.png){kind=icon}
Return
Data-driven identification-free approach for nonlinear structural dynamics
.png){kind=icon}
PRE
AI
DOI：10.1007/s11071-025-11163-7.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In the field of structural dynamics, accurately identifying and simulating nonlinear structures represents a significant challenge due to the inherent complexity of the phenomena involved. The traditional model-based approach, whether physics-oriented or data-driven, is inherently susceptible to epistemic uncertainties associated with the modeling and identification process. This work presents an alternative data-driven identification-free approach, inspired by the recently proposed data-driven computational mechanics (DDCM) paradigm. Unlike traditional model-based approaches that require explicit models of restoring forces for numerical integration, the proposed approach enables the simulation of nonlinear dynamical systems directly from measured restoring force datasets by formulating it as a double distance minimization between the dataset and discrete dynamic equilibrium constraints. In this study, we present the formulation and proof-of-concept of a novel data-driven solver based on a restoring force dataset to address nonlinearities in structural dynamics problems. A comprehensive numerical study of a Duffing oscillator with symmetric nonlinearity demonstrates the high prediction accuracy of the data-driven solver. The experimental application of the data-driven solver is also demonstrated for a Hybrid Nonlinear Energy Sink (HNES).
Keywords:
Data-driven computational mechanics
Restoring force
Data-driven solver
Hybrid nonlinear energy sink
Journal
IF:
6
Papers:
1.4W
Citations:
4.1W
Organization
Cited Papers
.png){kind=icon}
ShareSaveShareSaveDiscovering governing equations from data by sparse identification of nonlinear dynamical systems
OAAI
OAAIShareSave
ShareSaveShareSaveShareSave
View more