Carbon Steel

[Eldarion]

Would there be any problems with using carbon steel to build the body of a fusor? Is there anything different you would have to do while machining and welding it? It seems like it should work mostly the same, but I don't know for sure.

[Doug Coulter]

Starting from what are the ancient refs to this, the main objection would be its rather serious outgassing compared to what we're normally using. Otherwise the stuff has some good properties re fabricating and welding. It will just out gas like heck after you make it and try to use it. There exist references to coating the stuff with various things to reduce that, depending on whether hot ions are going to be hitting it or in general, whether it's going to get hot. Plain old iron oxide slowly ( eg challenges your patience) reduces to pure iron in our conditions. since we're blowing in H (or D), a reducing agent, this will mean water in the "vacuum" if you can tolerate that, pretty much every time you go for good base pressure. If you can handle that, and understand, no problem, else.....

Of course there are two reasons to use this (and some other iron alloys), one being fabrication ease, and the other cost. High temp strength isn't that great. you just have to work with what you're using.
Welds can be a source of serious contamination and weaknesses -- it's easy to make one that "looks perfect" that has voids, heat treat embrittlement etc. It takes some experience to tell if something like this works for you. I get away with it sometimes. Not always.

[John Futter]

Nelson

The biggest problem associated with carbon steel is if it gets magnetised.

If this happens your electron and ion trajectories will be modified by the field.

The inner surface can be modified by sputtering or evapouration to give it a better surface with regards to outgassing.
I have seen rather complex ultra high vac apparatus constructed out of ordinary steel and Nickel plated to protect the surfaces inside and out from rust.
FWIW

[Dan Tibbets]

If you haven't already looked the Belljar forum is very good for vacuum discussions. I recall some one rated stainless steel as having the best outgassing properties, followed by aluminum, then other iron/steel materials.

http://belljar.net/forum/index.php?board=1.0

A link from the Belljar forum lists various substances outgassing properties.

http://outgassing.nasa.gov/cgi/uncgi/se ... ectionb.sh

Dan Tibbets

[Chris Bradley]

Not sure that link works, Dan.

You might also be interested in seeing this link I came across the other day;

http://www.ee.ualberta.ca/~schmaus/vacf/outgas.html

[Doug Coulter]

I have done electroplating here (have a few grand invested in a large setup) and while with extreme (anal) preparation, have done things nice enough to put on say, an automobile and have it live, not good enough in general for vacuum uses. The sheer amount of pre preparation required utterly negates the material savings, and even nickel (which plates best of all the stuff I have, which is a long list) has issues when you start slamming hot charged particles into it, tiny voids and such that then outgas.
This is by the way, the reason having something plated is expensive. I've had people offer to do the prep themselves, but no one has ever done a job even close to what's required, and can't because on the trip from there to the tank -- it's contaminated again. It's amazing how much oil and other junk can get pretty deep into steel and come out just enough to mess things up.

Aluminum, whether deliberately anodized or not, has a honeybee comb surface structure -- lots of little deep hexagonal channels. And these not only hold gas, but also decompose back into aluminum + oxygen when slammed with ions. This one I've tested pretty thoroughly, and there's air coming out for a long time, even after some baking.

SS is pretty good, but of course, not very cheap. It's one of the easiest to TIG weld to very good quality, assuming the right tools.

Brass is pretty good if you don't heat it, else you get zinc which has a pretty high vapor pressure, this I've checked on the mass spectrometer. At room temp and below it's fine.

A constant cause of confusion here, is what constitutes "high vaccum" as sources vary where they draw that line. In operation, a fusor is running no where near anyone's definition of "high", and a base pressure of say, 1/100 of that should be just dandy in my opinion, somehwat backed up by having success with a diffusion pump system, no serious cold trap (water) and having that work just as well as my turbo system with a base pressure well below 1e-9 mbar. My gage, which reads high on D, says about 2.5e-2 mbar for comparison purposes when we are running on a sweet spot and the neutron are really coming out. I would guess if you thought you needed better than 99% purity, you'd have to have that kind of thing, which is hard to get even with the turbo system which like anything else outgasses for a long time after air exposure.

[Eldarion]

How would copper work if it was plated in that? I have the means to plate almost any metal with copper using electrolysis in water saturated with copper sulfate. I have enough copper sulfate to plate any amount of material I need, so I don't think that cost would be an issue. I think that it would hold the vacuum pretty well and damage should not be a problem, as I can plate it as thick as I want with enough time.

[Doug Coulter]

A good copper plate bath has a lot more than just copper sulfate (for acid copper, which, BTW won't go straight onto steel). There will be some sulfuric, but more important, a few ppm (30-50) of chloride ion and a generally proprietary leveling agent, reputed to be something like poly ethylene glycol of just the right molecular weight. Else you will just make loosely adherent dendrites etc, pretty but not too useful.

Caswell plating sells a kit for alkaline copper plating that will go straight on to steel without a flash of nickel first (like acid copper has to have) but it's not very cheap. I get a lot of things from them, and their book on plating is WAY worth the price, and they do stand behind their stuff. Here's their site:
http://www.caswellplating.com/

BTW that book used to have (and probably still does) some info on aluminum and the hexagonal structure it's surface gets when oxidized, eg within seconds of air exposure. There was a very pretty electron microscope shot there. For Ash's case, where I suspect he's working with something along the lines of semiconductor processing, this doesn't matter much -- they are running fairly high pressures and I bet usually always with the same gas, so if the walls pick up and release some, no problems. They tend to use things like "load locks" or airlocks and almost never bring a tank up with dirty shop air too.

All of these plates have considerable voids in them at a microscopic level, which in the case of copper, can be closed up with vigorous buffing, re preparation, then adding another layer -- repeat for a couple hundred hours, and you'd be there. Getting the waxy buffing compound off there is not simply a matter of wiping with acetone which does get all you can *see* but not good enough to accept the next plate layer. You might be able to put on a layer of copper, then gold or something ti fill the little voids, but hey, what I would do personally is just make the tank of good steel of some sort, and try it -- it might be fine. Chrome Moly may be cheap enough to use, and it's "nearly stainless", though it's tougher to work than plain old steel is.

I use McMaster Carr for metals, and their stainless isn't cheap, but compared to Lesker it is.
But! And this is important, only Lesker and guys like that have the exact right size to go into standard flanges....so it can be kind of a trick to use the less expensive stuff unless you machine your own something there -- flange, pipe end, whatever.

Copper is better than iron or a lot of other things for outgassing for sure, and thermally and electrically a lot better too.

I'd suggest you try it on a sample piece though, and try to wipe it back off, and check under a microscope at all stages and see what you see.

You cannot use an acid/sulfate Cu plating solution direct on iron. It dissolves iron, replacing it with Cu that doesn't stick well because the iron is dissolving out from under it. In other words, you've got some undesirable electrochemistry going on there before you even put the power on it.

I found the best way for me to go was just to buy a decent tig welder (Grizzly tools) learn how to use it (easy if it's a good one) then buy all the parts, even the expensive pipe from Lesker and go at it.
The things I built work fine and will never go bad, it's purely a one time cost and you can re coup that welding for people. It's those things that cost at a *rate* that kill you financially. A mere amount, no problem in the long run.

I have an aluminum tank I made -- it's never reached within a factor of ten the base pressure I get with an SS tank I made to replace it, which was bigger (6" dia vs 4"). Same everything else, on a good turbo and with baking to 150 c. Both pressures would probably have been fine for fusors, much lower than what we run at.

Clean iron need not be bad, but do bake it at like 400-500 deg F for a few hours before trying to pull vacuum, and then sand the rust off it or use oil free SS wool. It usually has oil pretty deep inside (grain boundaries). The more expensive metals are more uniform and cleaner that way, for some reason, I find and have less crud to them and fewer hard and soft spots.

[John Futter]

I thought that just water and copper sulphate gave a poor adhesion spongee type layer.

Copper sulphate and sulphuric acid is slightly better less porosity but bad thowing power.

commercially I'm sure they use copper cyanide which throws well with little or no porosity.

I'm sure Doug will correct me if i'm wrong

[Doug Coulter]

John, in my experience your are exactly right. Or nearly, cyanide is in "bad odor" in pro plating these days.

My commercial baths for acid copper are as mentioned above, and kind of "soapy" from whatever leveling agent they used -- trade secret darnit, else they couldn't charge $300 for what is *mostly* 5 bucks worth of CuSO4. That 30-50 ppm chloride content is key to it working, and kind of hard to hit by accident. Fails badly on either extreme of that range, it's really picky. They do use some sulfuric in that bath, but by golly, it will throw (badly) into the inside of a long skinny pipe and even clog it a foot into a pipe 3/8" id. How they got it to throw like that I dunno, but it throws out of the park in practice. Not that it sticks well inside the pipe, it flakes there. But in poundage it almost seems like more goes inside than outside, very strange.

In the other, alkaline, Cu bath, I think the complexing agent is ammonia or something like, I've almost duplicated that chemistry in testing here as it's good stuff and I'd love to have it cheap too. The leveling agent is past me, can't figure that out so far, maybe some grad student will spill the beans at some point and I'll find it that way (this turns out to be a good trick). A lot of this is like the chloride above -- works fantastic in a tiny range, or a tiny ratio range of two things, and hitting on it by just trying stuff is pretty unlikely. Neither bath will make good thick single coats, if you want thick you have to take it out and buff, re prep and go again and again.

CuSO4 alone stinks and makes pretty little coral trees that don't stick in all tests here, to anything. Any acid Cu solution won't go straight on iron, you have to flash it over with something else first, like nickel.

Else the iron displaces the copper in solution (autoplating) but in doing so, the iron is lost and the copper is left sticking to "nothing" and mostly just falls off. It just can't be made to work in my experience, and I've tried the obvious stuff, like stripping hard (reverse polarity) skipping that step, putting it into the tank with volts on already etc, nothing works.

None of my stuff is cyanide based, it does work very well, but not worth the risks, and (almost) nothing Caswell sells is, for that reason, Tort! Some fool would lick their fingers or something.
The whole plating business has gone away from it for that reason to the extent it's possible. Some still use it for silver, that's about it.

Hard chrome will go straight onto iron, I use it (rarely) for coating tools I make for lathe etc -- it's hard and slippery so they cut better and last longer. But that involves you with hexavalent chrome, about the nastiest stuff there is on earth (eg about as nice as Pu). It stains you, burns you chemically, poisons you chemically, then gives you cancer. And it takes such insane current density to plate at all that it foams hard and gets into the air as a mist. It's miserable as a process.
Avoid! Even Caswell will tell you that as they sell it to you reluctantly. They believe this hard enough that they sell cobalt plating solution as a replacement for that. The msds for Cr2O3 is quite evil to read.

Heck, I've done some plating that did live in a vacuum with ions, but it doesn't always fly, I'd just try the iron plain and do things that seemed necessary if they do later on.

You know, it might be good to coat that iron, but another way -- how about evaporation or sputtering?
Those work real nice in a decent tank, and stick well if the "Target" is very clean. Or evap a bunch of Ti and have it as getter too.
Al and Ti evap nice, Ti sputters well, Cu sputters well but slow, and all stick to tank walls very tightly when those are done. Those are the ones I've tried (some accidental) here so far. The Al I accidentally coated a bunch of my tank innards with (evap is like putting a bomb under a can of paint, you have to have shields as I know now) is resisting all but real serious effort to get it back off, I just gave up as it's not harming anything at this point, but did destroy a bunch of feedthrough type things until they were removed and acid or hydroxide etched to get the Al off, it really sticks well to almost anything (even plastic). I've made some nice optical mirrors with all 3 things, this kind of process is pretty easy and works pretty well, maybe that's the best idea of all. Certainly the cheapest.
I used the W wire coils from Lesker for evap, and rewound the secondary of a spet welder to get the current (lots of amps needed) at about 4-5v for a heater rated at about 8v, that makes it plenty hot for Al anyway. We have also taken the secondary off a MOT with a chisel and replaced that with #6 wire for a good low volt very high current source for that. We've done sputtering with very simple gear, no magnets or anything, just a target and some HV in argon. Works fine, that fancy stuff is for things like semiconductors and very controlled thickness and perfect source utilization etc -- not needed at all for this. I know, sputtering is actually hard to avoid sometimes, and it's a pain when the stuff goes to a bad place (like your HV feedthrough insulator).

[JohnCuthbert]

Am I missing something here? The data presented here
http://www.ee.ualberta.ca/~schmaus/vacf/outgas.html
kindly cited earlier give outgassing rates for mild steel of the order of 4E-7 to 1E-11
For stainless they give 4E-7 to 1E-13
Unless you are in a position to bake the whole system, under vacuum to 400C for ages there's not much difference between mild steel and stainless.
I have seen UHV systems with exposed copper and aluminium parts and they should outgas about as much as the mild steel, yet they work just fine.
We are not (as has been pointed out) trying to get UHV here.
Might it be better to concentrate the effort elsewhere?
(And, BTW, I suspect the only way to get an electroplated copper surface to stop outgassing like a baked bean fanatic the day after a real ale festival is to melt it.)

[Doug Coulter]

They probably specified "clean" iron, which isn't that bad -- and their numbers show it not being that bad. Other sources spec slightly rusted carbon steel perhaps 400 times worse. But -- hit it with some shop air, it's not clean anymore -- you got water and rust more or less within 10 seconds if my electroplating experience is anything to go by; you have to go fast from pickle through rinse to the plating tank or it just doesn't work, it's already contaminated again. That's usually the issue. Those numbers you are quoting are pretty extreme spreads, which is actually realistic. Metal history has a lot to do with it.

You could be right about the copper, I wouldn't trust it after reading Kohl's chapter on it -- gotta get it with zero phosphorous, no oxygen etc even when using it as a bulk in a vacuum tube.
Having said that, I use it a lot here (OFHC) and no special troubles, but my base pressure is limited by other stuff (some viton). I can definately tell right off when I put any not-special-treated substance in the tank, though, and am careful about that.

I agree, this isn't going to be anyone's main problem unless you let the steel get rusty, once it's been nicely baked once. Or magnetized, but small fields don't seem to hurt fusor much, or mine wouldn't work.

I think the sputter or evap idea would be fun, if it turned out to be needed, though.

[Chris Bradley]

For a fusor application, I think the concern might principally be that mild steel tends to suck up hydrogen (deuterium) less so than stainless. Actual outgassing of atmospheric content looks like it is much the same at non-UHV, otherwise. I say this from what I've read, not from what I know to be true.

[Quantum]

As I've previously stated, We use aluminium chambers at work.

We also use viton seals.

Vapour deposition runs at similar pressures to fusors.

Our chambers are basically just multi-layered sandwiches of ally and viton.

If you want to avoid welding, and have access to aluminium machining facilities, it's worth a thought.

[Dan Tibbets]

Link for a site with a number of vacuum topics

viewtopic.php?f=10&t=3344&hilit=http%3A ... htm#p21870

Dan Tibbets

[Quantum]

I think the ideal solution to overcome the drawbacks of carbon/alloy steel would be goldplating. (We use gold plated springs and other components in our chambers at work.)

This would also make an impressive looking chamber.