An experimental study of the relationship between joint spacing and layer thickness

Two methods for measuring joint spacing are described and compared. The area method is a constant for a given outcrop area and is not affected by joint distribution within that area; in contrast, the line method depends on the location of the linear traverse. Two kinds of joint sets are distinguished on bedding surfaces: (1) a poorly-developed set represents the early stages of development when typical joint lengths are less than typical spacing; (2) a well-developed set represents later stages when lengths are much greater than spacing. The area method is applicable to both poorly- and well-developed sets, whereas the line method produces inconsistent results for poorly-developed joint sets. Surface textures on many joints indicate point fracture origins and propagation parallel to bedding, whereas most numerical model studies of spacing assume linear origins and propagation perpendicular to bedding. The laboratory experiments described in this paper do not suffer from these restrictions. An important concept, confirmed during these experiments, is fracture saturation. When the applied strain reaches a certain value, fracture spacing stops evolving and remains nearly constant: these fracture sets are well-developed. Spacing at saturation is a function of layer thickness but is independent of strain, whereas spacing before saturation varies strongly with applied strain. Thus, plotting spacing vs thickness and comparing the slopes of lines fit to such data for poorly-developed joint sets in different layers is unlikely to be a diagnostic test for differences in material properties. On the other hand, spacing may be a sensitive indicator of strain for layers with poorly-developed joint sets. Assessing fracture saturation is a lirst-order consideration when gathering spacing data.