I have a video on NASA's current NTP program and a series on a bunch of different nuclear designs. And a video on using NTP designs to build a space tug.

This area is a bit of a hobby of mine.

The problem with NTR designs is that the old projects such as NERVA weren't really designed to build real engines. They did a lot of testing and came up with something that kindof worked but still had issues.

There have been numerous designs since then and a lot of assertions about how great NTR will be because "will you look at that Isp?".

There are two big obstacles.

The first is weight. The combustion chambers in chemical rockets are mostly empty space so those engines are light, while NTR engines have heavy nuclear cores. They also require shielding to keep from irradiating the payload, and their Isp benefit comes because they use liquid hydrogen, which requires larger tanks than chemical rockets (LOX is around 15 times more dense than LH2).

The second issues is durability. Chemical engines run at temperatures up to around 3300 kelvin, but NTR are typically designed at 2700K and a lot of the NERVA versions ran closer to 2200K. The problem is that it's really hard to design a reactor that runs with a very high thermal load and holds together, but you need that high temp or you don't get the Isp that you are hoping for. The nerva designs tended to spit their insides out during full power runs; this was somewhat fixed with plating the inside of the propellant channels, but they never achieved significant full-power runs with one engine.

You will find lots of assertions that new materials and new designs will fix these issues. NASA's NTR program gives us a chance of actually answering those questions, but they've set the bar fairly low in my opinion, which improves the chances of success - actually building and maybe testing a flight reactor - which is the right goal at this point of time.