This thesis documents the development of a new modelling approach for first-year ridge keel loads. The development involves a detailed review of previous and new ice rubble indentation and shear strength experiments. A systematic regression analysis of compiled laboratory data sets is used to establish a basic approach to keel load modelling, analogous to that for soil retaining problems. Experiments pioneering the modelling of first-year ridge keels with sand are also described. The dry sand tests afforded a high degree of control which led to the development of a new sand force prediction model that was adapted and calibrated for ice keel modelling.

The advantage of the new first-year ridge keel load model is that new effective structure width and keel shape models are utilized, ridge width is factored in and surcharge effects are considered. The model shows excellent agreement with a large body of new experimental data and the best field data available. Also, it is closed-form, has been successfully applied to both vertical and conical structures, and is based on fundamental earth pressure equilibrium mechanics as are other approaches already in the literature. A further advantage is that ice rubble shear strength yield criteria used in the model have been thoroughly examined so that associated parametric uncertainties are quantified and reduced. An in situ technique for testing the shear strength of ridge keels is developed and direction for future field work and modelling efforts is given.