Modified prediction models for heat and mass transfer characteristics of packed bed at low Reynolds numbers: Incorporating thermophysical property change and particle breakage effect

Traditional correlations for predicting flow, heat, and mass transfer exhibit significant deviations at low particle Reynolds numbers (Re_p ∼ 1), primarily due to neglected thermophysical property changes and particle breakage. Continuing our previous work on crushed particle packed beds, this study first evaluates commonly used prediction models based on particle breakage experiments and pore-scale simulations of packed beds with varying crushed fractions (0∼15%). It is found that, in the prediction of resistance pressure drop characteristics, when the particle Reynolds number exceeds 0.5, the relative deviation of traditional correlations from pore-scale simulation results can exceed 30%. For convective heat and mass transfer characteristics, classical correlations show large deviations under low Reynolds numbers, with the average prediction deviation of up to 6 times. To address these issues, the paper novelly introduced correction factors (ξ_f, ξ_h and ζ_h, ξ_m and ζ_m) considering thermophysical property changes. For intact particle packed beds, the modified models significantly improved prediction accuracy at low Reynolds numbers. The average relative deviations for resistance pressure drop, convective heat transfer, and convective mass transfer characteristics were reduced to 1.5%, 1.8%, and 2.7% respectively. Moreover, by incorporating the crushed fraction into the correction factors, prediction models for crushed particle packed beds with different crushed fractions were constructed. The average prediction deviations for resistance pressure drop, convective heat transfer, and convective mass transfer in these beds were 1.5%, 5.4%, and 3.0% respectively. Practically, the modified models identify a tolerable crushed fraction range (<10%) where heat and mass transfer efficiency could be enhanced by up to 75% while pressure drop remains controllable below 35%, providing quantitative guidance for system maintenance and operational optimization.