Laboratory characterization and modelling of the thermal-mechanical properties of binary soil mixtures

Abstract

Binary soil mixtures are extensively used in the construction of geothermal-related earth structures such as geothermal energy piles (GEP), ground source heat pumps (GSHP) and earth air tunnel heat exchangers (EATHE). An evaluation of the binary soil’s thermal-mechanical properties is the key process in determining the final performance of geothermal-related projects. Therefore, the thermal-mechanical properties of binary soil mixtures were systematically investigated in this paper. A series of thermal and mechanical property tests was conducted on five sand-kaolin clay mixtures with sand contents ranging from 0% to 100% by dry weight. The experimental results indicated that the sand-clay mixtures achieved the theoretically densest state when the sand content reached the critical threshold. The further the binary mixture’s sand content was from the critical threshold, the lower the mixture’s density was. As the sand content increased, the shear stress-strain curves gradually shifted from strain-softening behavior to strain-hardening behavior due to the decrease in suction stress. The relationship between the sand content and the shear strength of the mixtures exhibited an “S” shape, which is attributed to the interaction between the sand and clay particles and varied with the sand contents. The shear wave velocity of the sand clay mixtures was found to decrease continuously with the increase in sand content until the sand skeleton had formed. In addition, the thermal conductivity of the binary mixed soil changed linearly with the sand content, and the upper bound of the critical threshold interval (77%) was found to separate the two different heat conduction modes. Finally, an elastic shear modulus (G0) model, which correlated to the tangent elastic modulus of the binary mixture (Em), and a more generalized thermal conductivity (K) model were formulated for the binary sand-clay mixtures, and the effectiveness and feasibility of the proposed models were validated by comparing the values predicted with the model and the experimental data.