Polywell: We'll know in 7 months time?!

[chrismb]

It makes perfect sense [either way] and I don't see what the issue is.

If it meant 100MW then clearly the project will be base work for a future design concept, as there is no evidence yet of even a viable reaction.

If it meant 100mW then this might well be the achievable goal and I agree that this'd knock me off my lill'ol'stool and I'd go start baking some humble pie. 100mW = 5e11 reactions/s [for DD], or 2e11 reactions/s for p11B.

Seeing as noone has yet seen ONE reaction/s in a controlled p11B process, then jumping to 2e11 would be a staggering success that would justify a big follow-up spend.

[KitemanSA]

That is not established. A plant that produced 100mW of continuous NET power would be a great step forward.
...
How long could one of these WB research machines run before melting, if it generated 100MW net, i.e. above and beyond drive power? They aren't making any arrangement to extract the energy, so all the power in and power generated just heats the reactor, doesn't it? Maybe it really is milliwatts.

There are several problems with this interpretation:

1) They are calling it a "reactor" design, which connotes useful power. They make repeated reference to studying the suitabililty of Polywell as a reactor. It's very clear from context that by "reactor" they don't mean something that produces AA battery levels of power.

This is your interpretation. It could be that they just define a "reactor" as net power.

2) 100mW net power would be a very strange and arbitrarily precise requirement. It's like saying you want a vehicle to go 50.000001 mph. In a net power machine, 100mW isn't even a rounding error; you would have a hard time even measuring net output that small. One would expect the requirement to be simply "Q>1" or "net power" in such a case.

But then again they could be being dodgy and making use of all that past publicity to slip one by their sponsors too. After all, Dr. B. also advocated the development of two small scale units using non-circular magnets and they don't SEEM to be doing THAT either.

3) If they don't mean net power, 100mW makes even less sense. That's closer to what WB-6/7 produced than what we expect from WB-8. It makes no sense to ask for a design and feasibility study to produce WB-6 power levels after WB-8.

See above.

4) The plan all along, going back to Bussard's Tech Talk, was to build a 100MW Polywell reactor. It's been on EMCs web page from the beginning. It strains credulity to think we're on the path that was always envisioned, except for some reason we're delivering a design that produces 9 order of magnitude less power. After ~4 years of work.

IBID

It seems vastly more likely it is simply a typo. The technical writing in the award is a bit sketchy anyway.

I do hope you are right. I really would like you to be right. I have argued in the past that that "had to be" the correct interpretation. I'm just not so positive now. If it were Dr. B. running things, I would be totally convinced. But Dr. N. is much more methodical it seems, and I am just not positive anymore. We shall see. Either way, I'll be happy.

[Helius]

100mW would "prove" the WB. They'd have to do that before Jumping to a 100MB machine. They're still at the "proving wiffle ball" stage, and wouldn't go for a "Hail Mary" 100MW; At least I hope not. Do you really think they'd scale up between WB 8.0. and 8.1 to a Multi-meter sized machine?

100mW machine output, even without respect to Electrical input, would be VERY interesting in it's own right, IMHO. I don't think for a minute "100mW" is a typo.

[Robthebob]

hmmmm interesting, if the scaling laws work like they should. Building a core that's (2m)^3 is way way way more than enough for net power.

Making a mW machine and a MW machine may not even differ that much.

I'm thinking it's a typo.

[TallDave]

100mW would "prove" the WB. They'd have to do that before Jumping to a 100MB machine.

Bussard and Nebel both agreed we could have gone from WB-7 to trying a 100MW machine, so we can certainly get there from WB-8.1.

This is your interpretation. It could be that they just define a "reactor" as net power.

It's not just "my interpretation" it's the interpretation that makes more sense in context. Why spend millions to study the feasibility of the technology to produce AA battery levels of power? Generally, a feasibility study has a useful end product in mind. Also, if you go back and review Rick's use of the word "reactor" here on T-P, he is pretty much always referring to something producing useful levels of power.

But then again they could be being dodgy and making use of all that past publicity to slip one by their sponsors too.

I find it much more likely that the person typing up the award made a typographical error than the ONR review board (which rumor has it included Krall and other respected scientists) missed the difference between 9 orders of magnitude in output power.

It makes perfect sense [either way] and I don't see what the issue is.

100mW makes almost no sense. To believe this, you have to believe all of the unlikely things above PLUS that after years of talking about a 100MW reactor, they just happen to choose the exact 9 orders of magnitude difference that turns MW into mW (even though that arbitrary level of precision makes no sense in that context). It would be a huge coincidence sitting on top of a lot of unlikeliness.

I won't say 100mW is impossible but it's much, much more likely to be a typo.

[MSimon]

I don't think you can build a 100 MW reactor including shielding for less than $50 million first article. Unless you were doing 1 usec pulses.

Then you might be able to do it with the money the Navy has put up.

My guess is that it means 100 mW fusion. Which would be a nice step up if it was continuous.

=====

Alpha sputtering is a very serious problem. The first thing is to reduce their energy. Then line the collectors with stuff that won't poison the reaction. B11 might work in a pBj machine.

[TallDave]

Simon,

Sure, I agree you can't do it for less than $50M. But the contract only asks for a design for a 100MW reactor.

The report shall address the conceptual
requirements for a polywell fusion reactor capable of generating approximately 100mW. (A0001)

If WB-8 does well, then maybe they spend the big bucks for a reactor.

Let's take a brief look at the numbers from the perspective of Q>1 with 100mW net power. One would want to have at least a 1% greater output than input to be able to claim Q>1 with some statistical significance, but let's be really generous and give them an extra order of magnitude, so .1% gives us 100mw * 1000 = 100W inputs and output of 100.1W. Does anyone really think Polywells can break even at the power of a common household light bulb?

The nice thing about alphas are they're charged, so you can adjust their path. I can't wait to see the detailed reactor designs.

[Barry Kirk]

Been lurking on this site for several months now...

Keep up the great discussion... Sorry I don't have more to contribute.

[MSimon]

Simon,

Sure, I agree you can't do it for less than $50M. But the contract only asks for a design for a 100MW reactor.

The report shall address the conceptual
requirements for a polywell fusion reactor capable of generating approximately 100mW. (A0001)

If WB-8 does well, then maybe they spend the big bucks for a reactor.

Let's take a brief look at the numbers from the perspective of Q>1 with 100mW net power. One would want to have at least a 1% greater output than input to be able to claim Q>1 with some statistical significance, but let's be really generous and give them an extra order of magnitude, so .1% gives us 100mw * 1000 = 100W inputs and output of 100.1W. Does anyone really think Polywells can break even at the power of a common household light bulb?

The nice thing about alphas are they're charged, so you can adjust their path. I can't wait to see the detailed reactor designs.

Of course if Q>1 is the immediate goal. But if all they are doing is checking the scaling laws then going from 100 uW to 100 mW is probably good enough.

[chrismb]

EMC2 engineer to EMC2's lawyer; "err... we were gonna do a design for a 100MW reactor, but the Navy's put in 'mW' by mistake"

EMC2's lawyer to EMC2; "well, I don't know if that's good or bad, and I guess the Navy lawyer doesn't know the difference either. Is it easier to do 100mW or 100 MW?"

EMC2 engineer; "100mW is a tiny amount of power! It's 9 orders of magnitude on the easy side of 100MW!"

EMC2 lawyer; "no idea what 'orders of magnitude' means but it sounds much easier, so if that's the case my legal advice is to leave it well alone and sign off the acceptance of the offer letter!!.....why would you want to make the work more difficult for yourselves?! You engineers! You're freaks! Stop doing things for the benefit of mankind. Get greedy - No-one understands you whilst you're busy trying to make a success of things for the greater good!!"

[TallDave]

chrismb,

Heh. I always enjoy your visualizations.

Simon,

Way back before Talk-Polywell, Art and I argued over at the Polywell wiki about the PMT measurements and whether they meant anything. I did a rough calc to see if they could have expected to detect a flash from fusion, and came up with something around a milliwatt for Bussard's claimed fusion rate. (I think Kite recently did one too that was a little more accurate and we ended up agreeing on something like 1e9 fusions equalling .00282 watts.)

Anyways, if WB-8 is the same size at .8T versus .1T for WB-6, then with B^4 scaling we're looking at 8^4 = 4096 * .00282W = 8.192W -- and it's probably bigger. If it's twice as big (seems reasonable?) then we also have r^3 power scaling, which would give us 2^3 = 8 * 8.192W = 65.54W. So we should be way past milliwatts.

[KitemanSA]

Dave,

My calc was more like 0.6milliwatts, though I did use 2.5 neutrons (average of last two complete runs) and 0.4milliseconds rather than the more commonly quoted 0.25msec in my calc.

The statement of work for WB8 calls it a small scale unit, so there is no reason to believe that it is any bigger than WB7. The 8xfield can be easily achieved with copper tubing for the conductor carrying LHe as a coolant. At those temperatures, the copper can carry at least 8x the current.

Again, I'd be pleased as punch to be proven wrong!

By the way, if you are still arguing 100mW vs 100MW, the WB8 is nowhere near NET anything, so just making breakeven is a great step.

Please, oh please may I be wrong!

I must admit that your way makes more sense to me, but that proves nothing!

[TallDave]

Yes, I think I ended up liking your calc better.

I guess more T doesn't necessarily mean bigger, look what they drove that little teacup-sized guy at down in Sydney. 2*R is firmly in WAG territory. Fortunately 8W or 64W, it doesn't really affect my point.

WB-8 needs to show some beam-beam power scaling and loss scaling at least vaguely resembling Bussard's claims. Then hopefully WB-8.1 will show we can burn p-B11. And then, maybe we try to build a world-changing reactor.

I think the gradualist mindset is a result of the slow progress of all those big toks that never quite ignited. High-beta devices may jump right from 64W total power to 100MW net power.

[MSimon]

Anyways, if WB-8 is the same size at .8T versus .1T for WB-6, then with B^4 scaling we're looking at 8^4 = 4096 * .00282W = 8.192W -- and it's probably bigger. If it's twice as big (seems reasonable?) then we also have r^3 power scaling, which would give us 2^3 = 8 * 8.192W = 65.54W. So we should be way past milliwatts.

I did some BOE a while back (unpublished) that came up with similar numbers.

100 MW = 33.3 Amps of current at 3 MeV average per alpha, 9 MeV/fusion.

So 100 MW = 11.1 Amps * 6.24 E18 electrons/coulomb = 6.9 E19 fusions per MW*Second. 100 MJ = 6.9 E19 pB11 fusions. or 1E8 J = 6.9 E19 fusions.

1 J = 6.9 E11 fusions

.002 J = 1.4 E 9 fusions.

Which gives you ball park confirmation.

But think in lawyer terms. You expect 10 Watts. But things may not go as planned. If we want to keep the project running and it doest quite meet expectations why not set the improvement at a factor of 50 so serious progress is shown and yet not be too hard to meet.

Murphy is not your friend.

*** Corrected numerical error.

[MSimon]

I think the gradualist mindset is a result of the slow progress of all those big toks that never quite ignited. High-beta devices may jump right from 64W total power to 100MW net power.

Um. No. It is a result of too many contacts with Murphy. Go back to NASA Spaceflight. And look at the discussions Tom Ligon and I had about the engineering progression.

I think 1 mW - 10 W - 100 MW is about as extreme a I'd like to go.

The trouble with a progression like that is that the devices are so mechanically different that you can't build much off your previous devices. You go from a device where cooling is not a consideration to one where cooling is everything.