| dc.description.abstract | During lake/sea ice ridging and river ice jamming, broken ice blocks experience transient heat conduction. In present work, this process is represented by submerging a cold fresh-ice block into fresh water at freezing point. The thermal behavior of ice was investigated both experimentally in the cold laboratory of Norwegian University of Science and Technology and numerically with Finite Differential Method (FDM). Most thermal properties of fresh water and fresh ice are well known except the heat transfer between ice and water (convective flux). Ice pieces with varying initial thicknesses and temperatures were applied while ice growth and temperature histories were measured. The physical experiments were performed to obtain a heat transfer coefficient (h), which was essential to calculate the convective flux. The two dimensionless numbers, Fourier and Stefan numbers, were used to study the development of ice temperatures and the ice growth. During the process the convective flux has limited influence on ice temperature while it reduced the maximum ice growth. This is because new ice that forms on the water-ice interface insulates the original ice from the water. In other words, the dimensionless temperature development is governed by the Fourier number alone. The ice growth, on the other hand, is influenced by the convective flux, so that a higher initial thickness and/or a lower initial temperature decreases the ice growth fraction. A balance of latent heat, inertia and convective flux shows that an increase of convective flux directly reduces latent flux and consequently the ice growth. The overall results from simulation and experiments demonstrate that an areal scalar (convective flux) and a volumetric scalar (latent energy flux) cannot be scaled simultaneously. Therefore, the dimensionless ice growth is a function of Stefan number and modified by convective flux, which is parameterized by initial temperature and thickness. | en_US |
