Professional Documents
Culture Documents
h i g h l i g h t s
A robust methodology was developed to evaluate the storage stability of CRMB binders.
The instability mechanism was explored by the concept of dynamic asymmetry.
It is possible to manipulate raw material properties and interaction conditions to achieve a storage-stable CRMB blend.
a r t i c l e i n f o a b s t r a c t
Article history: One of the main drawbacks of crumb rubber modified bitumen (CRMB) is the storage stability issue. The
Received 9 December 2019 storage instability of CRMB impedes its further application. This study aims to develop a robust method-
Received in revised form 3 March 2020 ology to evaluate the storage stability of CRMB binders using both mechanical and morphological tests.
Accepted 20 March 2020
The effects of rubber contents (0, 5%, 10%, 15%, 22% by weight of bitumen) and different non-foaming
Available online 26 March 2020
warm-mix additives (wax-based and chemical-based additives) on the storage stability of CRMB were
investigated. Laboratory tests were also performed on the constituents of CRMB to have a deep under-
Keywords:
standing of the mechanism of storage instability. Standard tube separation tests were conducted on dif-
Crumb rubber modified bitumen
Storage stability
ferent binders. Both rheological tests and X-ray computed tomography (CT) scan tests were performed on
Warm mix asphalt the binder samples collected from different parts of the tube test. Separation indices were developed
Multiple stress creep recovery based on the difference in mechanical property and rubber content from the tube samples respectively.
Frequency sweep Results show that CRMB with a higher rubber content is more storage stable than that with a lower rub-
x-ray CT scan ber content. The addition of warm-mix additives is detrimental to the storage stability of the studied
Dynamic asymmetry CRMB. Rheological tests were performed on the individual constituents of CRMB (i.e., bitumen phase
and rubber phase) to understand better the dynamic asymmetry potentially existing within the unstable
CRMB binder. Results show that the residual bitumen becomes stiffer while the swollen rubber becomes
softer after interaction because of the preferential absorption of light components of bitumen by rubber.
The dynamic asymmetry existing between the bitumen phase and the rubber phase of CRMB results in
storage instability. When the bitumen phase has similar dynamic properties as the rubber phase, the
resulted binder system will be stable. It is possible to manipulate raw material properties and interaction
conditions to achieve the desired crossover between two phases of CRMB and hence obtain a storage-
stable CRMB blend.
Ó 2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
1. Introduction ing material [1,2]. The typical wet process used for bitumen mod-
ification is achieved by mixing bitumen and CRMs at 160–220 °C
How to appropriately treat and dispose end-of-life tires (ELTs) for a predetermined duration [3]. During this process, rubber par-
has been a concern for many years. A possible solution in civil engi- ticles are swollen to due to absorption of light fractions of bitumen.
neering is to incorporate crumb rubber modifiers (CRMs) produced Under excessively high mixing temperatures, rubber swelling will
from ELTs into bitumen and use the modified binder as a new pav- be finished in a short time and rubber degradation starts to occur
[4]. The interaction between bitumen and rubber has great effects
on the property development of crumb rubber modified bitumen
⇑ Corresponding author. (CRMB). Many factors, including the raw material properties and
E-mail address: haopeng.wang@tudelft.nl (H. Wang).
https://doi.org/10.1016/j.conbuildmat.2020.118840
0950-0618/Ó 2020 The Author(s). Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
2 H. Wang et al. / Construction and Building Materials 249 (2020) 118840
interaction conditions, have been shown to influence the bitumen- [18]. Optical observations are usually conducted using a fluores-
rubber interaction and hence the final binder properties [5–8]. cence microscope (FM). Images obtained from FM allows direct
Overall performance improvement and environmental benefits of and rapid visualization of the relative amount and morphology of
rubberized asphalt pavements using CRMB have been reported both polymer-rich phase and bitumen-rich phase based on their
for a long time [9]. However, there are two obvious drawbacks different UV excitation responses [10]. While FM is very popular
which impede its further applications. Firstly, for traditional rub- and successful in characterizing most PMB binders (colloidal dis-
berized asphalt mixtures, their mixing and compaction tempera- persion system), it is not very suitable for CRMB binders (colloidal
tures are usually higher than unmodified asphalt mixtures due to suspension system). This is because prepared samples subjected to
the excessively high viscosity of CRMB, which increases the energy FM observations usually require a thickness less than 0.1 mm.
cost. To solve this issue, warm mix asphalt (WMA) technologies are However, the size of CRM particles within the bitumen matrix is
innovated to effectively reduce construction temperatures as well usually bigger than the required sample thickness. Even though
as the emissions of volatile organic compounds (VOCs), polycyclic other sample preparation methods (e.g., ‘‘cooling and fracture”
aromatic hydrocarbons (PAHs) and total suspended particles (TSP), per EN 13632) may overcome this problem [10], a large proportion
etc. [2]. Another obvious drawback of CRMB which hinders its fur- of carbon black inside CRM impedes the fluorescent effect of rub-
ther application is the poor stability during storage at elevated ber which makes it difficult to obtain readable morphology. Gel
temperatures (140–180 °C). Storage instability is a prevalent issue permeation chromatography (GPC) has been adopted to obtain
existing in polymer modified bitumen (PMB) but more severe in the information of molecular weight distributions of bitumen-
CRMB. Since WMA has been widely utilized to improve the mix- rich phase and polymer-rich phase [19,20]. Through comparing
ture workability of rubberized asphalt pavements, its influence different phases at a molecular level, the molecular sizes of poly-
on the storage stability of CRMB is of great concern and has not mer and bitumen, which is an indicator of compatibility, can be
yet been fully investigated. predicted and hence the storage stability. Fourier transform infra-
Storage instability of PMB is actually a phase separation process red spectroscopy (FTIR) can also be utilized to determine changes
at a microscopic scale during which polymer modifiers separate in the polymer content by measuring the change of peaks of the
from bitumen matrix due to the different natures of polymer and corresponding functional groups [21]. However, inconsistent
bitumen [10]. If considering PMB as a binary blend which com- results from FTIR are often reported [22]. The complexity of equip-
prises a bitumen-rich phase and a polymer-rich phase, the ment, strictness of sampling procedure and difficulty in quantify-
dynamic asymmetry between these two phases directly drives ing results of these techniques hinder their widespread use.
the phase separation process. There are many factors controlling The most common way to evaluate the stability of modified bin-
this dynamic asymmetry within PMB. The density difference ders is the cigar tube test which simulates the phase separation
between polymer and bitumen is believed to be one of the impor- process in the field in the absence of continuous agitation. The
tant causes of storage instability in three-dimensional reality [11]. mechanical properties of samples from the different tube sections
More fundamental effects may come from the differences in after conditioning are measured and compared to give an
microstructural composition, physio-chemistry (molecular weight, indication of the storage stability. The actual performance of mod-
glass transition temperature, polarity and solubility) for polymer ified binders after storage or the effect of storage conditions on
and bitumen which contributes to the incompatibility [12]. Solu- their properties can be directly measured through mechanical test-
bility parameters are often used to interpret and predict the com- ing. In addition, since the densities of bitumen and rubber are dif-
patibility between bitumen and polymer modifiers [13]. In terms ferent, it is expected to distinguish these two phases by X-ray
of CRMB, more specifically, raw material properties (bitumen and computed tomography (CT) [23]. The obtained X-ray CT scan
CRM characteristics, CRM content, additives) and the interaction images can reflect the bitumen and rubber phases with different
conditions determine the storage stability. It is reported that CRMs greyscale levels, which can further give insights into storage stabil-
of small size obtained through ambient grounding result in more ity based on the distribution of rubber particles in different
storage stable CRMB binders [14]. Poor storage stability has nega- positions. This study adopted both mechanical testing method
tive influence on the overall binder properties and increases the and X-ray CT visualizing technique to evaluate the storage stability
handling difficulty of storage and transport, consequently influ- of CRMB binders.
ence the durability of asphalt pavements. It also results in difficul-
ties for the binder system to be transported by pipelines [14]. Since
storage stability directly influence the usage and performance of 3. Objectives and approach
CRMB, understanding the mechanism and influence of other fac-
tors (e.g., rubber content and warm-mix additives) is thus of great The present study aims to develop a robust methodology to
importance for both production optimization and quality control. evaluate the storage stability of CRMB with warm-mix additives
Besides controlling the preparation conditions of CRMB, there using both mechanical and morphological tests. Specific objectives
are generally two types of methods to improve storage stability include:
of CRMB. One way is to add various types of chemical compounds
into CRMB to form bonds between different components of binders investigating the effects of rubber content and warm-mix addi-
and/or form a polymer network structure [14]. The other way is tives on storage stability of CRMB binders. Both mechanical
achieved by surface treatment of CRM. Thermochemical, thermo- tests (dynamic shear rheometer, DSR) and morphological tests
mechanical, microwave, ultrasonic, biological, and plasma meth- (X-ray CT scan) were conducted on the samples from different
ods were used to activate the CRM [15,16]. Furaldehyde or acrylic parts of the tube after conditioning; and
acid was also reported to activate the surface of CRM [17]. having a preliminary understanding of the dynamic asymmetry
between the bitumen-rich phase (BP) and the rubber-rich phase
(RP). The glass transition temperature difference between bitu-
2. Methods for characterizing storage stability of CRMB men and CRM was analyzed by Differential Scanning Calorime-
try (DSC) test. Rheological tests were also carried out on the
Generally, three main techniques have been applied to charac- extracted liquid phase (i.e., BP) of CRMB and the swollen rubber
terize phase separation behaviours in bituminous binders: micro- phase (i.e., RP) to gain some insights into the dynamic
scopy, chromatography/spectroscopy and mechanical testing asymmetry.
H. Wang et al. / Construction and Building Materials 249 (2020) 118840 3
nously. The binder samples from different parts of the tube were
4. Materials and methods
subjected to both rheological tests by DSR and X-ray CT scan tests.
4.1. Materials
4.4. Extraction of the liquid phase of CRMB
Bitumen of Pen 70/100 was used as the base bitumen for mod-
ification. Its SARA fractions are respectively 7%, 51%, 22%, and 20%. During the interaction process, the swollen rubber phase (RP)
The technical information about CRM provided by a local company and the residual bitumen phase (BP) were formed, which are dif-
is shown in Table 1. ferent from the original properties of rubber particle and base bitu-
The morphological and elemental analysis of CRM were done by men. Therefore, it is necessary to investigate the rheological
Environment Scanning Electron Microscopy (ESEM) with an energy properties of both RP and BP to have a better understanding of
dispersive X-ray spectroscopy (EDX) in Fig. 1 [8]. It can be found the dynamic asymmetry between them. The BP of CRMB was
rubber particles have an irregular shape and multiaperture struc- extracted by filtering out the insoluble rubber particles from the
ture. They also contain different types of chemical compounds bitumen matrix through a mesh sieve (0.063 mm) in the oven at
(e.g., oxide, zinc, iron and silicon, etc.). Non-foaming warm-mix 165 ± 3 °C for 20 min [24]. The residual (drained) binder labelled
additives, including wax-based product (designated as W, syn- as CRMB-X-BP was collected on an aluminium pan for future
thetic microcrystalline wax) and chemical-based product (desig- testing.
nated as C, liquid chemical cocktail including surfactants, anti-
stripping agents, polymers, etc.), were added to CRMB binders. 4.5. Preparation of cylindrical swollen rubber sample
4.2. Binder sample preparation Since the remaining CRM particles on the sieve mesh from the
extraction test were loose and inhomogeneous, it is difficult to pre-
CRMB binders with four CRM contents (5%, 10%, 15% and 22% by pare a disk sample to perform the DSR test. To simulate the prop-
weight of base bitumen) were laboratory prepared at 180 ± 2 °C erty of the swollen rubber, cylindrical samples of rubber were cut
and designated as CRMB-5, CRMB-10, CRMB-15 and CRMB-22. from waste truck tire treads and were immersed in hot bitumen.
Detailed preparation procedure can be found in [8]. Two warm- The specially designed sample preparation process is shown in
mix additives were manually mixed with CRMB at 160 °C for Fig. 2 [25]. The dimension of the rubber sample is about 2 mm in
10 min. It is noteworthy that additives were added after the prepa- thickness and 8 mm in diameter, which is suitable to carry out
ration of CRMB and were not participating in the bitumen-rubber the DSR tests with the traditional configuration for bitumen test-
interaction process. Warm-mix additives were only added to CRMB ing. Laboratory swelling test of rubber in bitumen was carried
with a high rubber content in an attempt to reduce its excessively out by immersing the rubber cylinder into hot bitumen at 180 °C
high viscosity. The resultant binders were designated as CRMB-22- for 36 h for it to reach the swelling equilibrium [26]. The swelling
W and CRMB-22-C. The dosages of additives W and C were respec- equilibrium was defined as the point where the volume of rubber
tively 2.0% and 0.6% based on the recommendations from the tended to stabilize. The swollen rubber samples were cleaned by
respective manufacturer. wiping out the surface bitumen with solvents [26], which can sim-
ulate the properties of swollen CRM particles inside the bitumen
4.3. Storage stability test matrix. Through the mixing optimization process of CRMB, the
rubber particles inside the bitumen matrix were assumed to reach
The well-known ‘‘tube test” was employed as the storage stabil- the swelling equilibrium [8]. The swollen samples were trimmed
ity test. During the test, an aluminium foil tube with standard to 8 mm in diameter with a sharp tool to fit the DSR test. From
dimensions was filled with hot binder carefully to avoid bringing the viewpoint of material property, the mechanical properties of
in air bubbles. The tubes covered with a lid were vertically placed swollen rubber should not be influenced by the testing sample
in an oven at 163 ± 3 °C for 48 ± 1 h. After conditioning, settling geometry in principle. Therefore, an intact swollen rubber sample
tubes were moved to a freezer for 2 h. After cooling, the tubes were was used to represent the loose swollen rubber particles to per-
equally cut into three parts and samples from each part were col- form DSR tests to obtain the mechanical properties.
lected by removing the metal and stored after stirring homoge-
4.6. Experimental design and methods
Table 1
A detailed testing program of different binders to investigate
Basic properties and particle size distribution of CRM.
storage stability is shown in Fig. 3. There are two testing phases
Properties Description for the experimental design. At Phase 1, six types of CRMB binder
or value
samples at the fresh state were subjected to the tube storage sta-
Source Waste truck tyres bility tests. Multiple stress creep and recovery (MSCR) tests were
Processing method Ambient grounding
conducted at 64 °C on samples collected from different parts of
Morphology Porous
Density (g/cm3) 1.15 the tube after conditioning. The high-temperature performance
grade (H-PG) of the neat bitumen is 64 °C and all the H-PGs of
Chemical composition Rubber component 55 all CRMB binders are higher than 64 °C [8]. The choice of this test-
(natural and synthetic)
ing temperature for MSCR was expected to differentiate the
Carbon Black (%) 25
Processing agents (%) 20 mechanical properties of all collected samples. Storage stability
index was proposed based on the MSCR parameters. X-ray CT scan
Gradation Sieves (mm) Passing (%)
0.710 100
tests were also conducted on the collected samples from different
0.500 93 positions of the tube to gain more insights into the rubber particle
0.355 63 distribution in the bitumen matrix. At Phase 2, frequency sweep
0.180 21 tests were conducted on the liquid phase of CRMB and swollen
0.125 9
rubber samples, which represent the bitumen phase and rubber
0.063 2
phase respectively. The linear viscoelastic behaviours of the bitu-
4 H. Wang et al. / Construction and Building Materials 249 (2020) 118840
Table 2
Rubber contents of binder and separation index based on CT scan test results.
men phase and rubber phase were compared in an attempt to link 0.1 to 100 rad/s over a wider temperature range of 10–130 °C
to the phase separation phenomena. with an incremental step of 20 °C. The measurements were also
carried out under strain-controlled mode but at a higher strain
4.6.1. Frequency sweep test level of 1%, which is within the linear viscoelastic range of rubber
Frequency sweep (FS) tests of binder samples were performed as determined by strain amplitude sweep tests [25]. Before placing
over a frequency range of 0.1–100 rad/s at temperatures of 10, the rubber sample between the parallel plates of the DSR, a two-
30, 50 and 70 °C. A strain level of 0.1% under strain-controlled component adhesive (Evonik Rohm GmbH, Darmstadt, Germany),
mode was used for the measurement to ensure the linear vis- Plex 7742 and liquid Plexmon 801, was mixed and applied on
coelastic response. FS tests of rubber samples were performed from the surface of the bottom plate and the top surface of the rubber
H. Wang et al. / Construction and Building Materials 249 (2020) 118840 5
Fig. 4. X-ray CT scan test for binder samples after the storage stability test.
6 H. Wang et al. / Construction and Building Materials 249 (2020) 118840
2 r 2 Dq g
v¼ ð2Þ
9 g
Fig. 7. Schematic representation of CRM particles suspended in bitumen matrix. (Note: Different colours represent different states of CRM particle and bitumen; The lines
connecting rubber particles were imagined for illustration purpose).
segmentation, the measurement function was applied to calculate ber in the binder from different parts and the resulted separation
the area occupied by rubber particles. The rubber content was fur- indices. It can be seen clearly that the rubber contents in different
ther calculated by dividing the total area as displayed by the red parts of the tube are different due to the inhomogeneity of CRMB
circle. The rubber content in different parts of the tube sample binders. Because the density of rubber particles is larger than bitu-
was determined by averaging the results from 200 scan images men, rubber particles settle down in the tube during the storage
through batch processing using the above procedure. The calcu- stability test. That is why the rubber content in the bottom is
lated rubber content varied from different scan images and fluctu- higher. The varying degree of rubber content varies with the type
ated around the averaged value. The obtained rubber content from of binder. The average volume fraction of rubber for each binder
2D images maybe deviate from the reality but it is reasonable to is approximated twice as high as the original dosage. This confirms
compare the values from different samples since a consistent pro- the occurrence of rubber particle swelling in the bitumen matrix
cedure was used. [31]. From the SICT in Table 2, it gives similar findings as the previ-
Another separation index SICT based on the CT scan results is ous SI based on the mechanical properties (J nr ). The SICT value
defined in Equation (3). decreases with the increase of rubber content, indicating a more
storage stable binder. The addition of warm-mix additives has
j/t /b j adverse effects on the storage stability of CRMB binder as reflected
SICT ¼ ð3Þ
/avg by the increased SICT value.
Fig. 9 compares the separation index from the mechanical test
where /t and /b represent the rubber volume contents of the sam- and the CT scan test. There is a high correlation between these
ple taken from the top and bottom parts, respectively. /avg is the two separation indices, which confirms that the mechanical test-
averaged rubber volume content of samples from the three equal ing (MSCR in this study) is an effective method to differentiate
parts. The reason for using /avg rather than the original rubber vol- the difference between different parts of the sample to character-
ume content of each binder is because swelling of rubber particles ize the storage stability. It is noteworthy that the separation
occurring during bitumen-rubber interaction significantly increase indices from CT scan test are closer to each comparing to those
the volume of rubber particles. Similar to the previous separation from mechanical tests. This means J nr from the MSCR test is more
index, a smaller value of the separation index represents higher sensitive to the change of rubber content in CRMB binders. In
storage stability of the binder. addition, Fig. 10 shows the CT scan images of CRMB binders with
Based on the initial dosages of CRM and the densities of CRM different rubber contents. It verifies the previous assumption of
and bitumen, the original volume contents of rubber in CRMB-5, the status of rubber particles in the bitumen matrix as shown
CRMB-10, CRMB-15, CRMB-22 are 4.29%, 8.22%, 11.84% and in Fig. 7. When increasing the rubber content, a potential rubber
16.46% respectively. Table 3 summarizes volume fractions of rub- network can be established due to the interparticle interactions
8 H. Wang et al. / Construction and Building Materials 249 (2020) 118840
Table 3
Glass transition temperatures of bitumen and CRM.
Fig. 10. CT scan images of (a) CRMB-5, (b) CRMB-10, (c) CRMB-15 and (d) CRMB-22.
H. Wang et al. / Construction and Building Materials 249 (2020) 118840 9
ples were monitored and compared to those of the reference sam- Considering the reference temperature to build master curve is
ple (an empty aluminium pan in this study) for further derivations. 30 °C, the crossing point in the low-frequency region corresponds
The T g and related thermal parameters of bitumen and CRM are to high temperatures. This means the two phases have similar
shown in Table 3. The Delta C p parameter reflects the change of mechanical properties in the high-temperature region. CRMB with
heat capacity after glass transition. It is obvious that bitumen a low rubber content may be stable at lower temperatures (e.g.,
and CRM have quite different glass transition temperatures. How- ambient temperatures) because of the similarity between the bitu-
ever, since the large variation of CRM source and complex glass men phase and rubber phase as reflected by the crossing over at a
transition phenomena of bitumen, it is difficult to quantitatively relatively high frequency. It may be unstable at higher tempera-
characterize the dynamic asymmetry between bitumen and CRM tures (e.g., storage temperatures) because of the dissimilarity
based on their T g s. Bearing this in mind, the fundamental vis- between BP and RP. At a fixed low frequency (approximately cor-
coelastic properties of these two components were investigated responding to the storage temperature) as indicated by the green
in the following part to see what is the direct contribution to a arrow in Fig. 12, no crossing point is reached when the Rubber con-
dynamic symmetry or dynamic asymmetry. tent is low, which means the dynamic asymmetry remains. When
increasing the Rubber content to a certain value, the crossing point
appears and the dynamic asymmetry between the two phases dis-
5.2.3. Rheological properties of bitumen phase and rubber phase appears. At this moment, the CRMB binder exhibits as a stable
The bitumen-rubber interaction process alters the volume frac- blend during high-temperature storage. This dynamic symmetry
tions of rubber and the mechanical properties of both bitumen and between bitumen phase and rubber phase is the fundamental rea-
rubber. To investigate whether this interaction process has effects son why CRMB-22 is more storage stable. Since the crossing point
on the dynamic asymmetry between BP and RP, FS tests were per- of the master curves of two phases is highly dependent on the con-
formed on the extracted liquid bitumen phase and the swollen rub- stituents of CRMB and interaction conditions, it is possible to
ber phases. It is assumed the rubber particles in CRMB binders with manipulate these factors to achieve the desired crossover between
different rubber contents have similar mechanical properties as two phases and hence obtain a storage-stable CRMB blend. In gen-
simulated by the swollen rubber sample. This is because there is eral, bitumen-rubber interaction (mainly swelling) changes both
a sufficient amount of bitumen in the binder to allow swelling to the mechanical properties of bitumen matrix and rubber due to
occur [5]. The master curves of bitumen and rubber before and the swelling of rubber by absorbing light fractions of bitumen
after swelling were developed in Fig. 12. [31]. Therefore, by controlling the interaction conditions and
Comparing the dry rubber with the swollen rubber, it can be selecting the proper raw materials (bitumen and rubber), it is pos-
found that dry rubber exhibits obvious elastic behaviors while sible to achieve a resemblance between bitumen and rubber
the swollen rubber shows obvious viscoelasticity and becomes phases after the interaction.
softer than the dry rubber. The complex moduli of swollen rubber
are all lower than that of the dry rubber at low frequencies. At high
6. Conclusions and recommendations
frequencies, the moduli of different samples merges together. In
terms of the liquid phases of different CRMB binders, they are stif-
This study investigated the effects of rubber content and warm-
fer than the neat bitumen because of the loss of light fractions to
mix additives on the storage stability of CRMB binders and
rubber. Since the mechanical properties of both bitumen phase
explored the dynamic asymmetry between the bitumen-rich phase
and rubber phase are changed after interaction, it is of more inter-
and the rubber-rich phase. Various tests on the constituents of
est to analyze the relative relationship between these two phases.
CRMB were done to understand the mechanism of storage stability.
Comparing the bitumen phase with swollen rubber phase, the
The main findings are as follows:
swollen rubber sample has lower moduli in the high-frequency
region and higher moduli in the low-frequency region. This
The viscoelastic response of CRMB improves with the increase
explains the peculiar viscoelastic response of CRMB as shown in
of rubber content (i.e., higher complex modulus and lower
Fig. 11. It is important to notice that the crossing points of the mas-
phase angle). CRMB with a higher rubber content has a wider
ter curves of bitumen phase and swollen rubber phase are located
relaxation spectrum, which means a more gradual transition
in the relatively low-frequency region. When increasing the rubber
from the elastic behaviour to the viscous behaviour. A gradual
content, the crossing point shifts to the lower frequency side.
elastic-viscous transition is beneficial for the system stability
when encountering temperature changes.
The mechanical (or other) parameters of fresh CRMB binders
cannot be used as a reference in the separation index to evalu-
ate the stability after storage because bitumen-rubber interac-
tion and aging may continue occurring during the storage
period. Instead, a parameter taken from the stored samples
can serve this purpose.
Separation indices developed from both MSCR test and CT scan
test results reveal that CRMB with a higher rubber content is
more storage stable than CRMB with a lower rubber content.
The incorporation of both warm mix additives is detrimental
to the storage stability of CRMB-22. Based on the results in this
study, it is not suggested to add warm mix additives into CRMB
during the storage or transport stage.
When the bitumen-rich phase and rubber-rich phase of CRMB
exhibit similar mechanical behaviours, which can be called
dynamic symmetry, CRMB will be storage stable. The dynamic
Fig. 12. Complex modulus master curves of liquid bitumen phase and swollen asymmetry between binder components drives the phase sepa-
rubber phase. ration behaviour of CRMB. It is possible to manipulate raw
10 H. Wang et al. / Construction and Building Materials 249 (2020) 118840
material properties and interaction conditions to achieve the [10] G. Polacco, S. Filippi, F. Merusi, G. Stastna, A review of the fundamentals of
polymer-modified asphalts: asphalt/polymer interactions and principles of
desired crossover between two phases of CRMB and hence
compatibility, Adv. Colloid Interface Sci. 224 (2015) 72–112.
obtain a storage-stable CRMB blend. [11] J. Zhu, X. Lu, N. Kringos, Experimental investigation on storage stability and
phase separation behaviour of polymer-modified bitumen, Int. J. Pavement
For future studies, more systematic DSC analyses on CRMB and Eng. (2016) 1–10.
[12] J. Zhu, N. Kringos, Towards the development of a viscoelastic model for phase
its constituents can be carried out. The effect of warm-mix addi- separation in polymer-modified bitumen, Road Mater.Pavement Des. 16
tives on the storage stability can be extended to CRMB with differ- (sup1) (2015) 39–49.
ent rubber contents. The theory of solubility parameter can be used [13] J. Zhu, R. Balieu, H. Wang, The use of solubility parameters and free energy
theory for phase behaviour of polymer-modified bitumen: a review, Road
on the constituents of CRMB (rubber and bitumen) to investigate Mater. Pavement Des. (2019) 1–22.
the compatibility between rubber and bitumen, which can provide [14] M. Sienkiewicz, K. Borze˛dowska-Labuda, A. Wojtkiewicz, H. Janik,
a more fundamental understanding of the phase separation beha- Development of methods improving storage stability of bitumen modified
with ground tire rubber: a review, Fuel Process. Technol. 159 (2017) 272–279.
viours. Practical methods (e.g., adding chemical compounds and [15] S. Hosseinnezhad, S.F. Kabir, D. Oldham, M. Mousavi, E.H. Fini, Surface
surface treatment of CRM) can be applied to further improve stor- functionalization of rubber particles to reduce phase separation in
age stability of CRMB after considering the findings in this study. rubberized asphalt for sustainable construction, J. Cleaner Prod. 225 (2019)
82–89.
[16] F. Xiao, S. Yao, J. Wang, J. Wei, S. Amirkhanian, Physical and chemical
properties of plasma treated crumb rubbers and high temperature
CRediT authorship contribution statement characteristics of their rubberised asphalt binders, Road Mater. Pavement
Des. (2018) 1–20.
Haopeng Wang: Conceptualization, Investigation, Methodol- [17] S. Kocevski, S. Yagneswaran, F. Xiao, V.S. Punith, D.W. Smith, S. Amirkhanian,
Surface modified ground rubber tire by grafting acrylic acid for paving
ogy, Writing - original draft, Writing - review & editing. Xueyan
applications, Constr. Build. Mater. 34 (2012) 83–90.
Liu: Investigation, Supervision. Sandra Erkens: Funding acquisi- [18] H.U. Bahia, H. Zhai, A. Range, Evaluation of Stability, Nature of Modifier, and
tion, Project administration, Supervision. Athanasios Skarpas: Short-Term Aging of Modified Binders Using New Tests: LAST, PAT, and
Modified RTFO, Transp Res Record (1998).
Funding acquisition, Project administration, Supervision.
[19] B.J. Putman, S.N. Amirkhanian, Characterization of the Interaction effect of
crumb rubber modified binders using HP-GPC, J. Mater. Civ. Eng. 22 (2) (2010)
153–159.
Declaration of Competing Interest [20] S. Zhao, B. Bowers, B. Huang, X. Shu, Characterizing rheological properties of
binder and blending efficiency of asphalt paving mixtures containing RAS
The authors declare that they have no known competing finan- through GPC, J. Mater. Civ. Eng. 26 (5) (2013) 941–946.
[21] X. Yu, Z. Leng, T.Z. Wei, Investigation of the rheological modification
cial interests or personal relationships that could have appeared
mechanism of warm-mix additives on crumb-rubber-modified asphalt, J.
to influence the work reported in this paper. Mater. Civ. Eng. 26 (2) (2014) 312–319.
[22] B. Hofko, L. Porot, A. Falchetto Cannone, L. Poulikakos, L. Huber, X. Lu, K.
Mollenhauer, H. Grothe, FTIR spectral analysis of bituminous binders:
Acknowledgments reproducibility and impact of ageing temperature, Materials and Structures
51(2) (2018).
[23] M.E. Kutay, S. Varma, A. Jamrah, A micromechanical model to create digital
The financial support from the China Scholarship Council for the microstructures of asphalt mastics and crumb rubber-modified binders, Int. J.
corresponding author is also acknowledged. Special thanks go to Pavement Eng. 18 (9) (2015) 754–764.
Jirawat Buchagul for performing some of the tests. [24] A. Ghavibazoo, M. Abdelrahman, Composition analysis of crumb rubber during
interaction with asphalt and effect on properties of binder, Int. J. Pavement
Eng. 14 (5) (2013) 517–530.
References [25] A. Zegard, F. Helmand, T. Tang, K. Anupam, A. Scarpas, Rheological properties
of tire rubber using dynamic shear rheometer for fem tire-pavement
interaction studies, 8th International Conference on Maintenance and
[1] T. Wang, F. Xiao, X. Zhu, B. Huang, J. Wang, S. Amirkhanian, Energy
Rehabilitation of Pavements, Research Publishing Services, Singapore, 2016,
consumption and environmental impact of rubberized asphalt pavement, J.
pp. 535–544.
Cleaner Prod. 180 (2018) 139–158.
[26] H. Wang, X. Liu, P. Apostolidis, S. Erkens, A. Skarpas, Experimental
[2] H. Wang, X. Liu, P. Apostolidis, T. Scarpas, Review of warm mix rubberized
Investigation of rubber swelling in Bitumen, Trans. Res. Rec. J. Trans. Res.
asphalt concrete: towards a sustainable paving technology, J. Cleaner Prod. 177
Board (2020).
(2018) 302–314.
[27] H. Wang, P. Apostolidis, J. Zhu, X. Liu, A. Skarpas, S. Erkens, The role of
[3] M. Attia, M. Abdelrahman, Enhancing the performance of crumb rubber-
thermodynamics and kinetics in rubber–bitumen systems: a theoretical
modified binders through varying the interaction conditions, Int. J. Pavement
overview, Int. J. Pavement Eng. (2020) 1–16.
Eng. 10 (6) (2009) 423–434.
[28] J. D’Angelo, R. Dongré, Practical use of multiple stress creep and recovery test,
[4] M.A. Abdelrahman, S.H. Carpenter, Mechanism of the interaction of asphalt
Trans. Res. Rec. J. Trans. Res. Board 2126 (2009) 73–82.
cement with crumb rubber modifier, Trans. Res. Rec. J. Trans. Res. Board 1661
[29] A. Ghavibazoo, M. Abdelrahman, M. Ragab, Effect of crumb rubber modifier
(1999) 106–113.
dissolution on storage stability of crumb rubber-modified asphalt, Trans. Res.
[5] G. Airey, M. Rahman, A.C. Collop, Crumb rubber and bitumen interaction as a
Rec. J. Trans. Res. Board 2370 (2013) 109–115.
function of crude source and Bitumen viscosity, Road Mater. Pavement Des. 5
[30] J. Zhu, R. Balieu, X. Lu, N. Kringos, Numerical prediction of storage stability of
(4) (2004) 453–475.
polymer-modified Bitumen, Trans. Res. Rec. J. Trans. Res. Board 2632 (2017)
[6] S.-C. Huang, A.T. Pauli, Particle size effect of crumb rubber on rheology and
70–78.
morphology of asphalt binders with long-term aging, Road Materials and
[31] H. Wang, X. Liu, H. Zhang, P. Apostolidis, S. Erkens, A. Skarpas,
Pavement Design 9 (1) (2008) 73–95.
Micromechanical modelling of complex shear modulus of crumb rubber
[7] J. Shen, S. Amirkhanian, F. Xiao, B. Tang, Influence of surface area and size of
modified bitumen, Mater Design 188 (2020).
crumb rubber on high temperature properties of crumb rubber modified
[32] H. Wang, X. Liu, P. Apostolidis, T. Scarpas, Rheological behavior and its
binders, Constr. Build. Mater. 23 (1) (2009) 304–310.
chemical interpretation of crumb rubber modified asphalt containing warm-
[8] H. Wang, X. Liu, H. Zhang, P. Apostolidis, T. Scarpas, S. Erkens, Asphalt-rubber
mix additives, Trans. Res. Rec. J. Transp. Res. Board 2672 (28) (2018) 337–348.
interaction and performance evaluation of rubberised asphalt binders
[33] H. Soenen, J. Besamusca, H.R. Fischer, L.D. Poulikakos, J.-P. Planche, P.K. Das, N.
containing non-foaming warm-mix additives, Road Mater. Pavement Des.
Kringos, J.R.A. Grenfell, X. Lu, E. Chailleux, Laboratory investigation of bitumen
(2018) 1–22.
based on round robin DSC and AFM tests, Mater. Struct. 47 (7) (2013) 1205–
[9] D. Lo Presti, Recycled tyre rubber modified bitumens for road asphalt
1220.
mixtures: a literature review, Constr. Build. Mater. 49 (2013) 863–881.