Machine learning assisted calibration of a ductile fracture locus model

Research output: Contribution to journalArticleResearchpeer-review

Standard

Machine learning assisted calibration of a ductile fracture locus model. / Baltic, Sandra; Asadzadeh, Mohammad Zhian; Hammer, Patrick et al.
In: Materials and Design, Vol. 203.2021, No. May, 109604, 05.2021.

Research output: Contribution to journalArticleResearchpeer-review

Vancouver

Baltic S, Asadzadeh MZ, Hammer P, Magnien J, Gänser HP, Antretter T et al. Machine learning assisted calibration of a ductile fracture locus model. Materials and Design. 2021 May;203.2021(May):109604. Epub 2021 Feb 21. doi: 10.1016/j.matdes.2021.109604

Author

Baltic, Sandra ; Asadzadeh, Mohammad Zhian ; Hammer, Patrick et al. / Machine learning assisted calibration of a ductile fracture locus model. In: Materials and Design. 2021 ; Vol. 203.2021, No. May.

Bibtex - Download

@article{c224c7b4cb944ed3a302d999ef503971,
title = "Machine learning assisted calibration of a ductile fracture locus model",
abstract = "While several different specimen geometries are typically required to calibrate a ductile fracture locus model, this article presents for the first time a calibration methodology that uses one single specimen geometry. This is accomplished by a computational framework that combines finite element modelling (FEM) and artificial neural network (ANN). The combinations of the model parameters are used to generate the training database. The local displacement fields and global force-displacement histories are extracted throughout the complete numerical experiment and passed to the ANN. Therefore, the influence of the local stress state on the evolution of the local deformation is implicitly taken into account. The trained ANN is verified by evaluating its predictability of material parameters of FE simulations unseen in the training stage. The experimental data obtained from the shear tensile test using Digital Image Correlation is introduced to the trained ANN to identify the parameter set that predicts the real mechanical response of the shear specimen. Three different ANN architectures with distinguished input representations are studied. It turns out that all of them can acceptably describe the experimental behaviour of not only the calibration specimen but also the specimens not used for training the model.",
keywords = "Artificial neural network, Damage, Fracture",
author = "Sandra Baltic and Asadzadeh, {Mohammad Zhian} and Patrick Hammer and Julien Magnien and G{\"a}nser, {Hans Peter} and Thomas Antretter and Ren{\'e} Hammer",
note = "Publisher Copyright: {\textcopyright} 2021 The Author(s)",
year = "2021",
month = may,
doi = "10.1016/j.matdes.2021.109604",
language = "English",
volume = "203.2021",
journal = "Materials and Design",
issn = "0264-1275",
publisher = "Elsevier",
number = "May",

}

RIS (suitable for import to EndNote) - Download

TY - JOUR

T1 - Machine learning assisted calibration of a ductile fracture locus model

AU - Baltic, Sandra

AU - Asadzadeh, Mohammad Zhian

AU - Hammer, Patrick

AU - Magnien, Julien

AU - Gänser, Hans Peter

AU - Antretter, Thomas

AU - Hammer, René

N1 - Publisher Copyright: © 2021 The Author(s)

PY - 2021/5

Y1 - 2021/5

N2 - While several different specimen geometries are typically required to calibrate a ductile fracture locus model, this article presents for the first time a calibration methodology that uses one single specimen geometry. This is accomplished by a computational framework that combines finite element modelling (FEM) and artificial neural network (ANN). The combinations of the model parameters are used to generate the training database. The local displacement fields and global force-displacement histories are extracted throughout the complete numerical experiment and passed to the ANN. Therefore, the influence of the local stress state on the evolution of the local deformation is implicitly taken into account. The trained ANN is verified by evaluating its predictability of material parameters of FE simulations unseen in the training stage. The experimental data obtained from the shear tensile test using Digital Image Correlation is introduced to the trained ANN to identify the parameter set that predicts the real mechanical response of the shear specimen. Three different ANN architectures with distinguished input representations are studied. It turns out that all of them can acceptably describe the experimental behaviour of not only the calibration specimen but also the specimens not used for training the model.

AB - While several different specimen geometries are typically required to calibrate a ductile fracture locus model, this article presents for the first time a calibration methodology that uses one single specimen geometry. This is accomplished by a computational framework that combines finite element modelling (FEM) and artificial neural network (ANN). The combinations of the model parameters are used to generate the training database. The local displacement fields and global force-displacement histories are extracted throughout the complete numerical experiment and passed to the ANN. Therefore, the influence of the local stress state on the evolution of the local deformation is implicitly taken into account. The trained ANN is verified by evaluating its predictability of material parameters of FE simulations unseen in the training stage. The experimental data obtained from the shear tensile test using Digital Image Correlation is introduced to the trained ANN to identify the parameter set that predicts the real mechanical response of the shear specimen. Three different ANN architectures with distinguished input representations are studied. It turns out that all of them can acceptably describe the experimental behaviour of not only the calibration specimen but also the specimens not used for training the model.

KW - Artificial neural network

KW - Damage

KW - Fracture

UR - http://www.scopus.com/inward/record.url?scp=85102052960&partnerID=8YFLogxK

U2 - 10.1016/j.matdes.2021.109604

DO - 10.1016/j.matdes.2021.109604

M3 - Article

AN - SCOPUS:85102052960

VL - 203.2021

JO - Materials and Design

JF - Materials and Design

SN - 0264-1275

IS - May

M1 - 109604

ER -