Talk-Polywell.org

I dont know how or where I got the idea from, but I was thinking, because thermalization is such a huge problem, for the machine to run steady state, there has to be a way to deal with thermalization.

The resonance frequency of the ions and electrons are very different, so we can actually pick the right frequency microwave and heat what we want to heat. This may be the solution to dealing with thermalization.

My question is this, how quickly can the desired particle (I actually dont know which one to heat, I thought it's the ions, but I can be wrong) be heated compared to how fast the system thermalizes?

Robthebob wrote:I dont know how or where I got the idea from, but I was thinking, because thermalization is such a huge problem, for the machine to run steady state, there has to be a way to deal with thermalization.

The resonance frequency of the ions and electrons are very different, so we can actually pick the right frequency microwave and heat what we want to heat. This may be the solution to dealing with thermalization.

My question is this, how quickly can the desired particle (I actually dont know which one to heat, I thought it's the ions, but I can be wrong) be heated compared to how fast the system thermalizes?

There are a number of different types of thermalization, including

1. the electrons become isotropic
2. the electron energy distribution becomes Maxwellian
3. the ions become isotropic
4. the ion energy distribution becomes Maxwellian
5. the electron and ion temperatures becomes equal

If my memory serves me, the higher a process is on this list, the faster it is, or equivalently, the harder it is to prevent. The quantitaive formulas can be found in the chapter on "Collisions and Transport" in the NRL PLASMA FORMULARY. All of these processes have been discussed in relation to polywell physics.

You seem to have the last one in mind, presumably in relation to p-B11 fusion (which is not the only way to run a polywell, nor the easiest). In that case the ions must be hot enough for fusion, but the electrons cold enough to limit the power lost to bremsstrahlung. Thermalization in the sense that the ions tend to get colder and the electrons hotter is an obvious problem. If you heat one species, then it certainly must be the ions, and you have to do this one way or another anyway. Maybe you can use the power supplied to the electron and/or ion beams or to the magrid.

You could presumably also use microwaves. In a tokamak, you could use rf power at the ion cyclotron frequency. The choice is not so obvious in a polywell because in most of the plasma volume there is no magnetic field. But even if you found a way to do it, which I think you could, that doesn't change the basic problem of the power balance.

Art Carlson wrote: There are a number of different types of thermalization, including

1. the electrons become isotropic
2. the electron energy distribution becomes Maxwellian
3. the ions become isotropic
4. the ion energy distribution becomes Maxwellian
5. the electron and ion temperatures becomes equal

If my memory serves me, the higher a process is on this list, the faster it is, or equivalently, the harder it is to prevent. The quantitaive formulas can be found in the chapter on "Collisions and Transport" in the NRL PLASMA FORMULARY. All of these processes have been discussed in relation to polywell physics.

Art,

I read the wiki article on isotropy, didnt quite understand what it means. (if you would explain that to me) I dont know if my analysis is correct or not, I find that the thermalization of electrons to be just fine, in fact, if we can, the electrons should possess as little energy as possible, so if the distrubution becomes maxwellian with most of the population of the electron possessing low energy in order to minimize the interaction between ions and electrons, then it may be not be a bad thing. I dont know, I'm pretty sure we're trying for electrons not interacting with the ions, so the higher the energy difference, the better. But then you may have to worry about bremsstrahlung, um....

I'm actually not quite sure about the next part, when you use RF heating, how specific of the things you're heating be? I know you can selectively heat different populations, so you can choose to only heat electrons or ions. This next part is a bit weird, and if there isnt a way, which I was thinking about it, and theres probably not a way to do it. Is there a way to selectively heat only the lower end of the thermalized ion population? Meaning to only heat the ions with too little energy to fuse. I dont know...

Robthebob wrote:I read the wiki article on isotropy, didnt quite understand what it means. (if you would explain that to me) I dont know if my analysis is correct or not, I find that the thermalization of electrons to be just fine, in fact, if we can, the electrons should possess as little energy as possible, so if the distrubution becomes maxwellian with most of the population of the electron possessing low energy in order to minimize the interaction between ions and electrons, then it may be not be a bad thing. I dont know, I'm pretty sure we're trying for electrons not interacting with the ions, so the higher the energy difference, the better. But then you may have to worry about bremsstrahlung, um....

Isotropy means there is no preferred direction. For example, if the electrons are isotropized, and you look at those with a particular energy, then there will just as many going "north", as "south" or east".

For a fixed ion temperature, the rate of energy loss to the electrons is greater when the electron temperature is lower.

Robthebob wrote:I'm actually not quite sure about the next part, when you use RF heating, how specific of the things you're heating be? I know you can selectively heat different populations, so you can choose to only heat electrons or ions. This next part is a bit weird, and if there isnt a way, which I was thinking about it, and theres probably not a way to do it. Is there a way to selectively heat only the lower end of the thermalized ion population? Meaning to only heat the ions with too little energy to fuse. I dont know...

In some circumstances you can do a good job of heating mostly electrons or mostly ions. Sometimes you can even heat the low energy or the high energy populations preferentially, at least to some degree.

In the case of the polywell, nobody has shown exactly how to do preferential heating in just the way they want, and maybe it is not possible, or not possible at a reasonable efficiency. I've never gotten into it because you can show that it wouldn't help even if it were possible. For example, if you selectively heat low energy ions, those ions will re-thermalize a lot faster than they will fuse, so it costs you more energy than you gain through additional fusions.