Author(s): Iman El Gharamti; Jaakko Heinonen
Linked Author(s):
Keywords: Ice loads; Flexural failure; Croasdale model; Parametric study
Abstract: In ice-covered areas, ice interacts with offshore structures resulting in different failure modes. When interacting with vertical structures, ice fails by crushing. In contrast, conical structures lead to flexural failure of ice and lower loads than those from crushing. Various models have been developed to predict the ice loads. The Croasdale model, adopted in ISO 19906, calculates the flexural ice loads on conical structures as a sum of five distinct components, addressing the entire process: breaking the ice sheet in bending, pushing it through the rubble, riding-up of ice blocks along the cone, lifting the rubble on top of the advancing ice sheet, and turning of the broken ice. However, implementing this model demands several input variables, including mechanical ice properties and parameters specific to the interaction conditions. The values of these variables are not always readily available and need estimation. To assess the Croasdale model in predicting the flexural failure loads, this study incorporates it into a probabilistic framework using Monte Carlo simulation. A sensitivity analysis was conducted to examine the impact of the input parameters on the resulting predictions. The parameter ranges were selected based on literature data, revealing that some factors exerted minimal influence on load predictions. Generally, the parameters affecting the rubble accumulation process are the controlling ones. Notably, ice thickness, ice rubble ride-up height, cone-ice friction, rubble cohesion, and rubble repose angle significantly affected ice load outcomes. Hence, when employing the Croasdale model to calculate ice loads in a particular scenario or site, special attention should be given to the determination of these influential parameters. While the findings reported here are specific to the examined structure and might vary with different cone geometry configurations, this study provides valuable insights to estimate various ice and ice-structure properties.
Year: 2024