Efficient numerical methodology for the determination of thermal conductivity of asphalt mixtures

Abstract

Despite the importance of the thermal conductivity (k) to the study of pavement thermal behaviors, it is still a challenging task for engineers to come up with an efficient method for its determination without laborious laboratory measurements. Recently some computer-aided approaches were proposed to overcome the mentioned difficulty but still suffered from time-consuming and inefficient computation. This study presents a hybrid methodology for the determination of k value with a totally automatic program, employing genetic algorithm (GA) to achieving efficient model calibrations, coupled with the finite element simulation tomake estimations of the thermal conductivity of the asphalt mixtures. The improved performance of the proposed method is demonstrated by a 2-D row-column structure model example to achieve a higher accuracy and efficiency existing approaches under the same conditions.