Abstract
This thesis describes a study of the shear vane test which includes its history, a comprehensive review of the present state of knowledge concerning its analysis and interpretation and an investigation into the effect various factors have on it. A laboratory test program using a large instrumented shear vane allowed the shape of the shear stress distribution to be measured at the blade edges. A photoelastic and radiographic study on model vanes demonstrated the generation, size and shape of the shear zone around the mid-point of a vane. These measurements are shown to confirm analytical predictions. The influence of progressive failure on the interpretation of the vane test was assessed by means of a strain hardening - strain softening soil model. It is shown that the apparent strength derived from the conventional rigid - ideally plastic interpretation could be up to 18% in error, whilst the measurement of strength anisotropy by the two vane method may be overestimated by up to 30%. A field test study using a motorised vane with an electronic system to control rotation rate, examined the influence of testing rate whilst the strength anisotropy of the Brent Knoll clay was assessed using diamond vanes. The vane was controlled over a wide range of constant rotation rates, constant torques and constant torque application rates. It is shown that for Brent Knoll clay the strength is reduced with time to undrained failure by up to 40%. Finally, recommendations regarding the execution and interpretation of the vane test are presented. It is concluded that although the factors may combine often fortuitously to reveal an adequate prediction of field stability, they are uncontrolled and uncontrollable. Thus to ensure a more rational and consistent approach to design these factors should be assessed individually as well as collectively to determine their influence on the aquisition, interpretation and application of shear vane data.