<abstract xmlns="http://www.w3.org/1999/xhtml">

A method of creating a constitutive model of layered rocks based on an artificial neural network (ANN) is reported in this work. The ANN gives an implicit constitutive function ∑<italic>n</italic>+1=F(∑<italic>n</italic>,Δ<bold>E</bold> ), relating the new state of homogenized stresses ∑<italic>n</italic>+1 with the old state ∑<italic>n</italic> and with the increment of homogenized strains Δ<bold>E</bold>. The first step is to repeatedly run a strain- controlled homogenisation on an uni-dimensional finite element model of a periodic cell with elastic-plastic models (Drucker-Prager) of the components. Paths are created in (∑, <bold>E</bold>) space, from which, a set of patterns is formed to train the ANN. A description of how to prepare this data and a discussion on ANN training issues are presented. Finally, the procedure based on trained ANN is put into a finite-element code (ZSoil.PC) as a user-delivered constitutive function. The approach is verified by comparing the results of the developed model basing on ANN with a direct (single-scale) analysis, which showed acceptable accuracy.
</abstract>

A method of creating a constitutive model of layered rocks based on an artificial neural network (ANN) is reported in this work. The ANN gives an implicit constitutive function ∑n+1=F(∑n,ΔE ), relating the new state of homogenized stresses ∑n+1 with the old state ∑n and with the increment of homogenized strains ΔE. The first step is to repeatedly run a strain- controlled homogenisation on an uni-dimensional finite element model of a periodic cell with elastic-plastic models (Drucker-Prager) of the components. Paths are created in (∑, E) space, from which, a set of patterns is formed to train the ANN. A description of how to prepare this data and a discussion on ANN training issues are presented. Finally, the procedure based on trained ANN is put into a finite-element code (ZSoil.PC) as a user-delivered constitutive function. The approach is verified by comparing the results of the developed model basing on ANN with a direct (single-scale) analysis, which showed acceptable accuracy.

The modelling of layered rocks using a numerical homogenisation technique and an artificial neural network

Doctoral Studies, AGH University of Science and Technology

Technical Transactions

This work is licensed under the Creative Commons Attribution-ShareAlike 4.0 International License.

A method of creating a constitutive model of layered rocks based on an artificial neural network (ANN) is reported in this work. The ANN gives an implicit...

The modelling of layered rocks using a numerical homogenisation technique and an artificial neural network

Online veröffentlicht: 16. Mai 2023

Seitenbereich: -

Eingereicht: 26. Jan. 2023

Akzeptiert: 08. Mai 2023

DOI: https://doi.org/10.37705/TechTrans/e2023007

Schlüsselwörter
layered rocks, finite element method, homogenisation, artificial neural network

© 2023 Aleksander Urbański et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-ShareAlike 4.0 International License.

The modelling of layered rocks using a numerical homogenisation technique and an artificial neural network

Online veröffentlicht: 16. Mai 2023

Seitenbereich: -

Eingereicht: 26. Jan. 2023

Akzeptiert: 08. Mai 2023

DOI: https://doi.org/10.37705/TechTrans/e2023007

Schlüsselwörterlayered rocks, finite element method, homogenisation, artificial neural network

© 2023 Aleksander Urbański et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-ShareAlike 4.0 International License.

Schlüsselwörter
layered rocks, finite element method, homogenisation, artificial neural network