Resonant Cavities As Pulsed Electron/Ion Accelerators? / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Resonant Cavities As Pulsed Electron/Ion Accelerators?

Move Thread

LAN_403

jpsmith123

Wed May 26 2010, 07:33PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Well I'm just kind of thinking out loud here to bounce this idea around and maybe get some feedback.

Using a resonant cavity to accelerate particles is an attractive idea since it is generally 100% metal construction, which eliminates insulators and "triple points" and related field enhancement and flashover problems. Also, the metal is grounded, with the accelerating potential generally existing only inside along the beamline.

As I understand it, the problems with using a single cavity for, say, electron acceleration (in a "conventional" mode of operation) is that field gradients are limited by high power dissipation in the walls due to finite Q, and by electron multipactoring (a resonant RF discharge).

So here's the basic idea: What if two "suitable" resonant cavities were coupled together in a Tesla transformer type equivalent arrangement, operating in a single pulse mode?

If power dissipation in the walls and multipactor discharge were removed as limiting factors, the accelerating gradient would then be limited only by vacuum breakdown between the metal parts of the structure.

I happened to come across a paper that published some experimental (pulse) vacuum breakdown levels (efield) of various metals, as a function of pulse length. And for pulses on the order of say, 5 nsec or so, it seems copper will hold off about 2 MV/cm, and molybdenum, for example, is good for over 3 MV/cm. IIRC, anodized Al is good for even higher fields.

Unless I'm missing something, to me this implies that it might be possible to build something that's maybe a foot in diameter and a few feet long, that will output a pulsed electron beam with energy of maybe 5 MeV or so, at reasonable efficiency, with average power limited by the primary HV power supply.

It seems to me that whether this is possible or not depends on finding the right kind of resonant cavities and a reasonable design. For example I can imagine two coaxial quarter wave cavities, coupled together, where the "primary" cavity has bigger C and smaller L with respect to the "secondary" or output cavity.

The primary cavity can be shorter and filled with a material with a high dielectric constant or can even include a lumped capacitor, with the charged center conductor then discharged with a fast spark gap, feeding an intense pulsed magnetic field into the base of the secondary cavity.

I'm thinking that if the degree of coupling is in the right ball-park, and there are no degenerate or nearby unwanted modes, the voltage multiplication should happen exactly the same as it does in a conventional Tesla Coil, with a very high accelerating field appearing across the gap in the "secondary" cavity.

I'm trying to find an EM simulator that will simulate switched charged conductors to test the idea, but the only one I found so far doesn't want to run.

Anyway, does anyone see anything fundamentally wrong with this idea?

LutzH

Fri May 28 2010, 12:56AM

Registered Member #1721 Joined: Sat Sept 27 2008, 08:44PM
Location:
Posts: 136

Hello:

Its being done sort of in most new radiation therapy linacs. Most are a standing wave design which have a bunch of copper cavities, which are then essentially electro-plated together. They are fed via a waveguide from either a big magnetron, or for the large 16-25MEV ones: With a big Klystron. I even came across one design which is for a superconducting one, made from Niobium.

They pulse these at the 1MW+ level at up to 400 pps, the high peak power is needed due to the cavity surface loss, which I would assume would nix Mo, or most other less conductive metals. Silver should work great though.

For a traveling wave electron acellerator: You could use a large copper pipe, loaded with Cu washers which are spaced in the right way. This design would be much more simple to home build, and I wonder why no one has home made one yet? Maybe the reason is in part that you need very high power microwave pulses, so forget about parting out your microwave oven.

The other reason may be fear, because if it works you better run for the hills, the radiation is fierce with a half value layer of like 1 inch iof lead, so with a 1 inch thick lead wall you would only die half as fast.

For positive ions there is the RFQ design which is slick, and they have made working ones out of aluminum on these.

All of these units run at high voltage VHF-Microwave freq. so a Tesla type of resonant coil would not do the trick. Any of these would make a very worthy project for a determined and skilled person, maybe you could find an old mine, or munitions storage bunker to test it. Take Care :)

Fri May 28 2010, 01:01AM

Registered Member #2372 Joined:
Location:
Posts: 62

You can look up two beam accelerators as they basically do the same thing it sounds like you are talking about but they use a low voltage high current beam to transfer energy into a high voltage low current beam. You may also be interested in photonic band gap accelerating structures as these have some advantages over conventional pillbox type cavities.

Proud Mary

Fri May 28 2010, 09:25AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

A bit off-beam,

but Svetlana have the cutest little glass betatron tube in this year's X-ray tube catalogue, so there's no excuse nowadays for everyone not to have relative electrons on tap in the home.

1275038619 543 FT89877 Svetlana Betatron

jpsmith123

Fri May 28 2010, 02:29PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Hello Lutz,

I think a home-built RF linear accelerator of "conventional" design (either the SW or TW type) would be highly problematical in many ways. I think it's possible to do it, but considering things like the generally necessary accuracy of construction, the availablity of a suitable high power microwave source, etc., it seems it would be beyond the reach of most people, and I include myself in that group.

What I am talking about, basically, is taking a "dual resonant pulse transformer" (so-called "DRPT"), i.e., a Tesla transformer operated in a pulsed mode at a low duty cycle using a triggered spark gap, and scaling it up in "frequency" to the point where the lumped element primary and secondary become distributed elements, i.e., resonant cavities.

The difference between this mode of operation of the cavities and that of a conventional linac is that in this case, for every single input energy pulse, you get a single pulse of electrons out, that is, it operates with single micropulses vs. macropulses.

In a conventional RF linac, the structure is generally exposed to the high power drive fields for the length of the driving macropulse, e.g., hundreds of nsec or even a few usec depending on the details of design and operation. This being the case, as I understand it, the accelerating gradients of conventional RF accelerator cavities are limited not by "single pulse breakdown", but by "vacuum RF breakdown". While I don't pretend to understand the physics involved (and from what I read I don't thing anybody else fully understands it either), it seems that short pulse hold-off can be significantly greater than the RF breakdown limit.

So, it should be able to provide a few MV of acceleration, from a relatively simple device, relatively compact device, without needing a high power microwave source to drive it.

Here's a picture of one way I can imagine it:

1275055758 1321 FT89877 Pulse Accelerator

The short section on the left is filled with distilled water and charged up from, say, a Marx generator. When the center conductor charges up to several hundred kv, the gap breaks down, which delivers a high power pulse to the other cavity through the coupling slots in the wall separating the two cavities (which I was too lazy to draw). Also, there'd have to be an alumina plate or something to keep the water out of the other cavity.

And if the coupling was optimum, all the energy stored in the first cavity should appear across the gap in the second cavity, the accelerating gap, only at a much higher voltage due to the smaller value of C in this cavity.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob