Evaluation of anti-icing performance for crumb rubber and diatomite compound modified asphalt mixture

Construction and Building Materials 107 (2016) 109–116

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Evaluation of anti-icing performance for crumb rubber and diatomite compound modified asphalt mixture Haibin Wei, Qiuqi He, Yubo Jiao ⇑, Jiafeng Chen, Maoxu Hu College of Transportation, Jilin University, Changchun 130025, PR China

h i g h l i g h t s
Crumb rubber and diatomite compound modified asphalt mixture was prepared.
Mechanical properties of compound modified mixture were improved.
Evaluation of ice breaking rate by image processing technology was satisfactory.
Correlations of different factors with ice breaking rate were investigated.

a r t i c l e

i n f o

Article history: Received 10 October 2015 Received in revised form 24 December 2015 Accepted 5 January 2016

Keywords: Crumb rubber Diatomite Anti-icing Image processing Correlation analysis

a b s t r a c t In this paper, anti-icing performance of crumb rubber and diatomite compound modified asphalt mixture was evaluated. Firstly, compound modified mixture was prepared in laboratory. Corresponding volumetric parameters and mechanical properties were compared and verified. Secondly, anti-icing test specimens were produced and measured by self-designed apparatus. Ice breaking rate calculated from image processing technology was adopted to assess the anti-icing effect, and correlation analysis was used to evaluate the influence of factors including rolling time, ice thickness, testing temperature and load frequency. Finally, the effects of these factors on anti-icing performance of modified mixture were investigated. Results indicate that the mechanical properties of control asphalt mixture have been improved by crumb rubber and diatomite. The correlation of rolling time for ice breaking rate is strong positive, while it is strong negative of ice thickness and very strong positive of test temperature. Weak correlation between load frequency and ice breaking rate is also observed. The results can provide a reference for anti-icing pavement design using crumb rubber and diatomite compound modified asphalt mixture. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Bituminous materials have been widely used in highway pavement, and got satisfactory application effects [1]. As for conventional hot asphalt mixture, it should possess favorable properties to resist the influence of heavy traffic and environmental factors. However, the effect on durability of hot asphalt mixture is profound because of its viscoelastic property. The consequences are rutting or permanent deformation at high temperature, cracking at low temperature and fatigue at moderate temperature [2,3]. In order to overcome these deficiencies, various methods have been used to modify the asphalt by improving its viscoelastic behavior.

⇑ Corresponding author at: No. 5988, Renmin Street, City of Changchun, Jilin Province, PR China. E-mail address: [email protected] (Y. Jiao). http://dx.doi.org/10.1016/j.conbuildmat.2016.01.003 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.

Diatomite is a widely used mineral with low cost and considerable storage, which has high absorptive capacity and stability [4–6]. It has been used to improve the performance of asphalt. Cong et al. [7] investigated the effects of diatomite on properties of asphalt. The results indicated that no chemical reaction was found between diatomite and asphalt. Both viscosity and complex modulus increased after modification at high temperature. However, the complex modulus decreased below 5 °C, which would result in low temperature cracking of asphalt mixture. Song et al. [8] studied the absorption rule of diatomite and asphalt. The results suggested that diatomite could effectively absorb lower molecular group and form anchorage structure, which improved the property of asphalt. Cheng et al. [9] investigated the effects of diatomite on aging properties of asphalt. The results indicated that diatomite could improve the high temperature stability and antiaging property. However, it had adverse effects on ductility and low temperature property.

110

H. Wei et al. / Construction and Building Materials 107 (2016) 109–116

As shown from above results, diatomite-modified asphalt mixture has better high temperature performance and thermal physical property. However, its low temperature performance is negatively affected. Therefore, the pavement using diatomitemodified asphalt mixture is more easily to emerge low temperature cracking. This phenomenon will seriously influence the application of diatomite-modified asphalt mixture in low temperature areas such as seasonal frozen regions [3]. Crumb rubber is another bitumen modifier, which is produced from end-of-life tyres (ELTs). The increasing number of vehicles in worldwide generates millions of used tyres each year. The amount of ELTs is 290 million in 2003 in US, 355 million each year in Europe, and 200 million in 2010 in China. Moreover, the annual growth rate of used tyres is over 10% [10–13]. Millions of used tyres have been illegally disposed, which may pose great threat to human health and cause serious environmental risks. In recent years, crumb rubber obtained from ELTs is used in asphalt mixture, which can improve its low temperature, high temperature and fatigue performances [14–16]. Moreno et al. [17] analyzed the effect of crumb rubber modifier on asphalt. The results showed that crumb rubber modifier increased the stiffness modulus and creep modulus and improved the resistance to plastic deformation of bitumen. Xiang et al. [18] investigated the properties of crumb rubber on performance of modified asphalt. The results indicated that the crumb rubber modified asphalt have better performance than matrix asphalt. Dias et al. [19] evaluated the mechanical response of gap-graded asphalt rubber mixture produced through dry process. Comparative analysis with reference mixture indicated that asphalt rubber mixture not only is less sensitive to high temperature but also improves fatigue cracking performance. Palit et al. [20] compared the performances of crumb rubber modified asphalt mix with control one. It was found that the modified mix has better fatigue and permanent deformation properties, lower temperature susceptibility and greater resistance to moisture damage. Lots of studies have been conducted on the properties of diatomite-modified asphalt and crumb rubber-modified asphalt. However, the study on crumb rubber and diatomite compound modified asphalt is still limited. A more favorable pavement material can be obtained if the compound modified asphalt possesses the advantages of both diatomite and crumb rubber modified ones. Moreover, ice on pavement surface in cold winter is a great threat to traffic safety. Investigation data indicates that 15–30% of the traffic accidents are caused by snow and ice [12,21]. There-

Table 1 Physical properties of neat asphalt.

a

Property

Value

Technical criterion

Penetration(25 °C, 0.1 mm) Penetration index (PI) Softening point TR&B (°C) Ductility (15 °C, cm) Flash point (°C) Specific gravity (15 °C, g/cm3)

93 0.83 44.0 165.1 277 1.090

80–100 1.5 to +1.0 P42 P100 P245 –

After TFOT Mass loss (%) Penetration ratio (25 °C, %) Age ductility (10 °C, cm)

0.2 62.4 10

6±0.8 P57 P8

a

(%)

fore, anti-icing pavement is important and widely used in the world [21–24]. According to research results by Zhou and Tan [25], pavement using crumb rubber modified asphalt has better anti-icing performance than conventional one. In this paper, diatomite and crumb rubber compound modified asphalt mixture was prepared. Changes of volumetric parameters and mechanical properties of asphalt mixture before and after modification were investigated. Using image processing technology and correlation analysis method, the effects of rolling time, ice thickness, testing temperature and load frequency on antiicing performance of modified mixture were discussed and analyzed. 2. Materials 2.1. Raw materials Physical properties of AH-90 neat asphalt were tested and given in Table 1. Andesite mineral aggregate was chosen for its excellent adhesion with asphalt, and the physical properties of which were listed in Table 2. The mineral filler used was limestone powder. Physical properties of filler were shown in Table 3. Physical properties and particle distribution for diatomite were listed in Tables 4 and 5. Crumb rubber particle was obtained from the waste tire rubber, whose properties were shown in Table 6. 2.2. Mixture preparation Stone Mastic Asphalt (SMA) was widely applied in asphalt road construction for its good performance. Thus, the SMA-13 gradation was used in this investigation. The Course Aggregate Void Filling (CAVF) method was employed in the mixture design procedure for SMA-13 [25]. The design gradation was shown in Fig. 1. The diatomite content 15% by weight of asphalt was validated through experiment. And the crumb rubber content 3% by weight of aggregate was determined by volumetric parameters and Marshall stability of mixtures. In this research, the optimum asphalt aggregate ratio for the crumb rubber and diatomite compound modified mixtures was 6.3% according to Marshall design method. In order to make the crumb rubber and diatomite dispersed homogeneous in the mixture, the blend process was optimized. Firstly, aggregates and crumb rubbers are mixed in mixing pot for 90s in order to make crumb rubber fully dispersed in aggregate. That is because the density of crumb rubber is lower than aggregate, which is easy to agglomerate in the process of mixing, and lead to the reduction of workability for mixture. Then, asphalt is added into aggregate and crumb rubber for mixing 90s, crumb rubber modified asphalt mixture is obtained. According to our

Table 3 Physical properties of mineral powder. Property

Value

Hydrophilic coefficient

Apparent density (g/cm3)

Gradation Sieve size (mm)

Passing (%)

0.80

2.741

0.6 0.15 0.075

100 95.4 86.3

The technical criterion was based on Standard JTG F40-2004 [3].

Table 2 Properties of aggregate. Sieve size (mm) Apparent density (g/cm3) Bulk density (g/cm3)

13.2 2.866 2.795

9.5 2.867 2.785

4.75 2.894 2.783

2.36 2.705 –

1.18 2.640 –

0.6 2.697 –

0.3 2.637 –

0.15 2.502 –

0.075 2.529 –

111

H. Wei et al. / Construction and Building Materials 107 (2016) 109–116 Table 4 Properties of diatomite. Property

Color

PH

Specific gravity (g/cm3)

Bulk density (g/cm3)

Loss on ignition (%)

Content of SiO2 (%)

Content of Fe3O2 (%)

Value

Orange

8–10

2.152

0.37

60.25

P87.1

61.1

Table 5 Particle distribution of diatomite. Particle size (lm)

<5

10–5

20–10

40–20

>40

Percentage (%)

19.3

28.0

20.7

21.0

11.0

Table 6 Properties of crumb rubber particle. Property

Value

Sauer A hardness (°)

Apparent density (g/cm3)

Gradation Sieve size (mm)

Passing (%)

60

1.321

4.75 2.36 1.18

0.6 98.6 0.8

Fig. 2. Optimum blend process for crumb rubber and diatomite compound modified mixture.

Fig. 1. CAVF design gradation.

laboratory test results, distribution of asphalt, crumb rubber and aggregate is more uniform with the increasing of mixing time. The air void of asphalt mixture is the minimum when the mixing time is about 90s. If the mixing time continues to increase, mixture property declines because of aging of asphalt. Finally, filler and diatomite are mixed with crumb rubber modified asphalt mixture for mixing 90s, and diatomite and crumb rubber compound modified asphalt mixture is obtained. The reason for mixing 90s is similar with the preparation of crumb rubber modified mixture. If the time is too short, components are mixed uneven; on the contrary, asphalt is prone to aging. The optimum blend process was shown in Fig. 2. The compactions of Marshall samples (101 mm diameter and 63.5 mm height) and square slab specimens (300 mm  300 mm  50 mm) were performed by Marshall hammer method and Wheel Rolling Method, respectively [28]. Considering that adding of crumb rubber makes the compaction of specimens more difficult. Therefore, a two times molding process was adopted. For the compaction of Marshall samples, modified asphalt mixture was compacted for 25 times at each side. When the temperature of specimen was at 100 °C, the second compaction was taken and 50 times for each side. For slab specimen, modified asphalt mixture was firstly compacted for 4 times to make specimen smooth. Secondly, specimen was turned direction and compacted for 8 times. Finally, another 16 times compactions were taken at 100 °C of specimen.

Fig. 3. Anti-icing test specimen. 3. Experimental methods 3.1. Testing procedure In order to evaluate the anti-icing performance of crumb rubber and diatomite compound modified mixture, three testing specimens were firstly produced for each test. The production processes are listed as follows:
An ice specimen (size 120 mm  260 mm and thickness is variable) was made on a glass plate;

112

H. Wei et al. / Construction and Building Materials 107 (2016) 109–116
a slab specimen (size 300 mm  300 mm  50 mm) was compacted by rolling compactor and placed in the testing temperature for pre-freezing;
the ice specimen was taken from glass plate and placed on the pre-freezing slab specimen. Then, testing specimens was made by bonding ice and slab specimens together using water under freezing condition. The anti-icing test specimen is shown in Fig. 3. After the production of testing specimens, anti-icing experiments were conducted by a self-designed testing apparatus which can not only realize the temperature control but also apply dynamic load [26]. Dynamic load system of this apparatus (shown in Fig. 4) is consisting of wheel, weight block, reciprocating shaft and lifting device, etc. Rubber material is used for wheel. Weight block is applied to generate pressure; the pressure value for this apparatus is 0.7 MPa ± 0.05Mpa. Motion of rubber wheel along reciprocating shaft can be used to simulate vehicle motion. Motion frequency of wheel is variable, which can be adjusted through frequency converter. Moreover, internal temperature of this apparatus can be monitored and controlled by thermometer and air compressor, respectively. Anti-icing tests were carried out for different conditions of specimens. Rolling direction was along the long axis of ice specimen. Anti-icing pictures were collected by digital camera at the end of each experiment. Anti-icing field test is shown in Fig. 5. Fig. 4. Dynamic loading system. 3.2. Image processing technology In order to evaluate the anti-icing effect of crumb rubber and diatomite compound modified asphalt mixture, appropriate evaluation method and index for anti-icing performance are needed. Taking into account the time of laboratory test is relatively short, and the ice specimen does not appear large area of spalling. Therefore, image processing technology is adopted to obtain the ice breaking rate on surface of specimen. The detailed procedure is as follows: Firstly, the complete image for ice specimen was obtained (shown in Fig. 6a); Secondly, the damaged sections were identified through edge detection algorithm and the edge detection image was obtained (shown in Fig. 6b); Thirdly, Gauss algorithm for noise reduction was adopted to eliminate the noise in edge detection image and corresponding noise reduction image was obtained (shown in Fig. 6c); Finally, ice breaking rate can be calculated by percentage of white area in noise reduction image, which is shown in Eq. (1).

IBR ¼

AW  100% AI

ð1Þ

where IBR is the ice breaking rate; AW is area of white part in noise reduction image, while AI is area of entire image. In this paper, ice breaking rate is calculated at the end of each experiment. Three ice specimens are tested for each experiment, and the mean value is adopted to evaluate the anti-icing effect of asphalt mixture. 3.3. Correlation analysis Fig. 5. Anti-icing field test.

Correlation analysis has been widely used to assess the correlation between variables [27,28]. Assuming x ¼ ðx1 ; x2 ; . . . ; xn Þ and y ¼ ðy1 ; y2 ; . . . ; yn Þ are vectors with n parameters, n is the number of samples. Correlation coefficient r between x and y can be calculated by

Fig. 6. Procedure of image processing for ice specimen (unit: mm).

113

H. Wei et al. / Construction and Building Materials 107 (2016) 109–116 Table 7 Volumetric parameters and mechanical properties for control mixture, crumb rubber modified mixture and diatomite/crumb rubber compound modified mixture. Mixture type

Specific gravity (g/cm3)

Air void (%)

VMA (%)

VFA (%)

Marshall stability (kN)

Flow value (0.1 mm)

Control mixture Crumb rubber modified mixture Compound modified mixture

2.490 2.448 2.457

3.8 5.5 5.2

14.1 15.8 16.3

73.0 65.2 68.1

9.56 8.54 10.25

26.4 42.5 40.5

Pn i¼1 ðxi  xÞðyi  yÞ r ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pn 2 Pn 2 i¼1 ðxi  xÞ i¼1 ðyi  yÞ

ð2Þ

where xi and yi (i ¼ 1; 2; . . . ; n) are the ith sample of x and y, respectively; x and y are the mean values of x and y, respectively. The correlation coefficient r is a measure of relationship between two variables. Values of r are always between 1 and +1. A correlation coefficient between 0 and +1 indicates that two variables are in positive correlation, while a correlation coefficient between 1 and 0 indicates that two variables are in negative correlation. When r is 0, the variables have no correlation; r is 1, they are in perfectly positive correlation and r is 1, they are in perfectly negative correlation. Correlation is an effect size which can describe the strength of correlation using the following rules: If If If If If

the absolute value of r is between 0.00 and 0.19, correlation is ‘‘very weak”. the absolute value of r is between 0.20 and 0.39, correlation is ‘‘weak”. the absolute value of r is between 0.40 and 0.59, correlation is ‘‘moderate”. the absolute value of r is between 0.60 and 0.79, correlation is ‘‘strong”. the absolute value of r is between 0.80 and 1.00, correlation is ‘‘very strong”.

After the acquisition of correlation coefficient r; significance test need to be performed which can be used to suggest the evidence of correlation in the samples. Null hypothesis H0 is tested, there is no correlation in the sample; On the contrary, the alternative hypothesis H1 ; that there is correlation. Testing data will indicate the correctness of these opposite hypothesis. They can be expressed by



H0 : p ¼ 0 H1 : p–0

ð3Þ

4. Results and discussion 4.1. Mixture properties In order to investigate the properties of control mixture, crumb rubber modified mixture and crumb rubber and diatomite compound modified mixture, volumetric parameters and mechanical properties were tested. Firstly, content 3% of crumb rubber by weight of aggregate was added to modify the control asphalt mixture, volumetric parameters and mechanical properties before and after modification are listed in Table 7. As shown in Table 7, adding of crumb rubber into control asphalt mixture makes specific gravity, air void, VMA and flow value increase, while VFA and Marshall stability decrease. The reason lies in that adding of crumb rubber makes the skeleton structure of SMA mixture change from ‘‘stone-stone” into ‘‘stone-crumb rubber-stone”. Because crumb rubber is a kind of high elastic material, compactions of Marshall specimens are difficult. Therefore, air void and VMA are increased by addition of crumb rubber, which furtherly lead to the decrease of VFA under the same content of asphalt. ‘‘Stone-crumb rubber-stone” structure possesses lower stability, but higher elasticity, which causes the decrease of Marshall stability, but increase of flow value. Secondly, compound modified asphalt mixture with content 15% of diatomite by weight of asphalt and content 3% of crumb rubber by weight of aggregate was prepared according to the process shown in Fig. 2. Corresponding volumetric parameters and mechanical properties are also listed in Table 7. As shown in this table, air void increases, VMA decreases and VFA increases because of adding of diatomite. It is because the specific surface area of diatomite is large and it also has good

adsorption performance, which increase the structural asphalt and decrease the free one. Therefore, structural asphalt is more evenly filled in the aggregates. Moreover, modification effect of diatomite improves the consistency of asphalt, which makes the compound modified mixture with higher bonding strength. Thus, Marshall stability increases and flow value decreases after the adding of diatomite comparing with asphalt mixture modified by crumb rubber. However, Marshall stability and flow value have been effectively promoted comparing with the control mixture.

4.2. Effect of rolling time Testing temperature was 5 °C, ice thickness was 2 mm and load frequency was 40 times/min. Anti-icing tests under different rolling times were conducted. The anti-icing test of control mixture was used as the comparison test. After processing the ice breaking picture, ice breaking rates can be obtained. The results were listed in Fig. 7. As can be seen from the testing results, an initial breaking rate is presented for the ice specimen, which is because of the initial traces on ice surface during production process of specimen. The influence of rolling times can be evaluated by the growth of ice breaking rates. For matrix mixture, the ice breaking rate has no significant increase and ice specimen surface does not appear obvious breaking. As for the modified one, the ice breaking rate changes according to logarithmic rule. At the early stage, the growth is obvious, while it slows down at the middle stage, and it is stable at the last stage. The reasons for these phenomena lie in that crumb rubber is a kind of polymer material, which possesses high elasticity. Coarse aggregate is a kind of high stiffness material, adding of crumb rubber into control mixture improves its elastic property. Moreover, ice specimen produces a deformation under external load. The deformation ability for control mixture is weak before adding of crumb rubber. Ice specimen is not prone to break. However, the deformation ability is improved after adding of crumb rubber. Therefore, deformation between ice specimen and compound modified mixture is uncoordinated. Ice specimen is broken under the action of repeated deformation. Additionally, stress

Fig. 7. Relationship between rolling times and ice breaking rate.

114

H. Wei et al. / Construction and Building Materials 107 (2016) 109–116

Table 8 Correlations between rolling times and ice breaking rate. Factors Rolling times

Ice breaking rate

*

Table 9 Correlations between ice thickness and ice breaking rate.

Rolling times

Ice breaking rate

Factors

Pearson correlation coefficient Significance (2-tailed) N

1

0.668*

Ice thickness

9

0.049 9

Pearson correlation coefficient Significance (2-tailed) N

0.668*

1

0.049 9

9

Ice breaking rate

**

Significantly correlated at the level of 0.05 (2-tailed).

Ice thickness

Ice breaking rate

Pearson correlation coefficient Significance (2-tailed) N

1

0.955**

6

0.003 6

Pearson correlation coefficient Significance (2-tailed) N

0.955**

1

0.003 6

6

Significantly correlated at the 0.01 level (2-tailed).

20

20

Ice breaking rates (%)

Ice breaking rates (%)

18 16 12 8

y = 30.398e-0.475x R² = 0.9724

4 0

0

2

16 14

y = 1.609x + 18.585 R² = 0.956

12 10 8 6 4 2

4

6

Ice thickness (mm)

0 -12

-8

-4

Test temperature (

0

)

Fig. 8. Relationship between ice thickness and ice breaking rate.

Fig. 9. Relationship between test temperature and ice breaking rate.

concentration caused by crumb rubber is also an important cause for ice breaking. Uncoordinated deformation is obvious when rolling time is lower. However, cumulative fatigue deformation of ice specimen increases with increasing of rolling time. Therefore, ice breaking rate tends to be stable gradually. Correlation analysis were conducted and listed in Table 8. As shown from the results, Pearson correlation coefficient between ice breaking rate and rolling times is 0.668, the strength of correlation is ‘‘strong”. The significance (2-tailed) is 0.049, they are significantly correlated at the level of 0.05 (2-tailed). It illustrates that there is strong positive correlation between ice breaking rate and rolling times.

cantly correlated at the level of 0.01 (2-tailed). It indicates that there is very strong negative correlation between ice breaking rate and rolling times.

4.3. Effect of ice thickness Testing temperature was 5 °C, load frequency was 40 times/ min, and rolling times were 1000 cycles. Ice thicknesses were 1 mm, 2 mm, 3 mm, 4 mm, 5 mm and 6 mm, respectively. The effect of ice thickness on ice breaking performance of crumb rubber and diatomite compound modified asphalt mixture was investigated and testing results were listed in Fig. 8. From the test results, ice breaking rate decreases gradually in an approximate index rule with the increasing of ice thickness. When the ice thickness reaches 5 mm, the ice breaking performance of compound modified asphalt mixture is not significant. The reasons for these phenomena lie in that the influences of stress concentration of crumb rubber and uncoordinated deformation between compound mixture and ice specimen is weakened with increase of ice thickness. Therefore, the possibility for ice breaking reduces. Correlation analysis results are listed in Table 9. The correlation analysis shows Pearson correlation coefficient between ice breaking rate and ice thickness is 0.955, the strength of correlation is ‘‘very strong”. The significance (2-tailed) is 0.003, they are signifi-

4.4. Effect of testing temperature Ice thickness was 2 mm, load frequency was 40 times/min, and the rolling times were 1000 cycles. Test temperatures were 1 °C, 3 °C, 5 °C, 7 °C, 9 °C and 11 °C, respectively. The effect of temperature on ice breaking performance of compound modified asphalt mixture was studied and the results were listed in Fig. 9. From the test results, ice breaking rate decreases gradually with the decreasing of the testing temperature. When the temperature is below 9 °C, anti-icing performance of modified mixture is negligible. The reasons for these phenomena lie in that the strength of ice specimen is lower when the temperature is high. Ice specimen is more prone to break under the coupled effects of stress and deformation. On the contrary, strength of ice specimen increases with the decreasing of temperature, which lead to the ice specimen not easy to break. Correlation analysis results are listed in Table 10. The correlation analysis shows that the Pearson correlation coefficient of ice breaking rate and test temperature is 0.979, the strength of correlation is ‘‘very strong”. The significance (2-tailed) is 0.001, they are significantly correlated at the level of 0.01 (2-tailed). It indicates that there is very strong positive correlation between ice breaking rate and rolling times. 4.5. Effect of load frequency Test temperature was 5 °C, ice thickness was 2 mm, and the rolling times were 1000 cycles. Load frequencies were 20 times/

115

H. Wei et al. / Construction and Building Materials 107 (2016) 109–116 Table 10 Correlations between test temperature and ice breaking rate. Factors

**

Test temperature

Ice breaking rate

1

0.979** 0.001 6

Test temperature

Pearson correlation coefficient Significance (2-tailed) N

6

Ice breaking rate

Pearson correlation coefficient Significance (2-tailed) N

0.979** 0.001 6

1 6

Significantly correlated at the 0.01 level (2-tailed).

Fig. 10. Relationship between load frequency and ice breaking rate.

Table 11 Correlations between load frequency and ice breaking rate. Factors

Load frequency

Ice breaking rate

1

0.394 0.440 6

Load frequency

Pearson correlation coefficient Significance (2-tailed) N

6

Ice breaking rate

Pearson correlation coefficient Significance (2-tailed) N

0.394 0.440 6

Adding of crumb rubber into SMA mixture changes its skeleton structure, which causes the decreasing of Marshall stability, but increasing of flow value. Due to the good absorption performance of diatomite, compound modification of crumb rubber and diatomite will improve the Marshall stability and flow value of mixture comparing with the control one. Based on the acquisition of original ice breaking image, edge detection algorithm and Gauss noise reduction method can be used to calculate the ice breaking rate of specimen. As for the influences of rolling time, ice thickness, testing temperature and load frequency, the correlation of rolling time for ice breaking rate is strong positive, while it is strong negative of ice thickness and very strong positive of test temperature. As for the correlation between short-term effect of load frequency and ice breaking rate, it is weak. In summary, mechanical properties of control asphalt mixture can be improved by crumb rubber and diatomite. Therefore, crumb rubber and diatomite compound modified asphalt mixture can be used for pavement construction. Results of anti-icing performance for compound modified asphalt mixture, which can provide a reference for anti-icing pavement design in cold regions. Additionally, more research will be conducted to demonstrate the effect of longterm effect of load frequency in future. Acknowledgments

1 6

min, 30 times/min, 40 times/min, 50 times/min, 60 times/min and 70 times/min, respectively. The effect of different load frequencies on ice breaking performance of compound modified asphalt mixture was studied. The results were summarized in Fig. 10. From the test results, similar ice breaking situations under different load frequencies are observed. Ice breaking rates are between 12.15% and 12.35%, which are relatively stable. It reveals that the effect of load frequency on ice breaking rate is not obvious in this test. The reason for this phenomenon may lie in that the rolling times are fixed to be 1000 cycles, which is small and can only able to reflect the short-term effect of load frequency. Further research on effect of load frequency considering long-term effect will be investigated. Correlation analysis results are listed in Table 11. The correlation analysis shows that Pearson correlation coefficient of ice breaking rate and load frequency is 0.394, the strength of correlation is ‘‘weak”. The significance (2-tailed) is 0.440. The analysis shows that ice breaking rate has weak correlation with the load frequency. 5. Conclusions In this paper, crumb rubber and diatomite compound modified asphalt mixture was prepared and corresponding performances were investigated. The following conclusions can be obtained.

The authors express their appreciation for financial supports of National Natural Science Foundation of China under Grants nos. 51578263 and 51408258; China Postdoctoral Science Foundation funded project (nos. 2014M560237 and 2015T80305); Fundamental Research Funds for the Central Universities and Science (JCKYQKJC06), and Technology Development Program of Jilin Province. References [1] Q. Guo, Y. Bian, L. Li, Y. Jiao, J. Tao, C. Xiang, Stereological estimation of aggregate gradation using digital image of asphalt mixture, Constr. Build. Mater. 94 (2015) 458–466. [2] F.M. Nejad, P. Aghajani, A. Modarres, H. Firoozifar, Investigating the properties of crumb rubber modified bitumen using classic and SHRP testing methods, Constr. Build. Mater. 26 (1) (2012) 481–489. [3] Q. Guo, L. Li, Y. Cheng, Y. Jiao, C. Xu, Laboratory evaluation on performance of diatomite and glass fiber compound modified asphalt mixture, Mater Des. 66 (2015) 51–59. [4] S. Akin, J.M. Schembre, S.K. Bhat, A.R. Kovscek, Spontaneous imbibition characteristics of diatomite, J. Petrol. Sci. Eng. 25 (3–4) (2000) 149–165. [5] M. Reguerio, J.P. Calvo, E. Elizaga, V. Calderon, Spanish diatomite geology and economics, Ind. Minerals Rocks 306 (1993) 57–67. [6] Y.Q. Tan, L. Zhang, X.Y. Zhang, Investigation of low-temperature properties of diatomite-modified asphalt mixtures, Constr. Build. Mater. 36 (2012) 787–795. [7] P.L. Cong, S.F. Chen, H.X. Chen, Effects of diatomite on the properties of asphalt binder, Constr. Build. Mater. 30 (2012) 495–499. [8] Y. Song, J. Che, Y. Zhang, The interacting rule of diatomite and asphalt groups, Petrol. Sci. Technol. 29 (3) (2011) 254–259. [9] Y. Cheng, J. Tao, Y. Jiao, Q. Guo, C. Li, Influence of diatomite and mineral powder on thermal oxidative ageing properties of asphalt, Adv. Mater. Sci. Eng. (2015). [10] D.L. Presti, Recycled tyre rubber modified bitumens for road asphalt mixtures: a literature review, Constr. Build. Mater. 49 (2013) 863–881.

116

H. Wei et al. / Construction and Building Materials 107 (2016) 109–116

[11] H. Wang, Z. Dang, L. Li, Z. You, Analysis on fatigue crack growth laws for crumb rubber modified (CRM) asphalt mixture, Constr. Build. Mater. 47 (2013) 1342– 1349. [12] M. Zheng, J. Zhou, S. Wu, H. Yuan, J. Meng, Evaluation of long-term performance of anti-icing asphalt pavement, Constr. Build. Mater. 84 (2015) 277–283. [13] H. Liu, Z. Chen, W. Wang, H. Wang, P. Hao, Investigation of the rheological modification mechanism of crumb rubber modified asphalt (CRMA) containing TOR additive, Constr. Build. Mater. 67 (2014) 225–233. [14] M.B. Costa Liliana, M.R.D. Silva Hugo, R. Oliveira Joel, R.M. Fernandes Sara, Incorporation of waste plastic in asphalt binders to improve their performance in the pavement, Int. J. Pavement Res. Technol. 6 (4) (2013) 457–464. [15] C.T. Chiu, L.C. Lu, A laboratory study on stone matrix asphalt using ground tire rubber, Constr. Build. Mater. 21 (5) (2007) 1027–1033. [16] B. Yu, L. Jiao, F. Ni, J. Yang, Evaluation of plastic–rubber asphalt: engineering property and environmental concern, Constr. Build. Mater. 71 (2014) 416–424. [17] F. Moreno, M. Sol, J. Martín, M. Pérez, M.C. Rubio, The effect of crumb rubber modifier on the resistance of asphalt mixes to plastic deformation, Mater Des. 47 (2013) 274–280. [18] L. Xiang, J. Cheng, G. Que, Microstructure and performance of crumb rubber modified asphalt, Constr. Build. Mater. 23 (12) (2009) 3586–3590. [19] J.F. Dias, L.G. Picado-Santos, S.D. Capitão, Mechanical performance of dry process fine crumb rubber asphalt mixtures placed on the Portuguese road network, Constr. Build. Mater. 73 (2014) 247–254.

[20] S.K. Palit, K.S. Reddy, B.B. Pandey, Laboratory evaluation of crumb rubber modified asphalt mixes, J. Mater. Civil. Eng. 16 (1) (2004) 45–53. [21] L. Young, S. Durham, Performance of an anti-icing epoxy overlay on asphalt surfaces, J. Perform. Constr. Facil. 27 (6) (2013) 836–840. [22] Leahy R, Yang SSS, Vargas A, Schmalzer P, Effects of Deicing/Anti-Icing Chemicals (DIAICs) on Rubberized Asphalt Pavements (No. FHWA-AZ-14-617), 2014. [23] Z. Liu, M. Xing, S. Chen, R. He, P. Cong, Influence of the chloride-based antifreeze filler on the properties of asphalt mixtures, Constr. Build. Mater. 51 (2014) 133–140. [24] Z. Liu, S. Chen, R. He, M. Xing, Y. Bai, H. Dou, Investigation on the properties of asphalt mixtures containing antifreeze fillers, J. Mater. Civil. Eng. 27 (6) (2014) 04014180. [25] C.X. Zhang, Y.Q. Tan, Study on anti-icing performance of pavement containing a granular crumb rubber asphalt mixture, Road Mater. Pavement 10 (sup1) (2009) 281–294. [26] H. Wei, Y. Jiao, H. Liu, Effect of freeze–thaw cycles on mechanical property of silty clay modified by fly ash and crumb rubber, Cold Reg. Sci. Technol. 116 (2015) 70–77. [27] Y. Yu, G. Ma, Z. Hao, S. Wang, F. Niu, E. Zio, Correlation analysis for screening key parameters for passive system reliability analysis, Ann. Nucl. Energy 77 (2015) 23–29. [28] P. Xiao, W. Qu, H. Qi, Z. Li, Detecting DDoS attacks against data center with correlation analysis, Comput. Commun. 67 (2015) 66–74.