My recommendation for self-study would be to borrow or purchase a copy of "Fluid Simulation for Computer Graphics" by Robert Bridson. It is by far the most approachable resource on eularian fluid simulation that I have found.
I would warn, however, that it took me a nearly a month to build my fluid simulator[1] after reading. I didn't have any good reference implementations to follow when I built it, and there's a lot of little details to get right. My code still has an annoying bug that appears when building with optimizations on Haswell or later, and I need to clean up the code a little, but it should nevertheless be useful as a reference for the features it implements.
That said, being at a university really does help. There you're surrounded by other people who understand the linear algebra and vector calculus concepts you're working with... or, at least, in grad school you are. Having people you can talk with makes quite a difference when you're confused or unsure about something. It also makes this sort of study feel... normal.
He's also not going to achieve his goal without years of heavy study. Doing this in his off-hours may be fun and rewarding, but I think he's unlikely to successfully reach the point that he's impressing aerospace R&D.
Of course, I'm not sure an undergraduate degree in CS is really going to do it, either. A graduate degree maybe, but I'm not sure it would be easy to skip undergrad. If not, an undergrad degree in physics or mathematics might be a better complement to his existing experience.
Others have pointed out that the opportunity cost for going back to school is tremendous. They're right. For a basic estimate of the cost, he can multiply his current earnings by at least four. It's probably enough to buy a house, but that's what dreams cost.
Is it worth it? Well, financially no. The satisfaction from achieving that dream might be, though. That's up to him.
[1]: https://github.com/cgmb/euler/blob/master/README.md