Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Stress-Strain Modelling of Soils in Drained and Undrained Conditions Using a Multi-model Intelligent Approach
Abstract Appropriate modelling of soil behaviour is essential when dealing with issue related to soil mechanics and foundation engineering. In this paper, a new data-driven methodology is used for simulating and predicting shear behaviour of soils in both drained and undrained conditions. The proposed evolutionary based technique is capable of generating transparent and structured representation of the triaxial test data provided. Excellent agreements between the experimental data and the modelling results are observed in both cases. In addition, “feed-forward” algorithms are proposed separately for drained and undrained conditions, in order to simulate stress-strain paths using well trained machine learning-based constitutive models. It is shown that a machine learning-based constitutive model which has been trained to capture soil behaviour using a limited number of triaxial test results, can also be employed as a stand-alone tool to generate additional virtual triaxial data with a very high accuracy.
Stress-Strain Modelling of Soils in Drained and Undrained Conditions Using a Multi-model Intelligent Approach
Abstract Appropriate modelling of soil behaviour is essential when dealing with issue related to soil mechanics and foundation engineering. In this paper, a new data-driven methodology is used for simulating and predicting shear behaviour of soils in both drained and undrained conditions. The proposed evolutionary based technique is capable of generating transparent and structured representation of the triaxial test data provided. Excellent agreements between the experimental data and the modelling results are observed in both cases. In addition, “feed-forward” algorithms are proposed separately for drained and undrained conditions, in order to simulate stress-strain paths using well trained machine learning-based constitutive models. It is shown that a machine learning-based constitutive model which has been trained to capture soil behaviour using a limited number of triaxial test results, can also be employed as a stand-alone tool to generate additional virtual triaxial data with a very high accuracy.
Stress-Strain Modelling of Soils in Drained and Undrained Conditions Using a Multi-model Intelligent Approach
Rezania, Mohammad (Autor:in) / Ma, Guotao (Autor:in)
25.09.2019
10 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Undrained and drained (?) stress-strain response
| British Library Online Contents | 2000
Undrained and drained (?) stress-strain response
| Online Contents | 2000
The study of undrained and drained behavior of unsaturated soils
| TIBKAT | 1990
Similarity solutions for drained and undrained cavity expansions in soils
| Online Contents | 1994