The geometry and orientation of faults are examined using several field examples of small-scale (< lm displacement) fault systems. For isotropic rocks under triaxial compression, faults normally develop in conjugate sets about the maximum compressive stress (σ1), with dihedral angles usually of about 50°, as predicted by the Coulomb theory of failure. In layered rocks, the geometry of faults varies with the orientation of layering with respect to the stress field. Where σ1, is approximately normal to layering or anisotropy, conjugate faults also develop symmetrically about σ1. Where rocks have interbedded layers with different mechanical behaviours, however, faults tend to initiate orthogonal to the more brittle layers (i.e. originate as extension fractures sub-parallel to σ1), but oblique to the less brittle layers. As the fault steps through the layering, pull-aparts are developed which may reduce the dihedral angle. Where σ1 is oblique (c. 25–75°) to anisotropy, one set of faults is developed at a high angle to layering, with another at a low angle, usually showing a ramp-flat geometry. Large dihedral angles (up to 90°) may result and σ1, does not bisect this angle.
Where σ1 is approximately parallel to layering, two cases can be recognized. Where σ3 is approximately normal to layering, faults with layer-parallel flats and contractional ramps develop. Where σ2 is approximately normal to layering, conjugate faults develop which are symmetrical about σ1, the geometry resembling that in isotropic rocks. These observations are in agreement with rock deformation experiments which show the strong effect of anisotropy on fault orientation, but the observations incorporate the effects of layering of materials with different deformation characteristics.