Frank Nießen

Project title

Generative AI in crystal orientation space

What is your project about?

We are surrounded by materials whose properties are closely tied to their microscopic structure. Metals and their alloys, at this scale, consist of differently shaped and oriented crystals, collectively forming the microstructure. Engineering of this microstructure using different chemical compositions and manufacturing processes enables development of new alloys with unparalleled properties. Given this virtually infinite design space, rapid discoveries must rely on computational materials design that operates on digital representations of microstructures. To enhance the predictive capabilities of computational design workflows, this project will for the first time leverage generative AI in crystal orientation space to generate super-realistic digital microstructures. 

How did you become interested in your particular field of research?

Carrying out an apprenticeship at a stainless steel manufacturer while studying mechanical engineering in Germany, I for the first time got to experience the processing of molten steel, a powerful experience that peaked my interest for physical metallurgy ever since. My education in materials and manufacturing engineering at DTU made me appreciate the direct link between theory, engineering and industrial application, motivating me to follow a research career in the field. Based on emerging technology, I see huge potential for computation and data science to advance the field of materials engineering, which is why I increasingly focus on the interface between experimental and computational materials design. 

What are the scientific challenges and perspectives in your project?

While generative AI excels in generation of images, sounds, essays etc., its application to crystal orientation space is non-trivial and unexplored. Due to high-dimensional rotations that are subject to several symmetries, this space is extremely difficult to map. I am convinced that a break-through in the generation of super-realistic digital microstructures can be achieved by leveraging new crystal orientation parametrizations. Generated digital microstructures will then need to be validated with respect to their experimentally characterized counterparts using advanced microstructure analysis and virtual materials testing.

What is your estimate of the impact, which your project may have to society in the long term?

The project has the ambition of fundamentally changing how we design new metals and alloys. A success with this project would for the first time enable rapid digital computational materials design workflows with only minimal experimental input for validation. Such workflow would significantly accelerate and reduce the cost of materials design, allowing custom materials solutions for any given application. On the long term, society will therefore benefit from superior, cheaper and more sustainable products that are enabled by digital materials design.

Which impact do you expect the Sapere Aude programme will have on your career as a researcher?

Receiving the Sapere Aude grant as a researcher is a huge privilege. This grant will for the first time enable me to lead my own research group in a newly unfolding field filled with new and exciting possibilities. The Sapere Aude grant enables me to consolidate valuable international collaborations and opens the door to new collaborations at international top-level. It therefore marks a critical milestone in my career as an independent researcher.

Background and personal life

I live in Næstved together with my wife and daughter. Outside of my work I spend as much time as possible with my family, and I am dedicated to teaching my daughter German. To take care of my well-being, I engage in runs, meditations, and indoor climbing. During my commute to and from DTU I listen to podcasts on societal topics, psychology, and technology, as well as the latest Drum and Bass mixes.