOVA to be performed to determine the effect of TWC as well as FAC.
Results of the analysis are summarised and discussed later (as shown in Table 13). Except
for the TWC in mixture, some other factors may also affect the performance of mixtures, and
therefore, affect the OTWC. These factors need to be investigated before modelling OTWC.
4.2.2. Modelling
The material types with potential to affect mix properties (density, dry and wet ITSs) selected
for study were RAP type and binder type. Before investigating these two material types, the
best mixture property indicator to determine OTWC should be selected. Therefore, correlations
among density as well as dry ITS and wet ITSs were evaluated first, followed by a factorial study,
and the modelling of OTWC.
10
W. Ma et al.
Table 9. Experimental plan for determining effect of RAP type and binder
type on dry ITS.
No. of comparison
1
2
3
4
Downloaded by [UAE University] at 17:37 25 October 2017
1
2
3
4
5
6
Mixtures selected for comparing
Determination of effect of RAP type on dry ITS
Mix 2 vs. 3 at 2%FAC
Mix 2 vs. 3 at 3%FAC
Mix 5 vs. 6 at 2%FAC
Mix 5 vs. 6 at 3%FAC
Determination of effect of Binder type on dry ITS
Mix 1 vs. 4 at 2%FAC
Mix 1 vs. 4 at 3%FAC
Mix 2 vs. 5 at 2%FAC
Mix 2 vs. 5 at 3%FAC
Mix 3 vs. 6 at 2%FAC
Mix 3 vs. 6 at 3%FAC
4.2.2.1. Correlation between density and strengths Density, dry ITS, and wet ITS of six mixtures (nos. 1–6) were included in this analysis. Totally, 167 data points were used to correlate dry
ITS with density and 184 data points were used to correlate dry ITS with wet ITS. Based on the
level of correlation, dry ITS was found to correlate well with the two other performance results.
Therefore, it was used to determine and model OTWC. The results of correlations are discussed
later in this paper.
4.2.2.2. Factorial study Since dry ITS was selected to determine OTWC, factors affecting dry
ITS of cold recycled asphalt mixtures may also have effect on OTWC. Therefore, the effect of
the two material types (RAP type and binder type) on dry ITS was investigated. For simplicity,
additive (i.e. cement) was studied separately from these two material types. Cement was added
for all six mixtures at constant rate (1.0% by weight of dry RAP). Mixtures without cement (nos.
12 and 13) were studied later to evaluate the effect of cement on OTWC separately. The effects
were assessed by ANOVA for each pair of mixture combination. The TWC effect was included
in each ANOVA to compare with the other effects. Experimental plan for assessing RAP type
and binder type effects is shown in Table 9.
4.2.2.3. Regression analysis After the factorial study, the effects of factors affecting dry ITS
were known. Analyses results are discussed later in Tables 14 and 15. Both RAP type and binder
type were found to have a significant effect on dry ITS. Therefore, these two material types were
also included for modelling together with the TWC and FAC.
To construct a model for determining the OTWC, the measured OTWC was obtained first. The
measured OTWCs for six cold recycled foamed asphalt mixtures (nos. 1–6) were determined
using quadratic regression model (second-order polynomial), built by fitting dry ITS results at
different TWCs at each FAC. The measured OTWC corresponded to the maximum dry ITS on
the regression curve.
Then, the measured OTWC and three determining factors were fitted using a multiple linear
regression. In this model, binder type was treated as a categorical factor instead of a continuous factor to estimate the OTWC because these two binders differed in both grade and sources.
RAP type in the model was represented by the OWC, which was used as a continuous factor
since several RAP materials have different OWC levels in this study. Finally, the regression
Road Materials and Pavement Design
Table 10.
Mix no.
7
8
9
10
11a
12a
Downloaded by [UAE University] at 17:37 25 October 2017
a No
11
Experimental plan for validating using different mixtures.
Binder
RAP
FAC
58-34 (B)
58-34 (B)
67-22 (C)
67-22 (C)
58-34 (B)
58-34 (B)
VF (D)
VC (E)
VF (D)
VC (E)
DF (A)
DM (B)
2.5
2.5
2.5
2.5
3.0
3.0
cement was added in mix 11 and 12.
model was built by correlating one categorical factor (binder type) and two continuous factors (FAC and OWC) with the measured OTWC. Each binder type would have one regression
model.
4.3. Validation
The proposed model was validated using other six different mixtures. Also, dry ITS at predicted
OTWC based on the proposed model was compared to the dry ITS at other two OTWC levels
determined by typical methods.
4.3.1. Validating using different mixture types
Six additional foamed asphalt mixtures (nos. 7–12) were used to validate the two models.
Table 10 gives information of these mixtures, which are different from the previous mixtures
(nos. 1–6) in FAC, RAP type, and/or binder type. Also, mix 11 and 12 had no cement so that
effect of cement on OTWC can be evaluated. The mixtures were produced and tested for dry ITS
at different TWC levels. The measured OTWC of these mixtures corresponding to maximum dry
ITS was determined by quadratic regression.
4.3.2. Compare with typical OTWC determining method
The proposed OTWC determination method in this study needed to be compared with typical
methods to evaluate if it improves dry ITS. Mohammad et al. (2003) suggested using 100%
RAP OWC determined by the modified Proctor method (AASHTO 180) as OTWC. Wirtgen
GmbH (2008) suggested using the same procedure but a reduced RAP OWC as OTWC. Mixtures
tested at different TWC in this study allowed for the comparison of dry ITS at different water
contents, including proposed OTWC and other two OTWCs (100% RAP OWC by the modified
Proctor method and OTWC suggested by Wirtgen). Three mixtures were selected to compare the
proposed method with 100% RAP OWC by the modified Proctor method and six mixtures were
used for comparing with the Wirtgen method. The detailed mixtures’ information and OTWC
from each method is summarised in Table 18, together with discussions.
5. Results and discussion
The results from mechanism and modelling were analysed and discussed first, followed by the
validating results of proposed OTWC determining method.
12
W. Ma et al.
Table 11. Effect of water during mixing by ANOVA.
Effect on density
Factors
p-Value
Significance
Effect on dry ITS
p-Value
Significance
MWC
.016
Y
.144
N
TWC
Interaction
.034
.880
Y
N
.007
.987
Y
N
Table 12. Effect of water during compaction by ANOVA.
Downloaded by [UAE University] at 17:37 25 October 2017
Effect on density
Effect on dry ITS
Factors
p-Value
Significance
p-Value
Significance
TWC
< .001
Y
.479
N
FAC
.615
N
< .001
Y
Interaction
.303
N
.008
Y
5.1. Mechanism and modelling
5.1.1. Mechanism
5.1.1.1. Effect of water during mixing Table 11 summarises the effect of water on density, dry
and wet ITSs. MWC has a significant effect on density but is not on dry ITS. The interaction
of mixing water and total water has an insignificant effect on either density or dry ITS, which
means MWC and TWC have no combined effect. Effect on wet ITS was not analysed because
only three specimens in dry condition were prepared at each water content.
5.1.1.2. Effect of water during compaction TWC in Table 12 is the water content for compaction. As discussed in Section 4.2.1, this TWC composed of a fixed MWC and a varying
compaction water content. Densities at different TWC levels are significant different. TWC has
more influence than FAC and interaction on density, since p-value for FAC and interaction factor
are much higher than 0.05. However, dry ITS is not statistically different as compaction water
increases (or decrease). Meanwhile, FAC and its interaction with TWC significantly affect dry
ITS.
To separate water for mixing and compaction, water was added in two steps for the above analysis. However, for in-plant or in-place recycling in the field, water is added only once. Therefore,
the effect of water was further characterised using six cold recycled foamed asphalt mixtures.
Since water was added only once here, it facilitates both mixing and compaction process. The
effects can be evaluated through change in compacted density, dry ITS, and wet ITS.
5.1.1.3. Effect of TWC for each mixture Table 13 summarises effect of TWC on density and
strengths (dry and wet ITSs) for six mixtures. The effect of FAC was also evaluated to compare
with TWC factor. Interaction between two factors was included in analysis because their effects
may be dependent from each other. p-Value results from ANOVA are listed to determine the
significance each effect. p-Value less than .05 was considered significant, which means density
or strength is statistically different as the factor changes. Exact value of p-value below .001 was
not shown due to the decimal setting in Minitab software output.
Road Materials and Pavement Design
13
Table 13. Summary of water effect on density and strengths.
Effect on density
Effect on wet ITS
Factors
p-Value
2
TWC
FAC
Interaction
TWC
.270
.019
.156
.149
N
Y
N
N
.506
< .001
.153
.007
N
Y
N
Y
.134
.001
.576
< .001
N
Y
N
Y
3
FAC
Interaction
TWCb
.115
.152
.001
N
N
Y
< .001
.509
.194
Y
N
N
< .001
.184
.122
Y
N
N
4
FAC
Interaction
TWC
< .001
.552
.547
Y
N
N
< .001
.207
.048
Y
N
Y
< .001
.695
.997
Y
N
N
Mix no.
1
Downloaded by [UAE University] at 17:37 25 October 2017
Effect on dry ITS
Sig.a
5
p-Value
Sig.a
FAC
.680
N
.001
Y
.007
Y
Interaction
.394
N
< .001
Y
.106
N
TWC
< .001
Y
.405
N
.001
Y
FAC
< .001
Y
< .001
Y
< .001
Y
.018
Y
.492
N
.018
Y
Interaction
6
p-Value
Sig.a
TWC
< .001
Y
.001
Y
.041
Y
FAC
Interaction
< .001
.260
Y
N
< .001
.185
Y
N
< .001
.550
Y
N
a ‘Sig.’ stands for significance.
b p-Value of TWC effect on strengths
became close to .05 as interval doubled.
As shown in Table 13, TWC factor has significant effect on three mixtures’ densities (mix 3,
5, and 6), three mixtures’ dry ITS (mix 2, 4, and 6), and three mixtures’ wet ITS (mix 2, 5, and
6). Even though its effects on strengths (dry and wet) for mix 3 are not significant, as interval
between TWC levels doubled (from 25% to 50% of OWC), the effects became more significant
(p-value became .053 for dry ITS, and .060 for wet ITS). FAC has a significant effect on both
strengths of all the mixtures, as well as densities of four mixtures (mix 1, 3, 5, and 6). This effect
is more significant than the effect of TWC. The results in Table 13 indicate water has a significant
influence on half of these mixtures tested in terms of density, dry and wet ITSs. Therefore, TWC
needs to be selected carefully in mix design.
5.1.2. Modelling
5.1.2.1. Correlation between density and strengths Density and strengths are compared to
determined their relationships. If they have good correlations, one factor can be used to select
the OTWC without affecting other two factors. Dry ITS was compared to density and wet ITS to
determine correlations. Figures 5 and 6 show these two comparisons and their correlations were
determined based on linear regression. These two figures demonstrate positive correlations, indicating density and wet ITS increase as dry ITS increases. Dry ITS correlates better with wet ITS
than with density because of higher coefficient of determination (R2 ) of fitted model in Figure 6.
The data points in Figure 5 are more scattered and include several outliers, which resulted in
lower R2 value. The positive correlations indicate dry ITS can be used to select OTWC.
14
W. Ma et al.
5.1.2.3. Compare TWC with RAP type Two pairs of mixtures were used for comparison. With
same binder type and FAC, ANOVA was performed to determine the effects of TWC and RAP
type. Each pair of mixtures contains two RAP types – DM (medium gradation) and DC (coarse
gradation). Mixtures with DF RAP (fine gradation) was not included because the tested TWCs
were lower than mixtures with other two RAP types. Table 14 summarises the ANOVA results of
four comparisons. RAP type has significant effect on dry ITS in the first and third comparisons.
These two comparisons used lower FAC (2%). Other two comparisons at higher FAC (3%)
indicate RAP type is not significant. During modelling, RAP type was represented by OWC of
2100.0
Density (kg/m3)
2050.0
2000.0
1950.0
y = 0.4702x + 1898.9
R² = 0.16
1900.0
1850.0
0.0
50.0
100.0
150.0
200.0
250.0
300.0
350.0
Dry ITS (kPa)
Figure 5.
Correlation between dry ITS and density.
350.0
300.0
250.0
Wet ITS (kPa)
Downloaded by [UAE University] at 17:37 25 October 2017
5.1.2.2. Factorial study Since dry ITS was selected to determine OTWC of cold recycled
foamed asphalt mixtures, material types related to dry ITS need to be evaluated. The OTWC is
corresponding to the maximum dry ITS. Therefore, these material types may also affect OTWC.
Effect of TWC on dry ITS has been evaluated previously. Besides TWC, the RAP type and binder
type may also affect dry ITS, which influences the OTWC. Therefore, their effects on dry ITS
were evaluated and compared with TWC.
200.0
150.0
100.0
y = 0.7549x + 17.348
R² = 0.6405
50.0
0.0
0.0
50.0
100.0
150.0
200.0
Dry ITS (kPa)
Figure 6.
Correlation between dry ITS and wet ITS.
250.0
300.0
350.0
Road Materials and Pavement Design
15
Table 14. Summary of ANOVA results for TWC and RAP type on dry ITS.
Effect on dry ITS
Downloaded by [UAE University] at 17:37 25 October 2017
No. of comparison
Pair of mixes for comparing
Factors
p-Value
Significance
1
Mix 2 vs. 3 at 2%FAC
2
Mix 2 vs. 3 at 3%FAC
3
Mix 5 vs. 6 at 2%FAC
4
Mix 5 vs. 6 at 3%FAC
TWC
RAP
Interaction
TWC
RAP
Interaction
TWC
RAP
Interaction
TWC
RAP
Interaction
.256
< .001
.117
.319
.240
.437
.244
.027
.564
.789
.620
.835
N
Y
N
N
N
N
N
Y
N
N
N
N
RAP since it can distinguish RAP with different gradations. Compared to RAP type, TWC effect
is not significant to dry ITS since none of the comparisons shows significance of TWC and the
interaction. However, the analysis was based on limited types of mixtures with only two RAP
types (DM and DC) and two TWC levels (100% and 125% OWC) available for comparisons.
5.1.2.4. Compare TWC with binder type Three pairs of mixtures were evaluated at two FAC
levels. Each pair of mixtures has same RAP and FAC level. Binder types used for comparison
are 67-22(A) and 58-34(B). Two TWC levels (50% and 75% OWC) were used in comparison 1
and 2. Three TWC levels (75%, 100%, and 125% OWC) were used in comparison 3 and 4. And
three TWC levels (100%, 125, and 150% OWC) are used in comparison 5 and 6. ANOVA was
performed in each comparison to evaluate the effects of TWC and binder type and the results are
summarised in Table 15. Binder type has a significant effect on dry ITS in four comparisons and
TWC effect is significant in five comparisons. Binder type effect is not significant in comparison
4 and 6 where FAC is higher (3%). Therefore, TWC effect on dry ITS is as significant as binder
type. Mixtures with higher FAC (3%) tend to have same dry ITS regardless of binder type.
5.1.2.5. Determining OTWC As discussed before, maximum dry ITS is corresponding to
OTWC. Six foamed asphalt mixtures with combinations of three RAP types and two binder
types were tested for dry ITS. For each mixture, OTWC determined based on quadratic regression is termed as measured OTWC. Figure 7 shows an example of the regression relationship
between dry ITS and TWCs. The measured OTWC was determined at the peak ITS.
Then, the measured OTWC and three factors (binder type, RAP type, and FAC) were used to
build a mathematical model by multiple linear regression. The fitted model was used to calculate
the OTWC and compare with the measured OTWC. The results are summarised in Table 16.
Data in columns 4, 5, and 6 of Table 16 were used to build regression models. Two models built
are shown as Equations 1 and 2. Their adjusted R2 are 68.2% and 89.8%, indicating both models
have good fitting. Interaction term ‘OWC × FAC’ was included in model 2 because its effect is
significant when 58-34 binder was used. For all six mixtures, the measured OTWC ranges from
2.07 to 4.51 and calculated OTWC is found between 2.07 and 3.96, which is a little lower than
the measured range.
16
W. Ma et al.
Table 15. Summary of ANOVA for TWC and binder type.
Effect on dry ITS
No. of comparison
Mixes for comparing
1
Downloaded by [UAE University] at 17:37 25 October 2017
2
Factors
p-Value
Mix 1 vs. 4 at 2%FAC
TWC
.011
Y
Mix 1 vs. 4 at 3%FAC
Binder
Interaction
TWC
.009
.145
.004
Y
N
Y
Binder
Interaction
TWC
Binder
Interaction
TWC
Binder
Interaction
TWC
< .001
.238
.168
.001
.970
.047
.856
.165
.018
Y
N
N
Y
N
Y
N
N
Y
Binder
Interaction
TWC
Binder
Interaction
.002
.447
.047
.055
.037
Y
N
Y
N
Y
3
Mix 2 vs. 5 at 2%FAC
4
Mix 2 vs. 5 at 3%FAC
5
Mix 3 vs. 6 at 2%FAC
6
Mix 3 vs. 6 at 3%FAC
Significance
For mixtures with 67-22 (A):
OTWC = 0.143 × FAC − 1.383 × OWC + 7.68.
(1)
For mixtures with 58-34 (B):
OTWC = 3.520 × FAC + 2.750 × OWC − 1.260 × OWC × FAC − 4.37.
Figure 7.
Regression curve built to determine the measured OTWC (mix 6).
(2)
Road Materials and Pavement Design
17
Table 16. Summary of OTWC for each mixture.
Mix no.
1
2
3
4
5
RAP
OWC (%)
FAC (%)
Measured
OTWC (%)
Calculated
OTWC (%)
67-22 (A)
67-22 (A)
67-22 (A)
67-22 (A)
67-22 (A)
67-22 (A)
58-34 (B)
58-34 (B)
58-34 (B)
58-34 (B)
58-34 (B)
58-34 (B)
DF (A)
DF (A)
DM (B)
DM (B)
DC (C)
DC (C)
DF (A)
DF (A)
DM (B)
DM (B)
DC (C)
DC (C)
4
4
3
3
3
3
4
4
3
3
3
3
2
3
2
3
2
3
2
3
2
3
2
3
2.76
2.26
3.69
4.51
3.63
3.74
3.59
2.07
3.21
3.19
3.51
3.01
2.43
2.58
3.82
3.96
3.82
3.96
3.59
2.07
3.36
3.10
3.36
3.10
Correlation between measured and calculated OTWC was illustrated in Figure 8. The data
points are located near to the line of equality and a linear regression fits these data with R2 as
high as 86.7%. However, the capability of these two models built based on two binder types and
three RAP types to calculate OTWC of foamed asphalt mixtures needs further validation using
more types of mixture.
5.2. Validation
5.2.1. Validating with different mixture types
Table 17 lists information of the mixtures (nos. 7–12) used to validate the two models proposed
in this study. Models (Equations (1) and (2)) were applied to predict OTWC of each mixture.
Results of measured and predicted OTWC, as well as their differences are summarised in right
three columns of Table 17. Measured OTWC ranged between 2.74% and 5.01% while predicted
OTWC ranged between 2.07% and 3.89%. The two models under-predicted OTWC for these
mixtures but the differences are below 0.9% except for the two mixtures with binder 67-22(B).
This binder has the same PG as 67-22(A) but a different source. The foaming properties such as
5
Calculated OTWC (%)
Downloaded by [UAE University] at 17:37 25 October 2017
6
Binder
4
3
y = 0.8659x + 0.4355
R² = 0.8666
2
1
1
1.5
2
2.5
3
3.5
Measured OTWC (%)
Figure 8.
Comparison between the measured and calculated OTWC.
4
4.5
5
18
W. Ma et al.
Table 17. Summary of model validation using different mixtures.
Mix no.
7
8
9
10
11a
12a
RAP
FAC
Measured
OTWC (%)
Predicted
OTWC (%)
Difference
(%)
58-34 (B)
58-34 (B)
67-22 (C)
67-22 (C)
58-34 (B)
58-34 (B)
VF (D)
VC (E)
VF (D)
VC (E)
DFa (A)
DMa (B)
2.5
2.5
2.5
2.5
3.0
3.0
3.33
3.43
4.28
5.01
2.74
3.98
2.83
3.23
2.51
3.89
2.07
3.10
0.50
0.20
1.77
1.12
0.67
0.88
cement was added in mix 11 and 12.
ER and HL of this binder were also all less than 67-22(A). Therefore, this foamed asphalt may
need more water to distribute uniformly in the mixture. The model should be able to determine
OTWC with or without cement since removing cement does not significantly affect the prediction
accuracy for mix 11 and 12.
Additionally, the measured and predicted OTWC were plotted in Figure 9 for correlation.
These data points are basically parallel with and close to line of equality, except for one data
point representing OTWC of mix 9. The linear relationship between the two OTWC values has
an R2 equals 53.3%, which could be improved to 85.3% excluding mix 9. This indicates that
these two models may become more accurate if binder source is consistent.
5.2.2. Compare with typical OTWC determining methods
Table 18 summarises the information of mixtures available for comparison. OTWC1 indicates
100% RAP OWC by the modified Proctor method or the Wirtgen’s recommended OTWC.
OTWC2 represents the proposed OTWC determined from two models built. OTWC1* and
OTWC2* are the actual water contents used to find the peak dry ITS, which are the nearest to
OTWC1 and OTWC2 (within ± 0.5%), respectively. For comparison with 100% RAP OWC by
the modified Proctor method, eight mixtures had OTWC1 higher than the tested range of water
content in this study. Also, other two mixtures without cement were not included because the
models still need further calibration for no-cement mixtures. Therefore, only three mixtures are
6
5
Predicted OTWC (%)
Downloaded by [UAE University] at 17:37 25 October 2017
a No
Binder
4
3
y = 0.5718x + 0.7673
R² = 0.5332
2
Mix 9
1
1
2
3
4
Measured OTWC (%)
Figure 9.
Correlation between measured and model predicted OTWC.
5
6
Road Materials and Pavement Design
19
Table 18. Comparing proposed OTWC with 100% RAP OWC by the modified Proctor method and
Wirtgen’s method.
Mix no.
5
10
13
Downloaded by [UAE University] at 17:37 25 October 2017
2
2
8
9
10
13
Binder
RAP
FAC
OTWC1
OTWC2
OTWC1*
Compare proposed method with 100% RAP OWC by the modified Proctor method
58-34(B)
DM
3.0
3.8
3.1
3.8
67-22(C)
VC
2.5
5.3
3.9
4.9
67-22(C)
DF
2.5
6.5
2.5
7.0
Compare proposed method with Wirtgen (2008)
67-22(A)
DM
2.0
3.3
3.8
3.0
67-22(A)
DM
3.0
3.3
4.0
3.0
58-34(B)
VC
2.5
4.3
3.2
4.3
67-22(C)
VF
2.5
5.2
2.5
5.7
67-22(C)
VC
2.5
4.3
3.9
4.1
67-22(C)
DF
2.5
5.2
2.5
5.0
OTWC2*
3.0
4.1
2.7
3.8
3.8
3.5
2.7
4.1
2.7
available for comparison. For comparison with Wirtgen’s OTWCs, six mixtures with OTWCs in
the tested range were used.
Dry ITS of mixtures at proposed OTWC and at other two recommended OTWCs were compared in Figures 10 and 11. The percent changes in dry ITS are summarised at the top of columns.
Most mixtures show increases in dry ITS at the proposed OTWC except for mix 10. However,
statistical analysis did not show significantly difference between dry ITS at the proposed OTWC
and dry ITS at other two OTWCs. If the measured OTWC is used instead of the proposed OTWC,
dry ITS of mix 10 could be higher (about 258.3 kPa), which would be equal to that at 100% RAP
OWC by the modified Proctor method and higher than that at Wirtgen’s OTWC. Therefore, dry
ITS’ decrease of mix 10 shown in Figures 10 may be due to limitations of models’ predicting capability. Further calibration using more binder types and RAP types is needed to make
prediction more accurate.
Figure 10.
Method.
Comparing dry ITS at proposed OTWC and 100% RAP OWC by the modified Proctor
Downloaded by [UAE University] at 17:37 25 October 2017
20
W. Ma et al.
Figure 11.
Comparing dry ITS at proposed OTWC and Wirtgen’s OTWC.
6. Conclusions and recommendations
According to the results and analysis presented in this paper, the following conclusions and
recommendations are offered:
• A new compaction method was proposed to determine OWC of RAP using 30 gyrations
in a SGC. Compared to the modified proctor test method (AASHTO T180), the OWC
determined by proposed method is lower. The maximum dry densities determined by two
alternative methods are similar while those determined by proposed method are slightly
lower.
• The effects of water during mixing and compaction were investigated, respectively. The
result showed that mixing water has a significant effect on density but this effect is not
significant on dry ITS. When the MWC is the same, the compaction water has a significant
effect on compacted density and the effect is more significant than the effect of FAC. However, the effect of compaction water on dry ITS is not significant. Although this evaluation
only used one mixture, the method separating the two effects of water would be helpful to
future research.
• Evaluation of the effect of combination of water (including both mixing and compaction
effects) showed the TWC has a significant effect on density, dry and wet ITSs for three out
of six mixtures. Although this effect is not as strong as the effect of FAC, the OTWC needs
to be carefully determined for the mixture.
• Besides TWC and FAC, the material types (RAP type and binder type) were also found
to be significant factors to dry ITS. However, their levels of significance were found to
be dependent on the FAC in the mixture. For mixtures with higher FAC, these RAP type
and binder type were less significant. Both factors were considered in modelling OTWC.
Binder type was considered as a categorical factor and RAP type was represented by the
OWC as a continuous factor.
• Two models for determining OTWC using different binder types were built by linear
regression analysis. The relationship between OTWC and two determining factors (OWC
and FAC) was established. Prediction capability of the models were validated by six different mixtures and good correlation between the predicted and the measured OTWC was
found, although two of six mixtures had lower predicted OTWC due to different binder
source. Further comparisons were conducted to validate the improvement in mixtures’
Road Materials and Pavement Design
21
performance (dry ITS) using the predicted OTWC by the proposed method. Dry ITS at
two different OTWCs (100% RAP OWC and Wirtgen’s recommended OTWC) based on
the modified proctor test were used for comparison. Dry ITS results were found to increase
up to 22% at proposed OTWC. However, one mixture showed a decrease in dry ITS in both
comparisons, which may be due to change of binder source. This indicates the developed
model was not able to provide an accurate prediction for binders with same PG but different
sources. Therefore, further calibration using more binder sources is required.
Acknowledgements
Downloaded by [UAE University] at 17:37 25 October 2017
The authors thank Federal Highway Administration for supporting this research. The authors also acknowledge Buzz Powell, Jason Moore, Adam Taylor, and Jason Nelson from NCAT for their advice and assistance
during this research.
Disclosure statement
No potential conflict of interest was reported by the authors.
References
Asphalt Academy. (2009). Technical guideline: Bitumen stabilised materials (2nd ed.). Pretoria: Asphalt
Academy.
Asphalt Recycling & Reclaiming Association (ARRA). (2015). Recommended mix design guidelines for
cold recycling using bituminous recycling agents. Annapolis, MD: Author.
Chan, P., Tighe, S., & Chan, S. (2010). Exploring sustainable pavement rehabilitation: Cold in-place
recycling with expanded asphalt mix. 89th Annual Meeting Compendium of Papers, Transportation
Research Board.
Cross, S. (2003). Determination of Superpave® gyratory compactor design compactive effort for cold inplace recycled mixtures. Transportation Research Record: Journal of the Transportation Research
Board, 1819, 152–160.
Csanyi, L. H. (1957). Foamed asphalt in bituminous paving mixtures (Highway Research Board Bulletin
160). Washington, DC: Highway Research Board.
Diefenderfer, B. K., Bowers, B. F., Schwartz, C. W., Farzaneh, A., & Zhang, Z. (2016). Dynamic modulus of
recycled pavement mixtures. Transportation Research Record: Journal of the Transportation Research
Board, 2575, 19–26.
Jenkins, K. J. (2000). Mix design considerations for cold and half-warm bituminous mixes with emphasis
of foamed bitumen (PhD dissertation). University of Stellenbosch.
Mohammad, L. N., Abu-Farsakh, M. Y., Wu, Z., & Abadie, C. (2003). Louisiana experience with
foamed recycled asphalt pavement base materials. Transportation Research Record: Journal of the
Transportation Research Board, 1832, 17–24.
Muthen, K. M. (1998). Foamed asphalt mixtures: Mix design procedure (Contract report CR-98/077).
Pretoria: CSIR TRANSPORTEK.
Newcomb, D. E., Arambula, E., Yin, F., Zhang, J., Bhasin, A., Li, W., & Arega, Z. (2015). Properties of foamed asphalt for warm mix asphalt applications (NCHRP report 807). Washington, DC:
Transportation Research Board.
Sakr, H. A., & Manke, P. G. (1985). Innovations in Oklahoma foamix design procedures. Transportation
Research Record: Journal of the Transportation Research Board, 1034, 26–34.
Stroup-Gardiner, M. (2011). Recycling and reclamation of asphalt pavements using in-place methods
(NCHRP synthesis 421). Washington, DC: Transportation Research Board.
Thenoux, G., González, Á, & Dowling, R. (2007). Energy consumption comparison for different asphalt
pavements rehabilitation techniques used in Chile. Resources, Conservation and Recycling, 49(4),
325–339. doi:10.1016/j.resconrec.2006.02.005
Wirtgen GmbH. (2008). Suitability test procedures of foam bitumen using Wirtgen WLB 10S. Windhagen:
Wirtgen GmbH.
Wirtgen GmbH. (2012). Wirtgen cold recycling technology. Windhagen: Wirtgen GmbH.
Документ
Категория
Без категории
Просмотров
0
Размер файла
1 161 Кб
Теги
2017, 14680629, 1389088
1/--страниц
Пожаловаться на содержимое документа