The orientation of principal stresses in the continental crust can be inferred by using a number of different techniques. The more accurate of these, which are best used at outcrops or in mines, and have only become available during the last few decades (Jaeger and Cook, 1969), are based on the over-coring method. The orientation and magnitude of the least horizontal stress can be determined by using the hydraulic-fracturing technique. The orientation of the axis of greatest horizontal stress can also be obtained by ascertaining the orientation of the elliptical cross-section of a deep, unlined, originally circular-section bore-hole. The elliptical form is caused by the process often referred to as wall break-out. Other methods used to infer stress orientation can sometimes be obtained from seismic data and also from the theoretical analysis of geological structures.
In the over-coring method of stress analysis, a small, planar surface (of the required orientation) is selected or prepared. Strain-gauges, usually set at 60° to each other are affixed to this surface by adhesive
Figure 2.8 Strain gauges are cemented to a flat surface and wires from the gauges are taken up the barrel of a core-drill to circuits determining the expansion of the core. Over-coring to several inches (OC—dashed lines) enables the in situ stress to be released and gives rise to expansion of the 'core'. The dashed lines show, in exaggerated form, the strain release (Str).
(Figure 2.8). When the adhesive has set, the network of gauges gives the apparent zero-strain condition in the rock. The rock surface, on which the gauges are placed, is then isolated from the surrounding rock by over-coring. The free surface around the gauges permits the rock stresses in the cylindrical stub, beneath the strain-gauged surface, to induce strains which are recorded by the strain-gauges. The elastic moduli (Young's modulus E and Poisson's number m) for the rock are subsequently ascertained in the laboratory from the detached stub, or a sample taken from an adjacent site. The strain-data obtained by over-coring can then be translated into magnitudes and orientation of in situ, principal two-dimensional stress. By carrying out this exercise with sets of gauges arrayed at right angles to each other, the three-dimensional stress array can be inferred.
An interesting feature of the data obtained from such measurements is that, even in non-tectonic areas, down to a depth of 300 m, the horizontal stresses are always larger than the vertical stress (Figure 2.9) and may be larger than the vertical stress at depths of several kilometres. It has been argued (Price and Cosgrove, 1990) that the increase in relative magnitude of the horizontal stresses as the surface is approached can be attributed to the fact that, with uplift and erosion, the vertical stress is reduced more rapidly than the horizontal stresses. Consequently, the stress data obtained at, or near, the surface does not permit one to infer that the maximum horizontal stress, at some depth in the plate, is also the maximum principal stress. However, one can infer that the maximum principal stress, as determined at or near the surface, may well approximate to the orientation of the maximum horizontal stress at depth. Even this, however, is open to question, for it has been established that deep stress orientations may change, quite abruptly, with the age of the rock unit (Cowgill, 1994). Such changes presumably reflect earlier, different movement directions of the plate.
This technique requires a section of a borehole to be sealed and isolated (by the use of packers) above and below the rock horizon in which the least principal stress is to be determined (Figure 2.10). One of two main types of results may be expected.
Was this article helpful?