Historically, the Workload design method has served as the foundation for geotechnical designs. This approach relies on global safety factors which does not guarantee a sufficient margin of safety against every uncertainty involved in design. Limit states design accounts for probable failure by using partial safety factors for characteristic values to obtain a design value. Characteristic values are determined based on probability theory and statistical distributions. To obtain reliable probability density functions a large dataset of measurements is required to increase confidence in parameter values and statistical reliability. It is unfeasible to establish the 5th percentile of a soil parameter for most shear strength tests due to time and expense. This research evaluates three testing methods to obtain reliable probability density functions for more reliable data analysis and a less conservative characteristic value.
18th African Regional Conference on Soil Mechanics and Geotechnical Engineering (ARCSMGE2024)
Soil characterization