Circularly polarised RF field confinement.

[Chris Bradley]

Further to Christopher Strevens post [ viewtopic.php?f=8&t=1689#p8910 ]:

This work appears, prima facie, to be a 'regular' RF induction plasma though we await any further details of variations or permutations as may have been employed in his device.

To address my question on whether an RF field could, by itself and with no magnetic field, oblige an ion of fusible energy into a helical path a simple calculation should suffice:

Centripetal force required to maintain a particle in a circular (helical) orbit = mv^2/r
Force provided by an e-field=Eq
Energy=mv^2/2

So we can see that the energy (Q) that a particle orbiting due to circularly polarised fields (or by any other means, the superposition of two orthogonal RF fields in quadrature) will be Q=Eqr/2 [or Q=Er/2 in units of eV].

In the equipment as above, r would appear to be around 2cm and E is indicated as some 500V/m (?) so that Q=5eV, sufficient for partial ionisation.

(I am presuming here that the best case of a polarised field is generated.)

Clearly, to generate fusible energies in the 10keV range with an r max of 0.02m we'd need circularly polarised fields of 1MV/m.

A magnetic field could be employed to provide additional radial (centripetal) acceleration where applied in the 'correct sense' wrt ion motion. The application of a circularly polarised RF field and a crossed B field is along similar lines to the approach I am taking. This description of configuration has already been suggested. George Meacham made this suggestion in a patent application, and is shown below. Clearly, the inclusion of a B field will mean the due centripetal forces can be generated for fusion energies, but in the case of George Meacham's patent application I do not see how it would work (with any efficiency better than a fusor) as the ion's orbital axis could drift with considerable loss to the walls, as occurs with a cyclotron. Also, there doesn't appear to be a solution to avoid axial drifts (that is, in-and-out in the page) nor anything to put Coulomb scattered ions back into this circular current. These would be similar issues even if a 1MV/m circularly polarised field could be applied in Christopher Strevens setup.

I did speak with George Meacham on one occasion, as the application hadn't been taken through as a patent yet the search had not returned any prior art so it could have been so, which was curious. Apparently, George is a 'freelance' inventor, so to speak, providing patent services and this was one of his own ideas but I understood that he had other projects to take up his time and efforts, nor did he get any interest to support it from anyone else, to see it through to an issued patent.

[Todd Massure]

Hi Chris,
Is there anywhere that has more information on this? Specifically what the numbers are referring to.


Todd

[Chris Bradley]

You mean George's patent? It is US patent application 2005/0249324.

I'll stick a copy in the files section, it's not a very big pdf...

...viewtopic.php?f=19&t=8006#p57410

[Todd Massure]

Thanks.

[Quantum]

Chris, I've not read the link you've given, but My knowledge of magnetrons (mostly from Wiki, I think) whether for microwave cookers, or for radar, indicates that the electrons follow a helical path from cathode to anode.

This is what creates the RF.

My understanding is that you don't get RF without a 'helical path'.

[Chris Bradley]

Depends what you mean by 'helical', as in, what is the axis of this helix you're talking about.

Maybe you've read something and you've not quite got the picture - in a magnetron an electron is emitted from the central cathode and heads towards the outer anode. If the magnetic field and electrical potential between the anode and cathode are just at the 'magnetron' condition (Vcrit = [Q.B^2.s^2]/2m), then the electron won't reach the anode (viz will orbit around the cathode). But it is unbounded axially so it could execute helical motion if it also moves axially along the tube at the same time. Maybe this is what you've read.

At higher or lower potentials, or when the space charge wheel has not settled down (which it takes a moment to do, hence you can't turn these things on or off instantly or operate them very easily at part loads/voltages) the electrons can undergo epitrochoidal paths in the space between the cathode and the anode. If the e-field they are experiencing is lower that the magnetron condition, they crash back into the cathode. If higher, they may, epitrochoidally, spiral into the anode in a rather haphazard fashion.

[Quantum]

Yes

[Chris Bradley]

It goes to show that you have to think about the axial stability of orbiting ions, as well as just the 2D thought-invention as above.

There is also a device called an Orbitrap that exploits helical motion about a shaped central electrode so as to select q/m according to the time-of-flight behaviour in the axial, as well as orbital, motion.

[Doug Coulter]

Chris, in a cyclotron, the field lines are deliberately bent outwards via "shimming" to control the drift into the top and bottom of the Dees. This doesn't eliminate all monkey motion, but it does allow for a beam to form and be moderately stable with sinusoidal vertical motion as it goes around. The beam does go up and down (in a vertical H field machine) some, and there are tricks to control that which are kind of arcane, involving having an extra electrostatic variable width focus element on one of the Dee entrance sides, and even deflection plates to control this motion, which I believe is mistakenly called "betatron oscillation" (which makes no sense to me). At mere cyclotron energies it can be hard to damp this motion via photon-radiation losses, there just aren't enough.

This kind of stuff doesn't often make it into books we mortals can afford, of course, it tends to be institutional knowledge, a guild just like any other one that keeps its secrets. And, when it becomes "not cool to know" it's forgotten instead of released to the public, a loss to all.

You might surmise I'm not a fan of trade secrets. You'd be right.

[Chris Bradley]

Doug Coulter wrote:
> Chris, in a cyclotron, the field lines are deliberately bent outwards via "shimming" to control the drift into the top and bottom of the Dees.
Indeed; there are a few beneficial focussing effects that help the cyclotron on its way. Pole-face shaping at the edge of the field lines into concavity help a little but is, as I read it, still a bit weak. Good for electron focussing, poor for heavy ions.

The electrostatic fields generated between the dees is also helpful for axial focussing, as an ion spends more time on the 'side' of the fields pointing towards the central plane and less time in the region of fields pointing away from the centre line (because it'll have accelerated a little by then). I think both are regarded as 'weak focussing', and I'm not sure which have the higher contribution. A stronger focussing is done by varying the magnitude of the magnetic confinement fields azimuthally. This way the field is designed to 'bulge' in an appropriate manner to draw ions onto the centre plane.

I presume the magnetron gets away with it by piling up electrons at the ends from axially drifting electrons which may create [space] charging at the ends and reduces the extent of the helical motions. Or maybe it uses charged end-caps for the axial confinement - this is how a Penning trap does it.

That just about covers all the crossed-field devices invented to date (excepting the amplitron and trochotron which are both similar to the magnetron's principle).

I'd say that the best arrangement would be to configure e-fields where all the e-fields point towards the centre plane!.. Surely this would be the strongest axial focussing of all? That's what a Penning trap does, and it can have ion confinement times of *weeks*, not milliseconds!

[Quantum]

Does Fleming's 'right hand rule' not play a part here?

[Chris Bradley]

You probably mean left hand rule, but, in any case, to what is your comment aimed?

[Quantum]

I probably do mean left hand rule, Chris.

Any force will be at 90 degrees to the field, won't it?

[Chris Bradley]

magnetic fields are cross-product fields. They apply a 'force' that goes 'sideways' to their trajectory.... and.....so far, so what? Not following you.

A crossed field device like cyclotrons and magnetrons get charged particles in orbit because the centripetal acceleration needs to be orthogonal to the trajectory tangent, which happens to be the direction the magnetic field applies its 'force'. This is as shown above (I don't think the arrows 'down; the page are meant to be showing b field, but e-field at some instant), and as usually described in school science... Help me out here, is there a question or point in your posts somewhere that I'm missing?

These orbits can drift 'helically' because that magnetic force "doesn't care" if the particle is moving 'along' a magnetic field line, it'll just make it orbit the field line (a 'guiding centre') without regards for the particle's component of velocity parallel to the field line.

[Quantum]

I was thinking primarily about magnetrons.

Where is there a force to induce movement along the 'axis' of the field?

[Chris Bradley]

Just random motions. One has to anticipate small perturbations in a design. Only optimistic inventors think that they can hold something in a 2D plane if they only use forces in that plane. Disruptions and disturbances will always induce things to go in a directing in which they are not intentionally constrained. For magnetrons, the causes of axial instabilities is even more obvious. Even if there was a perfectly machined device, manufactured by God himself to a surface accuracy of a fractional atomic width, charge cannot occupy the same space and therefore as they mix some will get pushed 'up' and some 'down' by the charges ahead and behind them, thus will diffuse axially.

For similar reasons, the idea of ever aligning up a deuteron so it will 'hit another' head on, for improved fusion rates, is completely bonkers. (Sorry to offend, if anyone is trying that, but it is bonkers.)

Physicists seem to fall into the same trap aswell. I have seen the same errors embedded in bits of main-stream 'understanding' of plasmas and surface-liquid wetting physics - it is elementary to know that things are stable in a minimum energy null, and critically unstable at a maximum peak, but often you see points of inflexion (d2x/dy2=0) interpreted as a 'potentially stable' condition (just like axial motion in a cyclotron) and experiments that are actually performed to try them out. However, I guess I should be cautious with such a comment because sometimes what appears to be an unstable "d2x/dy2<0" condition can provide curiously stable behaviour, case in point are FRC plasmas.

[Quantum]

But surely only ten percent or so would be lost?

[Chris Bradley]

'Lost'? What do you mean? In a magnetron, it's the electrons orbiting the cathode that radiate the RF (at their orbital frequency). Whether they are bumbling up and down the device axially is irrelevant. As I said above, presumably either an e-field, or a build up of space and/or surface charge, at the ends of the device cause a restoring axial force, were the electrons to get that far.

There *is* a loss of electrons back into the cathode and outwards to the anode by virtue of azimuthal instabilities and space charge as well. Ideally there would be none, but, ultimately, it's 100% loss, of course. Hopefully, though, most of the electrons actual energy gets emitted through cyclotronic coupling with the anode but, I guess, some %age gets lost which is why it gets hot!

[Quantum]

I appreciate that there are losses in all ion sources, Chris.

I'm starting to design a source of my own. I'm assuming it's the high frequency AC field that ionizes the plasma.

I'm assuming that the advantage of an RF plasma source is that the ions have less velocity.

[Quantum]

Sorry, Chris, I think I was going a bit 'off topic' there.

There are losses in any system, Chris, the trick is to minimize them.

[Chris64Strev]

I read all your posts and I think I understand most of them.

The device is an LC tank circuit with a tube of low pressure hydrogen place inside the inductor which a simple 15 turn coil made from 4 mm enamelled copper wire. The capacitor is a 1000pF variable vacuum capacitor.

This is tuned to 3.7 MHz and I applied 800 watts from a radio transmitter at its resonant frequency.

The radio frequency currents in the coil heat the gas and ionisation takes place. At that point the gas conducts and an induced radio frequency current flows.

By the normal laws of induction the two curents repel each other and the ions are compressed towards the axis.

The ions are acted on by two electric fields, the longitudinal field and the circular field inside the coil. This tends to force the ion into a helical path that changes direction each time the electric field changes direction.

The ions travel very fast because they are light and pick up energy from the field during their passage in the helical path. The force of repulstion from the wires tends to keep the ions from going away from the axis.

I do not know how to calculate the ion path but the general idea is that since the mean free path is 1km at that pressure of 1E-8 Tor the energy will be 1000 x E where E is the E field in volts per meter. I do not know how to calculate the fields inside a coil carring AC but I believe this calculation is very difficult. However the field along the coil is about 7000 volts because of resonance so since the coil is only 0,06 meter long the E-field is 7000/0,06 = 116000 V/m so with a 1000 meter path length the collision energy will be 116 MeV. I know the path is helical but I cannot, as I say calculate the pitch or the number of revolutions in the length of the coil so I do not know if it can reach this energy. Even if the path length is only 10 meter the energy will be 70KeV.

I agree the yellow line was due to sodium mobilisation but there is a prominant green line.

The glass fluoresced during the burn (green).

After the burn (cooled and re-energised) the discharge had changed colour to be greenish from being blueish. But that is by eye.

I used a hand spectroscope used by chemists in the flame test mounted on a canon D45 digital camera.

More recently I used a tube filled with hydrogen at 1E-4 tor and excited it with 100 watt. No colour changes were observed but an external RF meter showed 130 watts given off. This started as 100 watt radiated power but it rose to 130 watt as the gas heated up.

The gamma signal increased a bit above background during this run.

I have also tried a 1 Tor tube and here the radiated power fell as the tube heated up.

The gamma detector registered only background.

The idea is that the higher pressure gas had a shorter mean free path and the ion did not receive as much energy during its flight before collision.

D has a much lower activation energy than protium so I would like to try this.

I'm hoping that I will get ignition at less than 100 watt exciter power.

I cannot be sure of the power emitted only that it rose during the run. The exciter power remained constant at 100 watt. I also had a calibrated RF E-field meter that indicated that aproximately 100 watt was being emitted but the meter is not precise enough to distinguish 100 w from 130 watt and this was estimated by another uncalibrated but more precise meter.

The fact that the radiated power dropped for the higher gas prssure indicated that in the lower pressure gas something different was happening and that power was being developed by the gas.

The idea is that the low pressure gas is acting as a negative resistor when fusion enefgy is being developed.

I do not know how much P+P > He4 (I know this has a low cross section) conversion took place per second to generate 30 watt but I would have thought it would be only a tiny amount. What would be the expected neutron flux and gamma flux?

I did a calculation to find the ignition point and I got 1000 Tor, and 1E9 watt.

The fusion power is transferred to the oscillating current by induction as the hot gas expands and changes the inductance slightly.

According to my fusion text book a similar D containing tube would generate 30 watt out with only 10 watt exciter power, if so we have ignition! This is by looking at the activation energy and proprtionatly scaling.

Is my proposed experiment dangerous?

[Dustinit]

You have some calculation misconceptions.
An ion accelerated through a field of 7000v gains 7000ev regardless of the
separation of the electrodes or the volts/meter.
Its unlikely you would be able to measure any fusion at these energies especially since they are all accelerated in the same direction alternately.
Ion on neutral is the best you can hope for at energies somewhat less than the 7kv applied.
Dustin.

[Chris64Strev]

Electric field strength misconception.

Oh yes - how silly of me! So I need a higher potential difference. I'll mull over that.

However what do you think of the apparent output power of 30 watt more than the input power?

I looked up the conversion: 1KeV ~ 1.16E7K so 7KeV~ 78 million K. From other readings I believe this is hot enough for fusion to take place.

The general Idea that I am trying is an induction coil with low pressure hygrogen inside.

By observation I can see that the hottest part of the gas is in the region of the mid point of the coil.

I read that the p-p reaction becomes important at 200 MK so the potential difference needs to be 17 KeV.

So to provide this a resonant LC circuit that generates 17000 volts needs to be made. I understand that the traditional Tesla coil could do this. So a single layer coil of 15 turns made of copper wire 4 mm diameter would be 6 cm long and 30 mm diameter tuned with a 550pF capacitor would give 17000 volts when excited with 3500 watt exciter running at 3.7 MHz. To withstand the high ohmic loss the copper coil would need to be cooled by forced convection or by liquid nitrogen. The current in the coil would be 220 amp.

For a pressure of about 1 Tor, According to my fusion book, the power density generated would be only about 4E3 Watt/cm^3 so with a reacting volume of about 3 cm^3 the power generated would be 4000 watt. This would be sufficient for ignition. The energy density is about 1.2 E10 watt/meter^3.

To carry out this experiment I would need a 1000pF vacuum capacitor for 20,000 volts and 300 amp.

If I used deuterium then the temperature needs to be about 10 MK so the collision energy would be about 1 KeV to my set up would do it as it stands.

With tritium the temperature would be even lower.

Although I have worked with accelerarators and on nuclear weapon design in the past in government research laboratories I am not a professional nuclear phyicist so I am in no way an authority on these subjects. And although I am a radio amateur I have only elementary radio skills.

I would like your comments and I would also like three sealed borosilicate or quartz tubes containing deuterium at a pressure of 1E-4 Tor to see if I can get my engine to ignite.

Or someone else could try it out and see if they can get ignition.

[Dustinit]

It is difficult to measure RF power into an inductively coupled plasma.
The power readings are only valid with a load impedance of 50 ohms.
The plasma varies its impedance with power (hotter is more conductive.)
Power out would be extremely difficult to measure as light heat etc are difficult to integrate.
I see no possible mechanism for any power created (if exists) to be coupled back as rf and be measurable in your setup and suggest you try measuring the reflected power to reconcile the difference.
Interesting experiments though.
Dustin.

[Chris64Strev]

The theory is that the power generated in the plasma heats it and so makes it expand. Because it is a conductor it therefore changes the inductance of the coil by reducing the area the flux of the magnetic field generated by the current in the coil available. This generates a current in the coil and increases the amplitude of the oscillating current.

The power I measured was the radiated RF EM power external to the tuned circuit. The reading of the RF meter increased from 36 to 46 as the gas heated up.

I have another E-field meter that is calibrated in KV/meter and by calculating the incident power from Poynting's vector and integrating all round the apparatus, assuming spherical symmetry found the approximate total radiated power at 100 watt. This meter is not precise enough to distinguish 100 watt from 130 watt. This was calculated by (46/36) x 100.

The reflected power also went up from zero to 5 watt during this period.

However I am not an Rf engineer so I do not fully understand it.