EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Using indirect reduced-order modelling for the optimisation of finite element models

Xiao Xiao, Thomas L. Hill

Nonlinear Dynamics · 2026-11 · 114 (22)

DOI: 10.1007/s11071-026-13151-x ↗

Kurz gefasst

Fachartikel in der Zeitschrift Nonlinear Dynamics, online erschienen am 19.11.2026. Verfasst von Xiao Xiao und 1 weiteren Autor:innen. Die Arbeit nennt 73 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: China Scholarship Council.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Optimising the dynamics of Finite Element (FE) models typically involves repeated dynamic simulations to search for optimal physical parameters. This process can be computationally expensive, especially for models with geometric nonlinearities. To alleviate the computational burden, Reduced-Order Models (ROMs) may be constructed, which can efficiently capture the dynamics of FE models using a small number of dynamically important modes. Among various model reduction methods, indirect applied force methods have emerged as an important class. These methods do not require access to the governing equations of a system; instead, only computationally efficient static datasets are needed during the ROM construction. Therefore, the applied force methods have been widely applied to FE models developed from commercial software. In recent works, the computational advantages of these methods have been comprehensively evaluated, indicating their potential in specific applications. In this paper, optimisation using an applied force method is demonstrated. An FE model, representing a wing, which exhibits complex dynamic modal coupling is introduced. Gradient-based methods are then utilised to optimise physical parameters in the model. The results show that force-based ROMs can capture the dynamics of this model, while parameter changes can alter the modal coupling and require different combinations of modes to be retained in the ROMs. To account for this, an efficient strategy is proposed to automatically determine if additional modes must be included in a ROM or if certain modes could be removed. Equipped with this strategy, using force-based ROMs in the optimisation is demonstrated to be approximately 3–4 times faster compared with direct FE-based optimisation in the wing-inspired model. Moreover, the computational efficiency of force-based ROMs is demonstrated to improve with an increasing number of design parameters, suggesting their applicability to more complex optimisation problems.

Förderung

  • China Scholarship Council

Zitieren

APA

Xiao, X., & Hill, T. L. (2026). Using indirect reduced-order modelling for the optimisation of finite element models. Nonlinear Dynamics, 114(22). https://doi.org/10.1007/s11071-026-13151-x

Vancouver

Xiao X, Hill TL. Using indirect reduced-order modelling for the optimisation of finite element models. Nonlinear Dynamics. 2026;114(22). doi:10.1007/s11071-026-13151-x

BibTeX
@article{Xiao2026,
  author = {Xiao, Xiao and Hill, Thomas L.},
  title = {Using indirect reduced-order modelling for the optimisation of finite element models},
  journal = {Nonlinear Dynamics},
  year = {2026},
  volume = {114},
  number = {22},
  doi = {10.1007/s11071-026-13151-x}
}
Kostenlos starten →