PhD Position: Multi-Objective Design of Structural Materials by AI-based Structure-Property Inversion in Zürich

The Advanced Manufacturing Lab (am|z) at the Department of Mechanical & Process Engineering develops advanced computational models alongside new manufacturing systems to overcome the limitations of today’s materials processing and production technologies.

The design of lightweight and efficient machine components relies on the optimal placement of material for a given load case. In many real-world applications, the loads span multiple physical domains, including structural, thermal, fluid, and magnetic. Multi-material additive manufacturing provides the ability to locally vary the mechanical structure (e.g. lattices, topology optimization) and material properties within the component to achieve multiple design objectives. What is lacking is a set of tools that leverage this capability and can apply it to the optimization of parts. To accomplish this task, we seek a highly motivated and talented PhD candidate who will support us to develop new and to extend existing design tools to design multi-material components satisfying multiple physical objectives.

Your main contributions to this research project will be to:

• develop finite element (FE) modeling tools for physics-based forward simulation of multi-material lattice structures and to derive data-driven surrogate models to improve computational efficiency;
• develop machine learning (ML) models to invert the design space for the allocation of materials and local geometries based on requirements;
• implementation of filters and cost functions to constrain the design space based on e.g. material compatibility, process stability, cost, and sustainability.

Besides benchmarking and programming duties, this PhD position entails publishing high-quality journal papers, presenting at international conferences, and teaching contributions. This interdisciplinary research is conducted in a modern working environment within a young and highly motivated team. The position is partially funded by an industrial partner through a collaborative research agreement.

• You hold a Master’s degree in Computational Science, Mechanical Engineering, Materials Science, or another relevant subject, with an analytical mindset and strong computer skills.
• You should have an excellent understanding of the FE method and the underlying mathematics, including solving partial differential equations, linear algebra, geometric analysis and numerical integration.
• Preferably, you will have experience using commercial FEM software such as COMSOL or ABAQUS in the manufacturing domain. Experience with physical modeling using continuum mechanics, fluid dynamics, and heat transfer is desired.
• You should also have an understanding of how to train and apply ML models for classification, prediction and optimization.
• Coding experience in C++/Python is required.
• You have the ability to work independently and in small cross-functional teams.
• Familiarity with collaborative coding and version control systems (e.g., git) is beneficial. Professional command of English (both written and spoken) is mandatory.

ETH Zurich is a family-friendly employer with excellent working conditions. You can look forward to an exciting working environment, cultural diversity and attractive offers and benefits.