The undrained shear strength of organic soils can be evaluated based on measurements obtained from the dilatometer test using single- and multi-factor empirical correlations presented in the literature. However, the empirical methods may sometimes show relatively high values of maximum relative error. Therefore, a method for evaluating the undrained shear strength of organic soils using artificial neural networks based on data obtained from a dilatometer test and organic soil properties is presented in this study. The presented neural network, with an architecture of 5-4-1, predicts the normalized undrained shear strength based on five independent variables: the normalized net value of a corrected first pressure reading (po − uo)/σ′v, the normalized net value of a corrected second pressure reading (p1 − uo)/σ′v, the organic content Iom, the void ratio e, and the stress history indictor (oc or nc). The neural model presented in this study provided a more reliable prediction of the undrained shear strength in comparison to the empirical methods, with a maximum relative error of ±10%.
Analysis of compaction parameters of the exemplary non-cohesive soil determined by Proctor methods and vibrating table tests.The purpose of the work is to analyze compaction parameters of non-cohesive uniformly graded soil -optimum moisture content (w opt ) and maximum dry density (ρ dmax ), obtained from Proctor tests and using vibrating table. The research was conducted on even-graded medium sand (Safgr), of uniformity coefficient C U = 3.10 and coefficient of curvature C C = 0.99. Compaction parameters were examined by using Proctor tests -standard Proctor tests (I and II) and modified Proctor tests (III and IV) in compliance with PN-B-04481:1988, and also standard (A+A and A+B) and modified (B+A and B+B) according to PN-EN 13286-2:2010, and by using a vibrating table in compliance with PN-EN 13286-5:2006 at four sample loading attempts. The moisture content of the samples increased by 1-2% in the range of about 0% to about 10%. On the basis of the analysis of data from soil studies with uniform grain size (poorly compactable soil), it can be concluded that the values of test results ρ dmax of medium sand with standard (or modified) Proctor tests according to PN-B-04481:1988 and PN-EN 13286-2:2010 are close to each other. It can be concluded that in the case of ρ dmax , the vibrating table method (with the assumed test conditions) allows to achieve results comparable to those of Proctor (mean relative difference 1.88%). Using the vibrating table the w opt values were lower than those obtained by Proctor tests (mean relative difference of 18.84%).
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.