Updating Uncertainties in Soil Shear Strength Parameters with Multivariate In-Situ and Laboratory Test Data

In recent years, geotechnical engineers have paid much attention on reliability-based design. The probability density functions (PDFs) of the uncertain soil parameters are required for the reliability-based design. Once these PDFs are obtained, reliability analysis can be conducted, and reliability-based design can be achieved. Among the soil parameters, soil strengths play important roles in geotechnical design, e.g. friction angle of sand and undrained shear strength of clay. It is essential to obtain the PDFs of the soil strengths to facilitate reliability-based design. How to obtain those PDFs based on all available information is an important research subject. It is possible to establish the PDFs of the soil strengths based on results from sophisticated laboratory tests, e.g. drained test for sand to obtain its friction angle, but those tests may be expensive and time consuming. On the other hand, some in-situ and laboratory tests are quick and convenient, but they usually do not provide direct information on the soil strengths. Nevertheless, they do usually provide indirect information of the soil strengths by correlation, e.g. SPT-N is positively correlated to friction angle of sand, and overconsolidation ratio (OCR) is also related to undrained shear strength of clay. Therefore, it is possible to update the PDFs of the soil strengths by using the in-situ and laboratory test data.