[hobscoop]

It sounds like they're trying to build a device with a distribution of ions in velocity space which is very far from the equilibrium Gaussian-type distribution ('Maxwellian'). This is difficult because there will be about a million collisions for every fusion reaction that happens, and collisions tend to redistribute velocities toward the equilibrium. Collisions between unlike particles (different species of ions, or ions vs electrons) will also cause diffusion in physical space. This could cause ions to leave the chamber much more quickly.

[ISL]

This is the key observation that isn't obvious about fusion. The capture cross-section is smaller than the scattering cross-section, so essentially any beam design is doomed to fail.

I'm an experimental physicist who recently looked hard at jumping into the fusion game. My approximate conclusion is that anything that doesn't involve some form of adiabatically heated plasma is extremely unlikely to work.

Fusion is an efficiency and engineering problem: how can you ensure that each particle gets enough energetic collisions to have a prayer of fusing before it loses energy?

[pmdulaney]

Hey, no fair -- you sound like you actually know what you're talking about...