Translating FRCs

Discuss how polywell fusion works; share theoretical questions and answers.

Moderators: tonybarry, MSimon

jmc
Posts: 427
Joined: Fri Aug 31, 2007 9:16 am
Location: Ireland

Translating FRCs

Postby jmc » Mon Sep 21, 2009 10:55 am

I read Art Carlson's post on FRCs in the news section with interest and I heard some talk of discussing the theory of FRCs in the theory section of our forum.

I've been sitting on this idea for ages and I've decided to post it to see what peoples thoughts on this are.

Some of you are familiar with Magnetized Target Fusion MTF researched in LANL and pursued by General Fusion. The idea is you have three phases

1) Formation
2) Translation
3) Compression

First you form a plasma with a temperature of 300eV and a density of 10^17 /cm^3 at 5 Tesla or so, then you translate this FRC configuration, or spheromak into a linar and compress it until its field reaches ~1000 Tesla, with densities of 10^19/cm^3 and temperatures of 10's keV, this plasma will then do a fusion burn in a microsecond or two and you repeat the process.

The catch is the compression stage, in the case of the linar in LANL it gets destroyed while the General Fusion design requires 100 mechanical pistons to impact a rotating liquid lithium metal vortex within 1 microsecond of each other.



This is my idea...

Why no focus on translating the plasmoid to ever higher kinetic energies through repeated magnetic squeezing at the rear and then let the FRCs own kinetic energy accomplish the compression stage.

I.e you'd start with a target plasma of density 10^17, field 5 Tesla temperature 300eV, you would then accelerate the FRC to a speed where each particle had a kinetic energies of 50 keV or so it would still be confined by the field of 5 Tesla because the random energy would be 300eV.

You then fire this FRC into a solid or liquid metal cone, the FRCs forward momentum causes it to compress itself and fuse. That way you don't need to worry about timing your pistons or exploding your linar.

A potential simplification in the design??

Art Carlson
Posts: 794
Joined: Tue Jun 24, 2008 7:56 am
Location: Munich, Germany

Re: Translating FRCs

Postby Art Carlson » Mon Sep 21, 2009 12:20 pm

jmc wrote:I read Art Carlson's post on FRCs in the news section with interest and I heard some talk of discussing the theory of FRCs in the theory section of our forum.

I've been sitting on this idea for ages and I've decided to post it to see what peoples thoughts on this are.

...

A potential simplification in the design??

The short answer is, I don't know. Some of the elements of your idea, including acceleration of FRCs to fusion-relevant kinetic energies and conversion of some of that energy into compression, have been discussed by John Slough and others for some time. Using the plasma motion rather than liner motion to create transient magnetic fields higher than obtainable with static (or pulsed) magnets - I don't know if that has been explored, although I expect it has. I don't see any fundamental problem, but I'm not familiar enough with the experimental and theoretical details to say much with confidence.

Solo
Posts: 261
Joined: Fri Jun 29, 2007 12:12 pm
Location: Wisconsin

Postby Solo » Wed Sep 23, 2009 1:31 pm

So you're saying, used the pulsed magnets to accelerate the FRC like a coilgun, then slam it into a tapered flux conserver and let that force the compression. Hmm. How fast is 50 keV for hydrogen? It might be hard to synchronize the accelerating coils to get something going that fast, that's probably a decent percent of c. Eddie currents might be a problem at that kind of speed. But you'd get away using a flux conserver to provide the compressive force, so no big multi-tesla magnets (pulsed or SC) around the reaction volume.

Art Carlson
Posts: 794
Joined: Tue Jun 24, 2008 7:56 am
Location: Munich, Germany

Postby Art Carlson » Wed Sep 23, 2009 2:26 pm

Solo wrote:How fast is 50 keV for hydrogen? It might be hard to synchronize the accelerating coils to get something going that fast, that's probably a decent percent of c. Eddie currents might be a problem at that kind of speed.

m_e c^2 = 500 keV
m_p / m_e = 1836

0.5*(2.5*m_p)*v^2 = 50 keV
=>
v/c = sqrt( (50 keV) / (1.25*1836*m_e*c^2) )
= sqrt( 50 / (1.25*1836*500) )
= 1 / 150
v = 3e10 cm/s / 150 = 2e8 cm/s

John is thinking something like baseball size at the end of the compression, so we will need ~ 1 cm precision on the coil timing.

(1 cm) / (2e8 cm/s) = 5 ns

Sounds like a tall order, but maybe not impractical. Maybe you could design some sort of circuit that results in a magnetic field pulse traveling down the tube at a predetermined (and accelerating) velocity. That sounds like a job for MSimon.

Eddie currents are what you are using to produce the field that compresses your FRC. I suspect on those times scales the resistivity of your coils and flux conservers is not going to produce significant drag.

MSimon
Posts: 14331
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Postby MSimon » Wed Sep 23, 2009 3:57 pm

I think I could do 5 nS with current technology - with feedback.

The trick is to make the device repeatable to 5 nS on 1 second time scales. Not too tough even for drive power on the order of 100 KW. I might be able to get repeatability to .1 nS with a lot of effort.

If I can get some kind of phase/location signal it would be a snap.
Engineering is the art of making what you want from what you can get at a profit.

Art Carlson
Posts: 794
Joined: Tue Jun 24, 2008 7:56 am
Location: Munich, Germany

Postby Art Carlson » Wed Sep 23, 2009 7:59 pm

MSimon wrote:I think I could do 5 nS with current technology - with feedback.

The trick is to make the device repeatable to 5 nS on 1 second time scales. Not too tough even for drive power on the order of 100 KW. I might be able to get repeatability to .1 nS with a lot of effort.

If I can get some kind of phase/location signal it would be a snap.

Sure, I'll give you a location signal. The FRC distorts (increases) the magnetic field between itself and the wall/coil, so you just need a field sensor that triggers a coil on the trailing edge as the FRC comes by.

What has me worried at the moment is the voltages involved. If I want flux changes of 0.1 Wb (e.g. 100 cm^2 and 10 T) on a timescale of 10 ns, then I need a loop voltage of (0.1 Wb)/(10 ns) = 10 MV. Ouch!

On the other hand, how hard do we really need to push things? After all, an FRC is actually a pretty good vacuum. And how fast do they shoot out if you form them in a conical theta-pinch? It can't be terribly difficult, but I don't yet have a grip on the physics of translating FRCs.

MSimon
Posts: 14331
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Postby MSimon » Wed Sep 23, 2009 8:04 pm

Art Carlson wrote:
MSimon wrote:I think I could do 5 nS with current technology - with feedback.

The trick is to make the device repeatable to 5 nS on 1 second time scales. Not too tough even for drive power on the order of 100 KW. I might be able to get repeatability to .1 nS with a lot of effort.

If I can get some kind of phase/location signal it would be a snap.

Sure, I'll give you a location signal. The FRC distorts (increases) the magnetic field between itself and the wall/coil, so you just need a field sensor that triggers a coil on the trailing edge as the FRC comes by.

What has me worried at the moment is the voltages involved. If I want flux changes of 0.1 Wb (e.g. 100 cm^2 and 10 T) on a timescale of 10 ns, then I need a loop voltage of (0.1 Wb)/(10 ns) = 10 MV. Ouch!

On the other hand, how hard do we really need to push things? After all, an FRC is actually a pretty good vacuum. And how fast do they shoot out if you form them in a conical theta-pinch? It can't be terribly difficult, but I don't yet have a grip on the physics of translating FRCs.


Even the 10 MV rqmt is not going to hurt much if the power is reasonable. Use a transformer.

Of course if the rqmt is 10 MV pulses at multi MW I'm going to have to do some research.
Engineering is the art of making what you want from what you can get at a profit.

jmc
Posts: 427
Joined: Fri Aug 31, 2007 9:16 am
Location: Ireland

Postby jmc » Thu Sep 24, 2009 8:06 am

I was thinking of having a corridor of coils that hold the FRC in which flux doesn't change along and then two opposing coils where the voltage changes, to reduce the inductance.

I've also considered the possibility of slamming a heavy FRC with say, a high density lead plasma at a temperature of 10 or 20eV and a kinetic energy per particle of 50-10keV into a stationary, less dense higher temperature (300eV) lighter DT FRC positioned at the base of a cone to squash the flux. Even if 1/4 of the kinetic energy of the heavy FRC was converted into compressing and heating the lighter fusion fuelled FRC it may well be enough to get net gain fusion and it would also relax the requirements for timing the acceleration.

Solo
Posts: 261
Joined: Fri Jun 29, 2007 12:12 pm
Location: Wisconsin

Postby Solo » Thu Sep 24, 2009 7:29 pm

Hmm? So actually have both coils turned on in opposite directions and then just stop one of them, instead of trying to start one? That's an idea. So maybe you could have a giant SC solenoid, and pulse these reversal coils to create a traveling mirror region. Then the back side of the mirror region is determined by the stopping of the coil current, rather than the starting of it.

As for the hammering things, wouldn't the two FRC's just link up by reconnection? I'd be afraid they'd mix and your high-Z driver would radiate all your energy away from the DT plasma.

This may be scatterbrained, but what about having a stationary mirror region and a tapered flux compressor that translates horizontally to compress the plasma? Or vice versa, a stationary tapered flux conserver and movable set of mirror coils outside? Problem is that for a reasonable taper angle, you'd have to get the conserver or the coils to move very quickly so that the compression happens on a short timescale. The coils and their power supplies would be a pain to move, but also having a flux conserver moving inside the vacuum chamber is probably not a good idea either.

TimTruett
Posts: 14
Joined: Sat May 02, 2009 6:26 pm
Location: Washington DC metro area

Plasmoid Acceleration

Postby TimTruett » Sat Sep 26, 2009 11:29 pm

About 10 years or so ago NASA funded some research on accelerating plasmoids. Some hardware was built. I remember seeing a picture of it. The idea was to develop an engine with a very high specifc impulse.

See http://www.niac.usra.edu/

Diogenes
Posts: 6958
Joined: Mon Jun 15, 2009 3:33 pm

Postby Diogenes » Mon Sep 28, 2009 1:44 am

MSimon wrote:I think I could do 5 nS with current technology - with feedback.

The trick is to make the device repeatable to 5 nS on 1 second time scales. Not too tough even for drive power on the order of 100 KW. I might be able to get repeatability to .1 nS with a lot of effort.

If I can get some kind of phase/location signal it would be a snap.



Sounds like a traveling wave tube. Just my two cents.

MSimon
Posts: 14331
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Postby MSimon » Mon Sep 28, 2009 2:44 am

Diogenes wrote:
MSimon wrote:I think I could do 5 nS with current technology - with feedback.

The trick is to make the device repeatable to 5 nS on 1 second time scales. Not too tough even for drive power on the order of 100 KW. I might be able to get repeatability to .1 nS with a lot of effort.

If I can get some kind of phase/location signal it would be a snap.


Sounds like a traveling wave tube. Just my two cents.


Yes it does. Except for the coil pulsing.
Engineering is the art of making what you want from what you can get at a profit.

93143
Posts: 1130
Joined: Fri Oct 19, 2007 7:51 pm

Re: Plasmoid Acceleration

Postby 93143 » Mon Sep 28, 2009 2:00 pm

TimTruett wrote:About 10 years or so ago NASA funded some research on accelerating plasmoids. Some hardware was built. I remember seeing a picture of it. The idea was to develop an engine with a very high specifc impulse.


http://www.aa.washington.edu/research/p ... s/elf.html

TimTruett
Posts: 14
Joined: Sat May 02, 2009 6:26 pm
Location: Washington DC metro area

link to Propagating Magnetic Wave Plasma Accelerator paper

Postby TimTruett » Mon Sep 28, 2009 6:00 pm

Propagating Magnetic Wave Plasma Accelerator (PMWAC) for Deep Space Exploration

" ... In these experiments, the acceleration
section was only 2 m long, yet velocities ~ 2.5x10E5 m/s were achieved."

http://www.niac.usra.edu/files/studies/ ... Slough.pdf

93143
Posts: 1130
Joined: Fri Oct 19, 2007 7:51 pm

Postby 93143 » Mon Sep 28, 2009 6:52 pm

That's quite a claim they make in that figure - 1 kN at 1e6 seconds Isp?

That's 5 GW. With inefficiencies taken into account, probably closer to 6 GW. Sound familiar?

Looks like Polywell won't run into a shortage of high-power electric propulsion ideas to pair it with...


Return to “Theory”

Who is online

Users browsing this forum: No registered users and 2 guests