Physical model tests are a powerful means of obtaining solutions to a variety of engineering problems. The applications in hydraulics and aerospace engineering are prominent, where the use of similitude and dimensionless numbers is well developed. The first step is to understand the mechanics of the process. In the case of ice, the theory has not been developed to the same degree as in fluid mechanics. The use of scale models in test basins has often focused on resistance to ship motion and on flexural failure of the ice. This has been reasonably well addressed. The properties of the model ice have often been modified to permit scaling of flexural strength as well as elastic modulus to achieve appropriate behaviour.
Extension of testing to situations where ice fails in compression or combined flexure and crushing leads to additional complication. At low rates of loading, ice creeps and also demonstrates enhanced rates of creep if the stress is sufficient to cause damage (microstructural change) in the ice. At higher rates of loading, fracture processes result in a localization of loading, and in the formation of high-pressure zones, which have their own special failure process.
In the paper a review of scaled ice testing is given, with associated mechanics including flexural failure. This is followed by a discussion of the failure processes in compression and related mechanics such as creep, damage and fracture. Suggestions as to scaling of these processes are made. An important aspect that is considered is the randomness of ice loads as measured in the full scale. Modelling this aspect and determination of appropriate extreme values is discussed. The Weibull modulus is suggested as an appropriate parameter.