ALBERT

Description / Table of Contents: Summary Exfoliation. Many attempts have been reported in the literature for setting up a theoretical model of the phenomenon of exfoliation which occurs very frequently in nature. Thus far, these attempts have been only very qualitative and no quantitative explanation of exfoliation has ever been proposed. In order to do this, it is first of all necessary to review the phenomenology of exfoliation in a systematic fashion. The latter encompasses features ranging from sheeting on domed granitic mountains of Australia, South America and Africa, to horizontal fissures in bedrock and to the splitting off of plates in sandstone. After the description of exfoliation, the quantitative observations are summarized so as to provide a basis for the critique of the explanations of exfoliation that have been proposed heretofore in the literature. The many investigations of exfoliation showed that the phenomenon is usually of preglacial age and that it is independent of the primary structure of the rocks. Furthermore, the thickness of the plates increases with depth below the surface (cf. Fig. 1) and the ordered orientation of the exfoliation joints which is strictly parallel to the surface, loses itself completely at a depth of about 50 m (in granite). Then, a critical review is given of the attempts at explanations of the exfoliation (mechanical effects of fire, freezing and vegetation; chemical weathering, tectonic forces and stress relief by removal of a load). It is noted that a stress relief theory is presented in the literature as the most likely explanation of exfoliation. Therefore, a simple model corresponding to this type of theory is analysed mathematically: The stability against buckling of an elastic plate under horizontal compression (caused by a former overburden) is investigated. The critical dimensions of the plate as a function of the pressure are shown in Fig. 2 for various possible boundary conditions. It is shown that this model is rather unsuitable even if the tensional strength of the plates is neglected. However, if the exfoliation is considered as an analogon to the well known multiple axial tension fractures in compression-tests (the tensional stresses are induced stresses), then a model is found which has great possibilities for the explanation of exfoliation. Tensional stresses are induced in an uniaxial compression at the boundaries of the (Griffith) cracks; it can attain (depending on the shape of the cracks) similar absolute values as the compressive stress. As soon as the induced tensional stresses exceed the tensional strength of the rock at the most dangerous points, a progressive tension fracture occurs which is aided by notcheffects. The result is the well-known multiple axial tension fracture. The occurrence of this fracture pattern can be prevented by a small lateral pressure. The pertinent calculations for the case of horizontal fissures in bedrock are given in the present paper; Fig. 3 shows the results obtained in graphical form. Our model of tension fracture by induced tensional stresses explains the parallelism between the exfoliation joints and the surface, it explains the increase in thickness of the exfoliated plates with depth, and it explains the disappearance of exfoliation altogether with depth. Finally, a special model is discussed which is applicable under specific conditions: this model explains the exfoliation of plates parallel to a vertical rock wall. Starting at the weak zone at the foot of a wall which is always induced by stress concentrations as shown by Sturgul and Scheidegger (1967), a tension fracture progress behind the wall parallel to its surface, see Fig. 4. Thus, new mechanical models have been found which explain quantitatively the phenomenon of exfoliation.