Numerical investigations into the mechanism of graben formation

Abstract

The formation of a graben has been investigated using finite element analysis. A new method of modelling faults has been developed which is based on calculating the shear stresses on the fault and, if they exceed the frictional strength, applying forces which cause frictional sliding. Both Newtonian visco - elastic and power law creep rheologies have been used for the lower lithosphere. The deformation patterns seen in the models are relatively insensitive to which one is used. Stress amplification is shown to result in normal faulting in the upper, brittle layer as a result of relatively small stresses of about 20 MPa being applied throughout the depth of the lithosphere. When a fault is introduced into the model the stresses adjacent to the fault are re-orientated and secondary faulting is predicted. The bending profile associated with the fault deformation results in a weakness where the stresses are most greatly modified. A second normal fault may form here. I f the fault movement is confined to the upper part of the brittle layer then the predicted graben width is between 5 and 15 km. For deeper fault movement and an underlying fluid the predicted width increases to 50 - 55 km. A more realistic theology for the underlying material is visco - elasticity. In this case the predicted width is about 25 km. The fault throw increases as the visco - elastic material relaxes but no significant change is seen in the width. The subsidence of a 50 km wide graben wedge has been examined. For applied stresses of about 50 MPa and coefficients of friction of less than about 0.1, subsidence of about 1 km is predicted. This does not include sediment in filling. The subsiding wedge causes large compressive stresses in the underlying material which may be long - lasting. The subsidence is controlled by the boundary faults and causes bending of the block which may result in internal deformation