Safe and efficient structural design requires knowledge of loading conditions. The importance of understanding the magnitude of the local ice-induced load on the ship structure is acknowledged due to the increasing maritime activities in waters with seasonal or year-round ice coverage. The stochastic nature of the ship-ice interaction process has emphasized the importance of full-scale measurements. The local ice loads are commonly determined by measuring hull response and utilizing inverse engineering to define load from the response. Shear strain difference-based measurements combined with the Influence Coefficient Matrix (ICM) to define strain-load relation are state-of-the-art techniques for indirect ice load measurements. However, this technique has shown limitations that can affect the measurement accuracy. To improve the understanding of the uncertainty related to the method, this paper studies the impact of structural details on the uncertainty of transversely-framed ships. At first, this paper defines the sufficient extent of the numerical model for the construction of ICM using FE analysis. The study shows that the extended model is less sensitive to the selection of boundary conditions. Then, the model is applied to investigate the outcome by changing the frame cross-sectional profile from the flat bar to bulb flat. It is observed that the asymmetric profile can introduce additional uncertainties, especially under the single-sided instrumentation. Finally, a discrepancy found from the previous research and an uncertainty reduction method are discussed. As the study focuses on the response of the structure within the linear elastic domain, the first-year ice class design criteria defined by the Finnish and Swedish Ice Class Rule (FSICR), the approach of decomposing the larger load patch to smaller ones becomes possible and serves

Peer reviewed