ABSTRACT
The present norms for determination of the undrained cohesion employing vane tests lead often to a classical value which overestimates the soil strength mobilized at the time of yielding. Based on the definition of the mobilized strength, it has been verified that the classical value is a lower bound of the undrained cohesion. In order to make a realistic determination of this characteristic using vane tests, we propose to limit the rotation of the vane at a given maximum value for which the torque value at yielding is retained. Applying this approach for “torque-amount of rotation” curves, from laboratory vane tests carried out for a Tunisian reconstituted clay, a lower value of undrained cohesion is deduced regarding the classical value corresponding to the maximum torque used by the present norm. Based on this recommendation, it has been verified that the classical value of undrained cohesion is overestimated up to 40%.
