This research focuses on the development of model order reduction methods to speed up the (repeated) multiscale prediction of permeability in composite manufacturing processes. The permeability of a composite material has a very large impact on the production process of components. It is therefore required to have an accurate knowledge and understanding of the permeability, which depends on the micro- and meso-scale geometry, in order to have a robust manufacturing process. One approach to analyze the permeability at the meso-scale (unit cell) level is through computational-fluid-dynamics techniques based on the Navier-Stokes equations. However, these simulations can be computationally demanding and, due to the inherent variability of the geometry of the microstructure for a typical material, it is practically infeasible to simulate a sufficient number of geometries in order to obtain a reliable average value of the textile permeability. To reduce this computational load into a feasible range, which will enable the use of these simulations in industry, this project aims to develop novel model order reduction techniques.
The research will focus on the following topics:
- Model reduction for steady-state and transient Navier-Stokes simulation
- Efficient spatial discretization for unit cells with complex geometries using model order reduction concepts
- Exploitation of parametric model order reduction methods for repeated simulations needed to capture different (in-space) distributed material configurations due to shearing and compaction
- Propagation of reduced order model response over multiple scales
The project will benefit from collaboration with Siemens PLM Software (CFD software Star-CCM+). This project (MOR4MDesign, http://www.sim-flanders.be/project/mor4mdesign-sbo) fits in the M3 R&D Program (http://www.sim-flanders.be/research-program/m3), funded by SIM.
Applications both for a PhD (4 years) or a post-doc (3 years) positions will be considered.
A successful candidate should have the following profile:
- Computational fluid dynamics theoretical knowledge and experience beyond simple use of commercial software
- Theoretical knowledge and experience in application of numerical methods and programming
- Ability to apply a CFD simulation to real life flow through porous or fibrous materials
Please write to: stepan.lomov [at] kuleuven.be (stepan[dot]lomov[at]kuleuven[dot]be)