ECCOMAS 2024

High performance computation of wave propagation through elastic and piezoelectric media using dolfin-hpc

  • Bhole, Ashish (KTH Royal Institute of Technology, Stockholm)
  • Hernández-Oliván, Javier (Instituto Tecnologico de Aragon (ITA))
  • Manuel Royo, José (Instituto Tecnologico de Aragon (ITA))
  • Calvo-Echenique, Andrea (Instituto Tecnologico de Aragon (ITA))
  • Hoffman, Johan (KTH Royal Institute of Technology, Stockholm)

Please login to view abstract download link

The main goal of the GENEX project [1] is to develop a novel end-to-end digital twin- driven framework for optimizing aircraft composite manufacturing and maintenance processes while ensuring aircraft safety. In this highly interdisciplinary project, we are interested in developing a methodology for advanced physical and data-driven simulations to predict structural damage. SHM technology uses piezoelectric transducers mounted on aircraft composite structures. In the context of the GENEX project, we are interested in developing HPC-scalable models, including piezoelectric transducers, for structural health prediction. The problem at hand translates to ‘wave propagation through elastic and piezoelectric media’. For our interest, the governing partial differential equations (PDEs) describing piezoelectric materials are written as a coupling between elastodynamic and electrostatic problems. The aircraft composite materials modeling (described by elastodynamics) is then a subset of piezoelectric modeling. While computational elastodynamics [2] is a very well-studied topic, the literature on computational piezoelectric media is fragmented. In this work, we present a systematic review of the latter topic. Dispersion relation preserving wave propagation simulations can be expensive to perform motivating the need for HPC. We consider the FEM to discretize PDEs and implement the resulting algorithms in the framework of the HPC-suitable open-source code: dolfin-hpc [3]. In the context of the GENEX project, our simulations will generate a dataset to train machine learning models for damage detection. REFERENCES [1] GENEX project, https://www.genex-project.eu/ [2] Thomas J. R. Hughes, The finite element method, Linear Static and Dynamic Finite Element Analysis. Dover Publications, 2000. [3] J. Hoffman, J. Jansson and N. Jansson, FEniCS-HPC: Automated Predictive High- Performance Finite Element Computing with Applications in Aerodynamics, Lecture Notes in Computer Science, Vol. 9573, pp. 356–365, 2016.