Journal of Cold Regions Engineering

The approach for designing a single vertically and laterally loaded pile rests primarily on the local moment capacity and on the overall instability of the pile section. Design codes generally recommend interaction formulas that accomplish some balance between allowable stresses in flexure and allowable axial stresses. For this reason the buckling behavior of the pile has to be investigated. In the past, the determination of the buckling load depended on an equivalent length factor and buckling length (the embedded pile was modeled as a fixed cantilever beam). It is the purpose of this paper to determine a general procedure whereby these quantities are calculated, thus facilitating the analysis in accordance with the existing design codes used in the industry.

Many northern communities rely on open reservoirs as a water source during winter. In the north, open reservoirs can experience ice thicknesses of over 2 m, which can result in a loss in usable storage of up to 60% of the total reservoir volume. This loss increases the required total size of the reservoir and cost. This paper evaluates available methods of reducing the loss of storage to ice. Methods considered included the geometrical optimization of the design, reducing ice thickness, and shortening the period that storage is required. To allow evaluation of many different options, the study concentrated on theoretical aspects, with many options being evaluated with a numerical ice growth model. It is recommended that field tests be conducted on the most promising options. Results indicated that of the geometric factors considered, depth was the most important. To reduce ice thickness, many methods were considered. However only accumulating snow on the surface and the placement of insulation on the ice surface proved to be feasible. The study also found the time period of storage could be slightly reduced by accelerating ice decay in the spring.

This paper presents an approach for incorporating reliability analysis in the prediction of excess pore pressures caused by thawing of frozen soils. The proposed approach is a risk‐based procedure that can be used to develop measures of reliability of the predictions based on the framework of the one‐dimensional theory of thaw consolidation. A first‐order uncertainty analysis is used to estimate the first two moments of the predicted values of normalized excess pore pressure. These statistical parameters are subsequently used to estimate the probability distribution function of the predictions. The types of uncertainties considered in this paper include parameter and model uncertainties. The validity of the first‐order analysis is checked by performing Monte Carlo simulations and comparing the approximate distribution functions. The proposed method is illustrated by numerical experiments with different input conditions and an example application for a realistic design situation.