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

Point out news stories, on the net or in mainstream media, related to polywell fusion.

Moderators: tonybarry, MSimon

ladajo
Posts: 6258
Joined: Thu Sep 17, 2009 11:18 pm
Location: North East Coast

Post by ladajo »

So we can dissect what this sentance really means:

"3.1.2 The design shall use circular coils around each main face cusp axis."

I think he is holding to the cube.
Also, we never did see what the "improved" coil connector stub tubes looked like for WB7.1. They may also have made a futrther imrovement with that, or eliminated them all together with wall mounted coils or something of the sort.

I have been toying with an idea to go with small bore copper tubing wound as a coil on each face, wall mounted and high pressure, high flow internally cooled for a continuous operation setup.

mvanwink5
Posts: 2143
Joined: Wed Jul 01, 2009 5:07 am
Location: N.C. Mountains

Post by mvanwink5 »

Just some thoughts and musing:
Scaling will illuminate different issues at different stages. Any thoughts as to how that would play out? (The earlier problems are discovered the lower the cost in solving them.) Scaling is dependent on science and engineering details, so at what point will engineering issues dominate? What is size is needed for illuminating the science? What will be the maximum reactor size and what will be the best reactor size? Will the tough issues surface before 1 MW? if not why not jump some steps?

Scaling stages matrix:
coil diameter
field strength
coil current (copper vs SC)
number of support struts needed per coil
strut cooling
coil cooling
methods of power generation (required for cooling and continuous operation)
conductor diameter as a percent of coil diameter
shielding requirements for SC, instruments, personnel
reactor voltage
cost



continuous operation - explore thermalization, fusion ash and contaminate exhausting, cooling method, coil degradation, fault current mechanical stress on coil supports
Counting the days to commercial fusion. It is not that long now.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

ladajo wrote: I have been toying with an idea to go with small bore copper tubing wound as a coil on each face, wall mounted and high pressure, high flow internally cooled for a continuous operation setup.
Many people have been toying with that concept and a number of cross-sections have been displayed. Indeed, it seems the best way to increase the field to .8T without going SC. There are other ways, but this seems simplest.

viewtopic.php?p=17665&highlight=#17665

Here is one such post.
Last edited by KitemanSA on Mon Feb 15, 2010 5:14 pm, edited 1 time in total.

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

Well, if you agree it is so then I have proved my point - it depends on the person doing the rationale for the meaning of *expensive*
No, actually, I've refuted your point with an objectively more relevant cost metric. Just as Rick says, these machines are far less expensive than toks at similar points on the same exponential neutron production scale.

I mean, really. This is like saying "WB-7 produced .002W of fusion for $2M, so using the same cost per neutron a 100MW reactor must cost 100,000,000/.002 * $2M = $100 quadrillion dollars!"

Economies of scale aren't terribly esoteric. This is something we see in finance and cost accounting all the time. You never want to make plans built around the assumption cost per thousand is the same as cost per million.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

TallDave wrote:
Well, if you agree it is so then I have proved my point - it depends on the person doing the rationale for the meaning of *expensive*
No, actually, I've refuted your point with an objectively more relevant cost metric.
Ehhhh!!!???? How on earth have you refuted my point that the calculation depends on the accounting decision of the person making the calculation when you go off and make your own accounting decision to perform the calculation!!


uughhh!!???

...and, yes, it really is like saying "WB-7 produced .002W of fusion for $2M, so using the same cost per neutron a 100MW reactor must cost 100,000,000/.002 * $2M = $100 quadrillion dollars!" Now you can see where WB sits in the scheme of things.

Show me this [supposedly] objective 'exponential' way of calculating things. If you don't like neuts/$ then what is your metric for tokamak versus Polywell, then?

Pick whatever metric you want and let's see the numbers...

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

Because cost accounting isn't done however someone feels like doing it. You DO NOT confound an exponential relationship with a linear one.
..and, yes, it really is like saying "WB-7 produced .002W of fusion for $2M, so using the same cost per neutron a 100MW reactor must cost 100,000,000/.002 * $2M = $100 quadrillion dollars!"


Yes, it is. And see what a ridiculous conclusion one reaches using that argument?
Pick whatever metric you want and let's see the numbers...
I offered some above.

Of course there aren't that many empirical data points for Polywell to compare to. What neutron count can you get from a $2M tokamak? For $10M?
Last edited by TallDave on Mon Feb 15, 2010 5:51 pm, edited 1 time in total.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

TallDave wrote: Yes, it is. And see what a ridiculous conclusion one reaches using that argument?
Why d'you think tok proponents think IEC is ridiculous!?
TallDave wrote:
Pick whatever metric you want and let's see the numbers...
I offered some above.
Where? I missed it. What value do you get for Polywell? It must be in some funny units of your own construction, like "neutrons/log($)" or something.

Is that your unit of comparison, the neut/ln$? I can't see it. You say you've done it, I can't see the number. Can't see it. Just say what it is. What is it? Why do you blither on without putting in some numbers when all you have to do is put some numbers in?

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

Why d'you think tok proponents think IEC is ridiculous!?
Because of losses, not because it would cost quadrillions of dollars. Unless you're implying they're too stupid to understand the scaling isn't linear to dollars.

Again, in case you missed the edit: What neutron count can you get from a $2M tokamak? For $10M? (and that's cost to build, no buying used toks!)

Better to choose similar points on the dollars-to-neutrons curve than to try to define the curve itself.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

MSimon
Posts: 14334
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

Just talking of the device.

$.5 million buys you 2 mW of fusion

$3 million buys you 100 KW

I think the function is: log (n)/$ = K, or something like that. Double the dollars and you get 100X as many neutrons. Or some such.
Engineering is the art of making what you want from what you can get at a profit.

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

Or. here's another way to ask the question: take a given tokamak, say JET. Now ask: for the equivalent inflation-adjusted dollars, how well (or badly) would a Polywell have to perform to produce the same fusion power for those dollars?

If you look at it this way, I'm guessing Polywell can underperform by about an order of magnitude from Bussard's expectations and still do better than toks. I will leave the math for later.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

You're still just not coming up with the numbers, are you!!

I mean - it's just that I have to do all the work for you..after a long day at work!

So you're trying to say that it's like $/[log(power)+k] relationship and we've got two data points saying 5e5/[log(0.002)+k]=3e6/[log(100000)+k]?

Is that your relationship? Can someone solve for k for Tall Dave, please...

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

So, err... back from tea and no-one's seems to know how TallDave is gonna "fix" his calculation?

TallDave, I've provided a calc and you've refuted it. You could've pointed out a numerical error, or you could have pointed out a better logical option. Neither have been forthcoming.

OK, so the only way to look at this is to say [and I estimate, from mental calc - improve accuracy as you wish]:

"you get 100uW for $0 (it's exponential, so you can't have a 'zero' valued quantity) and for each $1M you buy yourself another 3 orders of magnitude."

So what you were after was "Polywell = [ 3oom +100uW]/$1M"

Now let's see for JET with 16MW for $300M; "JET = [0.04oom+100uW]/$1M

Gee, will ya look at that - JET looks to be 100 times worse than Polywell's speculated performance, and I've done nothing more that some really straightforward everyday maths!!

....Personally, I'd tend to go for the 100uW version of Polywell, which should cost nothing at all, by TallDave's approach to the calculation.

clonan
Posts: 91
Joined: Tue Jun 10, 2008 4:16 pm

Post by clonan »

chrismb wrote: Now let's see for JET with 16MW for $300M; "JET = [0.04oom+100uW]/$1M
What is the 100uW cost for JET?

Please excuse the freshman level math but it seems that since JET and polywell are very different devices it is reasonable to assume the curves will be different. Since Polywell is early in development we are essentially guessing but JET already created 16 MW of neutrons, therefore we must know how much a machine that produces 100 uW of neutrons would cost.

If a 100 uW JET machine would cost $200M then wouldn't the curves really still favor polywell by a lot.

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

clonan wrote: What is the 100uW cost for JET?
Zero. It is a comparative figure so it must also cost nothing to build a 100uW tokamak, and if it isn't then it goes back to my point that such a calculation is in the eye [calculator] of the beholder, a point on which Tall Dave suggests I am wrong to imply.

I am presuming that he thinks there is some God-given objective exponential measurement system that I am missing, but he's a bit thin on his own details. God would give us the linear relationship to work with because He knows full well you get nuthin' for nuthin'. [He's omnicient, so He'd have to know that, I guess. "I help those who help themselves. I won't give you nuthin' for nuthin'". That's in the Bible somewhere, I think. It's the 0th commandment, i'n'it?]

There is no way of operating an exponential scale objectively, as TallDave seems to suggest, where there is a (0,0) point in it.

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

You're still just not coming up with the numbers, are you!!
I gave you two numbers. You haven't told me how many neutrons toks can produce for $2M and $8M. If the number is zero (i.e. can't be done) then Polywell wins at those dollar values.

Maybe there is some dollar level where toks outperform Polywells for the same dollars, but if so we won't know where until we spend that much on Polywell. In the meantime, it is quite silly to compare JET to WB-7, which obvious fact you seem intent on eluding.
You could've pointed out a numerical error, or you could have pointed out a better logical option. Neither have been forthcoming
I've given you a better alternative. I don't think you can "fix the equation" the way you're talking about. You can't write an equation for the Polywell curve without making an assumption about how it scales.

What might be more interesting is asking how Polywells would have to perform to match JET or ITER, and which can be calculated.
Last edited by TallDave on Mon Feb 15, 2010 9:41 pm, edited 1 time in total.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

Post Reply