Two large ice model test campaigns were performed in the period 2007–2010 as a part of a Joint Industry Project. The objectives of the project were to investigate different floater geometries and ice model test set-ups (model fixed to a carriage and pushed through the ice vs. ice pushed towards a floating model moored to the basin bottom) and their influence on the ice failure mode and structure responses in the various tested ice conditions. This paper presents the objectives and motivations for the project, the models tested, the target test set-up for the various tested configurations and the test matrix. Initial results from a fixed model tested in three first-year ice ridges with similar target ice properties are also presented and compared. Fixed models of both deep and shallow water platforms were tested in various ice conditions. All models except one had a downward breaking cone at the waterline. The influences of the waterline diameter, the angle of the downward breaking cone and the vertical cone height on the ice failure mode and the resulting ice load were investigated. Tests were conducted in level ice with a thickness ranging from 2 to 3 m and variable ice drift speeds ranging from 0.1 to 1.0 m/s in full scale values. The models were subjected to tests in managed level ice with varying speeds, ice concentrations and ice floe sizes. Fixed structures were also subjected to testing in typical first-year design ice ridge conditions with ridges of different depths and widths, as well as one multi-year ice ridge. One fixed model was also utilised for testing of the repeatability of scaled ice model testing. Moored models with the same waterline geometry as the fixed models were also tested. The moored models were tested in ice conditions similar to those of the fixed models with the objective of comparing their influences on the ice load due to structural responses.Copyright © 2011 by ASME