Accuracy evaluation of heat pulse method to determine ice thermal conductivity

Abstract

The heat pulse (HP) method is among the most widely used methods to measure thermal properties of frozen soil, snow or glaciers. The phase change or ice melting caused by the released heat of the HP method considerably affects the measurement accuracy. Accurate estimation of thermal properties at high subfreezing soil temperatures (e.g., −5–0 °C) remains as a challenge for decades, given that ice melting effects are too complicated to be accounted in the analytical solutions to heat conduction equation. Ice is the fundamental component that determines hydrothermal properties of frozen soil and snow/ice. Therefore, this study observed the different thermal responses of the ice to 6 heating strategies for the heat pulse (combination of 2 heating durations and 3 heating strengths). Meanwhile, the results of finite element method based on COMSOL numerical simulations and two classical analytical solutions were compared. The results showed that accuracy of the results was improved by optimizing the phase change parameters of latent heat term that is included in the governing equation in COMSOL. The main findings are: (1) The numerical simulations and SLS (short-time linear heat source) method well fit the observed temperature change as a function of time at a distance from the heater, ΔT(r, t) curves, with differences in phase change temperatures; (2) Modification of phase change parameters in COMSOL considerably optimized the ΔT(r, t) curve for heating duration over 15 s; (3) The numerical method effectively improved the calculation accuracy of ice thermal conductivity by 9.66% at −5 °C to −1 °C and 17.53% at −3 °C to −1 °C, respectively.