> The new technique involved repeatedly twisting a sample of 304 austenitic stainless steel in a machine in certain ways. This led to spatially grading the cells that made up the metal, resulting in the build-up of what the team describes as a submicron-scale, three-dimensional, anti-crash wall.

Interesting. Not a metallurgist but this takes advantage of stainless steels natural tendency to work harden. e.g. if you have ever broken a paperclip or other piece of steel by bending it back and forth until it fatigues, fractures, and beaks off. That happens in soft standard steels like A36 (edit forgot to finish this...) However, in stainless steel instead of a fracture forming at the bends crease, it hardens. As you try to bend it again, it bends in a new place as the original crease has hardened.

> Such improvements, the team claims, could allow products made using the metal to be up to 10,000 times more resistant to fatigue.

Very bold claim that if true is a game changer. My concern is how does this process scale to large complex structural pieces? Assuming since this internal structure will be ruined by annealing it must be performed after final shaping of the material. Welding should not be effected, especially low heat effect zone processes like laser and electron beam as you account for material alteration from welding during design.

I'd like to have some of these stainless steel paperclips. Sounds like a good fidget toy.