Journal of Cold Regions Engineering

Small cyclic loads, superimposed on a static load, will significantly increase settlement rates of friction piles embedded in frozen ground. An experimental study involving model steel piles with controlled surface geometry (lug size, shape, and spacing) provides information on displacement mechanisms operating at the pile/frozen‐sand interface. Settlement rates due to static and incremental static loads are analyzed and compared to rate increases resulting from superimposed cyclic loads. Variables controlled during these tests include cyclic load amplitude, load application frequency, static load magnitudes, pile surface roughness, sand density, ice fraction, and sand particle size. Typically, a small static load increase of magnitude X produces a displacement rate increase corresponding to a cyclic load with amplitude of $1.4X\text{.}$ Measured creep rates are independent of frequency of 0.1–10 Hz. Larger lugs (more than twice the maximum particle size) densify and partially crush sand particles at their leading edge. The onset of increasing pile displacement rates (tertiary creep or failure) is shown to be a function of lug height and spacing.

By reinforcing ice with readily available alluvium, an economic building material for remote Arctic locations may be obtained. Three‐point bending tests are performed on ice reinforced with sand ($\sim 80\%$ sand by weight) at three test temperatures ($-5°\text{,}$ $-10°\text{,}$ and $-20°C$). All samples fail in a brittle manner. Accordingly, a Weibull analysis of the results is performed. The test results show that both the mean bending strength and the Weibull modulus increase with decreasing temperature. This behavior may be explained by considering the variations of both unfrozen water content of the samples and the fracture toughness of the ice matrix with temperature. A simple model, based on failure arising from the propagation of flaws within the ice matrix, is presented, and it shows good agreement with the experimental results. Further work is needed to refine the model to account quantitatively for the variation of unfrozen water content with temperature.

This paper reviews permafrost conditions from the deeper water conditions offshore through the shallower water near‐shore area to the onshore colder permafrost conditions. Depending on the initial depth to the permafrost surface, some deepening of the thaw line beneath the pipe will result from operating the pipeline at temperatures much above the permafrost melting point. Some preliminary geothermal predictions indicate incremental thaw depths of 0–10 m because of pipe operation, depending on the initial depth to the permafrost table. Uniform thaw subsidence poses less of a concern than differential settlement, which might induce curvatures and strains in the buried line. Differential thaw subsidence resulting from thawing of a buried permafrost zone was modeled using a simple finite element program and a thermal strain analogy to represent the thaw contraction of a buried permafrost layer. The pipe was treated as a passive structural member as a first approximation to obtain an extreme upper bound on the curvatures that result at the pipe location. This approach was used to determine the minimum separation between the pipeline and the permafrost surface, above which thaw subsidence effects on the pipeline can be ignored.

The activation energy must be known in order to relate creep rate to temperature according to the Arrhenius equation. In this note the activation energy for man-made spray ice is determined from primary and secondary (minimum rate) creep data taken from settlement readings and in situ pressuremeter tests at Mars spray ice island. Spray ice is used to construct drilling platforms in the Arctic Ocean, and the creep settlement is an important engineering consideration in design. The value of the activation energy determined from secondary (minimum rate) creep of spray ice was found to be 78 kJ/mol, for a temperature range of -2 to -22°C. This activation energy is within the range of activation energies reported for the creep of polycrystalline ice. The activation energy determined from primary creep data was highly variable and had a value of about 50 kJ/mol.