Experimental Research on High-Temperature Stability of Asphalt Concrete Panels of Impermeable Layers

Abstract

In order to study the slope stability of an impervious layer asphalt concrete panel, in this study, the maximum aggregate size used was 19 mm, and a slope flow value test was carried out after changing the gradation index, filler content and bitumen aggregate ratio. The test results showed that the relationship curve between the slope flow value and the test time was mainly divided into three stages for the slope flow value: an almost linear growth stage, a gradual stabilization stage, and a stable stage. The grading index, bitumen aggregate ratio and filler content had an effect on the slope flow value of asphalt concrete. The slope flow value decreased with the increase in the grading index. A reasonable increase in the grading index can increase the slope stability of the asphalt concrete panel. The slope flow value increased with the increase in the filler content and bitumen aggregate ratio. When the filler content exceeded 13%, the slope flow value significantly increased. At the same time, it was also verified that the asphalt concrete slope with the maximum aggregate size of 19 mm had good thermal stability. On this basis, a prediction model of asphalt concrete slope flow value and test time was established. The model considered the effect of different parameters of mix proportion on the slope flow value. The calculation results were in good agreement with the test results.

1. Study on the Background

Hydraulic asphalt concrete has a good impervious performance and strong deformation ability; thus, it was widely used in relevant projects for pumped storage power stations [1,2,3,4,5]. Asphalt concrete is used as an impervious structure in key water conservancy projects, such as those in the upper reservoir of Baoquan Pumped Storage Power Station and the upper reservoir of Tianhuangping Pumped Storage Power Station [6].

So far, the research on the characteristics of asphalt concrete panels and core walls is more extensive [7,8,9]. Zhu et al. [10] found that adjusting aggregate gradation, filler and asphalt content can effectively improve the overall performance of asphalt concrete; Guo Haipeng [11] found that WG-I warm mix agent can improve the performance of hydraulic asphalt concrete, and its effect is better; Ning et al. [12,13] studied the dynamic compression of asphalt concrete at different temperatures and established a calculation model of related properties.

In areas where pumped storage power stations were built or are under construction in China, the temperature in summer is high, and the temperature difference between day and night is large. The asphalt concrete panel of the impervious layer is exposed to the sun, and the surface temperature of the panel can reach more than 70 °C (up to 76 °C). The high temperature causes the asphalt mortar in the impervious layer to soften. This results in the aggregate separating from the asphalt, causing flow, and decreasing the impervious performance of the panel. Early impervious asphalt concrete projects focused on the imperviousness of their materials. A high asphalt content and low gradation index were adopted. The low content of coarse aggregate leads to the deterioration of the stability of skeleton structure formed by asphalt concrete. Under a certain slope, the asphalt mortar flowed and the slope thermal stability deteriorated. Therefore, it is of great significance to study the slope stability of asphalt concrete in combination with the engineering environment of asphalt concrete impervious layer plates.

Yingbo Zhang [14,15] studied the performance of asphalt concrete by changing the filler content, and found that an increase in the filler/asphalt ratio was unfavorable for the thermal stability of asphalt concrete; Xiao Meng [16] explored the influence mechanism of asphalt concrete slope flow through a slope flow test, and found that the influence of mineral material gradation, asphalt content and filler content on the slope thermal stability of asphalt concrete varies; Zenghong Liu [17] compared, using a slope flow test, asphalt concrete evaluated by the European Union and China, and found that the method of using a Marshall specimen to conduct a slope flow test in China is still convenient. If the mix ratio is appropriate, the slope flow value meets the specification requirements (less than 0.8 mm); Yizhou Cai [18] analyzed the factors influencing the slope thermal stability of impervious asphalt concrete panels and concluded that the asphalt content had the greatest influence on the slope flow value.

These research results prompted further research on the slope stability of asphalt concrete slabs. However, there are relatively few studies on the influence of asphalt concrete slope thermal stability with a high-grade index and large aggregate size, and there is a lack of relevant calculation models.

Therefore, in this study, we conducted a slope flow test, using a maximum aggregate particle size of 19 mm and the advanced distribution index (0.35~0.45). The test plan was formulated to carry out the slope flow value test. The variation law of slope stability of asphalt concrete during the test time was studied and analyzed, and a prediction model of slope flow value and test time of asphalt concrete was established. The aim was to provide a reference for the study of slope stability of impervious asphalt concrete slabs in pumped storage power stations in the future.

2. Raw Materials Testing

The raw materials of hydraulic asphalt concrete used in this test included aggregate, filler and asphalt. The coarse and fine aggregates used in the test were limestone (the aggregate size was between 0.075 and 19 mm), the filler was ground from limestone and the asphalt was Karamay 70 petroleum asphalt.

The raw materials of the hydraulic asphalt concrete, such as coarse aggregate, fine aggregate, filler and asphalt, were tested and analyzed. The quality test results are shown in

Table 1. Coarse aggregate quality test results table.

Technical Specifications	Surface Density (g/cm3)	Adhesion to Asphalt/Level	Durability/%	Heat Resistance	Water Absorption Rate/%
specification requirement	≥2.6	≥4	≤12	-	≤2
test results	2.738	5	0.4	eligible	0.45

,

Table 2. Test results of fine aggregate quality.

Technical Specifications	Surface Density (g/cm3)	Water Stability Grade/Level	Sodium Sulfate 5 Times Cycle Weight Loss/%	Water Absorption Rate/%	Organic Matter Content/%
specification requirement	≥2.55	≥6	≤15	≤2	≤2
test results	2.726	9	0.45	0.6	0

,

Table 3. Filler quality test results table.

Technical Specifications	Surface Density (g/cm3)	Water Content/%	Hydrophilic Coefficient	Filler Grading Screening Results/%
				0.075	0.15	0.6
specification requirement	≥2.5	≤0.5	≤1.0	>85	>90	>100
test results	2.732	0.38	0.72	98.7	99.5	100.0

and

Table 4. Karamay No. 70 Petroleum Asphalt Quality Test Result Table.

Item		Quality Index	Test Result
Needle penetration (25 °C, 0.1 mm)		60~80	70.0
crisp point (15 °C)		≤−10	−20.8
Latency (15 °C, 5 mm/min)		≥150	>150
Latency (4 °C, 1 mm/min)		≥10	23.0
softening point (/°C)		48~55	50.6
solubility/°C		≥99.0	99.7
flash point/°C		≥260	310.0
density 25 °C		actual measurement	0.987
wax content/%		≤2	1.8
After the film overn	Quality change	±0.8	−0.1
	residual needle penetration ratio	≥61	80.6
	Latency	≥6	28.3

.

The results from the tests reported in Table 1, Table 2, Table 3 and Table 4 show that the test indexes of coarse and fine aggregate, filler and asphalt all met the requirements of the current specification.

3. Results Test Scheme

3.1. Mix Design

Hydraulic asphalt concrete is a dispersion system, formed by heating and mixing asphalt mortar. This mortar is composed of coarse and fine aggregate that forms a skeleton, along with filler and asphalt. In this paper, the coarse and fine aggregates were broken limestone (the maximum particle size of the aggregate was 19 mm). The filler was made of limestone, and the asphalt was Kelamayi No.70 asphalt. After the performance test of the raw materials, the indicators of the materials used met the requirements of China’s current specifications [19,20]. The gradation index refers to the aggregate ratio of different particle size ranges (0.075~19 mm).

The design gradation of hydraulic asphalt concrete mineral materials used in this paper was calculated by Ding Purong Formula (1):

$$P_i=P_{0.074}+(100-P_{0.074})\frac{{(d_i)}^r-{(0.074)}^r}{{(D_{\max })}^r-{(0.074)}^r}$$

(1)

In the formula: d_i is the sieve diameter; P_i is the passing rate of aggregate when the aperture is d_i; P_0.074 is the passing rate of aggregate when the sieve is 0.074 mm; D_max is the maximum particle size of aggregate; r is the gradation index.

The gradation index of the mixture ratio selected in the test was 0.36, 0.39, 0.42 and 0.45; filler content and asphalt content were 9~15%; 6.6~7.5%, and the slope flow test of asphalt concrete with different mix ratios was carried out.

3.2. Test Device and Test Scheme

Test device: slope flow meter, as shown in Figure 1a, multimeter, oven (200 °C, indexing value of 1 °C, can automatically adjust the temperature), displacement device (minimum indexing value of 0.01 mm), copper wire, thumbtack, iron sheet, high performance AB glue.

The asphalt concrete specimens were Marshall specimens, as shown in Figure 1b. The specimen size was 63.5 ± 1.3 mm high and the inner diameter was 101 ± 0.2 mm. After preparation, the porosity (≤2%) and density met the requirements of relevant specifications [19].

Test plan: The slope stability of asphalt concrete was obtained by the slope flow test. The slope flow value was the index. At the same time, combined with the actual project and the requirements of the Chinese code [20], the prepared Marshall specimen was stuck to the slope flow meter with high-performance AB glue. The slope of the slope flow meter was set to 1: 1.7, the test temperature was 70 ± 0.5 °C, and the test time was 48 h. The displacement of the specimen at 50 mm was recorded at intervals of 0 h, 1 h, 2 h, 4 h, 6 h, 8 h, 14 h, 24 h, 36 h and 48 h, respectively, and the slope flow value was calculated. The slope stability of asphalt concrete was analyzed by slope flow value. The calculation formula of slope flow value is as follows:

$$U=U_0-U_i$$

(2)

In the formula: U is the slope flow value; U₀ is the initial reading; U_i is a constant temperature i hour reading.

The test of each group of mixed proportion selected 3~6 specimens for parallel test. In this paper, the relationship between the slope flow value and the test time was a typical curve, and the other charts were the average value of the three effective specimens.

4. Test Results and Analysis

The slope flow test was conducted on asphalt concrete specimens with various mix ratios, and the slope stability was analyzed. The study aimed to explore the influences of different mix indexes, filler contents and bitumen aggregate ratios on the slope flow value and its change over time.

4.1. Influence of Gradation Index on Slope Stability

Through the slope flow test, the relationship between the slope flow value of asphalt concrete with different gradation indexes and the test time was obtained, as shown in Figure 2.

It can be seen from Figure 2 that under different gradations, the change law of slope flow value was similar to that of test time, which was mainly divided into three stages: the slope flow value was almost linear growth stage; slope flow value gradually stable stage; slope flow value stable stage. That is to say, the slope flow value increased greatly in the first 6 h (the completion rate of slope flow value reached 70~85%), the change of slope flow value tended to be stable in 6~12 h (the completion rate of slope flow value reached 85~100%), and the slope flow value was basically unchanged in 12~48 h. The soil structure theory [21,22] can be used to explain the change of slope flow value and the three stages of test time.

(1) In the initial stage of the test, the temperature in the oven increased, the temperature of the asphalt concrete specimen itself increased from room temperature (16 °C) to 70 °C, the asphalt mortar softened, and the original structure of the asphalt concrete gradually deformed. The stability of the asphalt concrete was mainly the skeleton structure formed by the coarse aggregate and the cohesion of the asphalt. After the asphalt was softened, it played a lubricating role. Under a certain slope, the specimen was affected by its own gravity, the skeleton structure of the asphalt concrete changed, and the secondary structure began to form. Due to the change in the primary skeleton structure of asphalt concrete, the secondary skeleton structure did not yet form, and the growth rate of slope flow value was large. Therefore, the curve relationship between the slope flow value and the test time was linear at the beginning.

(2) As the test continued, the original structure of asphalt concrete was gradually destroyed, and the secondary structure was gradually formed. The slope of the tangent line of the slope flow value and the test time curve decreased, showing a trend of growth weakening, because the secondary structure was gradually stable to resist slope flow, and the slope flow value was gradually stable.

(3) From the middle of the test to the later stage, the original structure of the asphalt concrete specimen was completely destroyed, and the secondary structure inside the asphalt concrete was completely generated. The specimen had stability, and the completion rate of the slope flow value can reach 95–100% of the final flow value.

(4) Different from the soil structure, asphalt concrete is a multi-level spatial network structure composed of aggregate, filler and asphalt. The properties of different materials play different roles in asphalt concrete, such as high bitumen aggregate ratio and high fluidity. With the increase in test time, as long as the aggregate in asphalt concrete can form a stable structure, even if the asphalt concrete may have a small displacement again, it will gradually stabilize, in general.

Figure 3 illustrates the relationship between the gradation index of asphalt concrete and the slope flow value. It can be clearly seen from the figure that the slope flow value decreased with the increase in the gradation index, and the thermal stability of asphalt concrete was enhanced. This was because as the gradation index of asphalt concrete increased, the proportion of coarse aggregate gradually increased, while the proportion of fine aggregate decreased. As the skeleton structure of asphalt concrete, the increase in coarse aggregate ratio can improve the slope stability of asphalt concrete. The selection of the gradation index was considered from two aspects of slope stability and asphalt concrete impermeability. Therefore, in the asphalt concrete slope flow test, the gradation index can meet the porosity requirements of the impervious layer, and the higher gradation index was selected.

4.2. Effect of Filler Content on Slope Stability

The variation law of slope flow value and test time of asphalt concrete specimens with different filler content and the relationship between slope flow value and filler content are shown in Figure 4 and Figure 5.

It can be seen from Figure 4 that the slope flow value of asphalt concrete specimens increased with the increase in filler content, and with the increase in filler content, the linear growth stage of slope flow value and test time curve increased greatly. The relationship between the filler content and the slope flow value was further explored. As shown in Figure 5, the slope flow value increased with the increase in the filler content. When the bitumen aggregate ratio was 7.5%, the slope flow value increased obviously. When the filler content increased from 13% to 15%, the slope flow value changed most obviously. When the bitumen aggregate ratio was 7.2%, the slope flow value increased from 0.991 mm to 2.139 mm, which was an increase of 115.7%; when the bitumen aggregate ratio was 7.5%, the slope flow value increased from 1.457 mm to 3.376 mm, with an increase in 131.6%. This was because when the filler content is appropriate, the asphalt mortar mixed with asphalt can better fill the voids generated by the skeleton formed by the aggregate, which increases the viscosity and strength of the asphalt concrete specimen, so that the asphalt concrete specimen has a strong cohesion and high stability to resist the slope. With the increase in filler content, too much filler and asphalt are difficult to mix evenly, and the asphalt mixture is dry and difficult to stir, so that the cohesion of asphalt mortar in asphalt concrete specimens decreases, the slope flow value increases, and the slope stability becomes worse. The test results showed that the filler content affects the slope stability of asphalt concrete, and the unsuitable filler content makes the slope stability worse. Therefore, the filler content in the impervious asphalt concrete panel is not easy to exceed 13%.

4.3. Effect of Bitumen Aggregateratio on Slope Stability

Figure 6 and Figure 7 depict the relationships between the slope flow value of asphalt concrete with different bitumen aggregate ratios, the test time, and the bitumen aggregate ratio.

Combined with Figure 6 and Figure 7, the influence of ratio on the slope flow value of asphalt concrete was explored. It can be seen that the slope flow value increased with the increase in bitumen aggregate ratio. When the filler content was 9%, the bitumen aggregate ratio increased from 6.6% to 7.5%, and the slope flow value increased from 0.461 mm to 1.086 mm, with an increase of 135.7%. When the filler content was 15%, the slope flow value increased from 0.983 mm to 3.376 mm, which was an increase of 243.4%. It can be seen that in the case of the same filler content, the larger the bitumen aggregate ratio, the greater the slope flow value, because with the increase in the bitumen aggregate ratio, part of the asphalt and the filler formed a structural asphalt, and the other part was free asphalt. When the filler content was constant, the bitumen aggregate ratio increased, the free asphalt increased, and the adhesion of asphalt concrete was relatively weakened. When the test temperature was 70 °C, the asphalt fluidity increased, the asphalt concrete specimen softened, the adhesion of asphalt concrete decreased, slipped, affected the stability of the specimen, and the slope flow value became larger.

4.4. Grey Relational Analysis

Grey relational analysis is a method and theory that employed the grey relational order to describe the strength, size, and sequence of relationships between factors. This approach was used to derive and predict incomplete information systems.

The steps of grey correlation analysis:

1. Let the dependent variable of the system be the reference sequence and the independent variable be the comparison sequence.

2. The grey correlation analysis uses the ‘initial value’ method to divide the test data and the initial value at the same level for dimensionless processing:

$$x_i\left(k\right)=\frac{x_i^{\prime }\left(k\right)}{x_i^{\prime }\left(1\right)}$$

(3)

In the formula: i = 0, 1, …, n, n is the number of influencing factors; k = 1, 2, …, m, m is the number of levels contained in each influencing factor.

3. Calculation of correlation coefficient

$${\mathsf{\zeta }}_i\left(j\right)=\frac{\left[\underset{i}{min}\underset{j}{min}\left|x_0\left(j\right)-x_i\left(j\right)\right|+\underset{i}{\mathsf{\rho }max}\underset{j}{max}\left|x_0\left(j\right)-x_i\left(j\right)\right|\right]}{\left[\left|x_0\left(j\right)-x_i\left(j\right)\right|+\underset{i}{\mathsf{\rho }max}\underset{j}{max}\left|x_0\left(j\right)-x_i\left(j\right)\right|\right]}$$

(4)

In the formula: $\left|x_0\left(j\right)-x_i\left(j\right)\right|$ is the absolute value difference between the corresponding elements of the comparison sequence and the reference sequence, and ($\underset{i}{min},\underset{j}{min},\underset{i}{max},\underset{j}{max}$ are the minimum difference and the maximum difference, respectively. ρ is the resolution coefficient, ρ (0, 1), generally 0.5.

4. Calculating correlation

$$r_i=\frac{1}{m_j}\sum _1^m={\zeta }_i\left(j\right)$$

(5)

5. Sorting the calculated correlation coefficients

The slope flow value is used as the reference sequence, and the gradation index, bitumen aggregate ratio and filler content are used as the comparison sequence to calculate the correlation coefficient. The calculation results are shown in

Table 5. Slope flow values of different influencing factors.

Slope Flow Value (mm)	Grading Index	Filler Content (%)	Bitumen Aggregate Ratio (%)
1.369	0.33	11	7.2
1.217	0.36	11	7.2
0.813	0.39	11	7.2
0.508	0.42	11	7.2
0.401	0.39	9	6.6
0.568	0.39	9	6.9
0.746	0.39	9	7.2
1.086	0.39	9	7.5
0.481	0.39	11	6.6
0.705	0.39	11	6.9
0.79	0.39	11	7.2
1.323	0.39	11	7.5
0.781	0.39	13	6.6
0.82	0.39	13	6.9
0.991	0.39	13	7.2
1.457	0.39	13	7.5
0.993	0.39	15	6.6
1.331	0.39	15	6.9
2.099	0.39	15	7.2
3.476	0.39	15	7.5

.

The correlation calculated from Table 5 is shown in

Table 6. Correlation ranking results.

Appraisal Items	Correlation Degree	Ranking
grading index	0.616012	3
filler content	0.686898	2
asphalt-stone ratio	0.703118	1

.

For the three evaluation items, the comprehensive evaluation of test temperature was the highest (asphalt-stone ratio: 0.703118), followed by filler content (correlation degree: 0.686898), and the gradation index was the lowest (correlation degree: 0.616012).

5. Prediction Model of Slope Flow Value and Test Time

5.1. Slope Flow Value and Time Prediction Model of Different Gradation Index

Taking the gradation index (r) as an example, the relationship between the slope flow value and the test time was analyzed. From Section 3.2, its characteristics were: (1) Slope flow value decreased with the increase in gradation index; (2) the growth trend of the curve was that the slope flow value rose rapidly in the first few hours of the test, and the slope flow value tended to be stable with the increase in the test time.

According to the variation trend of the relationship between the slope flow value and the test time, it conformed to the variation law of the exponential function BoxLucas1 model [23]. Therefore, the model was used to describe the variation of the slope flow value (U) with time (T). The curve model is as follows:

$$U_r=a\left[1-exp\left(-bT\right)\right]$$

(6)

In the formula: U_r is the slope flow value, mm; T is time, h; a, b are material parameters.

The fitting results of different gradation indexes are shown in Table 1, and the correlation coefficient R₂ of each fitting curve is greater than 0.9.

It can be seen from

Table 7. Different gradation index fitting results.

Grading Index (r)	Fitting Parameters		Error Analysis R2
	a	b
0.33	1.35	0.26	0.968
0.36	0.99	0.19	0.973
0.39	0.77	0.28	0.989
0.42	0.50	0.32	0.990

that the variation range of b value was small and there was no obvious rule. Therefore, the b value can be replaced by the mean value b in the calculation. The linear regression of the gradation index and the value a shows that a had a linear relationship with r when the bitumen aggregate ratio and the filler content were constant, as shown in Figure 8.

After fitting, it was found that the function relationship between a value and gradation index (r) was as follows:

$$a=\alpha +\beta r$$

(7)

In the formula: α and β are fitting parameters.

By substituting Formula (7) into (6), we obtain:

$$U_r=\left(\alpha +\beta r\right)\left[1-exp\left(-bT\right)\right]$$

(8)

Formula (8) is the BoxLucas1 correction model, and the calculation results were in good agreement with the experimental data. This formula can be used as a correction model for the prediction model of slope flow value under different gradation indexes. The model had three parameters, and the calculated results were α = 4.365, β = −0.923, b = 0.2625.

According to the calculation parameters, and the calculation model of Equation (8), the prediction model of slope flow value with different gradation index was obtained and fitted with the test data, as shown in Figure 9. It was found that the calculation results of this formula were in good agreement with the experimental data, and the variation law of slope flow value and test time can be seen. With the increase in test time, the slope flow value first increased rapidly and then, the upward trend slowed down. As the test continued, the slope flow value remained stable.

5.2. Slope Flow Value and Specimen Prediction Model with Different Filler Content

From Figure 4, the relationship between the slope flow value and the time of different filler content at the oil/stone ratio of 6.6% was obtained. According to the change rule of slope flow value and test time, the exponential function BoxLucas1 model was used to describe the change rule of slope flow value (U_F) with time (T):

$$U_F=a_2\left[1-exp\left(-b_{2}T\right)\right]$$

(9)

In the formula: a₂ and b₂ are material parameters.

It can be seen from

Table 8. Fitting results of different filler content.

Filler Content (F)/%	Fitting Parameters		Error Analysis R2
	a	b
9	0.45	0.38	0.968
11	0.51	0.42	0.973
13	0.73	0.36	0.989
15	1.05	0.41	0.990

that the variation range of b value was small and there was no obvious rule. Therefore, the b value can be replaced by the mean value b in the calculation. The results show that the relationship between a and F was an exponential function, as shown in Figure 10.

After fitting, it was found that the relationship between a value and F function was as follows:

$$a=exp\left({\alpha }_2+{\beta }_{2}F+\gamma F^2\right)$$

(10)

In the formula: α₂, β₂, γ are fitting parameters.

By substituting Formula (10) into (9), we obtain:

$$U_F=exp\left({\alpha }_2+{\beta }_{2}F+\gamma F^2\right)\left[1-exp\left(-b_{2}T\right)\right]$$

(11)

Formula (11) was the BoxLucas1 correction model, and the calculation results were in good agreement with the experimental data. The formula can be used as a slope flow value and test time correction curve model for different filler contents. The correction model had a total of four parameters. After calculation, α₂ = −0.432, β₂ = −0.156, γ = 0.013, b₂ = 0.393.

According to the experimental calculation parameters, according to the calculation model of Formula (11), the prediction model of slope flow value with different filler content was obtained. As shown in Figure 11, it was found that the calculation results of this formula were in good agreement with the experimental data.

5.3. Slope Flow Value and Time Prediction Model of Different Bitumen Aggregate Ratio

From Figure 7, the relationship between slope flow value and time at different bitumen aggregate ratios with 11% filler content was obtained. The exponential function BoxLucas1 model was used to describe the law of slope flow value (U_B) changing with time (T):

$$U_B=a_3\left[1-exp\left(-b_{3}T\right)\right]$$

(12)

In the formula: a₃ and b₃ are material parameters.

It can be seen from

Table 9. Fitting results of different asphalt content.

Bitumen Aggregate Ratio (B)/%	Fitting Parameters		Error Analysis R2
	a	b
6.6	0.73	0.29	0.975
6.9	0.83	0.23	0.986
7.2	0.91	0.34	0.995
7.5	1.47	0.24	0.996

that the variation range of b value was small and there was no obvious rule. Therefore, the b value can be replaced by the mean value b in the calculation. The bitumen aggregate ratio (B) and a value were fitted. The results showed that a and B had an exponential function relationship, as shown in Figure 12.

After fitting, it was found that the relationship between a value and B function was as follows:

$$a=exp\left({\alpha }_3+{\beta }_{3}F+{\gamma }_2{F}^2\right)$$

(13)

In the formula: α₃, β₃, γ₂ are fitting parameters. By substituting Formula (13) into (12), we obtain:

$$U_F=exp\left({\alpha }_3+{\beta }_{3}F+{\gamma }_2{F}^2\right)\left[1-exp\left(-b_{3}T\right)\right]$$

(14)

Formula (14) was the BoxLucas1 correction model, and the calculation results were in good agreement with the experimental data. The formula can be used as the correction curve model of slope flow value and test time for different bitumen aggregate ratios. The correction model had four parameters, and α₃ = 49.331, β₃ = −14.783, γ₂ = 1.1, b₃ = 0.275 can be obtained by calculation.

According to the experimental calculation parameters and the calculation model of Formula (14), the prediction model of slope flow value with different bitumen aggregate ratios was obtained, and it was fitted with the experimental data. As shown in Figure 13, it was found that the calculation results of this formula were in good agreement with the experimental data.

6. Discussion and Suggestions

According to the analysis of Section 4 and Section 5, the asphalt concrete specimen with the maximum aggregate size of 19 mm and the gradation index of 0.39 was prepared in the slope flow value test. When the filler content was 9~13%, the slope flow value changed little. When it exceeded 13%, the slope flow value suddenly increased, and the thermal stability of the asphalt concrete became worse. Therefore, the filler content of the impervious layer plate should not exceed 13%; As the bitumen aggregate ratio increased, the slope flow value gradually increased and the increase rate gradually increased, and the bitumen aggregate ratio should not exceed 7.2%. The variation law of slope flow value with different mix ratio parameters was similar to the slope flow test in references [9,10] (the maximum particle size of aggregate in asphalt concrete specimen was 16 mm). Additionally, the slope thermal stability was better when the maximum aggregate size of the asphalt concrete specimen was 19 mm and the high grade index was used. Therefore, when the maximum aggregate size in asphalt concrete panel was 19 mm, its thermal stability can be guaranteed. The grey correlation analysis was carried out on the three influencing factors of asphalt concrete. The slope flow value was used as the reference value, the gradation index, the bitumen aggregate ratio and the filler content were the evaluation indexes. The comprehensive evaluation of the three evaluation indexes was the highest (correlation degree: 0.703118), followed by the filler content (correlation degree: 0.686898) and, finally, the gradation index (0.616012).

7. Conclusions

The slope flow test of asphalt concrete with different mix ratios was carried out. The conclusions were as follows:

(1) The relationship between the slope flow value of asphalt concrete and the increase in test time is mainly divided into three stages: the slope flow value is almost at a linear growth stage, gradual stabilization stage, and basically remain unchanged stage.

(2) The flow value of asphalt concrete slope with the maximum aggregate size of 19 mm decreased with the increase in gradation index. Under the condition that the impermeability of asphalt concrete can be guaranteed, reasonable increase in gradation index will make asphalt concrete have better high-temperature stability. When the filler content exceeded 13%, the slope flow value changed significantly, and the slope stability deteriorates significantly. As the bitumen aggregate ratio increased, the slope flow value became larger. When the asphalt content exceeded 7.2%, the slope stability of asphalt concrete became worse.

(3) In the in-depth study of the relationship between the flow value and time of asphalt concrete slope, the prediction model of slope flow value and time was established. The model reflected the change rule of slope flow value with different gradation index and asphalt/aggregate ratio filler content with test time, which was in good agreement with the test results. Through the grey correlation method, it was concluded that the bitumen aggregate ratio had the greatest influence on the slope stability of asphalt concrete, followed by the filler content and, finally, the gradation index.

(4) In this paper, the slope stability of asphalt concrete under different mix ratio parameters was studied. In the follow-up study, the calculation model will be further optimized, and the effects of test temperature and asphalt color on the slope stability of asphalt concrete specimens will be further explored.