ABSTRACT
Most of the existing prediction models expressed soil thermal conductivity k as a function of water content, dry density, degree of saturation, or a combination of these factors. However, such models do not explicitly account for soil stress state and mineralogy. Based on thermal conductivity tests performed on specimens subjected to one-dimensional consolidation, recent research reported k values for saturated fine-grained soils with different mineralogical compositions. The present research employs available data to correlate k with consistency index I c at different values of water content along the normal consolidation line. Regression analysis of k values indicates distinct correlations for both transient-state thermal conductivity k tr and steady-state thermal conductivity k st with I c. Such a correlation allows to infer ground thermal conductivity profile through routinely available site investigation data. This paper also investigates comparative performance of few thermal conductivity models in predicting measured values of k for saturated fine-grained soils.
