| Description |
One of the most challenging aspects of geotechnical engineering is the characterization of subsurface material. In practice, engineers often base recommendations both directly and indirectly on index properties and corresponding correlations and models. However, these models should be calibrated for site-specific conditions prior to being used as the basis for any substantial design. This study includes details of a geotechnical case study. The objective of this study was to evaluate existing models used to estimate strength and deformation characteristics of intact rock and rock mass with respect to analyzing laterally loaded drilled shafts. Geological and geotechnical investigations were conducted for the given site. The field investigation employed sampling methods such as rock coring to obtain specimens that could be tested in the laboratory, as well as in-situ pressuremeter testing (PMT). Laboratory tests performed included uniaxial compression, point-load, Brazilian tension, rebound hardness, unit weight, and water content. The subsurface profile was shown to consist of alternating layers of claystone, limestone, and sandstone, overlain by a layer of clayey sand. To develop a basis for evaluating existing models, the variability associated with strength and index properties was assessed. In addition, site-specific correlations were developed between uniaxial compressive strength (qu) and the various index properties of the intact rock materials. The predictive capability of existing correlations to estimate qu based on index properties was examined. It was found that point-load index (Is(50)) provided the best estimate of qu among the index correlations studied. Additional rock mass correlations were evaluated on the basis of the strength and deformation characteristics derived from the PMT data. It was found that neither rock mass rating (RMR) nor geological strength index (GSI) produced reasonable estimates of rock mass deformation characteristics based on correlations examined. The method used by the computer program LPILE to develop p-y curves for weak rock (Reese 1997) was evaluated in terms of how well the predicted p-y curves developed compared with those from PMT data. It was found that the weak rock model highly overestimated the stiffness of the rock mass for this particular site compared to stiffness estimated from the PMT results. |
