A coparison of different techniques for determining in situ lateral stresses in a dense glacial till is presented. The site is located north of des moines, iowa, in the late wisconsinan drift plain. The till is generally composed of a cohesive matrix (P.I. equals 12; L.I. equals 0) and extends to a depth of 19 m. In situ tests conducted at the site included flat dilatometer (dmt), full-displacement pressuremeter (fdpmt), hydraulic fracture (hft), and prebored pressuremeter (pmt) tests. Push-in earth pressure cells were installed at various depths and were allowed to come to equilibrium to provide an upper-bound value of the in situ lateral stress conditions. Those data are also compared with dmt dissipation tests, which give an equilibrium value of sigma sub h. A coparison of the effective horizontal stresses obtained with the in situ tests is presented and is basedon field pore pressure measurements obtained from electric and pneumatic piezometers. Estimates of the at-rest coefficient of earth pressure, k sub 0, were made and are based on laboratory tests, using empirical relationships between k sub 0 and ocr. Ocrs were obtained from incremental loading laboratory oedometer tests. The work energy method was also used to predict in situ effective horizontal stresses, using results of oedometer tests on horizontally trimmed samples.Results indicate that full displacement tests, such as the dmt, hft, and fdpmt, tend to overestimate the in situ stresses while the non-displacement pmt provided results very close to the spade cell and laboratory data. A discussion of the results is presented concerningthe accuracy of each of the tests in predicting in situ stresses indense cohesive soils. This paper appears in transportation researchrecord no. 1278, Dynamic testing of aggregates and soils and lateral stress measurements 1990.
Abstract