Real-time reconstruction of thermal boundary condition of porous media via temperature sequence

Porous media are widely applied in numerous technological field due to its excellent thermal, mechanical and exchange properties. Because of involving multi-scale and multi-effect coupling, only a few studies are dedicated to heat flux estimation of porous media. In this study, on the basis of the finite volume method (FVM) and discrete ordinate method (DOM), a novel method is proposed to introduce the geometric analytical formulas of thermal properties into the multi-scale heat transfer problem of porous materials. Then, the temperature sequence in a short time are used as measurement information for inverse analysis. The time-varying heat flux of NiCrAl porous materials, copper and aluminum are reconstructed based on Kalman filter coupled with the recursive least square estimator (KF-RLSE). In addition, the sensitivity of the photothermal properties to temperature response information is analyzed. The effects of porosity, scattering albedo and specularity parameter on the accuracy and stability of the reconstructed heat flux are investigated. The numerical results with the noise level 0–0.3 show that the predicted boundary heat flux is less than 0.2 for NRMSE and more than 0.99 for the correlation coefficient, proving the validity of the proposed method.