Aging Behavior and Creep Characteristics of CACB

Yunliang Li, Yuze Liu, Lei Xu, Xin He and Haijiao Sun

(School of Transportation Science and Engineering, Harbin Institute of Technology, Harbin 150090, China)

Abstract: Cement asphalt composite binder (CACB) is a type of composite binder composed of cement, asphalt, water, and various additives through curing, whose creep properties are influenced by asphalt aging. By CACB creep tests at various aging times, the creep compliance curves of the material were obtained. Curves of the test results were fitted by the creep compliance of Burgers model revised by fractional derivative, and the effects of asphalt aging and mix proportions on the CACB creep properties were analyzed. This study demonstrates that as aging time increased, the ability of the CACB to resist instant elastic deformation and viscous flow deformation was enhanced, the amount of its elastic components increased but the viscous components were reduced, and the material deformability decreased, exhibiting deforming properties similar to elastic materials. The effects of asphalt aging on the CACB viscoelastic mechanical properties were great during the initial aging stage (within 12 h) and became weaker during the later aging stage (over 12 h). During the initial aging stage, the CACB creep deformability could be enhanced by increasing asphalt application. However, during the later aging stage, asphalt application had little effect on creep deformability.

Keywords: cement asphalt composite binder; aging; creep; viscoelastic properties

1 Introduction

Cement asphalt composite binder (CACB) is composed of cement, asphalt, water, and various additives through curing[1-3]. Possessing both cement and asphalt characteristics, it exhibits a certain degree of strength as well as significant flexibility and deformability[4-6]. It is thus widely used in transportation infrastructure construction.

As a semi-flexible binder, CACB can be used as binder for semi-flexible bases and sub-bases in highways to prevent reflection cracks[7-8]or as a sealant in pavement maintenance[9-10]. To serve the purposes of a binder, it demands a certain degree of flexibility and deformability. As a type of viscoelastic material, the deformability of CACB is closely related to creep properties. Furthermore, as typical viscoelastic material, the viscoelastic properties of asphalt change with asphalt aging[11-13], so do the creep properties of the CACB.

Asphalt aging and its effects on the properties of asphalt mixture have been extensively studied[14-16]. However, studies on the effects of asphalt aging on CACB creep properties are relatively scarce. Therefore, further research on asphalt aging and CACB creep properties is highly necessary.

2 Materials and Test Schemes

2.1 Raw Materials and Mix Proportions

Ordinary Portland cement with a strength grade of 42.5 and cationic emulsified asphalt with a solid content of 60% were used. The admixture consisted of polyurethane thickening agent, lignin water reducer, and organosilicon defoamer. The mass ratios of emulsified asphalt and cement (A/C) were 1.0, 1.2, and 1.4, while the water and cement ratio was 0.8. The technical properties of the cement and emulsified asphalt are listed in Tables 1-2, while the raw material mix proportions (where cement mass was used as the unit) are displayed in Table 3.

2.2 Test Methods

2.2.1Heatagingtest

An accelerated heat aging test was conducted. After a curing period of 28 d, specimens were placed in the oven for heating, and the test temperature was 150±2 ℃, while the aging times were 4 , 8 , 12, 18 , and 24 h.

2.2.2Creeptest

A dynamic shear rheometer was used for the CACB creep test. Both the original specimens and those after aging were tested. The test temperature was 25±0.2 ℃ and the duration of each test was 1 200 s.

Table 1 Technical properties of emulsified asphalt

Table 2 Technical properties of cement

Table 3 Mix proportions of CACB

2.3MechanicalModelandCreepEquationTaking the effects of aging on the viscoelastic properties of CACB into consideration, the Burgers model revised by fractional derivative was used to describe the mechanical behavior of the CACB. The viscoelastic mechanical model (Fig.1) is composed of Maxwell unit and revised Kelvin unit, in which an Abel dashpot replaces the Newton dashpot and five material parametersE1,E2,η1,η2, andrare used. The model differs from the Burgers model mainly in terms of the introduction of the Abel dashpot containing a fractional derivative, thus it is referred to as the Burgers model revised by fractional derivative. The constitutive relation of the Abel dashpot is as follows:

σ2=η2ε(r)

(1)

Fig.1 Burgers model revised by fractional derivative

The value of fractional derivativeris within the range of 0 to 1. The material becomes completely elastic-like whenrapproaches 0 and completely viscous-like whenrapproaches 1. The change inrduring the aging process can characterize the change in the elastic-viscous component ratio in the CACB, and the effects of aging on the CACB viscoelastic properties can be analyzed thoroughly usingr.

The model creep compliance consists of three parts, namely, compliance of the independent spring, independent dashpot, and revised Kelvin unit. The creep compliancesJ1(t),J2(t), andJ3(t) characterize the instant elastic deformation, viscous flow deformation, and delayed elastic deformation of the viscoelastic deformation, respectively. The instant elastic and viscous flow deformation are a result of the elastic and viscous components in the material respectively, while the delayed elastic deformation is due to the interaction between the two components. As typical viscoelastic material, the aging of asphalt influences its viscoelastic properties and composite binder creep properties. By analysis of the material creep compliance through aging, the effects of asphalt aging on the CACB creep properties can be further studied.

The creep compliance of the dependent spring of the Maxwell unit is as follows:

(2)

The creep compliance of the dependent dashpot of the Maxwell unit is as follows:

(3)

The constitutive equation of the revised Kelvin unit with the Abel dashpot is as follows:

(4)

Eq.(4) was transformed from Laplace into the equation below:

(5)

The Laplace transformation equation of the revised Kelvin unit with the Abel dashpot is as follows:

(6)

Eq.(6) was calculated by infinite series expansion and then inverse Laplace transformation, and the result is as follows:

(7)

The itemn= -1 was added to the infinite series and subtracted out of it at the same time, and the result is as follows:

(8)

The composite binder creep process is related to load duration. Consequently, Eq.(8) was approximated by assigningrin the Gamma function as 1 and retaining the parameterras the load duration index. Finally,Γ(n+1)=n! was put into the equation, and the result is as follows:

(9)

According to the equation below:

(10)

wherexin the above equation was assigned as:

(11)

Eq.(11) was then put into Eq.(10) and combined with Eq.(9). The creep compliance of the revised Kelvin unit with the Abel dashpot was established as follows:

(12)

According to the superposition principle, the creep compliance of the Burgers model revised by fractional derivative is as follows:

J(t)=J1(t)+J2(t)+J3(t)

(13)

The creep compliance of the Burgers model revised by fractional derivative was established as follows:

(14)

whereE1represents elastic modulus of the spring in the Maxwell unit,η1represents viscous coefficient of the dashpot in the Maxwell unit,E2represents elastic modulus of the spring in the revised Kelvin unit, andη2andrrepresent viscous coefficient and fractional derivative of the dashpot in the revised Kelvin unit, respectively.

3 Test Results and Fitting

The CACB creep compliance curves were obtained from the creep test on specimens with various aging times. The curves were fitted using Eq.(14). The fitting results are illustrated in Fig.2 and the fitting parameters are displayed in Table 4.

The fitting results demonstrate that Eq.(14) exhibited high fitting precision, and the Burgers model revised by fractional derivative effectively described the mechanical properties of the CACB after aging.

Fig.2 Fitting of creep curves

Table 4 Major fitting parameters of Burgers model revised by fractional derivative

Aging time(h)A/C = 1.0A/C = 1.2A/C = 1.4E1(MPa)η1 (GPa·s)rE1(MPa)η1 (GPa·s)rE1(MPa)η1 (GPa·s)r0187.8498.60.179 6142.2374.60.198 171.5114.90.235 14215.4716.00.169 0184.5496.30.175 1121.4185.30.193 68338.61 125.00.144 9309.4725.10.152 3223.7301.30.172 112689.31 334.20.139 5627.91 058.00.146 2405.3642.90.167 1181 167.11 701.30.129 21 017.11 472.60.139 9788.01 280.50.162 8241 722.02 732.00.125 71 593.82 541.00.136 61 301.72 162.30.157 4

4 Analysis of Test and Fitting Results

4.1 Effects of Aging on Deformability

The physical significance of the creep complianceJ(t) is represented by the material strain under a unit load at an action time of a certain loadt, and can characterize the material deformability. For the CACB with anA/Cof 1.4, the creep compliance at 200, 400, 600, 800, and 1200 s are illustrated in Fig.3, which shows that the material deformability decreased sharply as aging time increased. By reference to the creep compliance curve characteristic of an increasing function of load duration, it is found that the creep compliance increased with increasing load duration for the material that underwent the same aging time. Fig.3 further indicates that as the aging time increased, the material creep compliance gap at different load durations narrowed and the compliance values tended to be close, demonstrating that the material deformability decreased through the aging process.

Fig.3 Effects of aging on creep compliance

According to the fitting result of Eq.(14), the percentages of the instant elastic, viscous flow, and delayed elastic deformations in the viscoelastic deformation can be analyzed. For the CACB with anA/Cof 1.4, the effects of aging time on the relative proportions of the three deformation types at a load duration oft=1 200 s are shown in Fig.4.

Fig.4 Effects of aging on deforming properties

Fig.4 illustrates that with increasing aging time, the proportions of the instant elastic and viscous flow deformations were reduced, but that of the delayed elastic deformation increased. Furthermore, both the instant and delayed elastic deformation are essentially recoverable elastic deformation, thus their sum characterizes the material elastic component. The proportion of the elastic deformation sum increased obviously with aging time, which indicates that as asphalt aging deepened, the CACB deformability tended to be elastic-like.

According to Figs.3-4, the curves develop fully in two stages according to the aging time. When the aging time was within 12 h, the creep compliance and various deformation components varied sharply as aging time increased. When the aging time was more than 12 h, the variation trend of the curves slowed down. During the initial aging stage, a severe oxidation reaction occurred in the asphalt and a large amount of saturates and aromatics were converted into resins and asphaltenes, causing an obvious change in material properties. As the aging process continued, the light components in the asphalt reduced sharply and the oxidation reaction degree weakened, which made the material viscoelastic properties stable.

As asphalt aging progressed, the mechanical properties of the asphalt in the CACB changed from viscoelastic to elastic, which contributed to CACB stiffening. Furthermore, as asphalt aging deepened, the CACB deformability decreased and its mechanical properties tended to be elastic-like.

4.2 Effects of Aging on Parameters

According to the creep compliance fitting results of creep compliance, three major material parameters of the CACB with different mix proportions under various aging times were obtained, namelyE1,η1, andr, and their changes with increasing aging time are illustrated in Fig.5, which demonstrates that as aging time increased,E1andη1of the CACB increased, indicating that as asphalt aging deepened, the ability of the material to resist instant elastic and viscous flow deformation was enhanced and its deformability weakened. Moreover,rof the CACB decreased with increasing aging time, which shows that the elastic component proportion in the material increased and that of the viscous component was reduced, and the CACB properties tended to be elastic-like.

Fig.5 Effect of aging time on material parameters

Heat aging caused both the elastic and viscous deformability of the material to decrease, but the viscous deformability declined more sharply, thus the CACB tended to be elastic-like. For example, for the CACB with anA/Cof 1.2,E1,η1, andrwere 309.41 MPa, 725.14 GPa·s, and 0.152 3 at an aging time of 8 h, respectively, while these values were 1 593.8 MPa, 2 541.1 GPa·s, and 0.136 6 at an aging time of 24 h, respectively. Therefore, aging had a significant influence on the viscoelastic mechanical parameters of the CACB.

Fig.5 also indicates that the parameter variation trends developed fully in two stages according to the aging time, with the dividing point being an aging time of 12 h. This rule is in agreement with the above conclusion.

4.3 Effects of Asphalt Content on Deformability

Changes in elastic deformation in the sum deformation of the composite binder with various asphalt contents at a load duration oft=1 200 s are shown in Fig.6.

Fig.6 illustrates that when the aging time was within 12 h, the elastic deformation proportion in the sum creep deformation reduced with increasing asphalt content, while the viscous deformation proportion increased. The figure shows that within a certain aging time range, increasing asphalt content improved CACB deformability, which subsequently improved anti-aging ability. Thus, the CACB may still exhibit effective deformability after aging.

Fig.6 Effects of asphalt application on material deforming properties

When the aging time was over 12 h, the elastic deformation proportion in the sum creep deformation of the CACB with different asphalt contents was basically the same, which demonstrates that with a long aging time, asphalt content had less influence on the material deformability. Furthermore, increasing the asphalt content in this condition was not effective in improving the CACB deformability. When aging lasted for a long time, the CACB tended to become elastic-like due to the deep aging of the asphalt, thus asphalt content had less influence on the material deformability.

4.4 Effects of Asphalt Content on Parameters

The material parameters of the CACB with various asphalt contents after the same aging times are illustrated in Fig.7, which shows that after the same aging time,E1andη1of the CACB decreased with increasing asphalt content, the ability of the material to resist instant and shear deformation weakened, and the material deformability was enhanced. After the same aging time,rof the CACB increased with increasing asphalt content, and the viscous component proportion in the material also increased. For example, after 18 h of aging,rof the CACB with anA/Cof 1.0, 1.2, and 1.4 was 0.129 2, 0.139 9, and 0.162 8, respectively,E1was 1 167.1, 1 017.1, and 787.96 MPa, respectively, andη1was 1 701.3, 1 472.6, and 1 280.5 GPa·s, respectively. Taking aging into consideration, the material could be more viscoelastic, and its deformability could be maintained at a high level by increasing the asphalt content appropriately.

Fig.7 Effects of asphalt application on material parameters

4.5 Aging Characterization and Its Evolution Model

The fractional derivativerreflects the effect nature of aging on viscoelastic properties of the CACB to a large extent, which is the most representative parameter. Thus, an aging influence factor of CACB was established based on parameterr, which is defined as

(15)

whereDtrepresents aging influence factor of CACB afterttime of aging,rtrepresents fractional derivative oftof CACB afterttime of aging, andr0represents fractional derivative oftof CACB without aging.

Becauserreduced with increasing aging time, the range ofDtwas [0, 1] according to its definition formula, andDtmonotonically increased with increasing aging time. A largerDtindicates that after aging, the viscous component of CACB and deformability reduced more, and aging affected the material properties more seriously.

TheDtof CACB after different aging times were calculated by theirr, which are listed in Table 5.

Table 5 Aging influence factor of CACB

The evolution models of CACB aging characterization were established by the fitting of the aging influence factor curves. Three functions used in fitting are as below:

a) Power function:

y=axb

(16)

b) Exponential function:

y=c1(1-e-x/d1)+c2(1-e-x/d2)

(17)

c) Rational function:

(18)

wherea,b,c1,c2,d1,d2,f,g,hare undetermined parameters.

The fit curves of the CACB aging influence factor are illustrated in Fig.8, and the fitting parameters of power function, exponential function, and rational function are listed in Tables 6-8.

Fig.8 The fit curves of the CACB aging influence factor

Table 6 Power function fit curves of aging influence factor

A/CabR21.00.044 320.622 830.856 221.20.083 130.432 380.851 751.40.132 240.296 850.875 22

Table 7 Exponential function fit curves of aging influence factor

A/Cc1d1c2d2R21.00.183 8711.832 40.183 8711.830 50.863 051.20.164 927.318 390.164 927.318 090.956 581.40.162 354.976 410.162 354.976 230.983 20

Table 8 Rational function fit curves of aging influence factor

A/CfghR21.00.314 097.084 412.252 690.979 161.20.321 225.237 541.992 840.995 711.40.340 223.766 641.598 590.994 75

According to Tables 6-8, rational function fitted the CACB aging influence factor curve best, which is appropriate for the evolution model of CACB aging characterization. Fitting accuracy of three functions all increased with increasing asphalt content, which is in agreement with the conclusion proposed in the fitting part of the constitutive model.

In summary, the aging influence factorDcan characterize the effects of aging time and asphalt content well, which is applicable for the CACB aging characterization. The rational function fitted the CACB aging influence factor curve well, which is appropriate for the evolution model of CACB aging characterization.

5 Conclusions

1) As aging time increased, the creep compliance of the CACB decreased and its deformability weakened. As asphalt aging deepened, the instant elastic and viscous flow deformation proportions decreased, while as the delayed elastic deformation proportion increased, and the CACB deforming properties tended to be elastic-like.

2) As aging time increased,E1andη1of the CACB increased, whilerdecreased. As asphalt aging deepened, the ability of the material to resist instant elastic and viscous flow deformation was enhanced, the elastic component proportion in the material increased while the viscous component reduced, and the material deformability weakened.

3) The effects of asphalt aging on the CACB properties were related to aging time and developed into two stages. During the initial stage (aging time within 12 h), the aging time significantly influenced the viscoelastic mechanical properties of the material. During the latter stage (aging time over 12 h), aging time had less influence on the viscoelastic mechanical properties of the material.

4) During the initial aging stage, increasing asphalt content appropriately improved the CACB deformability. During the latter aging stage, the CACB deformability with various asphalt contents was basically the same.

5) The aging influence factorDcan characterize the effects of aging time and asphalt content well, the rational function fitted the CACB aging influence factor curve well, which is appropriate for the evolution model of CACB aging characterization.