Ice load is an important input for floating platforms in polar regions in terms of structural safety analysis and integrity management. As such, the establishment of fast and effective ice load forecasting techniques can guarantee the safe operation of floating platforms.

A high-performance discrete element method (DEM) used for sea ice modeling based on the CUDA-C parallel processing technique is established. The pack ice around a platform during operation is generated using the Voronoi algorithm, and the ice sheet is constructed from sphere elements considering the bond-break effect controlled by the failure criterion. The hull structure is constructed from triangular elements and treated as a six-degrees of freedom system under the action of buoyancy, ice load and restoring force of mooring line. The global ice load on the moored structure is calculated under the continuous action of ice.

The influence of ice thickness is studied statistically and the numerical results validated with field test results from Kulluk.

Ice thickness has a significant impact on the ice load of a floating platform during operation in ice-covered waters, so an ice-breaking service should be supplied to weaken the ice conditions in the operation area.