Aerospace Colloquium

The exposure to small quantities of hydrogen (in the ppm range) can reduce the ductility, fracture toughness, and fatigue crack growth resistance of metals by orders of magnitude. This remarkable phenomenon, so-called ‘hydrogen embrittlement', is pervasive across the transport, defence, energy and construction sectors, due to the ubiquitousness of hydrogen, and constitutes one of the biggest threats to the future of hydrogen as a clean energy carrier. However, understanding and predicting hydrogen-metal interactions have proven to be more difficult than first anticipated.

This talk will overview the author's cross-disciplinary efforts to shed light on the nature of hydrogen-assisted failures. Focus will be on the combination of experiments and modelling across scales and disciplines, to mechanistically resolve the physical processes underlying hydrogen embrittlement, from the (electro-)chemical uptake of hydrogen from gaseous and aqueous electrolyte environments, to the diffusion and transport of dissolved hydrogen in the crystal lattice, and to the hydrogen-assisted nucleation and growth of cracks. The insight gained has led to the development of a new class of physically-based electro-chemo-mechanical models for hydrogen embrittlement that can not only obtain an unprecedented level of agreement with laboratory experiments but also deliver finite element predictions at scales relevant to engineering practice, bringing the "Virtual Testing" paradigm to hydrogen-containing environments. Finally, focus will be placed on a recent, pressing challenge: with the upcoming use of hydrogen in aviation, there is a need to understand, from first principles, the behaviour of metals exposed to hydrogen and extreme environments (temperatures ranging from 20 K to nearly 1800 K). Given the small margins of the aviation industry, this requires a new way of thinking, whereby every hydrogen-material interaction must be appropriately described (and not fitted), from the atom scale, to the dislocation level, and to the macroscopic component performance, through a new generation of mechanistic quantum-mechanically informed, thermo-chemo-mechanical crystal plasticity models.