Pavement Performance Investigation of Asphalt Mixtures with Plastic and Basalt Fiber Composite (PB) Modifier and Their Applications in Urban Bus Lanes Using Statics Analysis

A new type of plastic and basalt fiber composite (PB) modifier, which is composed of waste plastic and basalt fiber using a specific process, was used for bus lanes to address severe high-temperature deformation diseases due to the heavy loads of buses. The dense gradations of asphalt mixture with a nominal maximum aggregate size of 13.2 mm (AC-13) and 19 mm (AC-20) were selected to fabricate asphalt mixtures. The impact of the modifier PB on the high-temperature rutting resistance, low-temperature crack resistance, and water damage resistance was investigated experimentally. The experimental results showed that adding the modifier PB could enhance the rutting resistance and water damage resistance of asphalt mixtures significantly while maintaining the low-temperature crack resistance. Then, PB-modified asphalt mixtures of AC-13 and AC-20 were employed into a typical pavement structure of a bus lane in Yangzhou city, China, and three types of designed pavement structures were proposed. On this basis, statics analyses of all of the designed structures were performed using the finite element method. The statics analyses revealed that, compared with the standard axle load, the actual over-loaded axle made the pavement structure of the bus lane suffer a 30% higher stress and vertical deformation, leading to accelerated rutting damage on the bus lanes. The addition of the modifier PB could make the pavement structure stronger and compensate for the negative effect caused by the heavy axle load. These findings can be used as a reference for the pavement design of urban bus lanes.


Introduction
More than 230 cities in China have created priority lanes or bus-only lanes, according to the preferential policies and measures providing priority for the development of public transport, to alleviate urban traffic congestion [1]. However, there is a lack of guidance for the design of the bus-only lanes. The current research on urban bus lanes is mainly focused on the route design, traffic flow calculation, traffic organization optimization, etc. The bus-running characteristics are seldom taken into account for the pavement structure design [2][3][4]. It is reported that the tire ground pressure of a bus normally exceeds 0.7 MPa, which is the standard axle load for urban pavement structure design in China according to the specification of CJJ 37-90. Subsequently, pavement distresses such as rutting, upheaval, etc., appear frequently on bus-only lanes due to the insufficient structure design [5]. Therefore, new material selection and (or) design structures of bus lanes are desired.
Waste plastics are considered as an effective modifier for asphalt since not only can they strengthen the pavement performance of asphalt mixtures [6][7][8], but they can also provide a solution to the environmental pollution caused by waste plastics [9,10]. Nouali et al. [11] The modifier PB was made of waste plastic and basalt fiber with a weight ratio of 1:1, and manufactured using a specific process. The main source of waste plastics is waste polythene, and the components and contents are shown in Table 1. Basalt fiber is produced by Jiangsu Tianlong Basalt Fiber Co., Ltd., Yangzhou, China, and the properties are listed in Table 2. The modifier PB has colorless transparent surface and a brown metallic luster inside as flat solid particles, as shown in Figure 1c, and can be stored at ambient or low temperature. As PB contains basalt fiber component, according to the previous finding that the optimum basalt fiber length for asphalt mixtures is 6 mm [24,26], the length of the modifier PB was determined to be 6 mm in this study. Its technical indicators are shown in Table 3. listed in Table 2. The modifier PB has colorless transparent surface and a brown metallic luster inside as flat solid particles, as shown in Figure 1c, and can be stored at ambient or low temperature. As PB contains basalt fiber component, according to the previous finding that the optimum basalt fiber length for asphalt mixtures is 6 mm [24,26], the length of the modifier PB was determined to be 6 mm in this study. Its technical indicators are shown in Table 3.   Limestone was used as the aggregates in this paper. The mineral filler selected was limestone powder, and the technical properties are shown in Table 4.

. Aggregates and Mineral Filler
Limestone was used as the aggregates in this paper. The mineral filler selected was limestone powder, and the technical properties are shown in Table 4. Both base asphalt and styrene butadiene styrene (SBS)-modified asphalt, produced by Tongsha asphalt factory in Jiangsu province, were adopted in this study. The technical properties of the two types of asphalts are shown in Table 5. Both base asphalt and styrene butadiene styrene (SBS)-modified asphalt, produce by Tongsha asphalt factory in Jiangsu province, were adopted in this study. The technic properties of the two types of asphalts are shown in Table 5.     Both base asphalt and styrene butadiene styrene (SBS)-modified asphalt, produced by Tongsha asphalt factory in Jiangsu province, were adopted in this study. The technical properties of the two types of asphalts are shown in Table 5.     According to the difference in asphalt and modifier, three types of asphalt mixtures with a gradation of AC-13 were fabricated, called base asphalt + AC-13, base asphalt + AC-13 + PB, and SBS-modified asphalt + AC-13, respectively. The same three types of asphalt mixtures with a gradation of AC-20 were also fabricated. It is worth pointing out that the "dry mixing process" was adopted to fabricate the PB-modified asphalt mixtures, which means that the modifier PB was mixed with heated aggregates at a mixing temperature of 175 • C for 90 s before adding asphalt according to JTG E20-T0702 [36]. The dosage of the modifier PB in asphalt mixtures was 0.6% by the weight of aggregates. The related volumetric properties and the optimum asphalt content (OAC) of all six types of asphalt mixtures were determined by the Marshall design method in accordance with JTG F40 [35]. The results are shown in Table 6. Wheel-tracking test and dynamic creep test were used to appraise the high-temperature stability of asphalt mixtures.
The wheel-tracking test was conducted in accordance with JTG E20-T0719 [36] under a test temperature of 60 • C and test tire ground pressure of 0.7 MPa. The dynamic stability (DS), relative deformation ratio (RDR), and comprehensive stability index (CSI) were adopted to assess the rutting resistance of asphalt mixtures, which were calculated by Equations (1)-(3), respectively. Generally speaking, higher DS, smaller RDR, and higher CSI values can guarantee a better rutting resistance of asphalt mixtures.
where: d 1 and d 2 are the deformation of the asphalt mixture at the running time t 1 (45 min), t 2 (60 min), mm; C 1 and C 2 are test coefficients, C 1 = C 2 = 1.0; N is the running speed of the wheel, usually 42 times/min; ∆L is the depth of rutting of the asphalt mixture under load at specific timing, mm; D is the total thickness of the specimen, mm. In addition, a dynamic creep test was performed in accordance with the method described in NCHRP9-29. Three test temperatures of 40 • C, 50 • C, and 60 • C were selected. The test would end until the cumulative permanent strain reached 50,000 µε or the cumulative loading time reached 10,000 s. The test process and typical strain-loading time curve are shown in Figure 4. The flow number and creep rate were served to assess the high-temperature deformation performance of asphalt mixtures. As shown in Figure 4b, the accumulated strain can be divided into three stages: creep migration (Stage I), creep The test would end until the cumulative permanent strain reached 50,000 με or the cumulative loading time reached 10,000 s. The test process and typical strain-loading time curve are shown in Figure 4. The flow number and creep rate were served to assess the hightemperature deformation performance of asphalt mixtures. As shown in Figure 4b, the accumulated strain can be divided into three stages: creep migration (Stage Ⅰ), creep stability (Stage Ⅱ), and creep failure (Stage Ⅲ ). Generally, the slope of the linear growth curve in Stage Ⅱ is regarded as the creep rate, and the flection point between Stage Ⅱ and Stage Ⅲ is regarded as the flow number.

Low-Temperature Performance Test
According to the Chinese test procedure JTG E20-T0715 [36], low-temperature bending beam test was adopted to explore the low-temperature property of asphalt mixtures. The loading speed was 50 mm/min and the test temperature was −10 °C . The flexuraltensile strength (RB), failure strain (εB), and flexural stiffness modulus (SB) were calculated according to Equations (4)- (6). Normally, the higher the flexural-tensile strain and lower flexural stiffness modulus are, the better the low-temperature cracking resistance of asphalt mixtures will be. The test process is illustrated in Figure 5. where: b is the width of the specimen, mm; h and L are the height and span diameter of the specimen, mm; PB is the maximum load on the specimen, N; d is the span deflection at failure point, mm.

Loading times
Stage Ⅰ

Flow Number
Stage Ⅱ Stage Ⅲ

Low-Temperature Performance Test
According to the Chinese test procedure JTG E20-T0715 [36], low-temperature bending beam test was adopted to explore the low-temperature property of asphalt mixtures. The loading speed was 50 mm/min and the test temperature was −10 • C. The flexuraltensile strength (R B ), failure strain (ε B ), and flexural stiffness modulus (S B ) were calculated according to Equations (4)- (6). Normally, the higher the flexural-tensile strain and lower flexural stiffness modulus are, the better the low-temperature cracking resistance of asphalt mixtures will be. The test process is illustrated in Figure 5.
where: b is the width of the specimen, mm; h and L are the height and span diameter of the specimen, mm; P B is the maximum load on the specimen, N; d is the span deflection at failure point, mm.

Water Stability Test
In accordance with the Chinese test procedure JTG E20-T0709 and T0729 [36], the water stabilities of asphalt mixtures were evaluated by immersion Marshall test and freeze-thaw splitting test. Residual stability (MS0) and freeze-thaw splitting tensile

Water Stability Test
In accordance with the Chinese test procedure JTG E20-T0709 and T0729 [36], the water stabilities of asphalt mixtures were evaluated by immersion Marshall test and freeze-thaw splitting test. Residual stability (MS 0 ) and freeze-thaw splitting tensile strength ratio (TSR) were used to estimate the water damage resistance of asphalt mixtures, which are defined by Equations (7) and (8), respectively.
where: MS 1 is the conditioned stability of the samples that endured hot water immersion (60 • C, 48 h), kN; MS is the unconditioned Marshall stability of the samples, kN.
where: R T2 is the average value of the splitting strength of the conditioned samples, MPa; R T1 is the average value of the splitting strength of the unconditioned samples, MPa.

Dynamic Modulus Test
The dynamic modulus test was conducted according to the Chinese specification of JTG E20-T0738 [36]. Since dynamic modulus is temperature and frequency-related, four different test temperatures of 5 • C, 20 • C, 40 • C, and 55 • C were selected in this study. With respect to the frequency, it is reported that the frequency caused by the vehicle moving is related to its speed, pavement evenness, etc., [37]. In accordance with the data provided by the Yangzhou Passenger Transport Management Office, the average speed of the bus is approximately 30 km/h, which caused a corresponding load frequency of around 0.1 Hz [38]. Therefore, a load frequency of 0.1 Hz was adopted in this study.
For all of the above mentioned tests, the average values of three duplicate samples were used for the results analyses and discussions.

Wheel-Tracking Test Results
The results of the DS, RDR, and CSI of the wheel-tracking test are shown in Figure 6. It can be drawn from Figure 6 that the addition of modifier PB can improve the DS and CSI significantly, while reducing the RDR to some extent. As for the asphalt mixtures with AC-13 gradation, the DS of the PB-modified asphalt mixtures increased by 3.6 times, while the CSI increased by 8.5 times and RDR decreased by 4.7 percentage points compared with the base asphalt mixtures. Furthermore, compared with SBS-modified asphalt mixtures, the DS and CSI of the PB-modified asphalt mixtures also increased dramatically by 30.5% and 35.1%, respectively, even though the RDR remained at the same level. As for AC-20-graded asphalt mixtures, the same trends of the wheel-tracking test results could be observed when the modifier PB was used. Those findings indicate that the modifier PB possesses a superior capability to improve the high-temperature stability of asphalt mixtures, even much better than the SBS-modified asphalt. One reason could be that waste plastics can increase the viscosity of the asphalt binder, leading to an improvement in the stiffness modulus of asphalt mixtures [10,11]. The other reason could be that basalt fiber can form a three-dimensional network structure, which restricts the plastic deformation of asphalt mixtures and enhances the shear resistance of the mixtures at high temperatures [24,25]. possesses a superior capability to improve the high-temperature stability of asphalt mixtures, even much better than the SBS-modified asphalt. One reason could be that waste plastics can increase the viscosity of the asphalt binder, leading to an improvement in the stiffness modulus of asphalt mixtures [10,11]. The other reason could be that basalt fiber can form a three-dimensional network structure, which restricts the plastic deformation of asphalt mixtures and enhances the shear resistance of the mixtures at high temperatures [24,25]. These findings infer that adding modifier PB can significantly strengthen the permanent deformation resistance of asphalt mixtures in terms of creep deformation. The higher the temperature, the greater the enhancement achieved. These findings infer that adding modifier PB can significantly strengthen the permanent deformation resistance of asphalt mixtures in terms of creep deformation. The higher the temperature, the greater the enhancement achieved.

Low-Temperature Crack Resistance
The results are illustrated in Figure 8. It can be observed that, when modifier PB was used, the failure strain values of both AC-13 and AC-20 asphalt mixtures increased by 10.3% and 6.0%, respectively, compared with the corresponding base asphalt mixtures. Meanwhile, the flexural stiffness moduli presented a slightly decreasing trend. It can also be seen that the failure strain values of SBS-modified asphalt mixtures for both AC-13 and AC-20 gradations were much higher than those of PB-modified or base asphalt mixtures. These findings indicate that modifier PB does enhance the low-temperature crack resistance of the mixtures to some extent. As mentioned before, the addition of the waste plastics modifier would only make the asphalt mixtures sensitive to cracking, despite the positive impact on the high-temperature performance. By combining waste plastics with basalt fiber, the new type of modifier PB eliminates that negative effect on the low-temperature anti-cracking performance. In addition, the failure strain of the PB-modified asphalt mixture exceeded 2000 με, which met the requirements of JTG F40 [35].

Low-Temperature Crack Resistance
The results are illustrated in Figure 8. It can be observed that, when modifier PB was used, the failure strain values of both AC-13 and AC-20 asphalt mixtures increased by 10.3% and 6.0%, respectively, compared with the corresponding base asphalt mixtures. Meanwhile, the flexural stiffness moduli presented a slightly decreasing trend. It can also be seen that the failure strain values of SBS-modified asphalt mixtures for both AC-13 and AC-20 gradations were much higher than those of PB-modified or base asphalt mixtures. These findings indicate that modifier PB does enhance the low-temperature crack resistance of the mixtures to some extent. As mentioned before, the addition of the waste plastics modifier would only make the asphalt mixtures sensitive to cracking, despite the positive impact on the high-temperature performance. By combining waste plastics with basalt fiber, the new type of modifier PB eliminates that negative effect on the low-temperature anticracking performance. In addition, the failure strain of the PB-modified asphalt mixture exceeded 2000 µε, which met the requirements of JTG F40 [35].

Water Damage Stability
Figures 9 and 10 illustrate the results of the water stability tests. It can be seen that the MS 0 and TSR of all types of asphalt mixtures exceeded 85%, which met the requirements in the Chinese specification of JTG F40. Furthermore, adding modifier PB not only improves the MS 0 values or TSR values of mixtures but also strengthens the absolute values of the Marshall stability or splitting tensile strength.

Water Damage Stability
Figures 9 and 10 illustrate the results of the water stability tests. It can be s the MS0 and TSR of all types of asphalt mixtures exceeded 85%, which met the ments in the Chinese specification of JTG F40. Furthermore, adding modifier PB improves the MS0 values or TSR values of mixtures but also strengthens the abso ues of the Marshall stability or splitting tensile strength.

Water Damage Stability
Figures 9 and 10 illustrate the results of the water stability tests. It can be seen the MS0 and TSR of all types of asphalt mixtures exceeded 85%, which met the req ments in the Chinese specification of JTG F40. Furthermore, adding modifier PB not improves the MS0 values or TSR values of mixtures but also strengthens the absolute ues of the Marshall stability or splitting tensile strength.  In terms of the mixtures of AC-13 gradation, compared with base asphalt mixtures, the MS 0 values of PB-modified mixtures increased from 87.0% to 91.3% while the TSR increased from 89.2% to 92.8%, which were comparable to those of SBS-modified mixtures. In addition, the unconditioned Marshall stability (MS) of PB-modified mixtures grew from 9.48 kN to 12.54 kN, while the unconditioned splitting tensile strength R T1 rose from 0.83 MPa to 1.11 MPa. In terms of the mixtures of AC-20 gradation, similar trends could be observed when modifier PB was used. These findings mean that adding modifier PB can significantly boost the water damage stability of asphalt mixtures.

Dynamic Modulus
The results of the dynamic modulus test are shown in Figure 11. It was noticeable that the dynamic moduli of the asphalt mixtures with modifier PB maintained the highest values within the test temperature range in terms of the same mixture gradation. For instance, at 55 • C, the dynamic moduli of asphalt mixtures with AC-20 gradation reached 97 MPa, 125 MPa, and 94 MPa for base asphalt mixtures, PB-modified mixtures and SBS-modified mixtures, respectively. These results indicate that the asphalt mixtures with modifier PB possess a superior deformation resistance at high temperature, which matches the results of both the wheel-tracking test and high-temperature creep test.
the MS0 values of PB-modified mixtures increased from 87.0% to 91.3% while the TSR increased from 89.2% to 92.8%, which were comparable to those of SBS-modified mixtures. In addition, the unconditioned Marshall stability (MS) of PB-modified mixtures grew from 9.48 kN to 12.54 kN, while the unconditioned splitting tensile strength RT1 rose from 0.83 MPa to 1.11 MPa. In terms of the mixtures of AC-20 gradation, similar trends could be observed when modifier PB was used. These findings mean that adding modifier PB can significantly boost the water damage stability of asphalt mixtures.

Dynamic Modulus
The results of the dynamic modulus test are shown in Figure 11. It was noticeable that the dynamic moduli of the asphalt mixtures with modifier PB maintained the highest values within the test temperature range in terms of the same mixture gradation. For instance, at 55 °C, the dynamic moduli of asphalt mixtures with AC-20 gradation reached 97 MPa, 125 MPa, and 94 MPa for base asphalt mixtures, PB-modified mixtures and SBSmodified mixtures, respectively. These results indicate that the asphalt mixtures with modifier PB possess a superior deformation resistance at high temperature, which matches the results of both the wheel-tracking test and high-temperature creep test.

Pavement Structure Design of the Bus Lane
Taking the pavement structure of the bus lane of Wenchang Road in Yangzhou City for example, the original pavement structure included five layers, namely (from top to bottom) an asphalt surface layer, asphalt binder layer, cement-stabilized gravel base layer, lime-soil base layer, and soil foundation. Asphalt mixtures with a gradation of AC-13 and AC-20 were used for the surface layer (4 cm thickness) and binder layer (8 cm thickness), respectively. PB-modified AC-13 and PB-modified AC-20 were used to replace the original materials of the asphalt surface or (and) binder layer. Therefore, three different types of pavement structures of the bus lane were proposed, as shown in Figure 12. The finite element software of ABAQUS 6.11 was employed to perform the statics analysis. The distribution of the corresponding stress and stain was analyzed for the designed structures of bus lane.

Pavement Structure Design of the Bus Lane
Taking the pavement structure of the bus lane of Wenchang Road in Yangzhou City for example, the original pavement structure included five layers, namely (from top to bottom) an asphalt surface layer, asphalt binder layer, cement-stabilized gravel base layer, lime-soil base layer, and soil foundation. Asphalt mixtures with a gradation of AC-13 and AC-20 were used for the surface layer (4 cm thickness) and binder layer (8 cm thickness), respectively. PB-modified AC-13 and PB-modified AC-20 were used to replace the original materials of the asphalt surface or (and) binder layer. Therefore, three different types of pavement structures of the bus lane were proposed, as shown in Figure 12. The finite element software of ABAQUS 6.11 was employed to perform the statics analysis. The distribution of the corresponding stress and stain was analyzed for the designed structures of bus lane.

Material Parameters
The material parameters selected for the statics analysis are listed in Table 7, of which, the density and Poisson ratio were provided by Yangzhou City Municipal Administration Department. It should be noted that the dynamic modulus of the asphalt mixture at 0.1 Hz and 20 • C was used as the modulus needed in Table 7, though the static modulus of the asphalt mixture should have been used herein. This is because there is a good correlation between the dynamic modulus and static modulus of asphalt mixtures, and the values of the two indexes become comparable, especially at low-frequency conditions (such as 0.1 Hz) [37].

Material Parameters
The material parameters selected for the statics analysis are listed in Table 7, of which, the density and Poisson ratio were provided by Yangzhou City Municipal Administration Department. It should be noted that the dynamic modulus of the asphalt mixture at 0.1 Hz and 20 °C was used as the modulus needed in Table 7, though the static modulus of the asphalt mixture should have been used herein. This is because there is a good correlation between the dynamic modulus and static modulus of asphalt mixtures, and the values of the two indexes become comparable, especially at low-frequency conditions (such as 0.1 Hz) [37].

Load Determination
The standard tire grounding pressure of 0.7 MPa is commonly used for pavement structure design in China. However, the tire ground pressure of a typical city bus normally exceeds 0.7 MPa. Therefore, the actual load was utilized for this study, along with the standard axle load. According to the data provided by Yangzhou City Bus Company, as shown in Table 8, a tire ground pressure of 0.83 MPa was considered as the actual load of a typical type of bus in Yangzhou city. Since a two-dimensional pavement model was used for the statics analysis, the applied load in the modeling process needs to be converted from the surface load (tire ground pressure) to line load. The relationship between the tire ground pressure and axle load can be expressed by Equation (9) [39,40].
Therefore, the axle weight of the vehicle can be calculated by Equation (10).
where: p i is the tire ground pressure, MPa; p is the standard tire ground pressure, 0.7 MPa; L i is the axle weight of the vehicle, kN; L is the standard axle load, 100 kN. In addition, the equivalent circle radius of the tire contacting area can be calculated by Equation (11), based on which, the line load can be calculated by Equation (12).
where: A is the tire contacting area, cm 2 ; r is the equivalent circle radius, cm.
where: q l is the line load, N/m. The results are listed in Table 9.

Model Establishing
A two-dimensional model of the cross-section of the bus lane was built for this study, with dimensions of 3.75 m (width) × 3 m (height). The CPE8R (reduced integral) unit was used. In order to accelerate the running process, the mesh was divided into a size of 0.1 m × 0.1 m for this calculation, along with meshes of 0.04 m × 0.1 m and 0.08 m × 0.1 m for the asphalt surface and binder layers, respectively. There are seven positions where the pavement is most likely to be damaged, which were marked as coordinate points in the model, namely the middle of the tire gap (Point a), the inner edge of the two tires' contacting area (Point b and b'), the middle of the two tires' contacting area (Point c and c'), and the outer edge of the two tires' contacting area (Point d and d'). The model establishment, mesh division, and loading and boundary conditions are shown in Figure 13.
namely the middle of the tire gap (Point a), the inner edge of the two tires' contacting area (Point b and b'), the middle of the two tires' contacting area (Point c and c'), and the outer edge of the two tires' contacting area (Point d and d'). The model establishment, mesh division, and loading and boundary conditions are shown in Figure 13.

Tensile Stress Distribution
Under both standard tire ground pressure (0.7 MPa) and actual tire ground pressure (0.83 MPa), the tensile stress clouds of the original pavement structure and the three types of designed structures are illustrated in Figure 14. The maximum tensile stress of each potentially damaged point (Point a to Point d), which appears at the bottom of each structure layer, is plotted with the different vertical positions (L1 to L5) in Figure 15. Due to the symmetry of the loading and the structure, only the data of one side (Point b, c and d) were used for analysis.

Tensile Stress Distribution
Under both standard tire ground pressure (0.7 MPa) and actual tire ground pressure (0.83 MPa), the tensile stress clouds of the original pavement structure and the three types of designed structures are illustrated in Figure 14. The maximum tensile stress of each potentially damaged point (Point a to Point d), which appears at the bottom of each structure layer, is plotted with the different vertical positions (L1 to L5) in Figure 15. Due to the symmetry of the loading and the structure, only the data of one side (Point b, c and d) were used for analysis.  As shown in Figure 15, the tensile stress values of all coordinate points were negative on L1 to L3, but turned positive on L4 and L5. These results indicate that the road surface (L1) and the bottoms of the asphalt surface layer (L2) and binder layer (L3) endure compressive stresses, whereas the bottoms of the cement-stabilized gravel base layer (L4) and  As shown in Figure 15, the tensile stress values of all coordinate points were negative on L1 to L3, but turned positive on L4 and L5. These results indicate that the road surface (L1) and the bottoms of the asphalt surface layer (L2) and binder layer (L3) endure compressive stresses, whereas the bottoms of the cement-stabilized gravel base layer (L4) and lime-soil base layer (L5) suffer tensile stresses. In addition, in the middle of the tire gap (Point a), the maximum compressive stress appeared on L2, as shown in Figure 15a. The maximum compressive stress at other coordinate points occurred on L1, as shown in Figure 15b-d.
It can also be observed that the axle load causes a significant impact on the tensile stress distributions. Taking the coordinate Point a of the original structure for example, as shown in Figure 15a, when the applied axle load increased from the standard tire ground pressure of 0.7 MPa to the actual tire ground pressure of 0.83 MPa, the tensile stresses on L1 to L5 increased by 28.3%, 25.3%, 27.9%, 28.0%, and 28.5%, respectively.
Furthermore, the use of PB-modified asphalt mixtures also resulted in a fluctuation in tensile stress distributions. As shown in Figure 15, compared with the original structure, the compressive stress on L1 to L3 increased to some extent when the PB-modified surface layer or (and) binder layer were designed, whereas the tensile stresses on L4 and L5 remained as the same value. This is due to the higher moduli of the asphalt mixtures with modifier PB, leading to the PB-modified layers bearing more stress. Since asphalt mixtures possess a superior compressive strength, the increasing compressive stress does not mean an accelerated deterioration of asphalt structure layers.

Vertical Deformation
Under tire ground pressures of both 0.7 MPa and 0.83 MPa, the vertical deformation of all of the potentially damaged points on L1 to L5 are illustrated in Figure 16. The vertical deformation of the asphalt pavement under load could reflect the deformation resistance of a pavement. in tensile stress distributions. As shown in Figure 15, compared with the original structure, the compressive stress on L1 to L3 increased to some extent when the PB-modified surface layer or (and) binder layer were designed, whereas the tensile stresses on L4 and L5 remained as the same value. This is due to the higher moduli of the asphalt mixtures with modifier PB, leading to the PB-modified layers bearing more stress. Since asphalt mixtures possess a superior compressive strength, the increasing compressive stress does not mean an accelerated deterioration of asphalt structure layers.

Vertical Deformation
Under tire ground pressures of both 0.7 MPa and 0.83 MPa, the vertical deformation of all of the potentially damaged points on L1 to L5 are illustrated in Figure 16. The vertical deformation of the asphalt pavement under load could reflect the deformation resistance of a pavement. As shown in Figure 16, the tire ground pressure impacted the vertical deformation hugely. Taking the coordinate Point a of the original structure for example, when the tire ground pressure increased from 0.7 MPa up to 0.83 MPa, the vertical deformation increased by 29.5~29.8% on L1 to L5.
In addition, the use of the PB-modified surface layer or (and) binder layer could reduce the vertical deformations effectively compared with the original structure. It was also noticeable that designed structure III presented the smallest vertical deformation, followed by designed structure II and structure I. This result indicates that structure III possesses a superior resistance to deformation.

Conclusions
A new type of self-developed plastic and basalt fiber composite modifier called PB modifier was used to fabricate asphalt mixtures. A series of laboratory tests, including the wheel-tracking test, dynamic creep test, low-temperature bending beam test, immersion Marshall test, freeze-thaw splitting test, and dynamic modulus test, were adopted to explore the influence of the PB modifier on the performance of asphalt mixtures. In addition, three types of urban bus lane pavement structures were proposed by using the PB-modified asphalt mixtures for the surface layer and (or) binder layer. Statics analyses were conducted using ABAQUS 6.11 finite element software. According to the results and discussions in this study, the following conclusions can be drawn: (1) Adding PB modifier can improve the dynamic modulus and high-temperature stability remarkably and reduce the creep rate of asphalt mixtures, presenting a superior high-temperature stability that is even better than SBS-modified asphalt mixtures. (2) By combining with basalt fiber, the PB modifier can compensate for the adverse effect on the low-temperature crack resistance of mixtures caused by the addition of waste plastics. (3) The PB modifier can not only improve the anti-water damage performance indexes of the residual stability and tensile strength ratio of mixtures, but can also strengthen the absolute values of the strengths, presenting a better water damage resistance. (4) The actual axle load of a bus will cause severe tensile stress and vertical deformation compared with the standard axle load. Using PB-modified asphalt layers for bus lanes can offset the negative impact caused by a heavy axle load. Using PB-modified asphalt mixtures for both the surface layer and binder layer (designed structure III) presents the best strengthen function.