Heat produced by Focus Fusion and cooling

[benf]

Sorry, I have my beams mixed up. I meant the ion beam from which electricity will be extracted at the solenoid. Will the solenoid get very hot?

[Brian H]

benf wrote: Sorry, I have my beams mixed up. I meant the ion beam from which electricity will be extracted at the solenoid. Will the solenoid get very hot?

The only sources of heating I can imagine are by direct conduction from the plasma and by resistance in the coils to the induced current flow. Are those what concern you?

[benf]

If I understand correctly, a charged particle beam is created by the fusion reaction in the plasmoid. The beam will exit the pinch and be directed to the solenoid, which converts the concentrated energy into electricity. I presume this will be hot, as with a lightning strike? And it will be focused into a beam which would mean a small surface area will be impacted. Can it be absorbed without damage to the solenoid, long term?

[zapkitty]

benf wrote: If I understand correctly, a charged particle beam is created by the fusion reaction in the plasmoid. The beam will exit the pinch and be directed to the solenoid, which converts the concentrated energy into electricity. I presume this will be hot, as with a lightning strike? And it will be focused into a beam which would mean a small surface area will be impacted. Can it be absorbed without damage to the solenoid, long term?

?…

I’d assumed that the power takeoff for the alpha particles would be via pass-through coils and thus no physical impact with anything until the helium ions had been slowed way down by those coils?

As this is a very predictable beam one doesn’t need plates scattered across the interior of the containment volume ala a polywell and thus no need for impacts.

It’s the electron beam heading in the opposite direction that would pose a threat to equipment.

[Brian H]

zapkitty wrote:

If I understand correctly, a charged particle beam is created by the fusion reaction in the plasmoid. The beam will exit the pinch and be directed to the solenoid, which converts the concentrated energy into electricity. I presume this will be hot, as with a lightning strike? And it will be focused into a beam which would mean a small surface area will be impacted. Can it be absorbed without damage to the solenoid, long term?

?…

I’d assumed that the power takeoff for the alpha particles would be via pass-through coils and thus no physical impact with anything until the helium ions had been slowed way down by those coils?

As this is a very predictable beam one doesn’t need plates scattered across the interior of the containment volume ala a polywell and thus no need for impacts.

It’s the electron beam heading in the opposite direction that would pose a threat to equipment.
Yes, the alpha (helium pos-ions) just passes through the coil tube (solenoid) on its way to the catchment chamber, being slowed and inducing current as it goes. It is neutralized in the chamber, leaving ordinary He4 gas.

The electron beam, as mentioned above, will be stopped/contained by the plasmoid itself according to Eric.

[Tulse]

Slightly off topic, and I’m sure this has been addressed before, but how much helium would a commercial FF unit produce? Would it be enough to be worth collecting and selling, especially given the forecast shortage of helium?

[Aeronaut]

zapkitty wrote:

If I understand correctly, a charged particle beam is created by the fusion reaction in the plasmoid. The beam will exit the pinch and be directed to the solenoid, which converts the concentrated energy into electricity. I presume this will be hot, as with a lightning strike? And it will be focused into a beam which would mean a small surface area will be impacted. Can it be absorbed without damage to the solenoid, long term?

?…

I’d assumed that the power takeoff for the alpha particles would be via pass-through coils and thus no physical impact with anything until the helium ions had been slowed way down by those coils?

As this is a very predictable beam one doesn’t need plates scattered across the interior of the containment volume ala a polywell and thus no need for impacts.

It’s the electron beam heading in the opposite direction that would pose a threat to equipment.

Also like the polywell, the electric output pulses will very likely have verry high voltages, which would lead to higher currents. Any way you slice it, this is a lot of power in a very small volume, I expect it to need hefty cooling, which the patent confirms near the end. The patent assumes that the onion is going to be the biggest set of challenges to commercializing FF as an electric generator. Cooling and manufacturing in any kind of volume are significant engineering challenges in their own right.

[Brian H]

Tulse wrote: Slightly off topic, and I’m sure this has been addressed before, but how much helium would a commercial FF unit produce? Would it be enough to be worth collecting and selling, especially given the forecast shortage of helium?

There was a discussion of that somewhere onsite at one point, and Eric weighed in saying it would be worthwhile, I think. But if you consider that the total boron required in a year for a 5MW generator would be around 5 lb., that’s near the upper limit of how much helium you’d get.

[jamesr]

I would expect the ion beam to have an angular spread of at least a few degrees. However since the Rogowski coil (NB which is a coiled-coil so not technically a solenoid) will just be slowing down the ions in the direction along the beam, as it slows the beam will spread out further. Until all the parallel motion is extracted from the ions, or they collides with the edge.

The perpendicular ion velocity also increases due to the like charges repelling. So you want the coil long enough & with enough turns to extract enough energy, before the beam blows up significantly. However I suspect you want it short enough that you can let it expand and dissipate the perpendicular component of its energy once safely out the other side.

[benf]

“Cool!”, that lends clarity to what takes place. A slow motion visual of this process would be nice too. In the Torulf Greek animation the beam looks laser like, but it’s possible to slow it down and basically strip it, imparting the energy progressively to the coil as electricity without too much heat. Interesting…

[Brian H]

benf wrote: “Cool!”, that lends clarity to what takes place. A slow motion visual of this process would be nice too. In the Torulf Greek animation the beam looks laser like, but it’s possible to slow it down and basically strip it, imparting the energy progressively to the coil as electricity without too much heat. Interesting…

I may be wrong, but it is my impression that solenoids are used almost ubiquitously, without noticeable heat issues. I’d have assumed that the resistance of the coil wiring is low enough to minimize that.

[jamesr]

Brian H wrote:
I may be wrong, but it is my impression that solenoids are used almost ubiquitously, without noticeable heat issues. I’d have assumed that the resistance of the coil wiring is low enough to minimize that.

If you think that we want to be getting roughly the same power induced in the rogowski coil as is put through the anode at the start (with the excess energy output coming from the x-rays), then the resistive heating issues could be the same order of magnitude as for the anode.

I suspect though since the pulse from the ion beam will be much shorter in duration than the current flow in the anode, the skin effect would be even more pronounced. If the induced voltage is much higher then the current, an therefore heating losses, will be correspondingly lower. There will be a trade off between having a large minor radius to increase the voltage, and a small enough minor radius in comparison to the major radius of the coil to reduce its inductance and so response to the high frequency pulse.

See wikipedia for the relavent formulae.

[Allan Brewer]

vansig wrote: …

skin depth depends on the frequency. we’re rather wide-band, here, with the rising edge in ~25 ns, and the pulse lasting microseconds. if 25 ns is 1/2 wave, then f=20 MHz; whereas the bulk of the current is delivered as though it is, say, a ~500 kHz pulse, and skin depth increases through the pulse, as δ = sqrt( (2 ρ) / (ω μ) ).
http://en.wikipedia.org/wiki/Skin_effect

Working on those figures the skin depth goes from 36 micrometres (at f=20 MHz) up to 820 micrometres (f=500KHz), the latter for the bulk of the current. This is equivalent (using notional 300 shots per second each of length 1 microsecond) to only 1.3KWatts (at f=20 MHz) down to 48Watts (f=500KHz), of copper anode resistive heating, the latter for the bulk of the current.

I apologise that I had been working on an erroneous skin depth of 20nm which I assumed from the discussion on added layers rather than looking at the proper calculated skin depth. This means I was completely wrong when I said “So its beginning to sound like the dominant source of heat would be resistive heating in the outer skin of the electrodes.”, and misled the ensuing discussion about electrode structure to reduce resistive heating, although it was interesting. Its still correct however that assuming the expected 10-20% inefficiency there would be 4-8MWatts of heat to remove from the system – which is challenging – its just that the heat is not arising from electrode resistance.

[Brian H]

Allan Brewer wrote:
…

Its still correct however that assuming the expected 10-20% inefficiency there would be 4-8MWatts of heat to remove from the system – which is challenging – its just that the heat is not arising from electrode resistance.

I’m curious what inefficiencies you envisage. Heat is directly generated in the plasmoid, and some losses would be expected in the ‘onion’. What others do you see?

[Allan Brewer]

Brian H wrote:

…

Its still correct however that assuming the expected 10-20% inefficiency there would be 4-8MWatts of heat to remove from the system – which is challenging – its just that the heat is not arising from electrode resistance.

I’m curious what inefficiencies you envisage. Heat is directly generated in the plasmoid, and some losses would be expected in the ‘onion’. What others do you see?

Well that’s really the question I have been asking during this thread. I started with Rezwan’s posting https://focusfusion.org/index.php/site/article/how_will_we_get_there_from_here/ which not unreasonably expects unspecified inefficiencies of that order. I suspect you are right that the 70KJoules of energy to form the plasmoid will not all come out with the fusion energy into the Rogowski coil, and the onion will be less than completely efficient capturing the X-rays’ energy.

[Author]

Posts