Abstract
Ice loads are core design criteria for hydropower dams in northern regions. They are responsible for a significant proportion of the design load of small dams that are typically used in Norway. In spite of their importance, the magnitude and regional variability of ice loads are still poorly understood and current regulations rely on limited amounts of field measurements. In order to develop tools and procedures for ice load predictions, numerical simulations have been compared with field measurements. This study presents preliminary results from 3-D numerical simulations of thermal ice loads in a small reservoir. Simulations are driven by measured air temperature data. The impact of inhomogenous boundary conditions is investigated with respect
to snow depth distribution, ice cover thickness, and confinement. Simulations are compared with temperature and stress measurements in the ice. The results show that 5-week period of stresses can be simulated without periodic re-initiation of the model, that periods of disagreement
between model and observations coincide with extreme events due to mechanical forcing or melting temperatures, that snow cover and boundaries have a significant but manageable
influence on ice stress, and that agreement between thermal stresses modeled with a thermoelastic model and measurements is promising.
to snow depth distribution, ice cover thickness, and confinement. Simulations are compared with temperature and stress measurements in the ice. The results show that 5-week period of stresses can be simulated without periodic re-initiation of the model, that periods of disagreement
between model and observations coincide with extreme events due to mechanical forcing or melting temperatures, that snow cover and boundaries have a significant but manageable
influence on ice stress, and that agreement between thermal stresses modeled with a thermoelastic model and measurements is promising.