Vacuum Pump Ideas

Every fusor and fusion system seems to need a vacuum. This area is for detailed discussion of vacuum systems, materials, gauging, etc. related to fusor or fusion research.
RealBorg
Posts: 27
Joined: Thu Mar 01, 2012 8:40 am
Real name: Thomas Zehetbauer
Contact:

Vacuum Pump Ideas

Post by RealBorg » Fri Aug 22, 2014 8:56 pm

My mechanical vacuum pump only makes it down to about 700 micron. Before I resort to buying a diffusion or turbomolecular pump from eBay, I want to experiment with some ideas:

1.) Sorption (already working down to 10 microns)
Activated carbon has a large surface which is in adsorption/desorption equilibrium according to pressure and temperature. Therefore when I heat the activated carbon during evacuation, seal the chamber and cool the active carbon the equilibrium is shifted toward a lower pressure. If I pump down to 700 micron, backfill with deuterium, pump down again and seal the chamber it should give me a pretty good deuterium vacuum.

2.) Mechanical Diffusion Pump
Standard pump oil already has a vapor pressure far below the vacuum obtainable with these pumps. If I use a mechanical pump for the oil and a long tube with a spray nozzle the falling drops should pull remaining molecules with them.

3.) Thermal Pump
A Crookes radiometer works an the principle, that adsorbed molecules get more impulse from a hot surface than they get from a cold. A careful arrangement of alternating hot and cold surfaces should therefore be able to create a flow of molecules. The construction might be as simple as having heated diagonal fins opposing a cold surface and could theoretically provide an almost 100% (ignoring radiation losses) efficiency.

4.) Ion Pump
I have read about ion pumps with a chemically active electrode. What if the electrode was not chemically active? I am thinking about ionizing remaining molecules with a needle or a heated filament and accelerate them toward the vacuum pump with an electrical field. I would probably prefer having molecules hit the anode and be accelerated toward the cathode than with the classic penning trap approach (which is better suited for high vacuum).

5.) Turbo Molecular Pump
While conventional turbo molecular pumps require precise engineering, a simple multi stage centrifugal pump should be able to exploit the same principle. Molecules enter the first chamber, become absorbed on the rotor, get a momentum toward the chamber wall and enter the next stage.

What do you think about these ideas?

Tom

George Schmermund
Posts: 262
Joined: Wed Aug 16, 2006 12:51 am
Real name: George Schmermund
Location: Carlsbad, CA

Re: Vacuum Pump Ideas

Post by George Schmermund » Fri Aug 22, 2014 9:23 pm

Talk's cheap and you'll find plenty of that here. Except for the 'possible' use of sorption pumping, I find your ideas more amusing than interesting.

Do some experiments with your ideas and report back when you have some results.
Anything obvious in high vacuum is probably wrong.

Doug Browning
Posts: 156
Joined: Sun Mar 11, 2012 2:19 am
Real name: Doug Browning

Re: Vacuum Pump Ideas

Post by Doug Browning » Sat Aug 23, 2014 3:10 am

Another novel Vacuum Pump thread that got similar comments:
viewtopic.php?f=10&t=9577

I cannot agree less with the usual comments that "talk is cheap" on these boards. Sure its easy to propose untested novel ideas, but some discussion is necessary to weed out the impractical, to fill in the engineering parameters that make or break an idea, and often can lead to some cross fertilization helpful for another idea or extension. It educates all involved as well.

The thermal surfaces idea is interesting, but I think constrained atoms in material surfaces will not develop sufficient impact momentum to give gas molecules much forward momentum. Now if you could come up with some engineered molecular coating that a laser could exite to swing a molecular tail appendage around....

The realization that only the surface needs be active rather than the whole volume ( like in a diffusion pump ) is important for efficiency considerations. However, the configuration must also prevent back diffusion thru the bulk volume.

User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Vacuum Pump Ideas

Post by Chris Bradley » Sat Aug 23, 2014 10:16 am

I think it is fine for someone to ask about the relative viability of a few ideas before trying them.

Could save a lot of time, especially with some of these ideas.

You need continuous pumping to operate a fusor. Your carbon adsorption idea will give neither control nor have sufficient effect. Micromazes have been used to augment the performance of two stage pumps - I think Richard did exactly this in his early setups - but it is only an augmentation to help purify the chamber volume and you still need good pumping. You need to get down to 10-5 torr or so just to start with so you get a higher quality fill when you start, so you have to drop a few orders of magnitude lower and a trap might help for a moment but will then get 'filled'.

You need to be at least approaching the molecular flow regime before diffusion or turbo pumps will work. You need to be under 100 microns before you start these ideas.

I don't think there would be enough knowledge on your 'thermal pump' idea. I can't see it myself - I guess it might work at extremely low pressures, but it just doesn't feel physically possible at the sorts of pressures a fusor works at. Again, if you are still in viscous flow regime then I can't see it stands a chance.

User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Vacuum Pump Ideas

Post by Chris Bradley » Sat Aug 23, 2014 10:21 am

smoking-amp wrote:Another novel Vacuum Pump thread that got similar comments:
viewtopic.php?f=10&t=9577
err... I can't see the comparison. George was actually building an idea. Any adverse/wayward comments were your own.

I can see that there might be room for some novel sort of diffusion pump. Some brain-work would, I am sure, come up with several alternatives to the 'standard' design, but I suspect the standard design has evolved into what it is for various trade-off reasons.

I've always thought that it might be possible to make some sort of plasma diffusion pump, in which a region of a device can push particles into a magnetic trap. In a way, you'd be dragging molecules out with the plasma, and letting the electrons escape to ground.

User avatar
Dennis P Brown
Posts: 1930
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Vacuum Pump Ideas

Post by Dennis P Brown » Sat Aug 23, 2014 3:21 pm

I'd think that before building or test talking about a number of high vacuum (HV) systems, you might first address the dismal performance of your mechanical pump. Most HV systems require sufficient performance from a two stage mechanical pump - generally under 100 microns is an acceptable starting point (under 10 microns is more typical.) Also, pumping volume (l/s) is of some importance.
As for throwing out random ideas, research some of these topic first - google is your friend. I have already addressed the issue of absorption pumps for you earlier.
Not clear why you want to explore an area that has already undergone massive research by so many hundreds of scientist and engineers over a number of decades on systems that already have been proven to perform so well; also, so many good systems with excellent performance at low cost are available on e-bay (used, of course.)
Exactly what is wrong with diffusion pumps or turbo's that need replacement for a fusor system?
I will not comment on your ideas because most do not correctly demonstrate understanding of current systems or physics that is given. Again, use the internet - especially the Wiki on the related systems you are referencing.
Best of luck

Aside: any system can be improved - just takes money and effort, For instance - in college my diffusion pump bottomed out near 1.5 *10^-6 torr; I added a refrigeration system to the cold trap (istead of water; also I salvaged all my parts from a old refrigerator!) Then I filled the system with propane (vastly more efficient than feron or its alternatives.) My DP got down into the low 10^-7 torr. Simple, cheap but did take some work. But off-the-self systems are dollars and effort ahead unless you just want to research HV systems. Not really a fusor topic.

George Schmermund
Posts: 262
Joined: Wed Aug 16, 2006 12:51 am
Real name: George Schmermund
Location: Carlsbad, CA

Re: Vacuum Pump Ideas

Post by George Schmermund » Sat Aug 23, 2014 6:36 pm

Don - Thanks for the bump on my DP thread!
Anything obvious in high vacuum is probably wrong.

Dan Tibbets
Posts: 578
Joined: Thu Apr 17, 2008 5:29 am
Real name:

Re: Vacuum Pump Ideas

Post by Dan Tibbets » Sun Sep 21, 2014 5:09 am

Vacuum pumping is a complex interaction. If you are only reaching ~ 700 Microns, first consider leakage. Valve off the pump, if your pressure then climbs rapidly, leakage is a problem and no amount5 of pumping will be satisfactory. What is rapid leakage? That is a vague value, but if pressure increases more than 20-30 Microns per second without active pumping, then major leakage and or outgassing is occurring. Good systems seem to leak at only small fractions of a Micron per second. My best is ~0.3 Microns/ second without a glow discharge. Out gassing will improve with prolonged pump down, and/ or repeated pumping without exposure to outside gas.There are methods for checking for major leaks, mostly touch, listening, and using some liquid to challenge a seal- alcohol wiped on a suspected leak site may lead to a change in the plasma color or pressure change rate. I have actually seen bubbles when a hose clamp was not tight enough and alcohol was swabbed on the join.

As mentioned both diffusion and Turbo molecular pumps work well a pressures below ~ 100 Microns. I have researched this and starting a diffusion pump at 200-300 Microns may be tolerable and increase overall pumping efficiency, though to a lesser degree. ~ 500 Microns seems to be a never to exceed pressure for heated oil in a Diffusion pump. Not only does it not work well, but the oil will react more rapidly with the proportion of oxygen in the chamber and basically fry the oil (turns to gunk). I don't know if using an inert gas prefill will modify this significantly. A Turbo molecular pump doesn't work well at these relatively high pressures either and it might be damaged by the roters chopping through the relatively dense gas (abrassive effects and overheating and possibly stress on the bearings).

First step after satisfying yourself that leaks are not a problem is to replace the oil in the vane pump. This may help a lot and the oil is not expensive, as compared to the liquid gold used in diffusion pumps.

Dan Tibbets

RealBorg
Posts: 27
Joined: Thu Mar 01, 2012 8:40 am
Real name: Thomas Zehetbauer
Contact:

Re: Vacuum Pump Ideas

Post by RealBorg » Thu Oct 16, 2014 12:48 pm

Leakage seems to be down to a few tens of microns per day, I guess it is just the cheap single stage rotary vane vacuum pump which cannot go down further.

I remember operating ultra centrifuges about 20 years ago where we switched on the diffusion pump or it was turned on automatically "below 1000 microns". These ultra centrifuges used to be a big vacuum camber with a big opening, provide excellent shielding (withstand rotor exploding at 100krpm ~ 250kg Dynamite) and reached a final pressure of or below 1 micron. I'd love to turn one of these into a fusor ;-)

A simplified turbo machanical pump could work well at these pressures. Instead of the fine blades one would use a rotor with a smooth surface and multiple stages where remaining molecules get momentum toward the exit from a rotating surface. Schematic:

Code: Select all

   Chamber
---          --- Stator
 ------------  Rotor

---          ---
 ------------

---          ---
 ------------

---          ---
   Pump
Tom

Doug Browning
Posts: 156
Joined: Sun Mar 11, 2012 2:19 am
Real name: Doug Browning

Re: Vacuum Pump Ideas

Post by Doug Browning » Thu Oct 16, 2014 2:25 pm

That's fairly close to what is called a "molecular drag" pump. Typically used after the turbo pump, but before the mechanical pump these days. Maybe want to Wikipedia it.

Here is one of the 2 stage, air cooled, diffusion pumps from a Beckman ultra-centrifuge (scroll down to see):
http://www.ebay.com/itm/Vacuum-Diffusio ... 7675.l2557

Post Reply