The work described in this paper forms part of a research project which has been evolved over a period of ten years in the Department of Mining Engineering at the University of Nottingham. The broader topic is concerned with establishing reliable theoretical analyses of the behaviour of Rock Masses when affected by Mining (see paper by Berry (1)). One of the major difficulties encountered in the development and use of a theory of ground movement lies in defining for the rocks the physical characteristics and failure criteria associated with the stresses involved. This study is confined to the investigation of the failure criteria associated with brittle materials particularly under the action of direct or induced tensile stresses. Such knowledge is of importance in the analyses since it then becomes possible to define the areas of ground in which fracture has taken place and hence where deformations are no longer in the elastic regime. Quite apart from such a study, there is an ever growing interest in the field of brittle materials testing. Much of the impetus comes from the need to know more about constructional materials and in particular, ceramic.s and concretes. The history of such testing techniques is now well known. Therefore it is not intended in this paper to refer to the development of the various tests, but to describe an attempt to make use of established techniques on coal measure rocks. One of the early snags to be encountered when dealing with brittle materials is that of specimen preparation. British coal measures are no exception to this general statement and much time was consumed in establishing an acceptable technique. This difficulty also had an influence upon the types of tests which could be carried out and it was soon found that with many of the rocks, only cylindrical specimens could be produced at all easily. Experience had already shown that the elementary analysis given in most text books for the diametrally applied load was not sufficient to explain away the phenomena observed. Since for such a protracted series of tests a large number of specimens would be required, it was decided to use hard stone-plaster specimens so that reproducibility of characteristics and quantity production could be achieved. The work has since proved of such magnitude that only after two years experimentation is the use of rock specimens being considered. Recent authors (2. 3) have stated that it is unlikely that strengths computed from different types of test (e.g. flexure, direct pull and diametral compression tests) will be compatible. Our work tends to confirm this view.
Because of the friability of many of the coal measures rocks from which the bulk of our rock specimens are to be taken a special system of specimen preparation was designed (4). This was based upon the fact that some of the rocks encountered break down in water and also that many others which do not actually break down do show a pronounced skin effect after wetting.
