Asphalt mixtures with high porosities (known as porous asphalt (PA) mixes) are becoming a popular choice among road authorities as it provides better skid resistance while also reducing tire-pavement noises. Towards the design and manufacture of PA mix pavement, the evaluation of the mechanical properties of PA mixes is of great importance. To predict the mechanical properties of PA mixes, micromechanical models have been considered as an effective tool. In most research studies, continuum-based micromechanical models, i.e. the Self-consistent model, the Mori-Tanaka model, etc. are widely used to predict the stiffness of asphalt mixtures. However, the limitation of these models is that they cannot describe the characteristics of individual particles and thus they cannot provide accurate predictions. On the other hand, the discrete-based micromechanical model (DBMM) which simulates a granular material as an assembly of bonded particles seems to be a promising alternative. Limited research studies have focused on studying the utilization and the applicability of this model for asphalt mixes. Therefore, this paper aims to propose a framework to use DBMM and to evaluate its performance in estimating a PA mix's stiffness. Based on the obtained results, both the merits and limitations of this model were highlighted.

Crumb rubber modified bitumen (CRMB) can be regarded as a binary composite system in which swollen rubber particles are embedded in the bitumen matrix. The current study aims to further improve the prediction accuracy of micromechanical models for CRMB by considering the interparticle interactions. To accomplish this goal, two different strategies were used. Firstly, the (n+1)-phase model was applied to the CRMB system by considering the multilayer properties of swollen rubber particles. Secondly, a new micromechanical scheme called the J-C model was used to account for the interparticle interaction issue. Results show that the (n+1)-phase models slightly increase the prediction accuracy but the underestimation of complex modulus at lower frequencies remains unsolved. The J-C model remedies the underestimation of modulus in the low-frequency range by other models and provides an overall improvement for the relative prediction accuracy by properly addressing the interparticle interactions from the perspective of particle configuration.

Crumb rubber modified bitumen (CRMB) can be considered as a binary composite system where rubber particles are embedded in the bitumen matrix. The bitumen-rubber interaction process (mainly swelling) significantly changes the mechanical properties of both bitumen and rubber phases. This study aims to predict the complex moduli of CRMB binders with more representative constituent parameters using micromechanical models. To achieve this goal, frequency sweep tests using a dynamic shear rheometer were performed on the liquid phase of CRMB and swollen rubber samples to represent the essential properties of bitumen matrix and rubber inclusion. In addition, the numerical swelling model was developed to estimate the effective volume concentration of rubber after swelling. Results show that the liquid phases of CRMB are stiffer and more elastic than the neat bitumen while the swollen rubber is softer and more viscous than the dry rubber. The effective volume concentration of rubber can increase to 2.126 times as the blend percentage based on the finite element analysis. Using the liquid phase of CRMB binder and swollen rubber properties as the micromechanical model inputs yield more accurate predictions. The used four micromechanical models predict well at higher frequencies while underestimating the complex modulus at lower frequencies.

The mechanical properties of asphalt mixture are always required for the evaluation of the durability of pavements. In order to obtain these properties without conducting expensive laboratory tests and using calibrated empirical models, research studies have been carried out to develop micromechanics-based models. Continuum-based micromechanical models (CBMM), which are developed based on continuum mechanics, have increasingly been utilized to estimate the mechanical properties of asphalt materials based on the fundamental properties of individual constituents. These analytical models are expected to provide reliable predictions without the need for extensive computational facilities. Although the utilization of CBMM has been presented by several past studies, most of the studies do not provide a concise and critical review of these models. Therefore, in this paper, a complete review of CBMM was presented. Commonly used CBMM were introduced in detail and their advantages and disadvantages were discussed and compared. Comprehensive summaries and critical discussions about their current utilization and limitations for predicting the mechanical properties of asphalt materials were given. Further modifications and new development for addressing the limitations were extensively described and discussed. In the end, research challenges were highlighted and future recommendations from different perspectives were proposed.

ZOAB (Zeer Open Asphalt Beton) is the most widely used asphalt mixture in the Netherlands. As a type of open asphalt mixture, it is known to suffer from raveling distress. In order to analyze the propensity of raveling, micromechanical models are considered effective. However, most of the research work about micromechanical models has focused on dense asphalt mixture and the application of these models on ZOAB mixes has not been paid adequate attention. Therefore, in this research study, the performance of various micromechanical models for predicting mechanical properties of ZOAB was evaluated. The predicted results were compared with the measured values from a dynamic uniaxial compression test. The analysis results showed that none of the applied micromechanical models could obtain acceptable predicted results of the dynamic Young’s modulus and phase angle of ZOAB. On one hand, the Dilute model, the Mori-Tanaka model, the generalized self-consistent model and the Lielens’ model provided lower values of dynamic Young’s modulus and higher values of phase angle, whereas, for the self-consistent model, the predicted results of dynamic Young’s modulus were higher, and the values of phase angle were lower. On the other hand, the shapes of the predicted master curves of both dynamic modulus and phase angle of ZOAB could not match well with the experimental results. The further research on the differential scheme method showed that at lower frequencies the predicted mechanical properties of ZOAB mixes by the applied micromechanical models could not be improved even by following this scheme.

Within the pavement engineering community, open graded mixes (OGM) are regarded as mixes capable of reducing noise and improving wet skid resistance. However, during their design life, these asphalt mixes are known to suffer from a particular distress type known as raveling. This results in a premature failure of a road network. In order to study the propensity of OGM to raveling, homogenization-based approaches are considered to be accessible and effective. One of the most widely accepted homogenization models for asphalt concrete is proposed by Christensen et al. Several studies related to homogenization techniques have been conducted in the past; however, to the best of the authors’ knowledge not a lot of attention has been paid to the study of OGM by means of homogenization models. The other limitation of the Christensen model is that some parameters are difficult to physically understand. Under the above realization, the objective of the paper is twofold: (1) to propose a modification of the Christensen model for OGM; and (2) to verify the modified model’s capability in predicting the mechanical properties of OGM. In general, it was found that once the proposed factor is calibrated for a given OGM by laboratory tests the obtained results are accurate.

Warm mix asphalt (WMA) technology has been increasingly utilised in rubberised asphalt pavements to reduce the production and compaction temperatures and the incidental fumes and odours. This study aims to investigate the high, intermediate and low-temperature performance of crumb rubber modified asphalt binders containing WMA additives. The asphalt-rubber interactions under various mixing combinations of temperature and time were investigated through both microscopic and mechanical methods to obtain the optimum mixing procedure. The effects of WMA additives (wax-based and chemical-based products) on the binder performance were investigated by multiple stress creep and recovery (MSCR) test, linear amplitude sweep (LAS) test and low-temperature frequency sweep test. Results show that rubberised asphalt binders significantly improve the binder performance of base asphalt at different temperature ranges. The effects of WMA additives on binder performance varied with base asphalt and rubberised asphalt binder. In addition, the nonrecoverable compliance difference was found not suitable to characterise the stress sensitivity of rubberised binders and the difference in the nonrecoverable compliance for an incremental change in applied stress was proved to be a more accurate alternative. For the cyclic LAS test, the failure energy was found to have a strong correlation with the predicted fatigue life using simplified viscoelastic continuum damage analysis and therefore can be considered as a simple indicator for binder fatigue performance ranking. Relaxation modulus and rate derived from low-temperature frequency sweep tests produced comparable results for ranking the low-temperature performance of different binders. It is feasible and promising to use a unified DSR test methodology to characterise the binder performance covering the whole service temperature range.