Engineered structures such as mines, shafts and tunnels, and storage caverns for hydrocarbons, chemicals and brine are being built in natural rock salt formations in increasing numbers. In addition, salt formations are being considered as potential hosts for mined, nuclear waste repositories. The popularity of these formations for engineering purposes stems from the many favorable commercial, physical, and thermal characteristics attributed to the material itself. Not only have the formations been stable and free of dissolution for hundreds of millions of years, but the salt is fairly easily mined and has very low permeability, low water content, and high thermal conductivity. Another characteristic of salt is its tendency to flow or creep when subjected to a shear stress. Although this behavior may be desirable in certain instances, it may be detrimental in other instances since the creep rate may be large enough to produce deformations that tend to reduce the volume of storage caverns or to restrict access to mining operations. In either case, the engineer is faced with the problem of accurately predicting the long-term structural deformations under various combinations of stress and temperature and is required to compare these deformations with long-term design criteria. In the laboratory, deformation-versus-time curves obtained at constant stress and temperature show that creep of salt is similar to creep of many other crystalline solids. Figure 1 shows a typical deformation-versus-time curve for salt. When a shear stress is first applied, the rate of deformation is high, but decreases monotonically. This initial behavior is called transient or primary creep. At some time, the deformation rate no longer decreases but continues at a constant rate. In some experiments, a third regime called tertiary creep follows the steady-state region. Tertiary creep usually occurs at low mean stress and low temperature and is characterized by accelerating creep rates and, ultimately, specimen failure (Wallner, 1981). The duration of transient creep is usually quite short, lasting only a month or two at most. The duration of steady state creep can be very long, however, especially if low mean stresses and low temperatures are absent. Therefore, accurate prediction of long-term deformations requires that the steady-state creep rate be known precisely as a function of stress and temperature. For cases when the load on the structure or temperature remain constant or change very slowly, the creep straining of salt can be approximated by a steady-state-only model. Elastic and transient creep strains associated with constant or slowly changing load and temperature are usually very small compared to strains that result from steady-state creep at the long times of interest for these structures. Constitutive models have been developed to relate the steady-state strain rate to the current stress and temperature. The values of the parameters in the models are determined by fitting the model to measured steady-state rates determined at different stresses and temperatures. This paper compares two constitutive models that have been proposed for the steady-state creep of salt in the ranges of stress and temperature of interest to geo-engineers.

This content is only available via PDF.
You can access this article if you purchase or spend a download.