Cluster Analysis for the Determination of the Undrained Strength Tendency From SPT in Mudflows and Residual Soils

Document Type

Article

Rights

Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence

Disciplines

Geology, Civil engineering

Publication Details

Bulletin of Engineering Geology and the Environment

This material is not open access. However, As part of the Springer Nature SharedIt initiative, you can access a full-text view-only version of this paper by using the following link: https://rdcu.be/bmLJB

Abstract

The standard penetration test (SPT) remains one of the most commonly used field tests to obtain the shear strength properties of soils, the undrained strength (cu) being one the most SPT-N-correlated parameters for geotechnical applications. The overall N-value-cu correlations show a direct relationship between them, characterized by presenting an equation formed by a constant value multiplied by N. More recently, the use of cu in geotechnical engineering has been of great interest in the evaluation of the influence of the undrained strength variability with depth on slope stability analysis. Therefore, an evaluation of the variability with depth of the N-value is made according to the geological origin. However, the (N1)60 values obtained from the SPT have different limitations due the possible “outside of tendency” data known as outliers caused by random factors such as: rock fragments content, weak zones and variations in the state of weathering, which may ultimately affect the estimation of the cu function with depth. Therefore, a cluster analysis of the SPT data was performed in order to identify the values that affect the best-fitting mathematical SPT-N function with depth. These analyses were implemented to the (N1)60 values obtained from multiple SPTs in two distinct geological units, mudflows and residual soils, in order to study the influence of the geological origin in the SPT and, therefore, of the shear strength tendency with depth. It was found that the best clustering method to identify the SPT tendency and the state of weathering in residual soils is the Ward method. For mudflows, the best cluster algorithm is the single method; however, it is concluded that for large areas, the use of a unique cluster method is not recommended. For most projects, the undrained shear strength showed a nonlinear tendency with a squared Z (where Z is the depth in meters) function being common among all geologies. The function gradient for residual soils is about twice when compared with that of the mudflows, mainly due to the overburden pressure and to the decrease in the state of weathering with depth, which increases shear strength in the former type of soils.

DOI

https://doi.org/10.1007/s10064-019-01472-8

Funder

Administrative Department of Science, Technology and Innovation of Colombia – Colciencias


Share

COinS