Applications of rock mechanics to the design of engineering structures in rock involve the assumption that the stresses are due entirely to the present applied loads. For instance, it is assumed that the vertical normal stress in undisturbed rock free of topographic effects is equal to the weight per unit area of the overburden. In fact, vertical dilatational strains associated with stresses of more than 100 times the present overburden pressure have been measured by strain-relief methods on outcrops1,2 . The notion of residual strain (stress) or latent stress 3 seems unassailable; and indeed this has long been recognized by metallurgists. It is difficult, however, to assess the residual component of the in-situ state of elastic strain with conventional strain relief methods employed in mines or down boreholes. That is, in principle, the in-situ state results from the superposition of present applied loads (natural or maninduced) and potentially recoverable elastic strains (stresses) that exist in a given volume of material even though there are no external forces across the boundaries of the body (Voight, 4 and Fig.. 1). Accordingly, only in unloaded blocks can the true residual strains be measured by strain relief methods.
Using X-rays, the elastic strains locked within the quartz grains or crystals of a rock can be measured. These strains are known to persist even when the rock is reduced to small chips but to vanish when the rock is disaggregated and the grains are powdered. Although we have just begun to investigate the mechanisms by which the strains are stored (see Preliminary Results), we believe that "pressure solution" at grain boundaries, secondary cementation under load, permanent ("plastic") deformation, and temperature changes are possible processes. In any event, in our study of naturally deformed rocks, it is significant that we are working with small chips removed from the outcrop and hand specimen. Clearly there are no stresses across the boundaries of these chips, and in all paleostresses.
Fig. l--Components of in-situ state of elastic strain. (Available in full paper)
X-ray diffraction can be used to measure elastic strains because it provides a means for accurately recording the spacings between layers of atoms that constitute a crystal structure. It is well known that layers of atoms will diffract X-rays if the Bragg condition is satisfied. Thus, if a polished surface of an aggregate of crystals is inclined at the appropriate angle ¿ to an incident beam of known wave length ¿, diffraction will occur from {hkl} planes of specific interplanar spacings . Here the diffracting {hkl} planes are oriented parallel to the sample surface (Fig. 2a). For uture reference it should be noted that the sample can be rotated with respect to the fixed incident beam so that the diffracting planes become inclined to the surface by a known amount (Fig. 2b). Lester and Aborn 5 were the first to show how d spacings change elastically when loads are applied to an aggregate. The distance d acts as a gage length, and changes in this length are directly proportional to elastic strain:
Equation (1) (Available in full paper)
