The diversity of methods available to engineers and geologists for studying rock masses is a reflection of the fact that, despite the rapid and encouraging progress made over the few years, there is still no unified and comprehensive method which can be used to predict the over-all behavior of a foundation.
Nevertheless, it is agreed that the fundamental characteristic affecting the engineering properties of rock is fissuration, and this paper describes a few new or improved methods of studying fissuration and its influence on major engineering factors such as permeability, deformability, or internal stresses in the volume.
Conventional tests on samples using thin slices can be improved by staining the slices before cutting with certain ionic dyes which have an affinity for certain types of surface. This shows up the microdiscontinuities.
For example (quoted terms are French trade names): "Turquoise Motylne JSA" will show up porous alteration of basic feldspars. "Vert brillant SA" is used for acid feldspars. "Fushine ASA" will show up voids in all rocks. A mixture of "Turquoise" and "Rhodamine BSA" will show up zones of alteration in all types of feldspars.
With this technique, it is easy to make an optical classification of the different structural types, e.g.: (1) rocks completely free from voids; (2) rocks with microfissures, which in turn can be subdivided into (a) rocks with microfissures secant to the mineral structure, (b) rocks with microfissures parallel to the mineral structure, and (c) rocks with randomly orientated microfissures; (3) porous rocks of the sandstone type; (4) rocks rendered porous by the concentration of small voids in the microcrystalline zones; and (5) rocks rendered porous by the concentration of small voids in large crystals.
Using the optical description of the microfissures (width, length, shape, number, etc.), proper experimental correlation was found between (1) the geometric characteristics of the matrix and (2) the numerical values of porosity and permeability, and their variation with stress. These values were obtained by using Jean Bernaix's method on samples1 which were subsequently stained and cut into thin slices. From the optical observation of a thin slice, it is then possible to predict the mechanical behavior of the sample.
For instance, the color intensity of a sandstone sample determines the porosity value. The dimensions of the stained pores give the permeability value with no marked variation with stress. Impregnation of the black mineral beds of a gneiss covering half the surface of the slice gives a permeability of 10-8 cm/s in the direction of the beds and a slight variation of this with stress.
A dendritic microfissure, 5 to 10 µ wide and uncemented, gives permeability of 10-6 cm/s, varying greatly with stress, whatever the direction in which it is applied.
A linear microfissure,3 µ wide and uncemented, gives permeability of 10 -s cm/s, and marked variation with stress when it is normal to the edges of the fissure.
Two cemented fissures give permeability of 10-9 cm/s with no variation with stress. The same studies have been made with samples of concrete in order to determine the causes of poor values after crushing tests.
