ABSTRACT: The key question addressed in this paper is the predictability of the time-dependent deformation of openings in salt. Constitutive laws are applied to predict time-dependent closure of boreholes in salt specimens subjected to various load configurations. Rheological, structural and empirical models have been formulated from creep tests, strain and stress rate controlled tests, cyclic loading' tests and constant strain-rate triaxial tests. Uniaxial, triaxial and polyaxial borehole closure testing has been performed to assess the validity of each model.
1 INTRODUCTION
The deformation of underground openings in salt is of importance in a number of applied rock mechanics problems. Salt remains a primary candidate for nuclear waste disposal. Salt deformation will determine loading of emplaced waste, will affect opening, backfill, and seal deformation and loading, and hence will influence post-closure isolation performance. Other examples of the significance of salt deformation include loading of borehole casing, stressmeters, shaft liners, and bulkheads. Given the likelihood of increased utilization of underground space in salt, with increasing performance demands on design, it is clear that the predictability of the deformational behavior of salt is a key question in rock mechanics. Considerable uncertainty remains about the description, the fundamental mechanics, and, in particular, the predictability of salt creep (e.g. Baar 1970, Mandin et al. 1984, Nelson and Kelsall 1984). Factors influencing the creep strain rate of natural salt are not only the current differential stress but also the temperature, moisture content, humidity, grain size and boundary, impurity, scale and stress history (Senseny 1984, Lindner and Brady 1984). These numerous and complex influences, coupled with inconclusive methods to derive multiaxial formulations make the prediction of the long-term in-situ behavior difficult. Quantitative corrections of these factors based on thermomechanical and physical concepts have been attempted and lead to formulation of very complex models. A fundamental, but rarely addressed, question is whether constitutive models developed from one type of experiment (e.g. conventional creep testing) can describe the behavior of salt samples, under the same scale, environmental and past history conditions, for other types of experiments (e.g. borehole closure testing). The objective of this work is to assess the validity of some constitutive models for salt by determining model parameters from experiments under one set of boundary conditions and applying them to predict behavior under a fundamentally different set of boundary conditions. The effort involves creep and rate-controlled tests, construction and formulation of constitutive models for salt, and, most importantly, borehole closure experiments for assessing the validity of a broad range of the models.
Figure 1. Experiments on salt used to formulate constitutive laws and to evaluate their validity.(available in full paper)
2 METHODS OF INVESTIGATION
Figure 1 summarizes the experimental program. Viscoelastic and viscoplastic models and the potential and structural laws are used to describe the mechanical behavior of salt. The data for the model formulation are obtained from creep tests, constant stress rate uniaxial tests, strain-rate controlled uniaxial and triaxial tests, and cyclic loading tests. Model parameters derived from various sets and subsets of experiments are compared, and the comparisons are used to determine the validity and limitations of the models under different stress levels, strain and stress rates and testing periods.
