A means to 'recycle' lost scattered energy.

[Chris Bradley]

I'm just kicking this thread over for fear that this discussion will disappear under some 'Polywell' title, wtih which the following bears no association whatsoever. This is quite different to what the Polywell is attempting.



JamesC wrote:
> So I think I get it. In the thermonuclear version energy lost to scattering is not lost but just transfered to the heat and essentially recovered later on to some particle in the high end of the maxwellian distribution so the scattering losses are recycled.
>
> However in a beam version those losses cannot be recovered and are genuine losses and since the ratio of scattering probability to fusion probability is so biased toward scattering the average scattering loss exceeds the fusion gain even in an ideal scenario.
>
> Hence impossible even if we construct a PERFECT beam power device. If I had to describe such a device it would seem to look a bit like a particle accelerator with two monoenergic ion beams heading directly towards each other with some sort of containment field that would magically correct any scattering to straigten up a particles direction after a scattering event and then also reinject the energy the particle lost to scattering to get it back to fusion energy. Then upon fusion the newly fused particle would instantly disappear and the fusion energy released recovered without loss. This magic collision chamber would be long enough to ensure all particles entering it would fuse before exiting the other end.[Edit: Of course this device lives in a perfect vacuum]
>
> Ok.. so in this dreamland. Could this device produce net fusion power?
>
> The device has to put energy into each particle and energy to compensate each scattering event. How would I go about calculating if this ideal device produces net power? I think you had a go at something like this in your scattering post.




Chris Bradley wrote:
> Exactly! Top o'the class!!
>
> ...in one go you've summed up the issue and most of the solution I have suggested may be a route forward. The missing bit is just that instead of the beams being linear, they are circular and thus have no ends!
>
>
> JamesC wrote:
> > Could this device produce net fusion power?
>
> Regrettably the answer is 'unlikely', but something different is a step along a path nonetheless. Whether it is the right direction along the path.... ?!?





JamesC wrote:
> Great! time to pull out the calculator then.. I just have zero feel right now for the energies/losses we are talking about but your other post on scattering seems to be a good start!
>
> For a real world device I think the circular motion instead of the linear "thought experiment" version popped into my head also. Maybe it would take care of one other problem also which is that any neutrals/fused atoms would fling off the circular track "Instantly disappearing". The only problem is that a typical circular particle accelerator seems to need to be the size of france and cost a few billion but hey'.. maybe an electrostatic version instead of magnets could make it a bunch smaller. Again I just have no idea on the magnitudes of the forces here. Actually as I think about it having electronic control over the beam might help target a circular electrostatically controlled beam.. hmmm.
>
> I remember reading somewhere something about laser wave particle acceleration creating impressive linear velocities on a desktop but in any case none of it will be any use if even a lossless thought experiement cannot produce a net positive yield so I guess I better start with that!
>
> Thanks again, for helping to "bootstrap" me into fusion

[Chris Bradley]

JamesC wrote:
> I just have zero feel right now for the energies/losses we are talking about but your other post on scattering seems to be a good start!

My preliminary calculations suggest that this whole process only shows potential net over-unity at over 200keV collision energy, and then it is only a very modest marginal gain - so in this configuration not likely to produce sufficiently 'over-unity' to positively drive a heat exchanger into net energy. Hence, even the tiny theoretical window of viability turns against the reality.

Still working on the numbers but my [potentially erroneous] calculations are currently saying that at that collision energy the *average* loss of energy given up to the other particle(s) is 17eV. (Some collisions are very large - unit fractions of the particle's energy - but there are only a couple of these likely to each chance of fusion, whereas there will be millions of little collisions.)

At this energy we are talking about 10E5 scatters with this average energy loss per fusion event, so you'd have to pump in some 2MeV of energy to keep that particle 'alive' [at its 200keV energy] long enough to get it to fuse.

You can see what is gained by this process of feeding back in the energy - a 200keV particle can only possibly loose 200keV in a fusor-type device, yet you need 2MeV to keep it going long enough to fuse even at these elevated particle energies. If a process can be set up that delivers this energy back into particles already accelerated within the device, so the particle might survive long enough to fuse.

> The only problem is that a typical circular particle accelerator seems to need to be the size of france

These are very very slow particles compared to those 'nucelon' smashers. At 200keV you need merely 10 cm or so of radius, and fractions of Teslas of magnetic field strength. The electric and magnetic fields required are all very 'do-able' - theoretically!!

[inflector]

It is precisely this sort of focusing of beams and collection of scattered energy that I hope to accomplish with the positively charged tetrahedrons in the fusor I described here:

viewtopic.php?f=14&t=6833#p42516

The basic idea being that the positive charge on the tetrahedral grid will serve to focus the scattered ions back into a line. The three forces coming in at 60 degrees from each other will serve to center the ions. The three corners of the tetrahedron will serve like an electric notch and attenuate circular motion around the axis.

Once the ions lose their energy because of the combination of the positive grid repelling force and the negative inner grid's attractive force, the ions will be lined up in the same center line which goes from the outer vertex of the tetrahedron to the center of the fusor's inner grid. Lower energy ions will stop closer to the inner grid, higher energy ions will stop further out.

It is my belief that some of the collisions will take place between the higher energy ions bringing the lower energy ions up to greater speed. Then somewhere closer to the center of the fusor, all these ions will start colliding with accelerated ions coming in the opposite direction from the tetrahedron on the opposite side. The shape of the tetrahedrons will serve as a constant aligning force nudging scattered ions back into position after minor collisions.

- Curtis

[JamesC]

Good stuff, so (theoretically) if a configuration can get these scattering particles back on track you will be investing 2MeV to yield a fusion from between 5 and 15MeV which looks like a (narrow) window. Of course that 17eV is the actual loss not what it might actually cost to get the particle back on track.

Ok some quick questions.

It seems like 200keV is not that hard to achieve so why not make it 300keV. Does the power needed go up exponentially?

At 200keV you said 10E5 scatters are needed. How is this calculated?

I want to try to figure out what the limits are of two highly focused beams, how do I do this. Lets say you throw everything possible at creating a highly focused beam, how can I figure out the probability curves for fusion. Clearly this is the first key parameter for this hypothetical "dream machine". I am going to try to build high level models of two versions of the dream machine, energy recapture and particle reguiding.

This is really interesting stuff. I wonder if I will ever get any other work done ever again!

[Chris Bradley]

Curtis Faith wrote:
> It is my belief that some of the collisions will take place between the higher energy ions bringing the lower energy ions up to greater speed.

This will happen, but remember that this slow particle has already given up some of its energy in its other [thermalising] interactions, so the new 'full' energy particle will then, in turn, have a little bit of its energy sucked out, etc., etc... this IS the process of thermalisation. To keep the energy in the particles requires an active input of energy, more than just the movement of charged particles in an electric potential gradient.

If you pour marbles into a bucket, the ones you are pouring in never impart their energy back to the slower ones such that they can bounce out to their original height again. They may bounce up and down for a while, with slowly diminishing energy until they lie 'energy dead' at the bottom. This is a precise facsimile of ions in the fusor. Both gravity and the electrostatic field cannot do overall 'work' on the 'particles', all such particles have a given kinetic+potential energy and, statistically speaking, this sum tends to go only one way! This is why Carl points out how difficult it is to pull particles back on a coherent track. Something 'active' needs to be done and to date those processes employed are inefficient.

But the actual direction of that track can, possibly, be influenced by the means you have described. I think it may be an improvement to the fusor for further reasons. I meant to get back to your original post on this at some stage..

[Chris Bradley]

JamesC wrote:
> Good stuff, so (theoretically) if a configuration can get these scattering particles back on track you will be investing 2MeV to yield a fusion from between 5 and 15MeV which looks like a (narrow) window.

or 2.5MeV in the case of recovering neutrons from DD. Very, very narrow.

In fact, I think too narrow, hence I say it may be 'a step' - but I would not go to say 'viable' until I see it working BETTER than the theory*!!!


>
> Ok some quick questions.
>
> It seems like 200keV is not that hard to achieve so why not make it 300keV. Does the power needed go up exponentially?

For DD I would guess it goes up approx. linearly through to 400keV, then actually ramps off, as per the reactivity curve I have plotted in another post. Remember this is collision energy, so 400keV would be equivalent 800kV drive voltage in a fusor. (i.e. half the drive voltage from what you see on that graph I plotted)

>
> At 200keV you said 10E5 scatters are needed. How is this calculated?

For Coulomb scattering losses, I've turned Rutherford scattering equation into an energy transfer equation between interacting particles then integrated in two axes the impact parameter up to the Bohr radius for hydrogen. I get a long winded equation with logs in it, then multiply out to get an equivalent cross-section to compare with the fusion cross-section. For recombination and ionisation losses I am using the cross-sections from 'Massey', which may be an old out of date text now, but I don't think the universe has changed much since his day. (That's not to say I've interpreted the data right.) There are multiple caveats in these approaches and I'm still mulling over it all.

>
> I want to try to figure out what the limits are of two highly focused beams, how do I do this. Lets say you throw everything possible at creating a highly focused beam, how can I figure out the probability curves for fusion. Clearly this is the first key parameter for this hypothetical "dream machine". I am going to try to build high level models of two versions of the dream machine, energy recapture and particle reguiding.

Theoretically, that's quite simple and you can read through;

http://fds.oup.com/www.oup.co.uk/pdf/0-19-856264-0.pdf

and you should be able to figure out the rest. You seem to be on the ball with this stuff.

best regards,

Chris MB.


*(Well, it is permissible to live in hope with such outcomes - it has certainly happened with tokamaks. No one yet really knows why tokamaks adopt the 'H-mode', which you can see in images where there is little mass transport at the edge and a tight 'surface'. This was a surprise to tokamak researchers but means tokamaks have been able to perform 'better than expected'.)

[JamesC]

>the motion of the fast particles must be continuous and cannot afford to reciprocate back and forth

Yes, at some point they would have zero speed as they oscillate which would seem an ideal time to form a neutral especially when it's living in an electron cloud! ( remember I have only very knowledge of these things which is why I am asking so many questions! )

So really to minimise the neutral count which seems by all accoutns to be a bad thing, you need to have a minimum number of electrons lurking around ( no electron cloud! ), as close to a vacuum as possible ( no gas! ), and continously moving fast ions.

I am slowly working up my "idealised dream" machine model and have some more questions.

This scattering loss which was calculated as ~17eV per scatter event on average in a previous post. This is derived by integrating the probability of a scattering angle vs the loss at a scattering angle .. something like the attached picture ? So after the collision the sum of the energies of both particles is less by the scattering loss which is the loss that needs to be actively injected again "somehow". Is that loss symmetrical? or will one particle loose more than the other. The change in direction should be symmetrical if both particles have the same energy/velocity?

What is the nature of the energy released, is this an X-Ray, a photon or something else. Is there any chance to recover any of this energy. Will this lost energy be a directed energy or lost in a random direction?

Thanks!

[Chris Bradley]

Your diagram is as I envisage the calculation, and the total loss at all angles is the integral under the curve. Yes. And, indeed, the fusion cross-section is comparable only with the extreme angles of incidence of the scattered particles.

In terms of 'energy loss', what I am aiming to calculate isn't 'loss' in the sense of a photon or whatever. I'm simply talking about how much energy is transferred to a stationary nucleus off of which the beam nucleus impinges and scatters. So the target picks up this energy. But you are right as this process is not 100% and there would be EM emissions. However, it'll be fairly minor (overall), excepting for the very large angle scatters associated with large energy losses which are rare.

I suppose it may be better to talk about 'energy transfers'. But I'm essentially playing safe and being pessimistic here, and assuming that this thermalisation loss (for this is what it is we are talking about) is lost and gone and will end up as low-quality heat somewhere in the system. It is this energy we have to replace by pumping energy in.

Now, in terms of getting rid of un-ionised materials, we are at a point where we can go two ways. We can a) either accept, and therefore aim and deal with, a situation where these fast ions are passing through a volume of, essentially, cold neutral gas and we have, then, both ionisation losses and coulomb scattering losses to accommodate, or b) where we try to keep the back ground as ionised and as low a density as possible.

In the case of a), we would then be describing a fusor. This is essentially the fusor's operation and is where most of the fusion comes from fast ions colliding with the nucleii of some background neutral medium. I am thinking that, again, elevating the collision energy limits, but never eliminates, the electron interactions but we are then talking >500keV for net energy gains by this treatment. It all gets yet further suspect and 'hairy'.

The other mode, b), which you are thinking for your 'dream machine' is where you keep two beams going that are brought together to collide and with no background materials. This is not possible and there have been recent threads discussing this very point. However, in the case we are talking about, you can, at least, do better than (a). So you may not get to have a 'pure' coulomb-collision-only device but you can bias it in your favour by running two colliding beams and pulling the vacuum as low as you can go.

Or the other option for the (b) scenario is where you hold up a continuous beam of ions running through a background medium of ionised fuel. Again, you get nothing for free, and here you still have the electrons running around and will happily generate fast neutrals to compromise your efficiency, but also you will then have to balance the degree of ionisation with the density (and thus the reaction rate) because this resultant plasma will screen electric fields. So if your device uses electric fields to maintain the beam, then it may stop operating under this (b) schema. I have put forward an argument, a few times on this forum, that a successful beam confinement system, such as we are talking about, will have to employ both electric and magnetic fields so I feel this would prove to be a limitation.

So, there are a few options there but I don't see any that I'd necessarily want in a 'dream machine'. Reality comes to bite - as usual!!

But keep thinking and questioning. Maybe there is a route through!!

best regards,

Chris MB.

[Carl Willis]

Hi James,

>This scattering loss which was calculated as ~17eV per scatter event on average

How come? This quantity is dependent on some unspecified particle energy, for one thing. But more important is that this kind of number seems to be fundamentally arbitrary to me, because it depends on an arbitrarily-chosen minimum scattering angle for the purposes of calculation. I put a handwritten note at the bottom of the page showing how you'd presumably calculate your ~17 eV number, to illustrate why the minimum scattering angle (and the particle energy) matter. Notation:

E is the incident particle energy. I presume it to be a single constant value.
E' is the scattered particle energy, ranging over E'max to E'min according to conservation of energy and momentum
<E'> is the average scattered particle energy.

So what you call the average scattering loss is then (E - <E'>), Line 1. This formulation you already described ("...derived by integrating the probability of a scattering angle vs the loss at a scattering angle") so I won't dwell on it other than to mention that my notation is sigma_s(E,E') for the differential cross-section with respect to E' (equivalent to the probability of scattering into a particular angle); and sigma_s(E) is the total scattering cross section at all viable angles. For charged-particle, Coulomb scattering like we are talking about, sigma_s(E,E') is called the "Rutherford cross section". It's come up a lot on here. One common formula for it is on Line 3.

The problem is that in the real world, the maximum scattered particle energy, E'max, is equal to E (or equivalently, the minimum scattering angle is zero degrees). The electrostatic forces responsible for Coulomb scattering act on EVERY incident particle, no matter what the impact parameter, because the forces have infinite range. This makes the total cross section infinite and the scattering loss undefined, unless you artificially put an upper bound on E'max even though lots of particles are scattered into any small delta(E') below E'max. If we didn't have a Coulomb potential here, if we were dealing with the scattering of neutrons by the strong force for instance, the problem would be more meaningful (in fact it often finds application in neutron-shielding calculations). But here, you're going to have to provide some serious convincin' that this 17 eV number actually means something!

-Carl

[JamesC]

>This scattering loss which was calculated as ~17eV per scatter event on average

>>How come? This quantity is dependent on some unspecified particle energy, for >>one thing.

I am not putting myself forward as a physicists and actually I have no idea why the loss is calculated as 17eV. I was just trying to get my head around the result suggested by Chris Bailey in this post.

viewtopic.php?f=15&t=7183#p51011

I thank you for your mathematical addition and I need to study that since the math doesn't come naturally to me. I am trying to boil it down to the simplest possible terms. Alot of the math seems to act over the statistical result of the simplest interactions. I want to understand that simplest interaction and build up from there.

Anyhow, actually in the last post by Chris I see my understanding of this is wrong and there is no actual scattering energy loss like I understood it.. which is a good thing but I am going to add to this in the next post.

[JamesC]

Chris Bradley wrote:
>The other mode, b), which you are thinking for your 'dream machine' is where you keep two beams going that are brought together to collide and with no background materials. This is not possible and there have been recent threads discussing this very point.

OK.. so I got it wrong. The scattering enegy is not a literal energy loss but an energy transfer and you have based your calculations based on the reaction rate results from documents like that fusion reaction document you linked to in the last email. Thats actually good because I couldn't figure out what that energy was, I thought it might be something like that brem thing where the deceleration/acceleration from the coulomb force generating some sort of x-ray. But then I thought in a gas collisions seem to be perfect as atoms bouncing off each other don't continously radiate energy? but your energy transfer description make sense.

This 'dream machine' does indeed have to 2 beams heading toward each other rather than one as a stationary into which another beam collides. Can you point me to the thread that says this is impossible? Why can't one beam go one way and the other the opposite way both focused on each other? Thats what particle accelerators do no? if we are just talking about 2 ions that other seems stationary to the first? So cant' you just consider one of the beams to be the stationary gas and the other to be the beam but with double the energy when doing the calculations?

Another question: How rare are those high angle deviations (order of magnitude ) vs the actual fusion rate. ie how do I actually create the real shape of that probability curve? Is it really just coulumb force + electric force. If I made a simulation where I modelled an ion with the electic force and coulomb force and simulated a couple hundred thousand ions shooting toward each other in a two way tight beam I could 'numerically' derive that probability curve? These are not at relativistic velocities or something are they? so basic newtonian forces should get most of the way there?

I have no idea if there is a way through.. highly unlikely as you say but if I question *everything* then maybe I can find out to my own satisfaction

[Chris Bradley]

As Carl says, the energy lost by a particle is dependent on its energy and angle, so to sum you have to define what are, effectively, arbitrary limits.

But you can look at the tiny scatters on a numerical basis. Let's do a 'number' example. You can play around with a Rutherford scattering calculator on;

http://hyperphysics.phy-astr.gsu.edu/hb ... rosec.html

if you wish. But be careful as this works out a higher scattering angle than in a D-D system because this assumes the target is stationary whereas in the DD case the target would move away, thus deflecting the incoming particle less, so the following is a pessimistic treatment of energy consumption:

Let's take an example of a 200keV collision between Z=1/1 for an angle of 1 degree. It gives a value of 5346 barns. That is, within that cross-section there will be a deflection of greater than 1 degree. 1 degree is equivalent to an energy loss of sin^2(theta/2) = 7.6^-5 = 15eV (classic mechanics of deflecting elastic collisions).

Now at 200keV, the cross-section of fusion is 35millibarn, or 1/152740th of the probability of the scatter. So let's say we attempt a fusion and each time happens to be scattered off at 1 degree. But after 152740 such scatters, the quantum statistical probability says it will likely have fused. So that would be an energy expenditure of 15eV x 152740 attempts = 2.3MeV.

If we had considered the particle lost after these scatters, we'd have needed to send 152740 such particles in for a fusing collision = 152740 x 200keV = 30GeV.

This is, therefore, why I am very keen on promoting my idea that pumping energy back into a scattered beam particle may get somewhere towards viability for over-unity beam fusion.

Now let's try that again at 0.1 degree. We get a scattering cross-section of 534600barns (a 0.1^2 factor), so the probability is now 1/15274000th, but the loss is now only 7.6^-7 (this is roughly a linear factor to the scattering cross-section because of the sin^2 term and because sin(theta)=theta for very small thetas).

So the numbers will work out the same as we've reduced both values by the same 0.1^2 term. So once we get into small angles, so long as it is true for one angle, it is true for all smaller angles that that.

But for the larger scattering angles, this simple maths doesn't hold up and we have to do a more complex calculation, as Carl began to lay out. But at least now we only have to worry about the big angles rather than the 'infinite extent' of smaller angle scatters.

Remember that once you've begun trying to integrate across a whole cross-section with the equation Carl suggests, you will also need the Jacobian - something I kept forgetting when I first started this calculation. (Actually, it makes the whole result much easier!)

I attach an earlier attempt at this. It follows a post I made with a yet earlier attempt. I never followed up posting this one because, I seem to recall, I decided there may be an error in the logic. But for the life of me I can't recall what that error was, it looks OK to me now, so I'll post it here, now, and wait for the critique!!

..and after all of this...don't forget the charge-sharing and ionisation cross-sections!!

best regards,

Chris MB.

[JamesC]

In the dream machine there are two beams, not one beam and a stationary target.

So does this thermalisation loss due to scattering apply to the dream machine?

Doesn't the energy transfered from Beam A to Beam B equal the energy transfered from Beam B to Beam A cancelling each other out?

In your cofiguration, you seem to be talking about a Beam going through a gas, transfering some energy to the gas on each pass but then you retarget the beam and add back some energy to make up for the energy lost to the gas for the next pass through the gas.

However if you do this same thing for 2 beams moving in opposite directions the energy is conserved for any particles that can be successfully retargeted after scattering and so this thermalisation loss wouldn't occur except for particles you couldn't "unscatter"? So "theoritically" particles would just keep going round and round until they fusioned. -- Stop laughing.. this is a dream machine remember!

[Chris Bradley]

You've just talked yourself straight into a thermonuclear fusion plasma of the likes of JET/ITER.

What you are talking about is a linearly disposed set of ions (fusion plasma) where there is a radial ion transport barrier, and where any local momentum exchange radially to that central path is withheld thermally in the energy of the particles. So it is a thermonuclear plasma because what you are describing is that the net momentum of the particles is zero (their CofM is stationary to the lab reference) and so there is never any momentum loss. This IS a definition of a thermalised plasma.

Now wrap your linear volume around so it has no ends to loose ions through and...you have a tokamak. Yup, your dream machine is currently being built - in France!!!

Now, recovering the lost, thermalising, scattering losses in a locally hot, generally cold environment is different to this, hence the discussion I am putting forward looks to you like I am presuming the loss of a lot of energy. Yes, exactly!

[JamesC]

Hmm. maybe you are right. Here is what I am thinking for my dreammachine so far.

Keep in mind I am ignoring reality for now..

Two counter rotating ion beams in a uniform magnetic field. ( I have to check rotation is correct etc ) impact in a very small cross section. The small cross section is deliberately small so that the resulting scattering is as predictable and uniform as possible. On each beam there is fitted ( the very magical ) "unscatter" which corrects the scattering by applying the inverse to the scattering profile. Because the scattering occurs in a pointsource this forms a smooth profile which may have a chance of being corrected.

So in this machine, most scattering events are corrected. Placed around the remaining cross section is an energy collection device such that untrapped ions pass through a coil generating a voltage which is used to partially drive the ion guns.

--

I appreciate what you are saying for the radial energy and radial containment. So that becomes a tokamak but in this device the idea is to correct the scattering from a pointsource by applying a "lens". If this were a laser beam the correcting device would look like a lense. This is a particle beam and I can see that any radial momentum would quickly balloon out via the coulomb force. So I don't know enough yet to know if it is possible to keep a beam focused like this and certainly if it were anymore than a pointsource collision region it would very quickly get out of control. Also I think if the beam is sparse enough it might be easier to control. So it's all about keeping control of the beam.

In any case this device is only ever going to generate a very moderate amount of energy due to the pointsource collision region needed to produce a predictable scattering pattern in the first place. Maybe enough for a small heater?

[Chris Bradley]

Looks good to me. (Rotation is correct for +ions)

but....

a) How are you going to ensure the particles are all running around at the same speed, mono-energetically? If they don't then they'll have different radii and so will be all over the place.

b) Why do they orbit in the same path? They could drift anywhere in that magnetic field whilst they undertake their circular path. Why do the particles orbit about the same point?

c) what happens to really close scattering events that take the particles off by more than a radian of deflection, which might be more than your refocussing kit can cover?

[JamesC]

I'll take the 'Looks good to me' part but I will have to get back to you on the rest! That sounds far too much reality for now but here is a first shot.

--

>How are you going to ensure the particles are all running around at the same speed, mono-energetically?

Umm. don't know. Hopefully a radially constrained particle beam thermalises it's velocity to a mono energetic stream

>b) Why do they orbit in the same path? They could drift anywhere in that magnetic field whilst they undertake their circular path. Why do the particles orbit about the same point?

Umm.. don't know. Because they are monoenergetic and that orbit would be somehow made to be the energy minima

I doubt the whole thing would live in a single magnetic field. Only the collider parts needs to be to create the point source. The rest can use whatever other tricks are necessary to confine and present a monoenergetic particle beam to the in and out of the collider region.

>c) what happens to really close scattering events that take the particles off by more than a radian of deflection, which might be more than your refocussing kit can cover?

I mentioned these would be lost and put through a coil to recover their energy as much as possible. See "Collector Pic" for a cross section arrangment. The idea is these coils can be placed in radial positions from the collision pointsource and since we can predict the velocities of the particles from the monoenergetic beam and angle we can tune these coils to hopefully do a good job of extracting the energy. Since this is a vaccuum and no free electrons around they shouldn't form a neutral and hopefully by the time they hit the black cap collector of the coil they hit with close to zero velocity and exhanged their energy with the collector which puts it back into the ion guns.

Actually this would seem to be the theoretical limit of the device because ultimately all particles either end up fused or captured in one of these coils due to a high angle deflection so you would so the key thing becomes the ratio of unrefocusable high angle deflection vs fusions. So the higher the deflection angle of the "unscatterer" the lower the ratio of lost vs fusions.

So in the end break even ( not inclusing other losses etc ) is where the integral of the unrecoverable scattering events x collector efficiency x ionisationEnergy = the integral of the fusion events x fusion Energy ( see pic )

So in this dream machine it comes down to the retargeting angle threshold and the efficiency of the collector.

[JamesC]

Of course you probably need to modulate the intensity of the ion beams a little to induce the change in current/magnetic field in the collectors. The little collector coils should act like transformers and they can be quite efficient I think.

Any other comments before I try to build a simulation of this. Anyone have any suggestions for a charged particle / beam optic software library that could be of use in building a simulation of such a thing.

[Richard Hull]

Somehow all this improvement stuff winds up back at thermalization of a reaction zone or outright hot plasmas. I have constantly harped on the fact that only thermal fusion will yield big power as this is what we see "attempting to work".

Still, no matter, for regardless of scheme it will all just ooze out between out fingers of containment once we start really doing something, anyway.

The beauty of theoretical machinations is that even those that attempt to account for seemingly every leak and every gotcha' will just uncover new issues should they ever find their way to hardware. It has been the case for 60 years now.

Using recycled scattered energy is a great concept as there is so damned much of it in the fusion biz. It is the big part of the maxwellian curve....We dwell on the little, long drawn out tail.

Richard Hull

[Dustinit]

Probably the best way to simulate such a power scavenging arrangement would be a purely electronic simulation such as pspice.
Microcap have a good eval version with a limited component count but still very useful.
You could replace the ion beam with a low value inductor as transformer primary.
Dustin.

[JamesC]

Actually knowing it wont work is kind of reassuring. Mostly I am trying to explore and understand the problem. Since there is nothing I can do to contribute to thermal fusion I am stuck with the small end of town or get back to my other projects.

I have started build a simulation first, pure Coulomb forces to begin with and then if I can contain that in a model of the system I will start adding in other forces and if I can contain that then I throw more forces at it and so on. I have this hopelessly naive visualisation of these little ions whizzing around my 3D animation, guided by virtual magnetic fields and loop coils, fusing or banging into collection coils. If I can throw everything I can think of at it in a simulation and it still holds together I will then try to build one - that's when the real fun will just only begin.

Anyhow I have my particles going now and am trying to figure out how to approximate a loop coil magnetic field so I can see what happens when my ion beam goes through it. Hopefully it will help contain the beam which starts diverging pretty quickly!

This arrangement is pretty unique in that it is about the most containable arrangement you could hope for ( I think ) Highly symmetrical and only a single point of deviation. The downside will be limited power generation but it makes for a facinating thought experiment.

I need to produce the efficiency vs recapture angle charts for break-even to see what I am shooting for.

Thanks for your input
James Caska

[JamesC]

Thanks, not a bad idea!.. I have a feeling that if the beam can be contained into a steady state operation these coils will work very much like high efficiency transformers. It's going to depend on how smoothly the beam intensity can be modulated. Hopefully it will end up being some sort of natural harmonic of the beam doing the modulating but I am just guessing really. This part could be experimented on directly by putting a coil in a beam from an ion gun and modulating the beam.

[JamesC]

OK, seems this cylotron radiation is going to ruin my day for this configuration. X-Rays produced from the magnetic field accelerating the ions.

Does the fusor also suffer this loss? Would the polywell suffer this loss as ions oscillate ( are continously accelerated ) toward the center.

[Chris Bradley]

If you're thinking of sub-MeV ions with the device being of cm /10s cm in size, then there will be no detectable cyclotron emissions. The ion's mass is too great for cyclotron emissions in this energy range. Same for the fusor. The accelerations are just too small, they're 'only a few 10^10 m/s2' !! But if you crash an ion into something, *then* you may get X-ray emitting accelerations.

However, you may well get emissions from excited electrons being gyro-ed in the field.

Don't worry about the cyclotron emissions, I think you need to worry more about these energy-recovery coils. I don't see how they are going to work - particularly as you have various magnetic fields, and now there is a potential induction coil being talked about aswell. If these coils are recovering much power, they will need cooling, but also the continuous bombardment. They'll also need to be positioned to accommodate ions of various speed taking different curved tracks either orthogonal to the magnetic field, or straight tracks travelling in line with it (scatters going out of the plane of the orbits), or some half-curved track in between! I don't think they would have much coverage of all the possible scattering angles.

This idea of direct energy recovery from fast ions is something often mooted by Polywell/p+11B proponents. I don't see how it is actually an idea that would work. Has anyone seen a demonstrator of such a thing?

best regards,

Chris MB.

[JamesC]

Well.. have read alot more and run some simulations I can see the inclusion of the ion guns alone in the scheme makes it hopeless.

I am having fun though and learning alot.

Now my simulation is just a basic coulomb multiparticle system and I haven't even tuned it yet to the exact forces etc but it is still extremely informative.

I can see for example that ions thermalise in the center very quickly even in a perfect system. On my basic simulator it looks a plasma containment system and now and then the thermalised core flicks off high speed ions ( tail of maxwellian ) curve which fly high up the potential wall and then speed back to the center with a chance of a fusion. Since the "plasma" if thats what you call it is near the inner electrode which just gets collided with constantly.

One surprise in my simulation is if I apply a magnetic field through the whole fusor for a while it induces angular momentum to the plasma and creates an 'eye' in the center. Actually I only need to turn on the virtual magnet for a few moments to get the spin going and then the central ball magically changes into a disk with the inner electrode in the center eye.

So my question is this: has anyone put a magnetic field through a fusor to induce angular momentum? Í wonder if this would convert to a physical effect creating an eye which a ( smaller inner electrode ) could fit and hence avoid the collisions in the center. Could also be my simulation doesn't know what it's talking about!