Thermal conductivity of crumb-rubber-modified mortar using an inverse meso-scale heat conduction model

Abstract

Abstract This paper describes a 3D meso-scale numerical model employed to estimate the effective thermal conductivity (ETC) of crumb-rubber-modified mortar composites, whose individual components are represented by random parking of poly-dispersed spherical particles. The meso-scale structure, characterizing the ETC, is calculated though a finite difference method, where the thermal conductivities of the individual components are obtained by a parameter estimation method. The results indicated that a simultaneous estimation of two parameters lead to an increase of the accuracy error from 8% to about 30%, because of an ill condition and strong correlation among them. Individual parameter estimation generally requires additional experimental tests. Therefore, two types of simplified mixtures were used: sand-agar-air and rubber-agar-air, to enable an individual estimation of the conductivity of quartz sand and crumb rubber components. Finally, the ETC of crumb-rubber-modified mortar is predicted for various replacements of sand by crumb rubber up to 15 vol%. The simulated results are validated by a transient hot disk measurement system. The results of the numerical model calibrated by the independent-parameter estimation method, showed a good agreement with the experimental data, and clearly outperformed the considered analytical models.