Author(s): A. Sjolander; S. Li; E. Nordstrom; A. Ansell
Linked Author(s):
Keywords: Prediction; Ice-structure interaction; Ice load measurements
Abstract: Determining a suitable design value for the ice load acting on a concrete dam is still an open research question for the engineering and scientific communities. One potential improvement to today’s use of deterministic loads is to develop a data-driven forecast model for the ice load based on measurements. Previous research has shown that the ice load to some extent depends on the type and size of sensors used. Moreover, the load shows typically large fluctuations over time and depends most likely also on the mechanical properties of the ice and interacting structure and other parameters, e.g., water level fluctuations and ambient temperatures of air and water. Therefore, accurately predicting the ice load poses a significant challenge but will be explored in this paper. Based on several years of collected ice load data using a large custom-made steel panel, different prediction models and their potential practical use will be evaluated. The accuracy of the model is evaluated by comparing predictions to measured data, while the practical use is evaluated based on which parameters are required to accurately predict the load. Tested models include analytical models and regression models.
Year: 2026