On the measurement of the undrained strength of discontinuities in the direct shear box

A series of unconsolidated, undrained shear box tests on a tectonic shear d iscontinuity in the London Clay Formation yielded unexpectedly high values of undrained strength. Measurement of the effective stress in the block sam ples prior to shear leads to the suggestion that the measured undrained str ength on the discontinuity is controlled by the post-sampling mean effectiv e stress, rather than the original in situ effective stress acting across t he discontinuity. This observation has the result that in any soil where K- 0 not equal 1.0, a rigid-boundary test apparatus (such as the shear box) th at does not re-establish the in situ stresses cannot measure the correct un drained strength on a discontinuity.