Author(s): Marnix Van Den Berg Raed Lubbad Sveinung Løset

Linked Author(s):

Keywords:

Abstract: Loads from broken ice may be regarded as the governing design load for structures supported by ice management. Ice-tank tests are often used in the assessment of loads from broken ice. This study uses measurements from an ice tank in combination with results from numerical simulations in order to assess the uncertainty in results from ice-tank tests with broken ice. The numerical simulations are performed with the Simulator for Arctic Marine Structures (SAMS). The ice-tank measurements belong to a test that was performed at Hamburg Ship Model Basin (HSVA) in 2012. The test involved a multi-leg structure in 60% (initial) ice concentration. The test was part of a test campaign supported by the HYDRALAB-IV European research grant. In this study, the ice-tank test is reproduced numerically for 300 times. Each time, a different initial floe position is used. The results show a strong dependency of the simulated ice loads and the governing interaction mechanisms on the initial conditions (i.e., the initial floe positions herein). A strong dependency of ice tank test results on uncontrolled initial conditions, such as the initial floe positions, leads to a high random uncertainty in the test results. This random uncertainty should be considered in the statistical interpretation of test results from ice tank tests with broken ice.

DOI:

Year: 2020