Testing and Evaluation of Hard Sandstone Aggregate in Hot Mix Asphalt

,


Introduction
In recent decades, with the continuous demand for a highquality asphalt pavement, there has been an increasing shortage of aggregates, such as basalt and dolerite, for use in asphalt mixtures, and as a result, their prices have risen.To increase the availability of aggregates for use in asphalt mixtures, a type of hard sandstone aggregate was introduced.Ouyang et al. [1] used sandstone aggregates in asphalt mixtures, studied their performance in highways, and indicated that these aggregates could be used in asphalt.Metcalf and Goetz [2] also found that sandstone aggregates could be used in the asphalt pavement.Hunsucker and Graves [3] found that the hot mix asphalt (HMA)-used sandstone aggregate is resistant to rutting and that the application of an asphalt sandstone base and surface mixtures has been very successful.
Zhang and Li [4] used the sandstone aggregate to produce sandstone concrete as a road base.Li et al. [5] studied the feasibility of replacing limestone aggregates with sandstone aggregates.Yılmaz and Tugrul [6] studied the efects of diferent sandstones on the concrete strength and found that concrete mixtures made of subarkose aggregates had approximately 40-50% greater compressive and splitting tensile strengths than those made of subarkose-arkose, sublitharenite-litharenite, and arkose aggregates.Yang et al. [7] studied the mechanical properties and durability of concrete made from sandstone aggregates.
It can be seen from the abovementioned literature that sandstone has been primarily used for cement concrete and less used for the HMA pavement.In contrast, limestone, diabase, basalt, and broken gravel have been extensively used in HMA.Teir properties are listed in Table 1.
Shi et al. found that the crushing value of aggregates increased as the temperature increased [12].
During the process of mixing, paving, and rolling HMA, the aggregates are consistently maintained at high temperatures (approximately 130-200 °C).When the HMA pavement is subjected to high-and low-temperature cycles, the aggregates in HMA go through the same cycle of high and low temperatures.Aggregates may be crushed by the paver and rollers during the paving and rolling process, so the aggregate gradation will change.Hence, the properties of aggregates, such as the crushing value, polished stone value (PSV), and Los Angeles abrasion loss, must be tested before and after conditioning since they infuence the long-term performance of the HMA pavement.
In order to apply hard sandstone aggregates widely and solve the aggregate shortage problem for HMA, their properties were tested before and after conditioning in this study.Te testing process was as follows.First, the properties of the hard sandstone aggregate were tested using the standard method.Second, the crushing values were measured after the hard sandstone aggregate was heated, dried, and went through freeze-thaw cycles.Tird, the polished stone and Los Angeles abrasion values were measured after the hard sandstone aggregate went through the freeze-thaw cycles.Fourth, the surface energy of the hard sandstone aggregate was measured with the capillary rise method (CRM) and compared with those of other types of aggregates.To assess the efect of paving and rolling on the combined aggregate gradation, simulations were performed using a Marshall compactor.Te moisture susceptibility of HMA containing the hard sandstone aggregate was evaluated by the indirect tensile strength ratio (TSR) and the modulus ratio (MR), and the polishing tests of HMA were conducted with a three-wheel polishing tester.Figure 1 illustrates the fowchart of the work performed in this study.

Testing Method and Aggregate Conditioning
2.1.Testing Method 2.1.1.Sampling Methodology.Te hard sandstone aggregate came from a pile of an aggregate plant's pile yard.We frst removed the unrepresentative part of the pile foot and then obtained a group of roughly equal parts from several different parts evenly distributed on the top, middle, and bottom of the pile.Te sample was representative of the condition and quality of the lot.
To obtain a representative sample, the quartering method was applied.Te aggregate was placed on a plate, mixed evenly in the natural state, roughly fattened, and then spread out from the center to the side in two perpendicular directions.Te sample was then divided into roughly four equal parts and two parts of the diagonal were remixed.Te abovementioned process was repeated until the amount of the material after the reduction was slightly greater than the amount necessary for the tests.

Aggregate Crushing Test.
Te aggregate crushing value of the coarse aggregate is the relative resistance of the aggregate to crushing under a gradually applied compressive load.First, coarse aggregates with particle sizes of 9.5-13.2mm were selected after being dried in an oven.Second, the amount of coarse aggregates required for testing was added in three layers to a cylinder that was 11.5 cm in diameter and 18 cm in height, where each layer was subjected to 25 strokes with a tamping rod.Tird, the surface of the aggregate sample in the container was carefully leveled, and a plunger was inserted.A total load of 400 kN was applied in 10 min.Finally, the sample was sorted with a 2.36 mm sieve, and the fraction passing through the sieve was weighed.Te aggregate crushing value was the percentage of the weight of particles with diameters less than 2.36 mm to the weight of the total sample.

Los Angeles Abrasion
Test.First, the sample was prepared and placed in an abrasion test machine.Second, a specifed number of steel spheres were then placed in the machine, and the drum was rotated for 500 revolutions at a speed of 30-33 rpm.Te sample was then sorted with a 1.70 mm sieve, and the aggregates larger than 1.7 mm were retained on a 1.70 mm (no.12) sieve, washed with water, and dried in an oven.Te percentage loss was obtained by calculating the relative diference between the retained material (larger particles) weight and the original sample weight.

Polished Stone
Test.Four curved test specimens were prepared from each sample undergoing testing.Fourteen specimens were clamped around the periphery of a rubber "road wheel" and subjected to two phases of polishing by the rubber wheel.Te frst phase was abrasion by #30 coarse diamond sand for 3 h, followed by 3 h of polishing with #280 fne diamond sand.Two of the 14 samples were control specimens.Te PSV of the hard sandstone coarse aggregate was then measured using a portable skid resistance tester with a special narrow slide.

Surface Free Energy Test.
Te surface free energy of the hard sandstone aggregate was measured using the CRM.Te CRM was described in detail by Dennis Sinkonde et al. [15].

Marshall Compaction.
Road rollers can crush the aggregate during compaction of an asphalt mixture, which would change the combined aggregate grading of the asphalt mixture.Hence, in the laboratory, the hard sandstone AC-13C with an asphalt-aggregate ratio of 4.6% was compacted with 75 blows on both sides using a Marshall compactor.
where P is the maximum load (N), t is the specimen height immediately before the test (mm), and D is the specimen diameter (mm).Te indirect TSR was calculated as follows: where S 2 is the average tensile strength of the moistureconditioned groups, and S 1 is the average tensile strength of the dry groups.
2.1.8.Simple Performance Test.Six cylindrical specimens with diameters of 150 mm and heights of 155 mm were formed with a rotary compacting instrument, with three in each group, and the void ratio was 7%.Specimens with diameters of 100 mm were drilled with a core drilling machine and then cut to lengths of 150 mm with a cutting machine.One group was followed by saturation of the sample by applying a vacuum of 730-740 mm Hg for 15 minutes and placed in a plastic bag containing 10 ± 0.5 ml of water and sealed.Te specimens were frozen at −18 ± 2 °C for 16 ± 1 h and then boiled in 60 ± 0.5 °C water for 24 h.Te simple performance was measured on a simple performance tester at temperatures of 5 °C, 20 °C, and 40 °C and frequencies of 0.01, 0.1, 1, and 10 Hz, as shown in Figure 2.
Te MR was calculated as follows: where M 2 is the average dynamic modulus of the moistureconditioned groups, and M 1 is the average dynamic modulus of the dry groups.
2.1.9.Tree-Wheel Polishing Test.Te three-wheel polishing tester consisted of various components, including three tires, weights, turning devices, water spraying devices, and control devices, as shown in Figure 3. Te loading speed was 75 times per minute.Te load size and the tire shape and type could be varied.Te tire rotation diameter was 300 mm.Te water-spraying device could wash of the particles, simulating wet environment conditions and also prolonging the service life of the tire.Water mainly reduces the surface temperature of the tire and reduces tire softening.In previous studies [25,26], it has been observed that infatable rubber tires are susceptible to wear and rupture in practical applications.To enhance the efciency of the aggregate polishing process, we opted for a polyester tire measuring  Advances in Materials Science and Engineering 3 150 mm in diameter and 50 mm in width, possessing a Shore hardness of 85 [27].A counterweight of approximately 46.9 kg was used, as depicted in Figure 4.
Te wheel track width generated during this test fell within the range of 30-35 mm.To test the tilting friction coefcient, a rubber sheet with dimensions of 37.75 mm × 25.4 mm × 6.35 mm was used, as depicted in Figure 5.When employing the rubber sheet size in the pendulum friction coefcient tester, the reading of the pendulum friction coefcient was observed within the range of 0-1.0 or 0-100 (beyond the scale line), as shown in Figure 6.According to the division in Figure 7, the British pendulum numbers at four locations were used to observe the friction variation characteristics of the hard sandstone asphalt mixture in the accelerated loading test using the three-wheel polishing tester.

Heated and Dried Using Natural Gas Combustion and Stored in an Oven at 180
°C (1) In the laboratory, the hard sandstone aggregate samples were frst heated and dried using natural gas combustion for 30 min to simulate the process of the aggregate being heated and dried by fuel in a drying drum, as shown in Figure 8. (2) After the hard sandstone aggregate samples were heated and dried, they were stored in the electric blast drying oven to simulate the storage of aggregate in hot bins.After they were stored for 30, 60, and 120 min in the electric blast drying oven at 180 °C, their crushing values were measured at 180 °C.

Freeze-Taw
Processing.Te freeze-thaw conditioning process mainly simulated the efect of high-and lowtemperature cycles on the aggregate in the natural environment.In the laboratory, the freeze-thaw cycles of the hard sandstone aggregate were performed as follows: (1) Te hard sandstone aggregate samples were placed in stainless steel basins and kept in an oven at 180 °C for 24 h.(2) Te stainless steel basins with the aggregate were flled with water when the temperature of the aggregate was 180 °C.At this temperature, large amounts of hot water vapor formed, as shown in Figure 9. (3) Te stainless steel basins with aggregates and water were kept in a refrigerator at −18 °C for 24 h.(4) After the stainless steel basins with aggregate had been frozen at −18 °C for 24 h, they were placed in an oven at 180 °C and dried to a constant weight.( 5) After ( 1)-( 4), a freeze-thaw cycle was completed.Advances in Materials Science and Engineering

Composition of Hard Sandstone
Te sandstone was identifed as gravel-bearing coarsegrained tufaceous sandstone with dark gray, gravelbearing, sand-like, and sheet-like structures.It was composed of chlorite and was mainly felsic.Te felsic substance was granular, with sizes between 0.01 and 0.1 mm.Sericite was in the form of colorless phosphorous fakes, and chlorite was in the form of green scales.Figure 10 shows an image of the hard sandstone under a polarized light microscope, and Table 2 shows the main mineral components.

Materials
4.1.Coarse Hard Sandstone Aggregate.First, the physical properties of the hard sandstone aggregate were measured using standard test methods [13], and Table 3 shows the results.Because the adhesion of the aggregate to asphalt has a direct efect on the water stability of HMA, the boiling water method [14] was used to evaluate the adhesion between the hard sandstone aggregate and the asphalt, as shown in Figure 11.Te asphalt flm on the hard sandstone aggregate surface was completely preserved, and the   percentage of the stripping area was close to 0. Hence, the adhesion level between the hard sandstone aggregate and asphalt was 5.

Fine Aggregate.
Te fne aggregate was limestonemanufactured sand, and Table 4 shows its properties.

Filler.
Te fller was a limestone grinding powder, and Table 5 shows its properties.

Results for Hard Sandstone Aggregate after Various Conditioning Treatments
To further study the efects of heating, drying, and hightemperature storage on the properties of the hard sandstone aggregate, the processes of heating, drying, storage, and rolling were simulated in the laboratory and then the crushing values, PSVs, and wear loss rates were measured.

Efect of Storage Time on the Crushing Value.
Te hard sandstone aggregate samples were subjected to heating and drying using natural gas combustion.Subsequently, they   were placed in an oven and maintained at a temperature of 180 °C for durations of 30, 60, and 120 min.Following this, the crushing values were determined at a temperature of 180 °C. Figure 12 shows the values.
Figure 10 displays the crushing values of the hard sandstone aggregate under various conditions.Te crushing values of the hard sandstone aggregate stored in a blast drying oven at 180 °C for diferent durations after burning   Advances in Materials Science and Engineering and drying with natural gas were as follows: 32.8% for 30 min, 31.7% for 60 min, and 32.3% for 120 min.To analyze the infuence of the storage time at 180 °C on the crushing value, a one-way analysis of variance (ANOVA) at the 95% confdence level (α � 0.05) was used.Table 7 shows the results, and it was found that the P value was greater than 0.05, which fully indicates that the storage time at 180 °C had no signifcant infuence on the crushing value.

Efect of Temperature on the Crushing Value.
During the initial compaction, recompaction, and fnal compaction stages for HMA, the temperature of the mixture was reduced by approximately 90 °C from 180 °C [19], which could lead to varying degrees of crushing of the hard sandstone aggregate.To replicate this phenomenon, the crushing values of the hard sandstone aggregate were assessed at 20 °C, 140 °C, 160 °C, and 180 °C. Figure 13 shows the results.A crushing value of 23.4% was obtained at room temperature (20 °C).Crushing values of 23.3% and 23.2% were obtained when the hard sandstone aggregate was only maintained for 2 h in an oven at constant temperatures of 140 °C and 160 °C, respectively.Furthermore, a higher crushing value of 32.3% was measured for the hard sandstone aggregate sample that was frst heated using natural gas combustion and then placed in the oven at 180 °C for 2 h.
According to Figure 11, the crushing values remained relatively consistent at temperatures of 20 °C, 140 °C, and 160 °C.Based on the results of the one-way ANOVA of the crushing values at 20 °C, 140 °C, and 160 °C, as shown in Table 8, the P value was larger than 0.05, which proved that the temperature had no signifcant efect on the crushing value when the temperature of the hard sandstone aggregate was 160 °C and below.However, when the hard sandstone aggregate was frst heated using natural gas combustion and then placed in the blast drying oven at 180 °C for 2 h, its crushing value increased to 32.3%.Tis indicated that the burning and storage of the hard sandstone aggregate in the blast drying oven at 180 °C for 2 h signifcantly afected its crushing value.

Efect of Freeze-Taw Cycles on the Crushing Value.
Te crushing values of the hard sandstone aggregate were measured after being subjected to zero, one, three, and fve freeze-thaw cycles.Figure 14 shows the infuence of the number of freeze-thaw cycles on the crushing value of the hard sandstone aggregate.Overall, the freeze-thaw cycles had little infuence on the crushing value.However, one-way ANOVA was performed for the data in Figure 7, and Table 9 shows the result.It was found that the P value was less than 0.05, which showed that the number of freeze-thaw cycles afected the crushing value.

Polished Stone Value.
To assess the infuence of the freeze-thaw cycles on the skid resistance of the hard sandstone aggregate, freeze-thaw cycle tests of the hard sandstone aggregate were conducted in the laboratory and  8 Advances in Materials Science and Engineering then an accelerated polishing machine was used to accelerate the grinding of the hard sandstone aggregate.Te PSVs were tested using a pendulum friction coefcient meter.Figure 15 shows the results.
Figure 15 illustrates the results obtained after zero, one, three, and fve freeze-thaw cycles, indicating that the hard sandstone aggregate exhibited a minimum value of 56 and a maximum value of 60.Interestingly, these PSVs did not Crushing value (%) Temperature (°C) Figure 13: Crushing values at various temperatures.Advances in Materials Science and Engineering exhibit a decrease as the number of cycles increased.After the freeze-thaw cycles, the PSVs of the hard sandstone aggregate were only slightly larger than that with no freezethaw cycles by 2-4 PSV units.However, it was found that the number of freeze-thaw cycles afected the PSVs with oneway ANOVA, as shown in Table 10.Te PSVs of the hard sandstone aggregate were greater than those of aggregates such as limestone, diabase, and basalt shown in Table 1, which indicated that the hard sandstone aggregate had a better wear resistance.

Los Angeles Abrasion Loss.
After subjecting the hard sandstone aggregate to freeze-thaw cycles, Los Angeles abrasion tests were conducted using a Los Angeles machine.Figure 16 presents the test results.Te Los Angeles abrasion values of the hard sandstone aggregate increased after one freeze-thaw cycle, but after three and fve freeze-thaw cycles, the Los Angeles abrasion values hardly changed compared with that after one freeze-thaw cycle.Te one-way ANOVA results shown in Table 11 revealed that the number of freezethaw cycles had a signifcant infuence on the Los Angeles abrasion values.

Surface Free Energy of Hard Sandstone Aggregate
To assess the adhesion between the hard sandstone aggregate and the asphalt, the surface free energy was measured using the CRM [15].Table 12 presents the results.In addition, the surface energy for the limestone aggregate was measured, and the results indicated that the limestone aggregate had good adhesion properties to asphalt [17].Table 12 also presents the results of the surface energy measurements obtained by Yi et al. [16] using atomic force microscopy.Teir fndings revealed that the surface energy of sandstone was higher than that of limestone.Te total surface free energy of hard sandstone shown in Table 12 was slightly greater than that of limestone.Tis could be attributed to the coarser surface of the hard sandstone compared with that of limestone [18].Consequently, hard sandstone demonstrated a remarkable ability to withstand moisture damage when used as an aggregate in asphalt mixtures.

Effect of Marshall Compaction on the Impact Value of the Hard Sandstone Aggregate
During the compaction process for HMA, road rollers can crush the aggregate, resulting in a change in the aggregate grading curve of the asphalt mixture.Terefore, in the laboratory, we conducted a series of tests to investigate this phenomenon.To begin, we compacted the hard sandstone AC-13C with an asphalt-aggregate ratio of 4.6% using the Marshall compactor.Tis involved subjecting the sample to 75 blows on both sides.Subsequently, we performed extraction and screening tests [14] to determine the passing percentage of each grade of aggregate.Table 13 shows the results of these tests.
According to the data presented in Table 13, the standard Marshall compactor was used to compact the AC-13C asphalt mixture consisting of the hard sandstone aggregate.Te passing percentage of aggregate particles in the size range of 4.75-16.0mm remained relatively stable after the compaction process.Furthermore, it was found that the passing percentage of aggregate particles larger than 4.75 mm fell within the acceptable range of ±6%, as specifed in the gradation requirements [19].Tis indicated that the hard sandstone aggregate could withstand compaction with the Marshall compactor.

Moisture Susceptibility of Hard Sandstone AC-13C
Tere are many highway performance characteristics for HMA, such as rutting, low-temperature cracking, fatigue, and moisture susceptibility.Aggregate properties are one of the main factors afecting the moisture stability of HMA.Hence, the moisture susceptibility of AC-13C containing 10 Advances in Materials Science and Engineering hard sandstone aggregate was investigated.First, the optimum aggregate-asphalt ratio of AC-13C was obtained with the Marshall mix design method.Ten, the moisture susceptibility of the hard sandstone AC-13C was tested with freeze-thaw cycle indirect tensile tests and simple performance tests.

Optimum Aggregate-Asphalt Ratio.
Trough aggregate sieve analysis, the combined aggregate gradation of the AC-13C asphalt mixture was designed, as shown in Figure 17.By using the Marshall mix design method [11], Figure 18 shows the curves relating the bulk specifc gravity, Marshall stability (MS), fow value (FL), air void content (AV), voids in mineral aggregate (VMA), and voids flled with asphalt (VFA) to the asphalt-aggregate ratio.Second, the four asphalt-aggregate ratios selected above were averaged as follows: Based on the results in Figure 18, frrst, the following optimum asphalt-aggregate ratios were determined: (a) asphalt-aggregate ratio at maximum bulk specifc density, a 1 ; (b) asphalt-aggregate ratio at maximum MS, a 2 ; (c) asphalt- Tird, the ranges of the asphalt-aggregate ratio, OAC min -OAC max , that meet the technical specifcations (excluding VMA) were determined and averaged as follows: Fourth, OAC 1 and OAC 2 were averaged as follows: Te optimum asphalt-aggregate ratio was 4.6%.

Moisture Stability Test of AC-13C with Hard Sandstone.
When an asphalt mixture pavement is subjected to water-temperature-load multifeld coupling, its strength and modulus will decay and various diseases will appear in the asphalt mixture pavement, such as cracking and raveling.

Freeze-Taw Indirect Tensile Test. Te immersion
Marshall test and freeze-thaw indirect tensile test are commonly conducted to evaluate the water stability of asphalt mixtures.Te standard Marshall specimens were prepared by double-sided compaction 50 times in the laboratory, and based on a vacuum saturation and freeze-thaw process, the indirect TSR of the hard sandstone AC-13C was determined to be 81.2%, which exceeded the corresponding technical requirements of 80%.

Simple Performance Test.
To further evaluate the water stability of the hard sandstone AC-13C, the dynamic modulus was measured with the simple performance test, and the water stability was evaluated by the residual dynamic MR [21][22][23][24].First, hard sandstone AC-13C samples were subjected to the vacuum saturation and freeze-thaw process and then the dynamic moduli were measured.Table 14 shows the dynamic moduli of the freeze-thaw and nonfreeze-thaw cycle groups.
From the results in Table 10, it was found that the dynamic modulus of the hard sandstone AC-13C decreased to 71-76% after one freeze-thaw cycle as the frequency increased.Based on the dynamic MR of 60% recommended by Jason Bausano et al. [24], the hard sandstone AC-13C had a good water-resistance ability.At the same time, the phase angle increased after one freeze-thaw cycle, as shown in Table 15.
From the results in Table 11, it was found that the phase angle increased by 5-10% as the frequency decreased, so the viscous component of the hard sandstone AC-13C increased, and the elastic component was reduced.Figure 19 shows the main curve of the hard sandstone AC-13C after one freeze-thaw cycle.Te main curve of the freeze-thaw groups was below the main curve of the nonfreeze-thaw groups over the entire frequency range, and as the frequency increased, the diference in the dynamic moduli between the two was greater.Terefore, after one freeze-thaw cycle, the hard sandstone AC-13C was damaged inside, and the dynamic modulus became small, which fully showed that the asphalt pavement was damaged by water.

Three-Wheel Polishing Test for Hard Sandstone AC-13C
A hard sandstone AC-13C sample with dimensions of 50 cm × 50 cm × 5 cm was made in the laboratory.Te AV was carefully maintained at 5.2% after compaction.Figure 20 shows the forming process.
As shown in Figure 21, the British pendulum number of the hard sandstone AC-13C sample gradually decreased with the number of polishing cycles.At the end of 160,000 three-wheel polishing device (TWPD) cycles, the British pendulum number was approximately 58, which was close to the PSV of the hard sandstone aggregate shown in Figure 15.
After 160,000 TWPD cycles, there was a ring wheel track on the sample, and it was found that some aggregate particles had white spots on their surfaces, indicating that the asphalt flm fell of and that the aggregate was gradually polished, as shown in Figure 22.Advances in Materials Science and Engineering

Figure 5 :
Figure 5: Rubber sheet (the left sheet was used in this test).

Figure 7 :
Figure 7: Test points of the British pendulum number measured with the British pendulum tester.

Figure 8 :
Figure 8: Heating and drying using natural gas combustion.

Figure 9 :
Figure 9: Freeze and thaw cycles of the hard sandstone coarse aggregate.

Figure 10 :
Figure 10: Image of sandstone under a polarized light microscope.

Figure 12 :
Figure 12: Aggregate crushing values after heating and storage at 180 °C.

Figure 14 :
Figure 14: Crushing value vs. the number of freeze-thaw cycles.

Figure 16 :
Figure 16: Abrasion value vs. the number of freeze-thaw cycles.

Figure 20 :
Figure 20: Forming process of the hard sandstone AC-13C.

Table 2 :
Main mineral components of hard sandstone.

Table 3 :
Properties of the hard sandstone coarse aggregate measured using standard test methods.

Table 7 :
One-way analysis of variance (ANOVA) on efects of storage time on the crushing value.

Table 8 :
One-way ANOVA on efects of temperature at 160 °C and below on the crushing value.

Table 9 :
One-way ANOVA on efects of the number of freeze-thaw cycles on the crushing value.

Table 10 :
One-way ANOVA on efects of the number of freeze-thaw cycles on the PSVs.

Table 11 :
One-way ANOVA results on efects of the number of freeze-thaw cycles on the Los Angeles abrasion values.

Table 12 :
Surface free energy of hard sandstone aggregate and limestone aggregate (mJ/m 2 ).

Table 13 :
Results of extracting AC-13C of the hard sandstone aggregate (passing percentage).

Table 14 :
Dynamic modulus of the hard sandstone asphalt mixture AC-13C at 20 °C.

Table 15 :
Phase angle of the hard sandstone asphalt mixture AC-13C at 20 °C.