Search

This work focuses on the use of a heterogeneous generation method in order to develop a micromechanical model of Hot Mix Asphalt (HMA) that takes into account the effects of characteristics at the interfacial zone (IZ), where aggregate surfaces come into contact with the mortar matrix. Numerical calculations are conducted to investigate the effect of IZ mechanical characteristics on both the overall viscoelastic response of HMA and the local scale behavior. First of all, the HMA is treated as a biphasic material composed of mortar and granular inclusions. Next, a third phase is added between these two at each point of aggregate/mortar contact. Elastic and viscoelastic behaviors are assigned to the aggregates and mortar, respectively. The mechanical behavior of IZ varies between that of the inclusions and the matrix. At the global scale, these numerical results demonstrate that the use of triphasic micromechanical models improves the predictive accuracy of the dynamic modulus. At the local scale, a significant strain variation was found in each model. These results justify our triphasic modeling approach whenever local zone characteristics are the subject of interest. [ABSTRACT FROM AUTHOR]

• Performing temperature sweep test to generate master curves for viscoelastic properties of bitumen. • Optimizing the number of frequencies and temperatures to develop master curves of asphalt mortar and mixture. • Reducing the time and cost of dynamic modulus test without missing the accuracy of master curves. This study aimed to assess the feasibility of substituting temperature sweep tests for frequency sweep tests and explore the potential of utilizing fewer temperatures and frequencies in dynamic modulus tests to formulate viscoelastic properties master curves. The results indicated that master curves derived from temperature sweep tests mirrored the trends and values obtained from frequency sweep tests, presenting a viable alternative for constructing viscoelastic properties master curves of bitumen. Furthermore, models fitted to dynamic modulus test results at three temperatures and three frequencies exhibited comparable primary parameters, values, and trends to those derived from the complete dataset. The model fitted on reduced data demonstrated robust performance with correlation and R-squared values exceeding 0.95. Considering a marginal increase of about 5% in Residual Standard Error (RSE) values when using the reduced dataset, generating master curves with fewer test results underscored the feasibility and reliability of the original master curves. [ABSTRACT FROM AUTHOR]

• Measuring LVE properties of bituminous composites through a multiscale simulation. • Multiscale simulation has more accuracy than the generalized self-consistent scheme. • 3.35 mm is the best mortar cutting size to find mixture behavior with NMAS of 19 mm. Viscoelastic properties of asphalt mixture were studied through a multi-scale approach, including bitumen, mastic, mortar, and mixture length scale. Biphasic and triphasic finite element modeling were employed to construct the heterogeneous models of bituminous composites, including matrix, inclusions, and air voids. The 2S2P1D model consisting of two springs, two parabolic creep elements, and one dashpot was fitted on the output of the numerical model, i.e., dynamic modulus (E*) and phase angle (δ), to accurately determine the Prony series. Then, the data of each scale was utilized as a matrix for the next scale's model. Three cutting sizes, including 3.35, 2.36, and 1.7 mm were selected based on the previous experimental study of Khodadadi et al. (2021) to generate the mortar model and find the best length scale of the matrix. In addition, the numerical results were compared with the results of the generalized self-consistent scheme model as an analytical approach. Both numerical and the generalized self-consistent scheme models yield adequate results; nevertheless, the numerical solution has more accuracy than the analytical model. The cutting size of the mortar scale had a considerable effect on the LVE properties of the mixture. The result showed that a cutting size of 3.35 mm is more suitable for a mixture with a nominal maximum aggregate size of 19 mm. These findings were consistent with the experimental results. [ABSTRACT FROM AUTHOR]

• Using a multi-scale approach to find viscoelastic properties of three scales: mastic, mortar, and mixture. • Threshold stress is evaluated to characterize linear viscoelastic properties of the mortar scale. • Mortar with the cutting size of 3.35 mm has the closest behavior to the asphalt mixture with an NMAS of 19 mm. The mortar component plays an essential role in forming different asphalt mixture distresses. The fine aggregate size affects the viscoelastic behavior of mortar. In this study, three cutting sizes of 3.35, 2.36, and 1.7 mm, on the aggregate grading curve were considered to construct the mortar scales, namely mortar (I), (II), and (III), respectively. With a unique loading mode, a frequency sweep test was performed on four different length scales, including bitumen, asphalt mastic, mortar, and asphalt mixture, to make a direct comparison through a multi-scale approach. Also, a stress sweep test was conducted to find the mortar scale's linear viscoelastic domain. The results indicate that different mortar types have variable linear viscoelastic domains. The master curves of dynamic modulus (E*) and phase angle (δ) show that adding filler to bitumen has more effect on the E* than the δ. Although scaling up from the mastic to mortar (II) causes a significant difference between the viscoelastic parameters, the results remained approximately the same. The main gap between the mixture and mentioned sub-scales is filled with mortar (I). While the Cole-Cole diagrams demonstrate that the bitumen to mortar (II) scales have the same behavior, the closest scale to the mixture is the mortar (I). Therefore, using a cutting size of 3.35 mm is recommended to fabricate mortar (for asphalt mixture with a nominal maximum aggregate size of 19 mm) instead of 2.36 mm, which was suggested in previous studies. [ABSTRACT FROM AUTHOR]