The same mathematical structures – coupled PDEs, multiscale phenomena, parameter uncertainty – appear across remarkably different physical systems. This talk explores how a unified computational framework, developed across nearly a decade of research, can travel from one application domain to another while retaining its predictive power and computational efficiency.
Starting from the numerical simulation of organic thin-film transistors (PhD), through the development of high-performance digital twins of the human heart, to current work on reduced-order modeling for PEM fuel cells, I will show how advances in finite element methods, matrix-free solvers, and model order reduction naturally extend to new challenges – including those posed by hydrogen energy systems.
I will close by discussing how uncertainty-aware reduced-order models can act as enabling technologies for the digital twin of a hydrogen plant: making simulations fast enough for real-time use, reliable enough for decision-making, and transparent enough for certification