Characterization of Vertical Spatial Variability of Soils Using CPTu Data Exploration

Geotechnical uncertainties can be categorized into epistemic and aleatory uncertainties. Epistemic uncertainty is caused by the lack of knowledge or data. Aleatory uncertainty refers to the intrinsic randomness of a phenomenon. In geotechnical engineering, the aleatory uncertainty can be characterized by inherent spatial variability, which is the focus of this paper. The state-of-the-practice is to characterize this spatial variability using the scale of fluctuation or correlation length from a statistical interpretation of limited samplings. These descriptors characterize the distance over which the parameters of a soil or rock are similar or correlated. This paper presents a study case of a dam located in the northern region of Brazil. The undrained shear strength of the foundation soil was spatially modelled using CPTu profiles. The data were decomposed between trend and residuals and an autocorrelation function (ACF) was defined. The results were then compared against different theoretical autocorrelation models (ACM), and the vertical correlation length was estimated from the fitted ACM.

Keywords: Geotechnical site characterization, reliability, uncertainty, spatial variability, probabilistic models.