A distribution of chord lengths, generated from random normal intersections of a series of parallel, circular shaped log-normal discs, shows good correlation with field joint-trace populations. Since the probability of intersection is heavily influenced by the size of the disc, a change in the standard deviation of the population of parent diameters affects the relative value of the derived logarithmic mean chord. This feature throws serious doubt upon the validity of applying a fixed positive weight factor to convert the field mean trace length of a joint set to an inferred mean diameter for the whole joint population.
Rock joints can be considered as extremely thin disc-like breaks which cut through rock masses. Individual joints, of a single orientation population, are generally neither distinctive nor strictly planar. As a consequence it is extremely difficult to determine the limits and shapes of individual joints. Excavation Of small joints, or extremely closely spaced drilling across joint planes, can provide information to define joint limits. In most cases, such methods of investigation would be inordinately time consuming. Where diversely oriented exposures are available it is possible to obtain indications of the shape of the joints and of the total length distribution of the traces. In the present study an attempt is made to use the indicators of shape and trace length to establish a representative numerical model for the derivation of trace lengths from known parent populations. Subsequent comparisons of field and model distributions are used to evaluate the nature of the actual parent population of joint discs.
Evidence has been presented (Ref. 1) to show that joint samples from the CSA Mine, Cobar, New South Wales, and de Beer''s Mine, South Africa (Ref. 2) are approximately equi-dimensional in outline.
