This research focus advances simulation‑driven engineering for steel construction. We combine high‑fidelity finite element modeling, optimization and scientific machine learning to enable efficient, reliable and manufacturable designs—particularly for metal additive manufacturing (AM).
Core themes
- Finite element simulations for metal AM (thermal and thermo‑mechanical): We model heat transfer, residual stresses and distortion, including build‑sequence effects and support interactions. Calibration and validation against experiments, sensitivity studies and uncertainty quantification ensure predictive accuracy across materials and process windows.
- Structural and topology optimization: We develop optimization workflows that integrate manufacturability constraints (overhang angles, minimum feature size, support minimization), build‑orientation selection and lattice/graded infill. Objectives span stiffness, weight, thermal performance and fatigue, with robust and multi‑objective formulations.
- Scientific Machine learning for engineering simulations: We create surrogate and reduced‑order models to accelerate FE analyses, employ physics‑informed and operator‑learning approaches for multiphysics problems, and use active learning and uncertainty quantification for trustworthy predictions. Applications include rapid process parameter studies, real‑time distortion prediction and digital twins for AM and structural systems.
Links to other research focuses of the institute
- Advanced structural materials: Data‑driven and mechanistic constitutive models for steels used in AM and conventional fabrication.
- Automated/robotic/additive construction: Process‑aware simulation, support design and print‑path planning informed by optimization and ML.
- Support structures for wind energy: High‑fidelity FE for fatigue‑critical, nodes; design optimization of support structures and nodes; fast, reliable surrogates.
