Soils have a variety of mineral compositions. Although a number of thermal conductivity models have been developed for soils, few quantitatively investigated the effect of mineral composition. In this study, the finite element method was employed to estimate the thermal conductivity of dry porous soils (k_dry) considering the impact of mineral composition and porosity. A generalized model is proposed to predict k_dry. The proposed model involves two steps. First, a modified form of Johansen’s model to evaluate the thermal conductivity of soil solid (k_s) was established. The modified form considers a large number of soil samples where the soil composition varies and an empirical formula is obtained. Second, k_dry is observed parabolically decreasing with porosity. Based on the relationship between k_dry and porosity, a generalized model to predict the thermal conductivity of dry soil is proposed, and the empirical parameters for various types of soils are also determined for the sake of engineering applications. The performance of the proposed model is validated by comparing the predicted results with experimental data. A working illustration is exemplified for application of the generalized model.

土壤具有多种矿物质成分。尽管已经为土壤开发了许多热导率模型，但很少有人对矿物成分的影响进行定量研究。在这项研究中，考虑到矿物成分和孔隙度的影响，采用有限元方法来估计干燥多孔土壤的热导率 ( k _dry )。提出了一个广义模型来预测k _dry。所提出的模型包括两个步骤。首先，约翰森的模型的修改形式来评估的土壤固体（热传导率ķ_小号）成立。修改后的形式考虑了大量土壤样品，其中土壤成分变化并获得经验公式。二、k观察到_干燥随孔隙率呈抛物线下降。基于k_干与孔隙度之间的关系，提出了预测干土导热系数的广义模型，并确定了各类土的经验参数，以方便工程应用。通过将预测结果与实验数据进行比较，验证了所提出模型的性能。举例说明了通用模型的应用。