A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Model Order Reduction (MOR) Techniques
In previous chapters, a finite element framework was presented for the treatment of EHL problems under different configurations. The proposed models extended the computational domain of the contacting solids into their depth, rather than remaining at the contact surface, as done by conventional EHL numerical models. Yet, employing FEM enabled the use of non‐regular non‐structured meshing, maintaining the size of these models relatively small and making them at least as competitive as state‐of‐the art ones, in terms of computational efficiency. Nonetheless, a major improvement is still possible since, in most cases, the elastic deformation of the contacting solids is not needed within their depth for the solution of the EHL problem. Only the surface deformation is needed. Thus, many elastic degrees of freedom are solved for in vain. In the current chapter, the use of Model Order Reduction (MOR) techniques is proposed as a remedy to this problem. Two techniques are developed: EHL‐Basis and Static Condensation with Splitting (SCS). The former allows a spectacular reduction in the size of the linear elasticity part to less than 30 degrees of freedom. However, it is rather complex and suffers from a loss of generality. The latter confines the linear elasticity part to the contact surface, involving a smaller reduction order than the former. However, it is rather simple and it preserves the generality of the solution scheme. Both techniques allow significant computational time speed‐ups for the FEM modeling of EHL problems.
Model Order Reduction (MOR) Techniques
In previous chapters, a finite element framework was presented for the treatment of EHL problems under different configurations. The proposed models extended the computational domain of the contacting solids into their depth, rather than remaining at the contact surface, as done by conventional EHL numerical models. Yet, employing FEM enabled the use of non‐regular non‐structured meshing, maintaining the size of these models relatively small and making them at least as competitive as state‐of‐the art ones, in terms of computational efficiency. Nonetheless, a major improvement is still possible since, in most cases, the elastic deformation of the contacting solids is not needed within their depth for the solution of the EHL problem. Only the surface deformation is needed. Thus, many elastic degrees of freedom are solved for in vain. In the current chapter, the use of Model Order Reduction (MOR) techniques is proposed as a remedy to this problem. Two techniques are developed: EHL‐Basis and Static Condensation with Splitting (SCS). The former allows a spectacular reduction in the size of the linear elasticity part to less than 30 degrees of freedom. However, it is rather complex and suffers from a loss of generality. The latter confines the linear elasticity part to the contact surface, involving a smaller reduction order than the former. However, it is rather simple and it preserves the generality of the solution scheme. Both techniques allow significant computational time speed‐ups for the FEM modeling of EHL problems.
Model Order Reduction (MOR) Techniques
Habchi, Wassim (author)
2018-05-14
42 pages
Article/Chapter (Book)
Electronic Resource
English
AI-Based Model Order Reduction Techniques: A Survey
| Springer Verlag | 2025
A Critical Exposition of Model Order Reduction Techniques: Application to a Slewing Flexible Beam
| Online Contents | 2019
Challenges of order reduction techniques for problems involving polymorphic uncertainty
| BASE | 2019
Automated Data Reduction Techniques for Network Model Management
| British Library Conference Proceedings | 1994
Model reduction techniques for simulating complex flow processes
| DataCite | 2020