Comment by matt-attack

I’d be curious to know if this sort of thing could’ve been predicted through computer modeling. And if not, does that mean, we have a gap in our fundamental fluid equations?

And if so, couldn’t we just have a model iterate on different surface patterns and optimize?