The non linear elastic and anisotropic behaviour found in many rocks has a direct effect on the stress intensity and orientation calculated from stress measurement data. The development of a stress calculation model that allows to take into account simultaneously non linear elastic and anisotropic behaviour gives the opportunity to quantify the error introduced in the evaluation of stresses if these behaviours are not accounted for. Stress measurement simulations on rocksalt and Barre granite allowed a validation of the proposed calculation model.


Les comportements elastique non lineaire et anisotrope que presentent de nombreuses roches ont un effet direct sur l"intensite et I"orientation des contraintes calculees à partir de donnees de mesure de contrainte. Le developpement d"un modèle de calcul permettant de tenir compte simultanement de la non linearite et de I"anisotropie des roches rend possible l"evaluation des erreurs que I"on introduit si l"on neglige cos caracteristiques de leur comportement mecanique. Des simulations de mesure de contrainte sur du sel gemme et du granite Barre ont permis de valider le modele de calcul propose.


Die nicht-linear elastischen und anisotropoachen Eigenschaften von vielen Gesteinen haben eine direkten Einfluss auf die Intensitat und Richtung der errechneten Werte von Felsspannungsmessungen. Ein mathematisches Modell wurde entwickelt welches es ermöglicbt, beide diese Eigenschaften in die Berechnungen einzubeziehen, und auch die inharente Ungenauigkeit durch die nicht- Beruecksichtigung derjenigen in den Berechnungen zu quantifizieren. Simulierte Spannungsmessungen mit Steinsalz und Barre Granit haben es ermöglicht, die Gueltigkeit des vorgeschlagenen mathematischen Modelles zu beweisen.

1- Introduction

The design methodology in rock mechanics requires that the mechanical properties of the rock and rock mass and the in situ stress field be known. The confidence one has in the determination or measurement of mechanical properties is general1y not questioned. Stress measurement results on the other hand, are often suspected of being erroneous because of a number of factors which can be divided in two categories: 1) technical factors; 2) theoretical factors.

If we consider stress measurement techniques requiring stress relief drilling, technical factors include all the possible sources of error related to the experimental measurement of strains or displacements. Means of detecting and correcting these errors have been dealt by many authors (Blackwood, 1978; Gill et al., 1987; Corthesy and Gill, 1990a).

Theoretical factors could be identified as the difference between the hypotheses on which the stress calculation model is based and reality, the usual hypotheses being that the rock is linear elastic, isotropic and homogeneous and reality being that the rock has non linear stress-strain relationships, is anisotropic and. heterogeneous.

The purpose of this paper is to show the influence of anisotropy and non linearity on the in situ stress tensor characteristics obtained from in situ measurements. This is done by comparing the calculated stress tensor using the usual hypotheses and introducing alternatively anisotropy and non linearity in the stress calculation model. The stress measurement technique used to illustrate this is the doorstopper technique.

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