ECCOMAS 2024

A plugin framework for large-scale multi-formulation topology optimization

  • Neiva, Mateus (Pontifical Catholic University of Rio de Jane)
  • Nale, Andrea (University of Ferrara)
  • Menezes, Ivan (Pontifical Catholic University of Rio de Jane)
  • Chiozzi, Andrea (University of Ferrara)

Please login to view abstract download link

Topology optimization is a design computational method that allows the maximization of performance within a given domain under prescribed constraints. The technique has found a wide range of applications due to its flexibility in addressing different problems. In particular, several formulations have been developed since the pioneering work of Sigmund(2001) and Bendsoe and Kikuchi(1988), which respectively handle single and multiple materials, and many educational codes Talischi(2012)-Sanders(2018) have been proposed. These codes demonstrate the potentiality of topology optimization in structural mechanics to generate optimized layouts. Advances and extensions have been proposed, including periodic Sanders(2021) and non-periodic Senhora(2022) microstructures. Motivated by these works, we propose a plugin-based framework that allows treating a multi-material formulation for large-scale structural mechanics problems. The formulation adopted here is the standard compliance minimization, i.e., maximization of stiffness under volume constraint, and the proposed implementation allows the handling of multiple materials with different stiffness properties, i.e., isotropic and anisotropic materials, and various constraints. The plugins are implemented to enable customization to the user for the specific problem to address. Moreover, the software allows adding features to the existing program to extend or change the formulation already implemented. Numerical examples are presented to demonstrate the capabilities and validate the main features of the proposed framework. Future developments will regard the implementation of more complex formulations, such as stress constraints and multiphysics problems.