reddit: We are nuclear fusion researchers, ask us anything

[ladajo]

Thanks for doing the math on the thread. I was trying to talk Joseph into it. I was going to show him a cite of it being done elsewhere, but I figured he would get more out of it if he did it himself.

Let's see what he ignores now.

[Joseph Chikva]

Thanks for doing the math on the thread. I was trying to talk Joseph into it. I was going to show him a cite of it being done elsewhere, but I figured he would get more out of it if he did it himself.

Let's see what he ignores now.

"Thanks for doing the math"
Such a math I did in middle school. But actually any thermal confinement scheme considers confinement of alphas in 5 T or weaker mag field.
For your reference #1: the term "ignition" means namely that confinement and then thermalization of plasma with the help of alpha's kinetic energy.
For your reference #2: projected on axis field for ITER 5.5 T


Now you talk about 2-3 T. But I have a quite good memory and remember that you told about 10 T superconducting magnet for Polywell.
While recently Kiteman told about 7 T.
And 2-3 T will confine some alphas too in case if dimesions of reactor has 1 m order or larger.

Good luck.

[ladajo]

Show me where I said polywell needed to be 10T.

[Joseph Chikva]

Show me where I said polywell needed to be 10T.

I remember this but could not find now using "search".

But I found the following:

Specifically, Dr. Nebel's team has calculated that if the polywell magnets are strong enough (10T?), the magnetic fields will even deflect the fusion-product alphas away from them; the alphas will bounce around inside the polywell until they finally find a cusp and escape, and the MaGrid will be spared the collision-induced spallation and heat load problems.

viewtopic.php?t=2666&highlight=10t+superconducting

Then Dan is saying:

Another thought...

In a WB 100 at 10 Tesla and 1.5 M radius. The cusps would stay the same area ( the B field is 100 times greater, but so is the Wiffleball surface area. So, the cusp loss area to total surface area would be 100X less , but the actual cusp area would remain constant in this situation (provided the drive voltage is not changed).
In terms of electron confinement transits, the electron containment would increase by ~ 100 transits without recirculation and > 1000 with recirculation. In terms of time this would be approaching 1 second with recirculation. I had seen that number mentioned and I now understand how it would be possible.

viewtopic.php?t=3108&highlight=superconducting

So, 10T superconducting was considered even by Dr. Nebel? Is that right?

[WizWom]

Thanks for doing the math on the thread. I was trying to talk Joseph into it. I was going to show him a cite of it being done elsewhere, but I figured he would get more out of it if he did it himself.

Let's see what he ignores now.

"Thanks for doing the math"
Such a math I did in middle school. But actually any thermal confinement scheme considers confinement of alphas in 5 T or weaker mag field.
For your reference #1: the term "ignition" means namely that confinement and then thermalization of plasma with the help of alpha's kinetic energy.
For your reference #2: projected on axis field for ITER 5.5 T


Now you talk about 2-3 T. But I have a quite good memory and remember that you told about 10 T superconducting magnet for Polywell.
While recently Kiteman told about 7 T.
And 2-3 T will confine some alphas too in case if dimesions of reactor has 1 m order or larger.

Magnetic fields do not and cannot cause a parallel track to modify; any alpha or electron going straight at the very wide faces sections of the device (in a coil loop), will merrily go on their way, unless the electrostatic charge is enough to modify their path from the \vec{v}\dot\vec{B}=0 condition. That is why this is not called Magnetic confinement: its ludicrous to think you could do so.

In a Tokomak, the magnetic fields are oriented perpendicular to escape, and the field will induce curving in the track of particles to move them back to the center. That's why they are called Magnetic confinement. Tokomaks could be fairly well served by charging their exterior wall to a high positive potential, which would both prevent arcing from the positive plasma, and would repel the plasma.

The timothy.d.power quote is, of course, not indicative of whiffle-ball trapping (where the particle undergoes a significant turn and reenters the core) but deflection - where the particle undergoes a slight turn in its path, and thus does not impact the magrid. Electron gyro radius being three orders of magnitude smaller than that of fusion product Alphas, we would certainly expect that electrons would be trapped while alphas are deflected.

And go ahead, consider 10T or even 20T fields... the gyro radius goes down to 1/2 or 1/4 those I calculated; Alphas would still have centimeter range gyro radii, whilst electrons would go down to tens of micrometers.

[Joseph Chikva]

Magnetic fields do not and cannot cause a parallel track to modify; any alpha or electron going straight at the very wide faces sections of the device (in a coil loop), will merrily go on their way, unless the electrostatic charge is enough to modify their path from the \vec{v}\dot\vec{B}=0 condition. That is why this is not called Magnetic confinement: its ludicrous to think you could do so.

In a Tokomak, the magnetic fields are oriented perpendicular to escape, and the field will induce curving in the track of particles to move them back to the center. That's why they are called Magnetic confinement. Tokomaks could be fairly well served by charging their exterior wall to a high positive potential, which would both prevent arcing from the positive plasma, and would repel the plasma.

The timothy.d.power quote is, of course, not indicative of whiffle-ball trapping (where the particle undergoes a significant turn and reenters the core) but deflection - where the particle undergoes a slight turn in its path, and thus does not impact the magrid. Electron gyro radius being three orders of magnitude smaller than that of fusion product Alphas, we would certainly expect that electrons would be trapped while alphas are deflected.

And go ahead, consider 10T or even 20T fields... the gyro radius goes down to 1/2 or 1/4 those I calculated; Alphas would still have centimeter range gyro radii, whilst electrons would go down to tens of micrometers.

"I calculated"
Polywell is not the first mirror machines invented by humans. And any other was intended to confine alphas for keeping the self-sufficient reaction. Gyro-radii less than geometric dimensions are required for this. Some alphas being in certain "losses cone" will escape reaction zone and some being outside of that cone will deflected back to the reactor thus heating plasma via multiple collisions.

[KitemanSA]

[/quote]

Increasing mag field 70-100 times in Polywell you will confine some products of reaction thus intensing thermalization.

OK, lets take this single point.
You electron injection is ~30keV. Lets play real easy and say we have a full coulomb net charge in the well, and we will pretend it is a point charge (worst case).

mean velocity of an electron (gaussian): 102e6 m/s
mean energy of an Alpha from D-D fusion: 23.8 MeV
mean velocity of an alpha: 33.9e6 m/s

Alpha from D-D? Very infrequent. Usually T or ³He. When talking alpha escape, the presumption is usually pB&J, not DD. The alphas from pB&J are much lower velocity, but still WAY too high for the well to confine.

assume the alpha is created in the exact center, and must escape from alcatr, er, the polywell.

a = F/m
F = k*Q1*Q2/x^2
where
k = 8.99e9 Nm2/C2
Q1 = 1 C
Q2 = 3.2e-19 C
m = 6.65e-27 kg

integrate from 0 to 1 m for total acceleration of 4.32e-32 m/s

conclusion: even mega-coulomb well depths will not cause significant deceleration of fusion products.

Now, right at the magnet, you are saying the magnetic force will "confine" the particle (even though magnetic force only prevents contact of the particle with the grid). So let's work out the cyclotron radius at the surface, assuming 5T

r = mv/qB
so r = 0.14 m

but for an electron, at 30keV r = 1.17e-4 m

So, the magnetic field will not prevent the alpha from impacting, its gyro radius is way too large.

With pB&J alphas and the best guess design for a 100MW reactor, the gyro-radius was more like 0.05m (by MSimon's calcs). This should be enough to reduce, if not prevent impact between the alpha and the MaGrid.

[KitemanSA]

While recently Kiteman told about 7 T.

I talk 7T because that is about where the top end of MRI magnets have gotten. They are about 1.5m in major diameter, so it seems that strength is doable for Polywell.

Until there is more info on the loss factors it would be foolish to finalize a design for a large Polywell at this time. But a fairly cheap unit COULD be made at ~7T. A 3T unit would be much cheaper still as there are a goodly number of that power MRI machines being replaced by 7T units and are available for a song, relatively speaking!

[KitemanSA]

And 2-3 T will confine some alphas too in case if dimesions of reactor has 1 m order or larger.

Yup, IIRC Dr. N. suggested the alphas may make a few passes before leaving the MaGrid. However, if the 5cm gr applies only at the maximum compression of the field near the MaGrid, the gr would be much larger near the point cusps and would leave MUCH more readily than the electrons. All this needs to be put into balance in the final designs.

[KitemanSA]

"I calculated"
Polywell is not the first mirror machines invented by humans. And any other was intended to confine alphas for keeping the self-sufficient reaction. Gyro-radii less than geometric dimensions are required for this. Some alphas being in certain "losses cone" will escape reaction zone and some being outside of that cone will deflected back to the reactor thus heating plasma via multiple collisions.

See? I knew you mis-understood the Polywell. Polywell is not a "mirror" machine, it is an IEC machine.

It has the happy happenstance of SOME mirror containment for the electrons. There is another happy happenstance in that the mag field also provide some protection for the MaGrid itself.

If you insist on thinking of this as a mirror machine, you will continuously get it wrong.

[Joseph Chikva]

See? I knew you mis-understood the Polywell. Polywell is not a "mirror" machine, it is an IEC machine.

You are only juggling with numbers and words. “This called so and because of it is so”. Does not matter if call penis as “finger”. Regardless name penis is penis. And not finger.
MaGrids are nothing more than magnetic mirrors intended to confine electrons. But they inevitably will confine alphas too in case if they would produce strong enough field. And 2-3T is strong enough.
Gyroradius is depends on momentum of particle. And yes, as electron has mass in about 7500 times lower than alpha and velocity about 10 times higher, thus electron's gyroradius at the same point of field (that is not uniform but curvilinear) will be about 750 times lower.
This only means that "losses cone" for alphas will be a little bit wider than for electrons.

[KitemanSA]

See? I knew you mis-understood the Polywell. Polywell is not a "mirror" machine, it is an IEC machine.

MaGrids are nothing more than magnetic mirrors intended to confine electrons.

Well, they were intended to protect the grid from the electrons while the static electric field confined them, but by happy happenstance, they did a pretty good job of improving the density ratio between the core and the rest of the chamber. This very much helps with limiting Paschen arcing. If arcing weren't an issue it would be still in question whether the wiffleball would be of interest at all.

But they inevitably will confine alphas too in case if they would produce strong enough field. And 2-3T is strong enough.
Gyroradius is depends on momentum of particle. And yes, as electron has mass in about 7500 times lower than alpha and velocity about 10 times higher, thus electron's gyroradius at the same point of field (that is not uniform but curvilinear) will be about 750 times lower.
This only means that "losses cone" for alphas will be a little bit wider than for electrons.

Yup, and that is one of the design points. Stronger field with more confinement of alphas or weaker field and the need to make a larger machine to compensate, vs., vs., vs.... We got to find out what all the "vs." are!

[ladajo]

Joseph likes to think in absolutes. Not in nuance.

[Joseph Chikva]

Well, they were intended to protect the grid from the electrons while the static electric field confined them, but by happy happenstance, they did a pretty good job of improving the density ratio between the core and the rest of the chamber. This very much helps with limiting Paschen arcing. If arcing weren't an issue it would be still in question whether the wiffleball would be of interest at all.

In any case MaGrids are the magnetic mirrors. Regardless to that how you call them.

[KitemanSA]

Well, they were intended to protect the grid from the electrons while the static electric field confined them, but by happy happenstance, they did a pretty good job of improving the density ratio between the core and the rest of the chamber. This very much helps with limiting Paschen arcing. If arcing weren't an issue it would be still in question whether the wiffleball would be of interest at all.

In any case MaGrids are the magnetic mirrors. Regardless to that how you call them.

Magnetic deflectors if it weren't for the desire for the wiffleball. Magnetic deflectors STILL for the ions?