Would an intersecting plasma in a fusion reactor increase yield due to the collision of particles in the plasma?

Question

It seems to me that having a plasma “stream” in a tokamak that intersected itself would increase the number of collisions happening between particles. Is this feasible or is it negligible in the face of the speeds that particles are individually moving at? Sorry if this is a stupid question.

Answer 1

So to make this work, you would in essence need two tokamaks, with the fuel orbiting in opposite directions, and then somehow make those intersect. Whether such a thing is possible while still being stable seems... unlikely.

However, it is possible to make such a system if you don't thermalize the plasma, and instead work with individual ions. This is called "colliding beam fusion", in which the answer to "does collision help" is "in theory, yes". The problem is that "in theory" part.

Consider a simple version with two particle accelerators aimed at each other, one firing D and the other T. Let us say they are focused and aimed so that the resulting beams are perfectly overlapped. To overcome coulomb repulsion, we only need a total energy of about 100 keV, which can be produced by tabletop devices.

The problem is that the scattering potential for D-T is 10^-9 times its collision probability. So while the fusion reaction returns ~15 MeV compared to the 100 keV put into them by the accelerators, the energy balance is... poor. Such a system is always net negative energy.

This leads to any number of proposals to recirculate the ions while retaining their energy so they have many chances to collide and thereby make the input to output ratio more favourable. One example is the Migma, another is TAE's solution. Neither has proven useful.

So let's consider a spectrum where at one end you have a purely collisional process like Migma, and at the other, a purely thermal distribution like a tokamak. In a tok, the energy is (ultimately) recirculated through the alphas from the reaction being captured by the confinement field and then thermalizing with the fuel, but no similar process exists in something like the Migma, where random thermal energy is precisely what you do not want.

So the question ultimately boils down to which effect is more useful, the additional energy from the collisional process, or the random thermal heating. Given the numbers above, I think the answer is clear, even if one were to roughly double the reaction rate because you have two toks interacting (assuming it's linear), the thermal heating, about 20% of the total output, is far, far more useful given the collisional rates.