Applicability of traditional ice load measurement systems on longitudinally framed ships

Abstract

This thesis evaluates the applicability of the traditional Influence Coefficient Matrix (ICM) method to ice-load measurement on a longitudinally framed ship structure. ICM is a matrix that creates a relationship between structural responses and external loads. A finite element model was developed in Abaqus and both shear-strain and normal- strain responses were used to establish ICMs. Two rectangular loading patch sizes were investigated, which are 400 mm (transverse) × 800 mm (longitudinal) and 200 mm (transverse) × 400 mm (longitudinal). The total equivalent load was 1 kN in each case. The errors varied systematically with whether the loading centre was located on the stringer, a longitudinal stiffener, a plating bay, or the transverse web frame. The normal- strain-based measurement also produced large positive and negative values. These results suggest that the ICM may not be applicable to loading cases different from those used in its construction. Local load transfer and the direct inversion process may also increase the measurement error. The present ICM formulation is therefore not reliable over all studied loading conditions, although accurate measurement remains possible at individual positions. Future work should construct the ICM using a wider range of loading positions and patch sizes; constrained or regularized inverse approaches are recommended.