Professional Documents
Culture Documents
Burguers
Burguers
A R T I C L E I N F O A B S T R A C T
Keywords: The Burgers model is ineffective in the representation of dynamic viscoelastic properties of asphalt binders; the
Asphalt binder existing fractional derivative Burgers (FDB I) model has limitations in its application and construction concept.
Master curve Consequently, this study started by deriving the constitutive equations of the FDB I model and proposed a new
Dynamic viscoelastic properties
application idea of Abel dashpot to construct the second fractional derivative Burgers (FDB II) model. Then,
Abel dashpot
dynamic modulus formulas for the FDB II model were derived. On this basis, the frequency sweep tests of base
Fractional derivative constitutive model
and two polyphosphoric acid (PPA) modified asphalt binders were carried, and the dynamic viscoelastic prop
erties of three different asphalt binders were described using three different models, and in-depth comparisons
and analysis are carried out. The results show that both fractional derivative Burgers (FDB) models can simul
taneously express the dynamic viscoelastic properties of asphalt binders with a set of parameters. Both FDB
models perform better than traditional Burgers model in the representation of dynamic viscoelastic properties,
and fitting results of master curves no longer exhibit oscillation properties like the traditional Burgers model. The
R2 of FDB models could reach above 0.99, while R2 of traditional Burgers model was only 0.9349 at minimum.
Among them, the FDB II model performs best. Five parameters of the FDB II model have relatively clear physical
meanings. The new application concept of fractional dashpots was verified its theoretical feasibility and appli
cation advantages, which will provide a new idea for further application and development of fractional deriv
ative constitutive models.
* Corresponding authors at: School of Highway, Chang’an University, Xi’an, Shaanxi 710064, China.
E-mail addresses: lixinzhou@chd.edu.cn (X. Li), ams@chd.edu.cn (A. Sha), jiaowenxiu@chd.edu.cn (W. Jiao), srm19988@163.com (R. Song), yscao@chd.edu.cn
(Y. Cao), 2020021043@chd.edu.cn (C. Li), zzliu@chd.edu.cn (Z. Liu).
https://doi.org/10.1016/j.conbuildmat.2023.133552
Received 4 October 2022; Received in revised form 29 July 2023; Accepted 25 September 2023
Available online 6 October 2023
0950-0618/© 2023 Elsevier Ltd. All rights reserved.
X. Li et al. Construction and Building Materials 408 (2023) 133552
creep and recovery response. You [21] found the Burgers model could constructed by replacing Newton dashpot with fractional derivative
better express the static creep properties of the asphalt mastics under element, just like fractional Maxwell model [25], fractional Zener model
repeated loadings. The Burgers model is considered as one of the best [26], fractional Burgers model [27,37,38]. However, the application of
models to express the static viscoelastic properties. Nevertheless, as the fractional derivative elements that directly replace Newton dashpot
research progressed, researchers find that the traditional models which element (e.g., the FDB I model) is very limited. The mathematical
are represented by the Burgers model are not in great agreement with definition of the fractional derivative elements makes it possible and
experimental results in describing dynamic viscoelastic properties of scientific to replace the spring element but not the Newton dashpot, and
asphalt binders, which means they could not truly represent the actual the actual effective should be explored.
dynamic viscoelastic properties of viscoelastic materials. In order to The aim of this study is to explore the concept of constructing the
solve this problem, fractional derivative constitutive models gradually fractional derivative Burgers (FDB) model and advantages in describing
caught the attention of researchers. dynamic viscoelastic properties of asphalt binders. Specifically, these
By adding different types of fractional derivative constitutive ele include: (i) investigating the effect of the FDB I model for dynamic
ments to traditional constitutive models, a series of fractional derivative viscoelastic performance prediction of asphalt binders; (ii) investigating
constitutive models were constructed, such as: Huet model [22], Huet- the new idea of applying the basic fractional element in the constitutive
Sayegh model [23], 2S2P1D model [24], fractional Maxwell model model construction; (iii) investigating the advantages of the new-
[25], fractional Zener model [26], fractional Burgers model [27,28], constructed model and the influences of model parameters on master
1S1A1D model [29], etc. C. Celauro [30] used a fractional derivative curves. To this end, this study starts with deriving dynamic viscoelastic
constitutive model to describe the linear viscoelastic behaviors of formulas for the FDB I model. Then, a new idea of applying the basic
asphalt binders with a very limited number of parameters. Zhou [31] fractional elements is proposed and the new fractional derivative Bur
introduced a generalized fractional derivative viscoelastic model to gers (FDB II) model is constructed. The analytical expressions of com
describe master curves of bending creep stiffness of different asphalt plex modulus for the FDB II model are deduced. The master curves of test
binders and carried out a study on the evaluation of the crack resistance data are constructed with three different asphalt binders, and parame
at low temperature of materials. Chiara Riccardi [32] explored the as ters determination method is introduced for characterizing all dynamic
sociation between bitumen content in mastic and characteristic time viscoelastic indicators simultaneously. Finally, the actual fitting effects
parameters based on the 2S2P1D model. Sun [33] verified that the HN of different models were compared and the influences of model pa
model provides a more comprehensive characterization of the linear rameters on the master curve were explored.
viscoelastic behavior of asphalt mixtures than the traditional sigmoidal
model. Yin [34] constructed a fractional derivative Kelvin model and a 2. Constitutive model
fractional three elements solid model to describe the dynamic visco
elastic behavior of asphalt mixtures, and the fractional three elements 2.1. Classical Burgers model
solid model can describe the dynamic viscoelastic properties of mixtures
more accurately compared with traditional models. Sheng [35] pro The Burgers model (shown in Fig. 1) consists of the Maxwell model
posed the generalized form (GFDZ) of FDZ model, which could better and Kelvin model connected in series, which can be used to describe the
express the static and dynamic viscoelastic properties of asphalt binders instantaneous elastic deformation, delayed elastic deformation and
and asphalt mastics than the mCAM model and 2S2P1D model. Yuan viscous deformation of viscoelastic materials [44]. The model was
[36] used the 1S1A1D model to describe the dynamic viscoelastic widely used in the rheological analysis of viscoelastic materials such as
properties of high-viscosity modified asphalt binder and verified that asphalt binders and asphalt mixtures because of the small number of
increasing the number of fractional elements has no great impact on the model parameters and their clear physical significance. For the Burgers
fitting effect of the model. M. Lagos-Varas [37,38] established a me model, constitutive equations of storage modulus and loss modulus are
chanical model by replacing Newton dashpot with Abel dashpot in Eq. (1) and Eq. (2).
Burgers model to research the creep and recovery phenomena of binders
and asphalt mixtures. p1 q1 ω2 − q2 ω2 (1 − p2 ω2 )
E′(ω) = (1)
From the review of literatures, it can be known that fractional de q1 ω2 + (1 − p2 ω2 )2
rivative elements are seen as the expression of viscoelastic properties
composited of Hooke spring element and Newton dashpot element. E″(ω) =
p1 q2 ω3 + q1 ω(1 − p2 ω2 )
(2)
Their advantages in describing viscoelastic characterization are obvious q1 ω2 + (1 − p2 ω2 )2
[39]. Furthermore, dynamic viscoelastic properties of asphalt binders
could be obtained by dynamic mechanical analysis which are closer to where, E′(ω) represents the storage modulus which is the real part of
the real loading state in the pavement, and could evaluate the visco complex modulus; E″(ω) represents the loss modulus which is the
elastic properties of binders in wider range of temperature and fre imaginary part of complex modulus;p1 = Eη11 + η1E+2η2 ; p2 = Eη11 ηE22 ; q1 = η1 ;
quency. Dynamic mechanical analysis is almost essential in the q2 = ηE1 η22 ; E1 , E2 , η1 , and η2 are parameters of the Burgers model as shown
preparation of modified asphalt binders [40,41] and in the rheology
in Fig. 1. Specifically, E1 and E2 are the elastic modulus of the spring
performance evaluation of base and modified asphalt binders [42,43].
spots in Maxwell model and Kelvin model respectively; η1 , and η2 are the
Consequently, better describing the dynamic viscoelastic properties of
viscosity coefficient of the Newton dashpots in Maxwell model and
the base and modified asphalt binders is crucial. Based on the analysis
above, the following could be noticed: (i) although the existing frac
tional derivative Burgers (FDB I) model has been proposed as shown in
Fig. 3, the FDB I model was only applied to the static viscoelastic be
haviors (stress relaxation, creep and recovery phenomena, etc.) of
asphalt binders and mixtures at this stage [37,38]. For one constitutive
model, its accurate description of static viscoelastic properties may not
mean that it would perform similarly excellently in describing dynamic
viscoelastic properties. There have been almost not reported in
describing dynamic viscoelastic properties using the FDB I model.
Therefore, the application fields of the FDB I model are worth further
exploration; (ii) Nowadays, many fractional constitutive models were Fig. 1. Representation of the Burgers model.
2
X. Li et al. Construction and Building Materials 408 (2023) 133552
F[Dα ε(t)] = (iω)α F[ε(t)] (6) Combining Eq. (10), (12) and (13), Eq. (14) could be derived.
E2 1 1 1 E2 1 E
iα = cos(απ /2) + isin(απ/2) (7) (∂αt +1 + ∂1t )ε(t) = ( ∂αt +1 + ∂αt + ∂1t + ∂ + 2 )σ (t) (14)
η2 E1 η1 η2 E1 η2 t η1 η2
E∗ (iω) = η(iω)α = ηωα [cos(απ /2) + isin(απ/2)] (8) Combining Eqs. (6) and (7), the Fourier transform of Eq. (14) gives
expressions of complex modulus E∗ (iω)(as shown in Eq. (15)), the stor
2.2.2. FDB I model age modulus E′(ω)(as shown in Eq. (16)), the loss modulus E″(ω)(as
At present, the FDB I model has been constructed by replacing the shown in Eq. (17)), and the phase angle θ(ω)(as shown in Eq. (18)).
Newton dashpot in the Kelvin model with Abel dashpot, as shown in
[A(ω)C(ω) + B(ω)D(ω)] + i[B(ω)C(ω) − A(ω)D(ω)]
Fig. 3. The FDB I model achieved quite satisfactory results in the E∗ (iω) = (15)
C2 (ω) + D2 (ω)
expression of static creep and creeping reversion of asphalt binders.
However, the FDB I model has almost not been used for the analysis of A(ω)C(ω) + B(ω)D(ω)
dynamic viscoelastic properties. Therefore, this paper first derives its E′(ω) = (16)
C2 (ω) + D2 (ω)
expressions of complex modulus using the differential operator and
related mathematical theory. The derivation process is as follows. B(ω)C(ω) − A(ω)D(ω)
The FDB I model consists of two models, the existing fractional de E″(ω) = (17)
C2 (ω) + D2 (ω)
rivative Kelvin (FDK I) model and traditional Maxwell model, in series.
Their differential constitutive equations are Eq. (9). E′(ω) A(ω)C(ω) + B(ω)D(ω)
tanθ(ω) = = (18)
E″(ω) B(ω)C(ω) − A(ω)D(ω)
3
X. Li et al. Construction and Building Materials 408 (2023) 133552
C(ω) = E1 E2 − η1 η2 ωα+1 sin(απ /2) + E1 η2 ωα cos(απ /2) (21) the storage modulus E′(ω)(as shown in Eq. (30)), the loss modulus
E″(ω)(as shown in Eq. (31)), and the phase angle θ(ω)(as shown in Eq.
D(ω) = E2 η1 ω + η1 η2 ωα+1 cos(απ/2) + E1 η2 ωα sin(απ /2) + E1 η1 ω (22) (32)).
Table 1
Properties of base asphalt binder.
Item Test Method Value
4
X. Li et al. Construction and Building Materials 408 (2023) 133552
Table 2 same series of parameters, the software “1stopt” and “L-M optimization
Properties of PPA and TOTM. algorithm + Universal Global Optimization” is used to fit the experi
Item Value mental results of storage modulus and loss modulus simultaneously.
PPA P2O5 content (%) ≥80 X = lg(ω)⇒ω = 10X (37)
SO−4 2 content (%) ≤0.005
Fe content (%) ≤0.05
TOTM Moisture content (%) 0.990
Y1 = lg[G′(ω)] = lg[G′(10X )] (38)
Boiling point (◦ C) 258
Viscosity at 20 ◦ C (mPa⋅s) 300 Y2 = lg[G″(ω)] = lg[G″(10X )] (39)
Density at 25 ◦ C (g⋅cm− 3) 0.988
5
X. Li et al. Construction and Building Materials 408 (2023) 133552
frequency range from 102 rad/s to 103 rad/s, the model fitting results of the actual compared with two modified asphalts. The elastic properties
storage modulus are slightly higher than test data. More specifically, the predicted are slightly higher than the actual in the frequency range from
viscosity properties of asphalt binders predicted by the FDB I model are 102 rad/s to 103 rad/s. It is obvious from Table 6 that R2 of all asphalt
almost identical to the actual. In the frequency range (about below binders can reach about 0.999, which proves that the FDB I model has an
10− 2.5rad/s), the elastic properties predicted are lower than the actual, excellent fitting effect overall.
and the predicted properties of SK70 show the greatest difference from
6
X. Li et al. Construction and Building Materials 408 (2023) 133552
Parameters of the FDB II model are listed in Table 7. Test data and the
Table 4
FDB II model fitting results for three different asphalt binders are shown
Shift factors lgαT for different asphalt binders (Tr = 40℃).
in Fig. 10.
Temperature SK70 TP1 TP2 As can be seen from Fig. 10, for the loss modulus curves, the FDB II
25℃ 1.3863 1.3379 1.6814 model fitting results are in great accordance with experimental results.
30℃ 0.9067 0.9022 0.9249 In the description of storage modulus main curves, the fitting accuracy
40℃ 0 0 0 using the FDB II model for the three asphalt binders is in the following
50℃ − 0.7769 − 0.7652 − 0.8087
60℃ − 1.4345 − 1.4164 − 1.4879
order: TP1 > TP2 > SK70. The fitting accuracy of all asphalt binders is
70℃ − 1.9994 − 1.9782 − 2.0766 fine in the high frequency range, and that of TP1 is the best. The fitting
results of SK70 deviate from test data in the low frequency range (from
10− 3 rad/s–10− 1 rad/s) where test data of modified asphalt binders can
4.1.3. Description by FDB II model be better fitted. Meanwhile, it can be also proved by R2 values in Table 7.
Master curves fitting were performed through formulas of storage Overall, the FDB II model achieved great effects for fitting test data of
modulus (Eq. (24)) and loss modulus (Eq. (25)) derived in section 2.2.3.
7
X. Li et al. Construction and Building Materials 408 (2023) 133552
8
X. Li et al. Construction and Building Materials 408 (2023) 133552
Table 6
The parameters of the FDB I model.
Asphalt binder E1 /Pa E2 /Pa η1 /(Pa⋅s) η2 /(Pa⋅s) α R2
9
X. Li et al. Construction and Building Materials 408 (2023) 133552
c
Fig. 9. Test data and the FDB I model fitting results.
Table 7
Parameters of the FDB II model.
Asphalt binder E1 /Pa η1 /(Pa⋅s) η2 /(Pa⋅s) η3 /(Pa⋅s) α R2
the FDB I model. Thirdly, fitting results of FDB I model and test data 4.2. Effect of parameters on master curves
show some deviation in the high frequency band (roughly over 101.5
rad/s), which will seem larger in the Cartesian coordinate. At this band The order of this section is to explore effect on master curves and the
the FDB II model has smaller errors than FDB I model. Therefore, the physical meanings of the FDB II model parameters. According to the
FDB II model has better fitting effectiveness for the partial data, and fitting parameters for SK70 in Table 7, the master curve functions con
results of the FDB II model are closer to test data in general than those of structed by changing one parameter are drawn in two different coordi
the FDB I model. As a whole, the fitting accuracy of three models is in the nate systems, as shown in Fig. 13.
following order: the FDB II model > the FDB I model > the Burgers In Fig. 13(a), the value of the parameter E1 significantly affects peak
model, whether storage modulus curves or loss modulus curves. value of storage modulus and loss modulus, and peak width of loss
modulus curves in the high frequency range (low temperature). As
parameter E1 increases, the peaks of storage modulus and loss modulus
10
X. Li et al. Construction and Building Materials 408 (2023) 133552
Fig. 10. Test data and the FDB II model fitting results.
become larger, and peak widths of loss modulus become wider while the mater curves in high frequency range (low temperature); Parameter η1
peaks move right. In Fig. 13 (b), the value of parameter η1 mainly affects affects the magnitude of master curves peaks in low frequency range
the magnitude of master curves in the low frequency range (high tem (high temperature); Parameter η2 affects the horizontal positions of
perature), the position of storage modulus sudden increasing, the value master curves; Parameter η3 affects the growth trends of main curves in
and position of loss modulus peaks. As parameter η1 increases, master low frequency range (high temperature); Parameter α affects the oscil
curves of storage modulus and loss modulus are shifted left and their latory characteristics of storage modulus master curves and the peak
peaks increase in the low frequency range, but loss modulus peaks variations of loss modulus master curves.
slightly reduce. In Fig. 13(c), the value of parameter η2 significantly
affects the horizontal position where the storage modulus curves reach
their peaks, the magnitude and horizontal position of loss modulus 4.3. Analysis of fractional derivative Burgers model
peaks. As parameter η2 increases, the muster curves of storage modulus
peaks early, and the peaks of loss modulus curves show increasing and In order to better analyze and evaluate the FDB models, this section
left-shifting trend. In Fig. 13(d), the value of parameter η3 affects the analyzes the theoretical reasonability of model construction concept and
magnitude of storage modulus in the low frequency range and changing advantages of new models. It is mainly attributed to the following
trends of all master curves in the middle frequency range. As parameter aspects.
η3 increases, the storage modulus decreases in the low frequency range,
the loss modulus peaks decrease, but all master curves grow faster in the (1) Theoretical reasonability of model construction concept
middle frequency range. The most complex influence on master curves
among these parameters is α, which no longer simply affects the peaks The theoretical reasonability can be verified by connections between
and horizontal positions of master curves, as can be seen from Fig. 13(e). constitutive equations of three different Burgers models. The reason for
For storage modulus, master curves exhibit different oscillatory prop replacing other elements with Abel dashpot mentioned in the previous is
erties in logarithmic coordinate system as parameter α varies within 0 to that these dashpots are inter-convertible when α is 0 or 1. For the FDB I
1, and the oscillatory properties are minimized when α = 0.6–0.7. As model, if parameter α = 1, it can be found that the Eq. (16) and Eq. (17)
parameter α increases, storage modulus master curves reach peaks in section 2.2.2 will transform into Eq. (1) and Eq. (2). At this point, the
earlier. For loss modulus, as the parameter α increases, the master curves FDB I model converts into Burgers model. Similarly, for the FDB II model
peaks decrease (reaching a minimum around α = 0.7) and then increase, constructed in this paper, the Eq. (30) and Eq. (31) will become Eq. (1)
accompanied by a leftward shift. and Eq. (2), if parameter α = 0. As a result, two fractional derivate
In general, five parameters of the FDB II model have relatively clear Burgers models are the evolution forms of traditional Burgers model in
physical meanings, and influences of parameters mainly include the essence. This also verifies the reasonableness of the concept to replace
following aspects: Parameter E1 affects the peaks characteristics of Hooke spring element with Abel dashpot and the derivations of formulas
in the previous works are accurate.
11
X. Li et al. Construction and Building Materials 408 (2023) 133552
(2) Advantages of fractional derivate constitutive models trigonometric functions with parameter α to constitutive equations of
the Burgers model, which reduces the oscillation amplitude of master
Firstly, due to the special mathematical properties of fractional order curves. And this analysis is in agreement with the influence of parameter
calculus, the introduction of trigonometric functions and powers α on master curves in Section 4.2. Furthermore, the powers in consti
increasing in constitutive equations make the fitting accuracy improved tutive equations of the FDB II model are larger than them in constitutive
significantly. The constitutive equations of the FDB models add equations of the FDB I model (the power increases from 2(α + 1) to 4 as
12
X. Li et al. Construction and Building Materials 408 (2023) 133552
maximum), which can further enhance fitting effect of the FDB I model. parameters of fractional derivative models used in this paper have clear
Secondly, two FDB models require only five parameters to describe physical meanings that have been verified in section 4.2. In addition, it
dynamic viscoelastic properties satisfactorily, which is difficult to ach can be observed from model representations that fractional derivate
ieve using traditional differential constitutive models. At the same time, Burgers models is able to describe instantaneous elastic deformation,
compared with traditional mathematical models (e.g., CAM models), the delayed elastic deformation and viscous deformation of viscoelastic
13
X. Li et al. Construction and Building Materials 408 (2023) 133552
Fig. 13. Master curves constructed by different parameters of the FDB II model.
14
X. Li et al. Construction and Building Materials 408 (2023) 133552
deformation, similarly to the Burgers model. About this, it can be veri accurately derived, and the applied concept of fractional deriv
fied by parameter η1. It is known from previous studies [40] that the ative element is feasible in theory.
deformation resistance of three asphalt binders is in the following order:
SK70 < TP1 < TP2. In Table 6 and Table 7, the fitting parameter η1 of CRediT authorship contribution statement
two fractional derivative models have the same pattern with this.
However, this pattern has not be found in fitting results of the Burgers Xinzhou Li: Writing – review & editing, Writing – original draft,
model in Table 5, and the reason is considered to be the huge oscillation Methodology, Formal analysis, Data curation, Conceptualization. Aimin
and error in fitting dynamic viscoelastic properties by the Burgers Sha: Funding acquisition, Conceptualization. Wenxiu Jiao: Writing –
model. review & editing, Software, Investigation. Ruimeng Song: Writing –
review & editing, Conceptualization. Yangsen Cao: Investigation. Chao
5. Conclusions Li: Validation. Zhuangzhuang Liu: Validation.
15
X. Li et al. Construction and Building Materials 408 (2023) 133552
[6] C. Koch, K. Georgieva, V. Kasireddy, et al., A review on computer vision based [30] C. Celauro, C. Fecarotti, A. Pirrotta, An extension of the fractional model for
defect detection and condition assessment of concrete and asphalt civil construction of asphalt binder master curve, Eur. J. Environ. Civ. Eng. 21 (1)
infrastructure, Adv. Eng. Inf. 29 (2) (2015) 196–210. (2017) 78–93.
[7] D. Yuan, C. Xing, W. Jiang, et al., Viscoelastic behavior and phase structure of high- [31] J. Zhou, X. Chen, G. Xu, et al., Evaluation of low temperature performance for SBS/
content sbs-modified asphalt, Polymers 14 (12) (2022) 2476. CR compound modified asphalt binders based on fractional viscoelastic model,
[8] W. Wang, M. Jia, W. Jiang, et al., High temperature property and modification Constr. Build. Mater. 214 (2019) 326–336.
mechanism of asphalt containing waste engine oil bottom, Constr. Build. Mater. [32] C. Riccardi, A. Cannone Falchetto, M. Losa, et al., Rheological modeling of asphalt
261 (2020), 119977. binder and asphalt mortar containing recycled asphalt material, Mater. Struct. 49
[9] H. Wang, I.L. Al-Qadi, Importance of nonlinear anisotropic modeling of granular (10) (2016) 4167–4183.
base for predicting maximum viscoelastic pavement responses under moving [33] Y. Sun, J. Chen, B. Huang, Characterization of asphalt concrete linear viscoelastic
vehicular loading, J. Eng. Mech. 139 (1) (2013) 29–38. behavior utilizing Havriliak-Negami complex modulus model, Constr. Build. Mater.
[10] M. Ameri, M. Malakouti, P. Malekzadeh, Quasi-static analysis of multilayered 99 (2015) 226–234.
domains with viscoelastic layer using incremental-layerwise finite element [34] H. Yin, Y. Li, N.Z. Wang, Research on fractional derivative viscoelastic constitutive
method, Mech. Time-Dependent Mater. 18 (1) (2014) 275–291. relation of asphalt mixture[C]//Advanced Materials Research, Trans Tech
[11] Z. Qi-sen, X. Xin, Research review on constitutive model and microstructure of Publications Ltd 446 (2012) 2560–2566.
asphalt and asphalt mixture, China J. Highway Transp. 29 (5) (2016) 26. [35] S. Liang, R. Luo, W. Luo, Fractional differential constitutive model for linear
[12] L. Wen-bo, L. Sheng, Z. Yong-jun, Fractional differential constitutive model for viscoelasticity of asphalt and asphalt mastic, Constr. Build. Mater. 306 (2021),
dynamic viscoelasticity of asphalt mixture, China J. Highway Transp. 33 (2) (2020) 124886.
34. [36] D. Yuan, W. Jiang, Y. Hou, et al., Fractional derivative viscoelastic response of
[13] Q. Li, X. Wang, X. Liu, et al., Review on constitutive models of road materials, high-viscosity modified asphalt, Constr. Build. Mater. 350 (2022), 128915.
J. Road Eng. (2022). [37] M. Lagos-Varas, A.C. Raposeiras, D. Movilla-Quesada, et al., Study of the
[14] P.S. Divya, C.S. Gideon, J.M. Krishnan, Influence of the type of binder and crumb permanent deformation of binders and asphalt mixtures using rheological models
rubber on the creep and recovery of crumb rubber modified bitumen, J. Mater. Civ. of fractional viscoelasticity, Constr. Build. Mater. 260 (2020), 120438.
Eng. 10 (1061) (2013) 438–449. [38] M. Lagos-Varas, D. Movilla-Quesada, A.C. Raposeiras, et al., Viscoelasticity
[15] J. Yi, S. Shen, B. Muhunthan, et al., Viscoelastic–plastic damage model for porous modelling of asphalt mastics under permanent deformation through the use of
asphalt mixtures: Application to uniaxial compression and freeze–thaw damage, fractional calculus, Constr. Build. Mater. 329 (2022), 127102.
Mech. Mater. 70 (2014) 67–75. [39] X.u. Yanan, L. Shan, S. Tian, Fractional derivative viscoelastic response model for
[16] L. Yongliang, K. Xiangming, Z. Yanrong, et al., Static and dynamic mechanical asphalt binders, J. Mater. Civ. Eng. 31 (6) (2019) 04019089.
properties of cement-asphalt composites, J. Mater. Civ. Eng. 25 (10) (2013) [40] K.e. Shi, F. Ma, J. Liu, et al., Development of a new rejuvenator for aged SBS
1489–1497. modified asphalt binder, J. Clean. Prod. 380 (2022), 134986.
[17] J. Hu, Z. Qian, D. Wang, et al., Influence of aggregate particles on mastic and air- [41] N.H.C. Gómez, M. Oeser, O. Fleischel, Chemical modification of bitumen with
voids in asphalt concrete, Constr. Build. Mater. 93 (2015) 1–9. novel isocyanate-based additive to enhance asphalt performance, Constr. Build.
[18] Y. Zhao, Y. Ni, W. Zeng, A consistent approach for characterising asphalt concrete Mater. 301 (2021), 124128.
based on generalised Maxwell or Kelvin model, Road Mater. Pavement Des. 15 (3) [42] Z. Zhang, J. Sun, M. Jia, et al., Study on a thermosetting polyurethane modified
(2014) 674–690. asphalt suitable for bridge deck pavements: Formula and properties, Constr. Build.
[19] C. Wu, L. Li, W. Wang, et al., Experimental characterization of viscoelastic Mater. 241 (2020), 118122.
behaviors of nano-tio2/caco3 modified asphalt and asphalt mixture, Nanomaterials [43] Z. Hossain, M.S. Alam, G. Baumgardner, Evaluation of rheological performance
11 (1) (2021) 106. and moisture susceptibility of polyphosphoric acid modified asphalt binders, Road
[20] N. Saboo, A. Mudgal, Modelling creep and recovery response of asphalt binders Mater. Pavement Des. 21 (1) (2020) 237–252.
using generalized burgers model, Pet. Sci. Technol. (2018) 1–9. [44] Yin Y, Zhang X. Study on constitutive relation of asphalt mixtures based on
[21] Q. You, B. Wei, J. Ma, et al., Interpreting the creep behavior of asphalt mortar at dynamic creep test[J]. Journal of Functional Materials, 45(23):23020-23024.
high temperature through experimental and numerical methods, Constr. Build. [45] S. Samko, Fractional integration and differentiation of variable order: an overview,
Mater. 258 (sup1) (2020), 120317. Nonlinear Dyn. 71 (4) (2013) 653–662.
[22] M. Oeser, T. Pellinien, Computational framework for common visco-elastic models [46] X.J. Yang, General fractional derivatives: theory, methods and applications[M],
in engineering based on the theory of rheology, Comput. Geotech. 42 (2012) Chapman and Hall/CRC, 2019.
145–156. [47] R. Song, A. Sha, K. Shi, et al., Polyphosphoric acid and plasticizer modified asphalt:
[23] E. Bocci, A. Graziani, F. Canestrari, Mechanical 3D characterization of epoxy Rheological properties and modification mechanism, Constr. Build. Mater. 309
asphalt concrete for pavement layers of orthotropic steel decks, Constr. Build. (2021), 125158.
Mater. 79 (2015) 145–152. [48] M. Jia, A. Sha, Z. Zhang, et al., Effect of organic reagents on high temperature
[24] N. Yusoff, D. Mounier, G. Marc-Stephane, et al., Modelling the Theological rheological characteristics of organic rectorite modified asphalt, Constr. Build.
properties of bituminous binders using the 2S2P1D Model, Constr. Build. Mater. 38 Mater. 227 (2019), 116624.
(JAN.):395–406 (2013). [49] Y. Yin, Research on dynamic Viscoelastic Characteristics and Shear Modulus
[25] S.W. Katicha, A.K. Apeagyei, G.W. Flintsch, et al., Universal linear viscoelastic Predicting Methods for Asphalt Mixtures Based on Dynamic Mechanical Analysis
approximation property of fractional viscoelastic models with application to (DMA) Means[D], South China University of Technology, 2010.
asphalt concrete, Mech. Time-Dependent Mater. 18 (3) (2014) 555–571. [50] N.W. Tschoegl, The phenomenological theory of linear viscoelastic behavior: an
[26] L. Gu, W. Zhang, T. Ma, et al., Numerical simulation of viscoelastic behavior of introduction[M], Springer Science & Business Media, 2012.
asphalt mixture using fractional constitutive model, J. Eng. Mech. 147 (5) (2021) [51] F. Zhang, Viscoelastoplastic Cotinuum Damage Constitutive Model of Asphalt
04021027. Mixtures under Compression Loading[D], Inner Mongolia University of
[27] M. Lagos-Varas, D. Movilla-Quesada, J.P. Arenas, et al., Study of the mechanical Technology, 2021.
behavior of asphalt mixtures using fractional rheology to model their [52] Y. Yin, W. Huang, J. Lv, et al., Unified construction of dynamic rheological master
viscoelasticity, Constr. Build. Mater. 200 (2019) 124–134. curve of asphalts and asphalt mixtures, Int. J. Civil Eng. 16 (9) (2018) 1057–1067.
[28] A.S. Okuka, D. Zorica, Fractional Burgers models in creep and stress relaxation [53] Li J, Huang C, et al. Dynamic Shear Performances of Adhesive Layer between
tests, App. Math. Model. 77 (2020) 1894–1935. UHPC and Asphalt Surface. Journal of Hunan University,202249(5):82-91.
[29] J. Shi, J. Li, et al., Asphalt viscoelastic model and its universality under dynamic
and static loading conditions, J. China Univ. Pet. 45 (4) (2021) 168–175.
16