Simulation-guided PGD-Stereocorrelation for hybrid testing
- authored by
- Xuyang Chang, Bastien Bodnar, Jan Grashorn, François Hild, Fabrice Gatuingt, Stéphane Roux
- Abstract
Hybrid testing aims to evaluate the dynamics of a large scale structure from which an isolated part is experimentally tested, while the rest is simulated in the control loop for prescribing earthquake-induced motions at boundaries. One of these tests, led in real-time on a three-story reinforced concrete structure to quantify damage due to successive multiaxial seismic loadings, and focusing experimentally on a single column, is considered in this study. Stereocorrelation (SC) using high speed cameras is an appealing technique to further analyze such tests, providing 3D surface displacement fields and damage indications. Assuming a spacetime separation of the kinematics and constructing a tailored temporal basis, Proper Generalized Decomposition SC (PGD-SC) is well-suited to handle long image series efficiently. Furthermore, it aligns perfectly with the philosophy of model (hyper-)reduction on the simulation side as needed for real-time control. The present study explores the potential of using temporal modes extracted from both existing simulations and image series for SC analyses. It is shown that the simulation-guided PGD-SC outperforms the classical PGD-SC in terms of computation time with a minimal implementation effort. The kinematic measurements also highlight the direction in which the numerical model should be refined to better account for the observed non-linearities. The perspective of using such simulation-guided PGD-SC into hybrid testing protocols (including pseudo-dynamic experiments) appears credible.
- Organisation(s)
-
Institute for Risk and Reliability
International RTG 2657: Computational Mechanics Techniques in High Dimensions
- External Organisation(s)
-
University of Science and Technology Beijing
École normale supérieure Paris-Saclay (ENS Paris-Saclay)
Conservatoire National des Arts et Metiers
- Type
- Article
- Journal
- Computer Methods in Applied Mechanics and Engineering
- Volume
- 435
- No. of pages
- 22
- ISSN
- 0045-7825
- Publication date
- 15.02.2025
- Publication status
- Published
- Peer reviewed
- Yes
- ASJC Scopus subject areas
- Computational Mechanics, Mechanics of Materials, Mechanical Engineering, General Physics and Astronomy, Computer Science Applications
- Electronic version(s)
-
https://hal.science/hal-04830346v1/file/CMAME2025-ccsd.pdf (Access:
Open)
https://doi.org/10.1016/j.cma.2024.117672 (Access: Closed)