A cycle of laboratory measurements, constitutive modeling and validation calculations is described as a basis for excavation design in rock salt. The results of triaxial creep tests were fitted by a creep model with exponentially decaying transients and power law thermally activated steady state creep which are functions only of deviatoric stress and temperature. The corresponding three-dimensional constitutive equation, including elastic terms, is then applied in finite element simulations of the known closure of an isolated mine drift.


Ein Zyklus von Labormessungen, konstitutiver Heschreibung und Rechnungen wird vorgestellt als Grundlage fuer den Entwurf untertaegiger Hoghlraeume in Salzgesteinen. Die Ergebnisse triaxialer Kriechversuche wurden im Ausgleichsverfahren zur Bestimmung eines Kriechgesetzes behandelt. Danach ist das Uerbergangskriechen als Exponentialfunktion und das stationaere Kriechen als Potenzfunktion mit thermischer Aktivierung gegeben. Ein entsprechendes dreidimensionales Stoffgesetz wird dann unter Einschluss elastischer Terme mit Hilfe des Verfahrens der finiten Elemente angewendet, um die Konvergenzen in einer isolierten Strecke untertage zu berechnen. Die errechneten Werte werden mit bekannten in situ Messungen verglichen und bestaetigt. RESUME: Cette communication decrit un cycle de mesures en laboratoire, des modèles constitutifs, ainsi que les calculs confirmant et servant de base au dimensionnement dune excavation dans le sel-gemme. Les resultats d''essais triaxiaux de fluage furent ajustes à un modele mathematique comportant une phase transitoire declinant exponentiellement, ainsi qu''un fluage thermique permanent, suivant une puissance. Ces deux lois sont fonctions seulement de la contrainte deviatorique et de la temperature. L''equation tridimensionnelle constitutive correspondante, contenant aussi les termes elastiques, fut alors appliquee à des simulations par elements finis dune galerie de mine isolee dont la convergence etait connue.


The design of a radioactive waste isolation facility and of oil storage caverns in the U.S. is supported by thermostructural design calculations using a variety of finite element codes. This paper describes three prerequisites for the successful application of such calculations. 1. Laboratory measurements of the *This work performed at Sandia National Laboratories supported by the U.S. Department of Energy under Contract Number DE-AC04–76DP00789. thermomechanical properties of rock salt, 2. Constitutive modeling of rock salt creep, and 3. A comparison of design calculations with in situ measurements. Descriptions of laboratory measurements and material modeling include a discussion of important implications and uncertainties of the proposed creep model which bear on its use for arbitrary stress and temperature histories. The uncertainties may also affect the extrapolation of the model to low stresses which are probably encountered around underground structures after long times.


Thermomechanical properties of rock salt are measured in specially designed machines using specimens which measure up to 10.8 cm in diameter and 21 cm in length. Details of the apparatus and procedures are described elsewhere (Wawersik and Preece, 1982; Wawersik, 1979). Short-term tests are conducted to determine the elastic properties of salt, failure envelopes and short-term inelastic properties. Short-term tests are carried out in hydrostatic compression (al = 02 = 03), triaxial compression (al > 02 = 03) and in triaxial extension (al = a2 > 03) at constant stress rate up to 6.5 MPa/s, at constant strain rate up to 10 -2 1/s and at constant stress. These tests include stress paths which are not ordinarily realized in axisymmetric experiments, for example, principal stress variations at constant mean stress and confining pressure variations at constant maximum compression 01 (Wawersik and Hannum, 1980). The maximum confining pressure and the maximum temperature are 70 MPa and 250°C. Creep experiments are performed to ascertain the long-term response of salt. Creep experiments are conducted at approximately constant true stress by means of frequent adjustments of the applied loads. In past tests, stress variations of up to ± 10% were experienced in confining pressure and deviatoric stress, either because of fluctuations in confining pressure or because of rapid creep with attendant changes in sample dimensions. Recent improvements limit confining pressure variations to ± 1% (±.04 MPa) and deviatoric stress variations to t 3% (±.2 MPa). Experimental measurements include test time, principal stress difference T = (al - 03), confining pressure 03, true (logarithmic) axial and radial strains el and e3. Strains are determined indirectly by monitoring changes in sample length and sample diameter. Maximum strains are limited to 35 percent. Finite element calculations have confirmed that the strains inferred up to these levels are good measures of the homogeneous strains under ideal conditions with no frictional restraints (Wawersik and Preece, 1982). Creep tests fall into two categories. 1. Single stage tests where each sample is subjected to one constant stress and temperature. 2. Multistage experiments or "change tests" where principal stress difference, confining pressure or temperature are changed from time to time. These changes may occur upward or downward. Accordingly, for example, a distinction is made between stress increment tests and stress decrement tests.

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