Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach

Autores
Miller Branco Ferraz, Franz; Sztangret, Lukasz; Carazo, Fernando Diego; Buzolin, Ricardo Henrique; Wang, Peng; Szeliga, Danuta; dos Santos Effertz, Pedro; Macio, Piotr; Krumphals, Alfred; Poletti, Maria Cecilia
Año de publicación
2023
Idioma
inglés
Tipo de recurso
artículo
Estado
versión publicada
Descripción
During the thermomechanical processing of titanium alloys in the β-domain, the β-phase undergoes restoration phenomena. This work describes them by a mean-field physical model that correlates the flow stress with the microstructural evolution. To reduce the computational time of process simulations, metamodels are developed for specific outputs of the mean-field physical model using Artificial Neural Network (ANN) and Decision Tree Regression (DTR). The performance of the obtained metamodels is evaluated in terms of the coefficient of determination (R²), the root-mean-square error (RMSE), and the mean relative error (MRE). No significant difference was observed between R2training and R2testing, meaning that all the metamodels correctly generalise the overall behaviour of the outputs for a wide range of inputs. The evolution of the metamodel outputs is compared with the model predictions in two different situations: 1) at a constant strain rate and temperature, and 2) during Finite Element (FE) simulations of the hot deformation of a hat-shaped sample, where temperature and effective strain rate vary at each element during deformation. The evolution of the outputs at constant and non-constant strain rates and temperature demonstrated the robustness of the metamodels in predicting the heterogeneous deformation within a workpiece. The computational time required by the metamodels to calculate selected outputs can be more than 100 times less than that of the model itself at a constant strain rate using MATLAB® and up to 19% less when coupled with FE simulations. The simulation results combined with microstructural analysis are used to visualise the different restoration mechanisms occurring in different regions of the hat-shaped sample as a function of the local thermomechanical history. The changes in strain rate and temperature during deformation influence the evolution of the wall dislocation density and the immobilisation rate of mobile dislocations at subgrain boundaries, leading to different kinetics of microstructure evolution.
Fil: Miller Branco Ferraz, Franz. Graz University Of Technology.; Austria
Fil: Sztangret, Lukasz. AGH University of Science and Technology; Polonia
Fil: Carazo, Fernando Diego. Universidad Nacional de San Juan. Facultad de Ingeniería. Instituto de Mecanica Aplicada; Argentina. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina
Fil: Buzolin, Ricardo Henrique. Graz University Of Technology.; Austria
Fil: Wang, Peng. Graz University Of Technology.; Austria
Fil: Szeliga, Danuta. AGH University of Science and Technology; Polonia
Fil: dos Santos Effertz, Pedro. No especifíca;
Fil: Macio, Piotr. AGH University of Science and Technology; Polonia
Fil: Krumphals, Alfred. No especifíca;
Fil: Poletti, Maria Cecilia. Graz University Of Technology.; Austria
Materia
ARTIFICIAL NEURAL NETWORK
DECISION-TREE REGRESSION
HOT DEFORMATION
MEAN-FIELD MODEL
METAMODEL
TITANIUM ALLOYS
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/223398

id CONICETDig_8f89f702cbd568aef0793ef186edbde8
oai_identifier_str oai:ri.conicet.gov.ar:11336/223398
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approachMiller Branco Ferraz, FranzSztangret, LukaszCarazo, Fernando DiegoBuzolin, Ricardo HenriqueWang, PengSzeliga, Danutados Santos Effertz, PedroMacio, PiotrKrumphals, AlfredPoletti, Maria CeciliaARTIFICIAL NEURAL NETWORKDECISION-TREE REGRESSIONHOT DEFORMATIONMEAN-FIELD MODELMETAMODELTITANIUM ALLOYShttps://purl.org/becyt/ford/2.5https://purl.org/becyt/ford/2During the thermomechanical processing of titanium alloys in the β-domain, the β-phase undergoes restoration phenomena. This work describes them by a mean-field physical model that correlates the flow stress with the microstructural evolution. To reduce the computational time of process simulations, metamodels are developed for specific outputs of the mean-field physical model using Artificial Neural Network (ANN) and Decision Tree Regression (DTR). The performance of the obtained metamodels is evaluated in terms of the coefficient of determination (R²), the root-mean-square error (RMSE), and the mean relative error (MRE). No significant difference was observed between R2training and R2testing, meaning that all the metamodels correctly generalise the overall behaviour of the outputs for a wide range of inputs. The evolution of the metamodel outputs is compared with the model predictions in two different situations: 1) at a constant strain rate and temperature, and 2) during Finite Element (FE) simulations of the hot deformation of a hat-shaped sample, where temperature and effective strain rate vary at each element during deformation. The evolution of the outputs at constant and non-constant strain rates and temperature demonstrated the robustness of the metamodels in predicting the heterogeneous deformation within a workpiece. The computational time required by the metamodels to calculate selected outputs can be more than 100 times less than that of the model itself at a constant strain rate using MATLAB® and up to 19% less when coupled with FE simulations. The simulation results combined with microstructural analysis are used to visualise the different restoration mechanisms occurring in different regions of the hat-shaped sample as a function of the local thermomechanical history. The changes in strain rate and temperature during deformation influence the evolution of the wall dislocation density and the immobilisation rate of mobile dislocations at subgrain boundaries, leading to different kinetics of microstructure evolution.Fil: Miller Branco Ferraz, Franz. Graz University Of Technology.; AustriaFil: Sztangret, Lukasz. AGH University of Science and Technology; PoloniaFil: Carazo, Fernando Diego. Universidad Nacional de San Juan. Facultad de Ingeniería. Instituto de Mecanica Aplicada; Argentina. Consejo Nacional de Investigaciones Científicas y Técnicas; ArgentinaFil: Buzolin, Ricardo Henrique. Graz University Of Technology.; AustriaFil: Wang, Peng. Graz University Of Technology.; AustriaFil: Szeliga, Danuta. AGH University of Science and Technology; PoloniaFil: dos Santos Effertz, Pedro. No especifíca;Fil: Macio, Piotr. AGH University of Science and Technology; PoloniaFil: Krumphals, Alfred. No especifíca;Fil: Poletti, Maria Cecilia. Graz University Of Technology.; AustriaElsevier2023-06info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttp://purl.org/coar/resource_type/c_6501info:ar-repo/semantics/articuloapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/223398Miller Branco Ferraz, Franz; Sztangret, Lukasz; Carazo, Fernando Diego; Buzolin, Ricardo Henrique; Wang, Peng; et al.; Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach; Elsevier; Materials Today Communications; 35; 6-2023; 1-162352-4928CONICET DigitalCONICETenginfo:eu-repo/semantics/altIdentifier/url/https://linkinghub.elsevier.com/retrieve/pii/S2352492823008395info:eu-repo/semantics/altIdentifier/doi/10.1016/j.mtcomm.2023.106148info:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2025-09-29T10:20:59Zoai:ri.conicet.gov.ar:11336/223398instacron:CONICETInstitucionalhttp://ri.conicet.gov.ar/Organismo científico-tecnológicoNo correspondehttp://ri.conicet.gov.ar/oai/requestdasensio@conicet.gov.ar; lcarlino@conicet.gov.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:34982025-09-29 10:20:59.665CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach
title Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach
spellingShingle Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach
Miller Branco Ferraz, Franz
ARTIFICIAL NEURAL NETWORK
DECISION-TREE REGRESSION
HOT DEFORMATION
MEAN-FIELD MODEL
METAMODEL
TITANIUM ALLOYS
title_short Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach
title_full Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach
title_fullStr Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach
title_full_unstemmed Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach
title_sort Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach
dc.creator.none.fl_str_mv Miller Branco Ferraz, Franz
Sztangret, Lukasz
Carazo, Fernando Diego
Buzolin, Ricardo Henrique
Wang, Peng
Szeliga, Danuta
dos Santos Effertz, Pedro
Macio, Piotr
Krumphals, Alfred
Poletti, Maria Cecilia
author Miller Branco Ferraz, Franz
author_facet Miller Branco Ferraz, Franz
Sztangret, Lukasz
Carazo, Fernando Diego
Buzolin, Ricardo Henrique
Wang, Peng
Szeliga, Danuta
dos Santos Effertz, Pedro
Macio, Piotr
Krumphals, Alfred
Poletti, Maria Cecilia
author_role author
author2 Sztangret, Lukasz
Carazo, Fernando Diego
Buzolin, Ricardo Henrique
Wang, Peng
Szeliga, Danuta
dos Santos Effertz, Pedro
Macio, Piotr
Krumphals, Alfred
Poletti, Maria Cecilia
author2_role author
author
author
author
author
author
author
author
author
dc.subject.none.fl_str_mv ARTIFICIAL NEURAL NETWORK
DECISION-TREE REGRESSION
HOT DEFORMATION
MEAN-FIELD MODEL
METAMODEL
TITANIUM ALLOYS
topic ARTIFICIAL NEURAL NETWORK
DECISION-TREE REGRESSION
HOT DEFORMATION
MEAN-FIELD MODEL
METAMODEL
TITANIUM ALLOYS
purl_subject.fl_str_mv https://purl.org/becyt/ford/2.5
https://purl.org/becyt/ford/2
dc.description.none.fl_txt_mv During the thermomechanical processing of titanium alloys in the β-domain, the β-phase undergoes restoration phenomena. This work describes them by a mean-field physical model that correlates the flow stress with the microstructural evolution. To reduce the computational time of process simulations, metamodels are developed for specific outputs of the mean-field physical model using Artificial Neural Network (ANN) and Decision Tree Regression (DTR). The performance of the obtained metamodels is evaluated in terms of the coefficient of determination (R²), the root-mean-square error (RMSE), and the mean relative error (MRE). No significant difference was observed between R2training and R2testing, meaning that all the metamodels correctly generalise the overall behaviour of the outputs for a wide range of inputs. The evolution of the metamodel outputs is compared with the model predictions in two different situations: 1) at a constant strain rate and temperature, and 2) during Finite Element (FE) simulations of the hot deformation of a hat-shaped sample, where temperature and effective strain rate vary at each element during deformation. The evolution of the outputs at constant and non-constant strain rates and temperature demonstrated the robustness of the metamodels in predicting the heterogeneous deformation within a workpiece. The computational time required by the metamodels to calculate selected outputs can be more than 100 times less than that of the model itself at a constant strain rate using MATLAB® and up to 19% less when coupled with FE simulations. The simulation results combined with microstructural analysis are used to visualise the different restoration mechanisms occurring in different regions of the hat-shaped sample as a function of the local thermomechanical history. The changes in strain rate and temperature during deformation influence the evolution of the wall dislocation density and the immobilisation rate of mobile dislocations at subgrain boundaries, leading to different kinetics of microstructure evolution.
Fil: Miller Branco Ferraz, Franz. Graz University Of Technology.; Austria
Fil: Sztangret, Lukasz. AGH University of Science and Technology; Polonia
Fil: Carazo, Fernando Diego. Universidad Nacional de San Juan. Facultad de Ingeniería. Instituto de Mecanica Aplicada; Argentina. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina
Fil: Buzolin, Ricardo Henrique. Graz University Of Technology.; Austria
Fil: Wang, Peng. Graz University Of Technology.; Austria
Fil: Szeliga, Danuta. AGH University of Science and Technology; Polonia
Fil: dos Santos Effertz, Pedro. No especifíca;
Fil: Macio, Piotr. AGH University of Science and Technology; Polonia
Fil: Krumphals, Alfred. No especifíca;
Fil: Poletti, Maria Cecilia. Graz University Of Technology.; Austria
description During the thermomechanical processing of titanium alloys in the β-domain, the β-phase undergoes restoration phenomena. This work describes them by a mean-field physical model that correlates the flow stress with the microstructural evolution. To reduce the computational time of process simulations, metamodels are developed for specific outputs of the mean-field physical model using Artificial Neural Network (ANN) and Decision Tree Regression (DTR). The performance of the obtained metamodels is evaluated in terms of the coefficient of determination (R²), the root-mean-square error (RMSE), and the mean relative error (MRE). No significant difference was observed between R2training and R2testing, meaning that all the metamodels correctly generalise the overall behaviour of the outputs for a wide range of inputs. The evolution of the metamodel outputs is compared with the model predictions in two different situations: 1) at a constant strain rate and temperature, and 2) during Finite Element (FE) simulations of the hot deformation of a hat-shaped sample, where temperature and effective strain rate vary at each element during deformation. The evolution of the outputs at constant and non-constant strain rates and temperature demonstrated the robustness of the metamodels in predicting the heterogeneous deformation within a workpiece. The computational time required by the metamodels to calculate selected outputs can be more than 100 times less than that of the model itself at a constant strain rate using MATLAB® and up to 19% less when coupled with FE simulations. The simulation results combined with microstructural analysis are used to visualise the different restoration mechanisms occurring in different regions of the hat-shaped sample as a function of the local thermomechanical history. The changes in strain rate and temperature during deformation influence the evolution of the wall dislocation density and the immobilisation rate of mobile dislocations at subgrain boundaries, leading to different kinetics of microstructure evolution.
publishDate 2023
dc.date.none.fl_str_mv 2023-06
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
http://purl.org/coar/resource_type/c_6501
info:ar-repo/semantics/articulo
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/11336/223398
Miller Branco Ferraz, Franz; Sztangret, Lukasz; Carazo, Fernando Diego; Buzolin, Ricardo Henrique; Wang, Peng; et al.; Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach; Elsevier; Materials Today Communications; 35; 6-2023; 1-16
2352-4928
CONICET Digital
CONICET
url http://hdl.handle.net/11336/223398
identifier_str_mv Miller Branco Ferraz, Franz; Sztangret, Lukasz; Carazo, Fernando Diego; Buzolin, Ricardo Henrique; Wang, Peng; et al.; Metamodelling the hot deformation behaviour of titanium alloys using a mean-field approach; Elsevier; Materials Today Communications; 35; 6-2023; 1-16
2352-4928
CONICET Digital
CONICET
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/https://linkinghub.elsevier.com/retrieve/pii/S2352492823008395
info:eu-repo/semantics/altIdentifier/doi/10.1016/j.mtcomm.2023.106148
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:CONICET Digital (CONICET)
instname:Consejo Nacional de Investigaciones Científicas y Técnicas
reponame_str CONICET Digital (CONICET)
collection CONICET Digital (CONICET)
instname_str Consejo Nacional de Investigaciones Científicas y Técnicas
repository.name.fl_str_mv CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicas
repository.mail.fl_str_mv dasensio@conicet.gov.ar; lcarlino@conicet.gov.ar
_version_ 1844614195377602560
score 13.070432