Van de graaff accelerator - do-able?

[jshine]

van de graaff accelerator -- do-able?

(posted in the wrong forum -- apologies)

 

[jshine]

van de graaf / fusion neutrons -- do-able?

After a bit of preliminary reading, I'm wondering about the feasibility of using common (off-the-shelf) Van de Graaf generators as a source for a delta-V large enough to accelerate protons (or deuterons) to energies high enough to engage in simple nuclear reactions.

Commercial Van de Graaf generators can be found for ~$500 that produce up to ~500kV, and since (apparently) their polarity can be adjusted, a pair of opposite-polarity 500kV generators could produce a delta-V of ~1MV for ~$1k. If the positive end of this set-up contained an ion source, and the two terminal spheres were separated by a ~2m glass tube evacuated with a turbopump (connected in the middle, at ~ground potential), and the negative end contained a deuterium-rich target (a bit of commercially-available LiD?), it seems like it may be possible to get some detectable fusion neutrons.

Perhaps 1MV is overkill though -- after all, commercial neutron generators appear to operate in the ~100-200kV range ( http://en.wikipedia.org/wiki/Neutron_generator ).

My question is basically one of feasibility. I'm no physicist, so I'm not familiar with the design considerations & possible pit-falls of this sort of setup. Obviously there's a bremsstrahlung issue to consider, as well as any residual activity produced by the neutrons. ...but what are the actual magnitudes of these? Are they curiosities or actual hazards? How would a suitable target be prepared? Would loose LiD be sufficient, or would one need to adsorb the deuterium onto a metal surface (and, if so, how is this done)? What is the simplest possible ion source that could be suitable (probably battery-operated, since it needs to live in one of the Van de Graaf sphere terminals)? And would a simple evacuated glass tube (long enough to avoid arcs between the Van de Graaf terminals) be sufficient to carry the beam, or does it absolutely require some sort of focusing electrodes along the way?

What are the thorniest issues likely to be?

Any references / suggested reading / etc. would be appreciated too!

 

[bcrowell]

This was discussed within the last year or two on this forum -- a search should turn it up.

 

[jshine]

This was discussed within the last year or two on this forum -- a search should turn it up.

Do you mean this thread: https://www.physicsforums.com/showthread.php?t=315752?

I read through it before posting, but this project seems to have been terminated at a very early stage, so it doesn't provide much guidance or any results. I've seen a number of other more speculative threads, but nothing that gave any clear direction (or a clear reason why such a project would be destined to fail).

I know that Van de Graaf generators can make good particle accelerators with sufficient support equipment (multi-story tanks of SF6 under pressure, fancy glass/metal columns, etc.), but I'm curious how many of these refinements can be stripped away w/o running into some insurmountable barrier.

 

[bcrowell]

I think the thread I was thinking of was this: https://www.physicsforums.com/showthread.php?p=3023812

But basically here's the deal. To get fusion, you need a center of mass energy equal to ke^2/r, where k is the Coulomb constant, e is the fundamental charge, and r is the sum of the radii of the nuclei -- roughly 2 fm for some isotope of hydrogen on some isotope of hydrogen. For a non-tandem van de graaf, the lab-frame energy you're going to get for hydrogen is eV, where V is the voltage. The center of mass energy is lower than the lab energy. If the beam and the target have equal mass, then the c.m. energy is half the lab energy.

All of this is for fusion above the Coulomb barrier. Below the Coulomb barrier, you get a cross-section that decreases exponentially.

1H on 1H is not going to give you appreciable fusion -- the cross-section is super small. Getting a deuterium beam and/or a deuterium target does not sound easy to me as a DIY project.

[EDIT] Corrected an error above -- c.m. energy is 1/2, not 1/4.

 

[Drakkith]

What are your goals for something like this jshine? I am currently attempting to build a Fusor at home. A fusor is similar to what you are thinking of, but it is not beam-target or beam-beam. I've got a 40kv power supply that should be fine for getting detectable fusion out of a Fusor. (Well, as long as it works...got it off ebay...)

Try www.fusor.net for more info on this or look it up on Wikipedia.

 

[jshine]

1H on 1H is not going to give you appreciable fusion -- the cross-section is super small. Getting a deuterium beam and/or a deuterium target does not sound easy to me as a DIY project.
[EDIT] Corrected an error above -- c.m. energy is 1/2, not 1/4.

Hypothetically -- for the sake of argument -- if one didn't mind the expense of D2 ( http://www.sigmaaldrich.com/catalog...RAND_KEY&N4=368407|ALDRICH&N25=0&QS=ON&F=SPEC ), are there any other considerations that would be absolute barriers?

The idea came from these: http://en.wikipedia.org/wiki/Neutron_generator It seems that these devices actually operate at a lower voltage than what a typical Van de Graaff generator would create (albeit at a higher current, most likely).

 

[jshine]

What are your goals for something like this jshine? I am currently attempting to build a Fusor at home. A fusor is similar to what you are thinking of, but it is not beam-target or beam-beam. I've got a 40kv power supply that should be fine for getting detectable fusion out of a Fusor. (Well, as long as it works...got it off ebay...)

Try www.fusor.net for more info on this or look it up on Wikipedia.

I'm familiar with fusors & the goal is essentially the same -- production of a detectable neutron flux.

 

[bcrowell]

Hypothetically -- for the sake of argument -- if one didn't mind the expense of D2 ( http://www.sigmaaldrich.com/catalog...RAND_KEY&N4=368407|ALDRICH&N25=0&QS=ON&F=SPEC ), are there any other considerations that would be absolute barriers?

I outlined a calculation for you in #5. Have you done the calculation?

Neutron detection is pretty difficult to do. The Pons-Fleischmann fiasco happened partly because Pons and Fleischmann didn't have sufficient expertise in neutron detection. It requires sophisticated electronics to tell a neutron pulse from another type of pulse. You will have a ton of x-ray and gamma background from Coulomb excitation when the beam hits the beam stop.

I guess it's true in theory that you can add the voltages of van de Graaff generators in series to produce a bigger voltage. In reality, I suspect you'll find it impractical due to sparking. This is why high-voltage van de Graaffs are large, use SF6, and have sophisticated electrode designs.

 

[jshine]

I outlined a calculation for you in #5. Have you done the calculation?

Sure, my take-away from #5 was the sentence "If the beam and the target have equal mass, then the c.m. energy is half the lab energy.".

Even taking this into consideration -- suppose 500kV was available (1 standard VdG, forget the VdG-in-series idea for a moment), given the cross-section vs. energy diagram for DD fusion presented Fig. in this paper:

http://escholarship.org/uc/item/1jn4c41k

it seems that there is still a decent cross-section in the low 100s' of keV -- maybe 1 order of magnitude less than at the peak of 2MeV.