Applicability of the Walder-Hallet frost fracture model to laboratory cyclic uni- and bi-direction freeze-thaw of limestone and sandstone

Abstract

Due to changing climate conditions, new geographic areas are being impacted by diurnal and seasonal freezing and thawing conditions. Many geologic materials in far north latitude areas, that had not undergone significant freeze and thaw previously, are now expected to experience those conditions more often. The bedrock in these northern regions is often used as the foundation for many buildings and other infrastructure, and so it is extremely important to understand the integrity of this material with the changing conditions. Walder and Hallet created a theoretical model to analyze the temperature of fine cracks within a rock when subjected to freezing conditions, and the growth of fine cracks due to thermally-induced water migration followed by freezing. More recently, Dr. Murton conducted multiple cyclic unidirectional and bidirectional freeze-thaw experiments in the laboratory that simulated active layer rocks both with and without permafrost below. These experiments measured temperature and pore pressure of the rock, and monitored the formation and growth of macroscopic (i.e. observable) cracks. Using Walder and Hallet's model as a starting point, I have created a numerical model to analyze the cyclic fluctuating temperature conditions used by Murton in his experimental work, but is not accounted for in the original Walder and Hallet model. I then compared his laboratory results with the numerical model predictions of temperature and location of observable crack growth. This required adjusting some model parameters used by Walder and Hallet to correspond with the experimental conditions and geological materials used. I found that the model predicts the correct location of maximum cracking, and the time of observable crack growth, for the unidirectional experiments. However, it predicted nearly the opposite of the laboratory results for the bidirectional experiments. To obtain these numerical results, I had to adjust parameters that attempt to describe the flow resistance within a fine-grained freezing rock material; a difficult and little understood phenomenon. Future work should focus on improving some of the original model assumptions that do not apply to most experimental situations including those of Murton. These include the angle of cracks, and the potential interaction between adjacent cracks. While the results of this numerical model did not predict all the observed results of Murton's experiments, it has shown what portions of the numerical model appear to work correctly, and what assumptions from the original theoretical model by Walder and Hallet need to be adjusted and improved.