Deriving stress/strain relationships from the contraction phase of pressuremeter tests in sands

ABSTRACT: Inserting a pressuremeter or dilatometer into sand inevitably alters the initial stress state, obscuring the early part of the test with a disturbance contribution. In extreme cases where a device has been pushed into the ground, the only part of the subsequent cavity expansion capable of straightforward analysis will be data obtained following a stress reversal. This approach is already used to derive high quality modulus measurements from unload/reload cycles (Hughes, 1982). The same argument applies to the final cavity unloading. In addition to stiffness, it is possible to derive sensible strength parameters as the unloading continues beyond the elastic range and material yields in the reverse sense. Closed form solutions that assume the form of the stress/strain response for the undrained contraction case have been widely used for some time, but for drained contraction, these give ambiguous results. This paper uses a finite difference approach based on Manassero (1989) to examine contraction data with minimal assumptions and no restrictions on the form of the stress/strain curve. Results from the unloading of self-bored, pre-bored and pushed pressuremeter tests in various dilatant materials are presented in the form of stress/strain curves with peak shear stress and friction angle identified. It is apparent that the shear strain is different for these two events, the peak shear stress occurring when the mobilized friction has reduced to the constant volume condition. Differences between the three pressuremeter insertion methods for the purposes of deriving strength are small.