Logotipo do repositório
 

Publicação:
On the use of Stochastic Bouc–Wen model for simulating viscoelastic internal variables from a finite element approximation of steady-rolling tire

dc.contributor.authorRaqueti, Rafael da S [UNESP]
dc.contributor.authorTeloli, Rafael de O
dc.contributor.authorda Silva, Samuel [UNESP]
dc.contributor.authorBussetta, Philippe
dc.contributor.authorCunha, Americo
dc.contributor.institutionUniversidade Estadual Paulista (UNESP)
dc.contributor.institutionUniv. Bourgogne Franche-Comté
dc.contributor.institutionMichelin - Manufacture Française des Pneumatiques Michelin
dc.contributor.institutionRio de Janeiro State University
dc.date.accessioned2023-07-29T12:28:41Z
dc.date.available2023-07-29T12:28:41Z
dc.date.issued2022-01-01
dc.description.abstractReducing tire rolling resistance and energy loss is a topic of interest to the tire industry. Understanding and modeling these phenomena are essential to approach this problem and propose robust solutions. This work suggests a reduced-order model based on the Bouc–Wen model to simulate internal variables from viscoelastic constitutive laws. Furthermore, sensitivity analysis is performed on the Bouc–Wen parameters to evaluate their influence on the system response and capture the full range of possible values that improve the predictive ability of the reduced-order model. This task is accomplished by calculating the Sobol’s indices estimated from a Polynomial-Chaos expansion. Once the range of feasible model solutions is established, the reduced-order model is calibrated through Bayesian inference. Finally, the uncertainties are propagated, and the reduced-order model is validated using data of viscoelastic internal variables from the finite element approximation of a steady-rolling tire. Satisfactory results are obtained, as the reduced-order model can simulate viscoelastic internal variables with a reduced computational cost for some branches of interest. Its responses are in agreement with the experimental data.en
dc.description.affiliationDepartment of Mechanical Engineering UNESP - São Paulo State University
dc.description.affiliationFEMTO-ST Institute Department of Applied Mechanics Univ. Bourgogne Franche-Comté
dc.description.affiliationMichelin - Manufacture Française des Pneumatiques Michelin
dc.description.affiliationDepartment of Applied Mathematics Rio de Janeiro State University
dc.description.affiliationUnespDepartment of Mechanical Engineering UNESP - São Paulo State University
dc.identifierhttp://dx.doi.org/10.1177/10775463221125038
dc.identifier.citationJVC/Journal of Vibration and Control.
dc.identifier.doi10.1177/10775463221125038
dc.identifier.issn1741-2986
dc.identifier.issn1077-5463
dc.identifier.scopus2-s2.0-85139007356
dc.identifier.urihttp://hdl.handle.net/11449/245987
dc.language.isoeng
dc.relation.ispartofJVC/Journal of Vibration and Control
dc.sourceScopus
dc.subjectBayesian inference
dc.subjectBouc–Wen hysteresis model
dc.subjectpropagation of uncertainties
dc.subjectSobol’ indices
dc.subjectviscoelastic internal variables
dc.titleOn the use of Stochastic Bouc–Wen model for simulating viscoelastic internal variables from a finite element approximation of steady-rolling tireen
dc.typeArtigo
dspace.entity.typePublication
unesp.author.orcid0000-0002-8962-4850[1]
unesp.author.orcid0000-0001-6430-3746[3]

Arquivos

Coleções