The application of Mixed Hybrid FEM in the frost heave prediction verified by FEM and indoor frost heave experiments

- We establish a mixed model for frost heave evaluation by Mixed Hybrid FEM and FEM successfully.
- We point out the advantage of MHF compared with FEM by two types of indoor frost heave experiments.
- We demonstrate the applicability of MHF both in constant freezing rate and fixed boundary temperature conditions.

We apply a modified Finite Element Method which is called as Mixed Hybrid FEM (MHF) for the thermal analysis of frost heave amount estimation, since MHF allows simultaneous computation of objective variable and its derivative with the same accuracy. Freezing rate is one of the key factor which affects the accuracy of frost heave prediction, accurate thermal analysis is highly significant for the frost heave evaluation. Therefore, in this study, we apply MHF for the thermal analysis which considers the phase change and latent heat influence. In order to demonstrate the advantage and the applicability of MHF in the application of frost heave simulation, we conduct two indoor frost heave experiments to verify the numerical simulation results obtained by FEM and MHF respectively. Experiment results demonstrate that, under constant freezing rate condition, MHF and FEM have similar results for both thermal analysis and frost heave amount. However, for the fixed boundary temperature condition, MHF has a slower freezing rate and a much more continuous freezing period than FEM. MHF shows a better accuracy for the frost heave amount evaluation compared with FEM as well. And we believe that this difference results from the accurate thermal analysis by MHF, and we also conclude that the advantage of MHF in the thermal analysis would be much more obvious when it is applied to complicated field conditions. Consequently, we recommend adopting the MHF in the thermal analysis of frost heave evaluation due to its high precision and wide applicability.