Large natural resources in the Arctic region will in the coming years require significant shipping activity within and through the Arctic region. When operating in Arctic open water, there is a significant risk of high-energy encounters with smaller ice masses like bergy bits and growlers. Consequently, there is a need to assess the structural response to high energy encounters in ice-infested waters. Experimental data of high energy ice impact are scarce, and numerical models could be used as a tool to provide insight into the possible physical processes and to their structural implications. This paper focuses on impact with small icebergs and bergy bits. In order to rely on the numerical results, it is necessary to have a good understanding of the physical parameters describing the iceberg interaction. Icebergs are in general inhomogeneous with properties dependent among other on temperature, grain size, strain rate, shape and imperfections. Ice crushing is a complicated process involving fracture, melting, high confinement and high pressures. This necessitates significant simplifications in the material modeling. For engineering purposes a representative load model is applied rather than a physically correct ice material model. The local shape dependency of iceberg interaction is investigated by existing representative load material models. For blunt objects and moderate deformations the models agree well, and show a similar range of energy vs. hull deformation. For sharper objects the material models disagree quite strongly. The material model from Liu et.al (2011) crush the ice easily, whereas the models from Gagnon (2007) and Gagnon (2011) both penetrate the hull. From a physical perspective, a sharp ice edge should crush initially until sufficient force is mobilized to deform the vessel hull. Which ice features that will crush or penetrate is important to know in order to efficiently design against iceberg impact. Further work is needed to assess the energy dissipation in ice during crushing, especially for sharp features. This will enable the material models to be calibrated towards an energy criterion, and yield more coherent results. At the moment it is difficult to conclude if any of the ice models behave in a physically acceptable manner based on the structural deformation. Consequently, it is premature to conclude in a design situation as to which local ice shapes are important to design against.