reddit: We are nuclear fusion researchers, ask us anything

[mvanwink5]

Joseph,
WB-8 was built to test the scaling to see what does happen in real life, so your point is shared. We won't know though until even larger machines are built with higher fields (although higher fields may not be required for D-D, as I understand it).
Best regards

[Joseph Chikva]

Joseph,
WB-8 was built to test the scaling to see what does happen in real life, so your point is shared. We won't know though until even larger machines are built with higher fields (although higher fields may not be required for D-D, as I understand it).
Best regards

Thanks. And when first results will be?
Higher field machines in case if scaling law works are more attractive as provide more higher power density and, so, acceptable power at resonable size.

[ladajo]

Jospeh,
It would appear that EMC2 has already had some results from WB8, and thus the further funding and timeline adjustments we are seeing now.

I was waiting to see what you thought of the secondary reaction D-T power capture idea. Obvisouly, it needs to be adjusted a bit for cross-section and confinement times based on the well depth. But none-the-less, interesting to note that to date we had not really considered the magnitude nor relativeness of power out from further down the chain.

I also think that Joel Rogers goofed his math in his last paper. I will have to pull it out and look at it again. But based on our discussion, I think he did.

[Joseph Chikva]

I was waiting to see what you thought of the secondary reaction D-T power capture idea. Obvisouly, it needs to be adjusted a bit for cross-section and confinement times based on the well depth. But none-the-less, interesting to note that to date we had not really considered the magnitude nor relativeness of power out from further down the chain.

I also think that Joel Rogers goofed his math in his last paper. I will have to pull it out and look at it again. But based on our discussion, I think he did.

In conventional fusion reactors running on DD fuel and Tritium nuclei created in situ as result of D+D=T+p reaction should react with D due to much higher cross-section of DT reaction.
It is well known.
But for this we should have enough confinement for reaction products. And I mentioned this possibility.
But for Polywell intended to confine ions with the help of potential well of virtual cathode and the desired depth of which is much lower than their kinetic energy (tens of keV vs. several MeV) potential well will not confine tritium. So, after think on it, I see that in Polywell this possibility is lower.
But as I know magnetic field of MaGrids should reach 7-10 T. So, depending on direction (commonly reaction products have random direction) some of them should be confined by mag field. And probability of DT reaction will not be near zero (but lower than for other concepts).
That time was very early morning, I was going to go running and thought that it is easy to calculate. And I referred to laziness. But now I see that it is difficult or impossible to calculate DT fusion rate.
In any time each fusion event with confined reaction charged products will take participation in thermalization. And MaGrids inevitably will confine some of them.

[MSimon]

If all goes well the next device will be WB-D (1,000 in Roman Numerals if that is significant.) 100 MW fusion prototype. Not a system.

microwatts
watts
megawatts

The scaling between devices seems about right to me.

[Joseph Chikva]

If all goes well the next device will be WB-D (1,000 in Roman Numerals if that is significant.) 100 MW fusion prototype. Not a system.

microwatts
watts
megawatts

The scaling between devices seems about right to me.

pico, nano, micro, milli, deci, deca, kilo, mega, giga, tera
This is scaling? Every developer says the same if all will ok, I will give you happiness. Fusion researchers of different approaches say this during 60 years.
I used 2 Tesla and further will use 6 and power will increase in 27 times. But power doesn't want to increase.
I have a simple question: is this difficult to increase mag field in existing device e.g. WB6. As I understand there is used rather weak 0.8 T magnets. Is that so difficult to increase that till for example 2 T by simple increasing of ampere-turns?

[mvanwink5]

Joseph,
Reason for choosing .8T was that it was all that was required to demonstrate scaling holds and sufficient to justify the next level WB-D if successful. At least that is how I understand it. In addition, they did an intermediate project labeled WB-7.1 between WB-7 and WB-8 that was more fully instrumented than WB-7.
Best regards

[KitemanSA]

I used 2 Tesla and further will use 6 and power will increase in 27 times. But power doesn't want to increase.
I have a simple question: is this difficult to increase mag field in existing device e.g. WB6. As I understand there is used rather weak 0.8 T magnets. Is that so difficult to increase that till for example 2 T by simple increasing of ampere-turns?

If you go from 2T to 6T the scaling would suggest that the power would go up 81 times (B^4), all else being equal.

And no, WB6 had ~0.1T fields. That is one reason they went to WB8 at 0.8T fields. This allows investigation of a goodly degree of scalability.
From field alone, the 8X increase in field should allow power increases up to ~4000 times WB6. But the size scaling will also be involved.

Oh, and just increasing the amp-turns will either greatly increase the minor diameter vs the major diamer of the MaGrid making it an ungaily machine, or it would make the resistivity of the coil much greater... unless you ALSO go to cryo-cooled or even superconducting magnets.

From the published graphics, it seem HIGHLY likely that WB8 is in fact a cryo-cooled machine with between about half and the same number of turns but much lower resistivity allowing greater amperage.

[Joseph Chikva]

If you go from 2T to 6T the scaling would suggest that the power would go up 81 times (B^4), all else being equal.

And no, WB6 had ~0.1T fields. That is one reason they went to WB8 at 0.8T fields. This allows investigation of a goodly degree of scalability.
From field alone, the 8X increase in field should allow power increases up to ~4000 times WB6. But the size scaling will also be involved.

Oh, pardon, really 3^4=81 And what?
If WB6 provided fusion rate 2.5E9 events/sec, so, they could strengthen the magnets from 0.1 T to 2 T (20 times higher field).
It is very easy if to take into account small confinement time (milliseconds order). As a rule the problems with higher fields begin from 2 T. For this only additional ampere-turns are required.
And then to show commparison fusion rate 2.5E9 for 0.1 T vs. 4E14 events/sec for 2 T (160000 times higher). As I understand they did not do that.
4E14 events/sec corresponds to fusion power aprox 240W.
What do you think, why this was not done?
Difficult to make stronger magnets?
Nope.

[Joseph Chikva]

Oh, and just increasing the amp-turns will either greatly increase the minor diameter vs the major diamer of the MaGrid making it an ungaily machine, or it would make the resistivity of the coil much greater... unless you ALSO go to cryo-cooled or even superconducting magnets.

My friend, I saw this just now.
0.1 T is so laughably weak field, that technically there is not any problem to strengthen that. You can use copper-silver alloy wire, you can use hollow conductors with central flow of coolant, to immerse coil into an insulator resin e.g. epoxy or polyether and to increase current by increasing of applied voltage. Till about 2 T it is quite easy.
Also recall that required time for holding of that field’s value does not exceed several milliseconds. So, cooling challenge would not be so critical.

[KitemanSA]

[/quote]

Oh, and just increasing the amp-turns will either greatly increase the minor diameter vs the major diamer of the MaGrid making it an ungaily machine, or it would make the resistivity of the coil much greater... unless you ALSO go to cryo-cooled or even superconducting magnets.

My friend, I saw this just now.
0.1 T is so laughably weak field, that technically there is not any problem to strengthen that.

Not bad for a hand-wound toroidal form magnet set and research on it for under a $M and <~8 months.

You can use copper-silver alloy wire, you can use hollow conductors with central flow of coolant, to immerse coil into an insulator resin e.g. epoxy or polyether and to increase current by increasing of applied voltage. Till about 2 T it is quite easy.
Also recall that required time for holding of that field’s value does not exceed several milliseconds. So, cooling challenge would not be so critical.

Hence WB8 which seems to do everything you want to do, except go for 2T field. But they went for 8X which allows for a lot of study without huge expense.
Dr. B. suggested that it was almost useless to go to another intermediate scale above WB6, that the effort should just jump to a 100MW core. Seems the "expert panel" that reviewed WB7 results thought there was benefit to going to at least one more intermediate scale experiment, WB8.
There is also talk of what might be an even larger intermediate scale (WB9) that, depending on how you read it, would NET approximately 100mW (breakeven) though others think that value (100mW) is a typo and was intended to read 100MW (i.e., full scale).

[Joseph Chikva]

Not bad for a hand-wound toroidal form magnet set and research on it for under a $M and <~8 months.

I am not saying bad or good. But only saying that it is easy to make much stronger. "hand wound" - all can be made by hands.

[KitemanSA]

Stronger = more cost. Why stronger if strong enough?

[Joseph Chikva]

Stronger = more cost. Why stronger if strong enough?

Yes, more cost on several hundreds dollars. Even schoolboys build similar in father's garage.
"Strong enough" for allowing you to speak about scaling "B^4"
When you'll make really stronger you will see that this scaling does not work for any device.

[KitemanSA]

Stronger = more cost. Why stronger if strong enough?

Yes, more cost on several hundreds dollars. Even schoolboys build similar in father's garage.
"Strong enough" for allowing you to speak about scaling "B^4"
When you'll make really stronger you will see that this scaling does not work for any device.

They made WB8 stronger. They should tell B^4 from that, at least until a full scale unit is built. They also made it ~2X bigger. They should be able to tell r^3 scaling from that, at least until a full scale unit is built.

You want more? Build it. As you say, a couple hundred dollars. So build one.