3.1. Effect of Different Filler Types on Water Permeability
The results of the permeability tests provided valuable insights into the impact of permeability on moisture damage resistance in different mix designs with identical air voids (
Figure 5). The findings indicate that the mix designs containing AC60/70 asphalt and coconut peat filler had the lowest permeability coefficients compared to those containing limestone and bagasse fillers. This suggests that the AC60/70 and coconut peat mix had a higher resistance to water penetration than other mixes with the same air voids.
Furthermore, the results also indicate that the AC60/70+CB mixes generally had lower permeability coefficients than the AC60/70 mixes, with the lowest value observed in the mix containing AC60/70+CB and coconut peat filler. However, the k values for AC60/70+CB mixed with limestone filler and bagasse filler were higher than those of AC60/70+CB with coconut peat filler.
These findings are important because they highlight the importance of selecting the right asphalt and filler types in mix designs to enhance moisture damage resistance. The results suggest that using coconut peat as a filler can significantly reduce the permeability of a mix and improve moisture damage resistance. Further research could investigate the impact of other filler types and asphalt grades on the permeability and moisture damage resistance of mix designs with identical air voids.
Permeability tests were conducted on porous asphalt mixtures with air voids of 20% for all the mixtures (
Figure 6). The results show that the AC60/70+SBS with the coconut filler mixture had the lowest k value of 1.508 × 10
−5 cm/s, which was lower than the k values of AC60/70+SBS mixtures with limestone and bagasse fillers, which had k values of 1.834 × 10
−5 cm/s and 1.933 × 10
−5 cm/s, respectively.
This suggests that the coconut filler in the AC60/70+SBS mixture may be more effective at reducing permeability compared to the limestone and bagasse fillers. The lower permeability in the AC60/70+SBS with coconut filler mixture helped improve the moisture damage resistance of the mixture, which is an important factor in asphalt pavement performance. Overall, the results of the permeability tests on the porous asphalt mixtures provide valuable insights into the performance of different mixtures and could inform the selection of optimal mix designs for asphalt pavement applications.
3.2. Relationship between Water Permeability and Moisture Damage Resistance
A prior research study demonstrated that the type of mineral fillers used can impact the resistance of asphalt mixtures to moisture damage [
16]. However, it has not been confirmed if varying filler types and water permeability, while maintaining the same air voids, affect moisture damage resistance. When moisture infiltrates asphalt mixtures, two primary mechanisms can occur: adhesive failure between the asphalt binder and the aggregate surface, or cohesive failure of the asphalt mastic bond due to interaction with moisture.
To examine the potential effect of varying permeability on moisture damage resistance, the tensile strength ratio (TSR) was calculated as an indicator from the indirect tensile strength (ITS) testing.
Figure 7 displays the ratio of tensile strength after wet conditioning of asphalt mixtures with limestone aggregate for all filler types. In dense gradation, the results indicate that the TSR for asphalt mixtures with limestone filler and both types of asphalt binders consistently produced values of 0.83 and 0.86 for AC60/70 and AC60/70+CB, respectively, which exceeds the minimum threshold of 0.7 for typical specifications for TSR. Similarly, the TSR values for AC60/70 with coconut peat were above the threshold value at 0.76 and 0.92 for AC60/70 and AC60/70+CB, respectively. However, AC60/70 with bagasse presented lower TSR values than the specification, while samples of AC60/70+CB with bagasse exhibited good resistance to moisture damage at TSR of 0.89. These values were subsequently analyzed in relation to the permeability coefficient.
In porous asphalt mixtures, the large amount of air voids could be easily compromised by water, but the TSR results of the samples provide insight into their moisture damage resistance. The sample of AC60/70+SBS with coconut peat filler exhibited a TSR result of 0.71, which was relatively good compared to AC60/70 with bagasse and even limestone fillers, as illustrated in
Figure 8.
Table 5 displays the coefficient of permeability (k) for all mixes of dense and porous asphalt gradation samples. The dense asphalt mixture samples with AC60/70 binder and coconut peat filler exhibited the lowest k value of 0.056 × 10
−5 cm/s, while those with limestone and bagasse fillers had k values of 0.064 × 10
−5 and 0.079 × 10
−5, respectively. Similarly, the trend of k values for porous asphalt mixture samples with AC60/70+SBS binder was consistent with dense asphalt mixture samples with AC60/70 binder. The porous asphalt mixture samples with AC60/70+SBS binder and coconut peat filler exhibited the lowest k value of 1.508 × 10
−5 cm/s, while those with limestone and bagasse fillers had k values of 1.834 × 10
−5 and 1.933 × 10
−5, respectively.
However, the asphalt mixture samples of dense gradation with AC60/70+CB binder exhibited a different trend from the others. The k values from lowest to highest were observed in samples with coconut peat (i.e., 0.010 × 10−5), bagasse (i.e., 0.013 × 10−5), and limestone (i.e., 0.041 × 10−5) fillers. Although it is commonly believed that the coefficient of permeability would be similar with the same air voids and asphalt binder and aggregate gradation, the filler type may influence the orientation of aggregates in asphalt mixtures, leading to different permeability.
To investigate the impact of coefficient of permeability (k) on moisture damage resistance, a correlation between TSR (tensile strength ratio) and k was established (
Figure 9). It was hypothesized that a higher k would result in a lower TSR. The findings reveal a robust linear association between TSR and k in both dense and porous modified asphalt binder gradations, with R
2 values of 0.79 and 0.74, respectively. This indicates that TSR and k are related to the same gradation.
The results of this study suggest that varying filler types and water permeability could impact the moisture damage resistance of asphalt mixtures. The TSR results show that the samples with coconut peat filler consistently exhibited good moisture damage resistance, while the samples with bagasse filler presented lower TSR values. The k values of the samples also show that the filler type may influence the permeability of the asphalt mixture.
The correlation between TSR and k demonstrates a strong linear relationship in both dense and porous asphalt mixtures, indicating that they are related to the same gradation. These findings are important for improving the design of asphalt mixtures to enhance their resistance to moisture damage. By selecting fillers that result in lower permeability, the moisture damage resistance of asphalt mixtures could be improved, ultimately leading to longer-lasting and more durable roads.
3.3. Effect of Contact Length on Water Permeability
In this section, laboratory-produced mixes were utilized for the examination of various indices. A total of nine mixtures were created, incorporating three different binders (AC60/70, AC60/70+CB, and AC60/70+SBS), three different fillers (limestone, coconut peat, and bagasse), and two different gradations (dense and porous). The volumetrics and gradation details of each mix are presented in
Table 5 and
Figure 1, respectively.
Previous research by Sefidmazgi et al. (2012) demonstrated a correlation between IPAS aggregate skeleton indices and flow number, with two indices (number of aggregate contacts and total aggregate contact length) exhibiting strong correlation [
6]. Based on this finding, the present study hypothesized that the aggregate skeleton may also influence the permeability of the mixes. The results of the analysis are presented in
Table 6.
To enhance the visibility of the correlation, linear line charts were constructed. The coefficient of determination (R
2) was utilized to assess the degree of correlation between the indices and coefficient of permeability. The correlations between the aggregate skeleton indices and permeability are depicted in
Figure 10 and
Figure 11.
According to the findings, there was a significant linear correlation observed between the number of contacts and contact length in the aggregate skeleton and the coefficient of permeability. The average value of R2 for the dense and porous gradations, indicating the correlation between the number of contacts in the aggregate skeleton and permeability, was 0.933. Similarly, the average value of R2 for the correlation between the total aggregate contact length in the aggregate skeleton and permeability was 0.93. Nonetheless, upon combining the same gradation for analysis, the relationship between the number of contacts in the aggregate skeleton and permeability remained significant. However, there were no associations observed between the total contact length in the aggregate skeleton and permeability. These findings suggest that the number of contacts may be a more important factor than contact length in determining permeability. These insights can help engineers design more effective and efficient permeable pavements and other applications that utilize aggregate materials.