Fusion Message Board

In this space, visitors are invited to post any comments, questions, or skeptical observations about Philo T. Farnsworth's contributions to the field of Nuclear Fusion research.

Subject: Re: Neutron counters and other parts.
Date: Apr 05, 5:16 pm
Poster: Pierce Nichols

On Apr 05, 5:16 pm, Pierce Nichols wrote:

>Nathan - as regards mechanical pumps the chief concerns are pumping speed, ultimate pressure and backstreaming of oil vapors. The pump speed is "probably" a bit less important than ultimate pressure.

Unless you are backing a diffusion pump, and then it makes a great deal of difference.

> If your roughing pump is one of the newer direct drive pumps and in good condition, getting to 1 micron or even below that on a tight clean system is not too difficult. The key words are tight and clean. Also a molecular sieve on the inlet side will help keep backstreaming of pump oil to a minimum.
>You should expect some backstreaming of roughing pumps below 100 microns.

Mechanical pumps are not the only roughing option. Although I have not built yet, my conceptual favorite is a pair of sorption pumps backed by an air aspirator (also called a venturi) vacuum pump. This system is almost 100% clean, but has a few disadvantages. First, you need four high vacuum valves instead of one. Second, you need LN2 to cool the sorption pumps. Third, they require a bit more care to operate -- you can pretty much just turn on a mechanical and leave it to run. Fourth, you need a source of compressed air (but good compressors can be rented inexpensively or borrowed).

>However, along the lines of the "Deceleration of ions.." discussions of a few days ago, you really want to minimize collisions between the ions you make and neutral gas molecules. this means the longest mean free path (MFP) possible.. which means the lowest pressure possible.. before D2 is put in the fusor. Since MFP is inversely proportional to absolute pressure... you need to go substantially below roughing pump vacuum to get the innards clean and gas free. Ion pumps are a bit pricey but very nice, turbos are great, but also pricey. Personally I like the good old diffusion pumps. They are relatively cheap...and with an LN2 cold trap, quite acceptable...Oil backstreaming from long term use can be an issue. But they are bulletproof and the price is right. The ideal diff. pump would be air cooled. Water cooling usually requires 1/2 to 1 gpm... which if you can avoid.. is one less expense.

I mostly agree with you, except for a couple of serious caveats. Diffusion pumps require a lot of care to run unless you want to coat the entire inside of your vaccuum system in oil. LN2 traps are *NOT CHEAP*, although they can be constructed. If the power goes out or your backing pump fails and you don't have a gate valve between the diff pump and chamber that closes instantly you are going to be scrubbing oil out of your vacuum chamber for a long time, and you may wreck vacuum instrumentation.

My personal favorites are ion pumps and an exotic critter called a non-evaporative getter (NEG). I have seen ion pumps relatively inexpensive -- ~$300 for a surplus pump in the 30 l/s range, with magnets, no power supply, but the power supplies are easy for anyone with electronics clue. Also, that sorption pump combo will pull enough vacuum to start it -- no roughing/backing pump!! NEGs are a little less known, but extremely cool. You heat the block of NEG material up to an activation temp, whereupon it suck up stray gas at a decent clip, especially hydrogen, which is hard for everything else to pump. It's nice for a fusor b/c it allows you to control the reactant pressure faster b/c excess reactant gets sucked up faster. Check www.saesgetters.com for more info.

Note that my suggestions are for a slightly higher end system. I am waiting to start until I have the money to build a system capable of pushing the boundary. I suggest getting a book called "A User's Guide to Vacuum Technology" for more info on vacuum system design and options.