A Galerkin/hyper-reduction technique to reduce steady-state elastohydrodynamic line contact problems

A novel model order reduction technique for the elastohydrodynamic problem consisting of the Reynolds equation and the elasticity equation is proposed that drastically reduces the dimensionality of the problem and the computing time. The method is developed for stationary Newtonian isothermal line c...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computer methods in applied mechanics and engineering 2021-12, Vol.386, p.114132, Article 114132
Hauptverfasser: Scurria, Leoluca, Fauconnier, Dieter, Jiránek, Pavel, Tamarozzi, Tommaso
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:A novel model order reduction technique for the elastohydrodynamic problem consisting of the Reynolds equation and the elasticity equation is proposed that drastically reduces the dimensionality of the problem and the computing time. The method is developed for stationary Newtonian isothermal line contacts. It employs the Galerkin projection method for the structural part combined with a hyper-reduction technique for the Reynolds equation. The model order reduction strategy consists of two phases. A first offline phase where snapshots of the full order model solution are computed using a selective refinement technique; and an online phase where the precomputed bases are used to compute the reduced order model. The method leads to a large reduction in the dimensionality of the full order model which results in a considerable reduction of the computing time, in a o(103) speedup factor, with an excellent correlation to the results obtained with the full order model. •Model order reduction technique for EHL line contacts in steady-state conditions•The MOR technique combines a Galerking projection and a hyper-reduction method•The proposed methodology drastically reduces the dimensionality of the problem•The reduced dimensionality translates to a speed up factor of o(103)•The speed-up factor can be further increased using a pre-computed Boolean matrix•The Reduced Order Model shows an error lower than 1% for the case analysed
ISSN:0045-7825
1879-2138
DOI:10.1016/j.cma.2021.114132