Fiction welding of sphere halves

[Nanos]

I was wondering whether fiction welding the vacuum chamber sphere halves together would be doable in a DIY setting, thoughts anyone ?

[Starfire]

Friction welding of large spheres would require a large machine, but it will certainly work if you have access to such a machine. Securing the hemisphere on the mandrel for spin-up will present problems.

[Nanos]

I wonder if using concrete to cast both outside and inside (held in place with a bolt going right through the middle of the concrete and sphere to hold it down, with the outside concrete to act as both mass for the flywheel aspect and to keep the shape intact.

At a guess, a small motor (perhaps from a washing machine maybe.) could spin it up to speed before pushing the two halves together.

Though I've no real idea how much speed, or flywheel mass, or pushing force required, except through trial and error.

[Starfire]

Friction welders are large and complex machines. They work by spinning up one part of the work piece attached to a mandrel to about 5000 RPM and then bringing the rotating part into direct contact with a stationary second part and applying considerable force by hydraulic ram to create sufficient heat to force a weld as the rotating part grinds to a halt. The work parts turn white hot at the point of contact. These machine’s are used in mass production lines and often where two dis-similar metals have to be welded. They are invariably used for parts that can be spun-up easily.

I don’t think this is the best route for welding SS hemisphere’s or flanges. Argon arc TIG welding is much simpler and a lot cheaper. It would be very difficult to make a rig to friction weld just a few parts.

[Nanos]

I'm less concerned with it being the best approach, only an approach that has not yet been tried on a DIY basis, and at first glance, appears rather a cheap method material wise.

At a guess, the RPM required might go down as the mass of the flywheel goes up. Though it interests me what kind of RPM could reinforced concrete manage before it became dangerious, but going by how its used in washing machines to around 1,500 RPM it gives me some hope that a larger model could perhaps be used. (Though I have realisation that the larger the diameter of the flywheel, the faster its going at the edges, that is correct isn't it ?)

I had pondered one might use the weight of the flywheel itself and allow gravity to provide the force required to push the two sphere halves together, though again I don't know how much force is actually needed. At a guess, because the points your trying to heat up and melt are thin, of a wide diameter, you may not require much in the way of speed or force compared with say trying to friction weld two solid rods.

I hadn't thought about how to fix the flanges yet still reading up about here, though I have my stainless steel half spheres courtesy of Ikea.

[Richard Hull]

Very interesting.

I must admit, if I were forced to imagine the most improbable and most nearly impossible way for any amatuer to join two anythings in metal to a vacuum tight fit, friction welding would come to mind first.

I know it will not be done to a vacuum seal level, of course, but all the best of luck.

Richard Hull

[Nanos]

What reasons do you see for it not being possible to do it to a vacum seal level ?

At a guess, I imagine perhaps because the material is too thin ?

[Richard Hull]

It is certainly possible to do it to a vacuum seal level with professional gear worked in professional hands. The amatuer attempting this is at a great loss and will suffer some great expense to do this and ultimately surely have to have leaks sealed by TIG welding.

I truly suggest doing it and reporting back on the effort once you attempt to pull your first vaccum.

Richard Hull

[Nanos]

I wonder if Friction Stir Welding might be slightly easier to attempt, thoughts anyone ?

http://www.msm.cam.ac.uk/phase-trans/2003/FSW/aaa.html

Otherwise, the silver soldering/brazing method I read about you mentioning in another thread appeared the most promising, though I'm not quite clear yet if one needs more than a blowtorch to manage that.

I am keen though to try and push the envelope a little in my design if I can, and want to try and do all the work I can myself at very low cost, but also to learn the nuts and bolts of how it all works, so prepaired to spend man hours, rather than greenbacks to gain an education. But at the same time, quite prepaired to go with what is tried and tested after having explored other possibilities.

[Richard Hull]

Blow torch!!? No Way! I haven't seen one of those in years! I assume you mean propane or MAPP gas torch. They won't work.

To silver solder (SILVER BRAZE) you will have to have an oxy-acetylene setup. Brazing is best done on small components or very thin material. Too many linear inches and too much warp strain in a seam for a silver solder seal on a pair of normal, fusor sized, SS hemispheres.

Richard Hull

[Donald McKinley]

Nanos,

You might want to think in terms of the experiments that you want to try when you get to actual building. I'll wager you'll want to dismantle your equipment and try different grid configurations inside or just repair potential damage which often occurs. Your idea to weld or braze will require you to cut open your fusor which will take longer and make more headaches for you than it solves.

If you are good mechanically, there are a thousand things you can do to economize on the project. I would recommend that you use your creativity to design a very fast closure on your fusor chamber and a good bakeout setup possibly with integrated medium watt density heaters. If you do this you have the potential to try out lots more design changes than the regular bolted fusor design.

PS Oxygen and City gas gets hot as H8@!% & is probably cheaper for your brazing. Use a small victor torch made for oxygen service.

Don

[Nanos]

I wonder if its still the case that you need a licence to own an oxy-acetylene setup in the UK, as I remember in my dads day some years ago him telling me that was the case.

[Nanos]

Yes I had been wondering that it would be best to be able to open it up again easily, from ploughing through the messages here that does indeed look like a sensible approach to take.

Any thoughts on a quick release closure ?

(I'd assume the sphere halves would each have to be most probably welded to something else, though not necessarly..)

(Ponders compression system..)

Are there reasons why one cannot have the entire chamber in a water tank that is pressurised (perhaps frozen?) to help push the two sphere halves together ?

Or just a big press (be it screw or hydrolic.) to push the two halves together ?

[Starfire]

Why try to re-invent the wheel? - the standard flange's are designed for vacuum seal and easy assembly or dismantling, be they conflat, ISO or CF/KF

There is much wisdom to be had by reading the threads in the Vacuum Techonolgy forum.

[Donald McKinley]

>Any thoughts on a quick release closure ?

I have previously published on this topic.

viewtopic.php?f=6&t=2683#p16182

In my opinion, most of our builders here are sacrificing a lot in the area of convenience in implementing experiments because there usually has not been sufficient effort invested in open and close-ability. On the other hand they have built fusors and I haven't yet.

DM

[Chris Trent]

I hate to say it guys, BUT...

While I admire the innovative approaches to vacuum vessel construction, it appears that it would just be easier and faster to deliver pizza's for a few nights and just pay for the TiG welding and some standard flanges..

Yes, I know I won't learn as much delivering Pizza's, but I will probably have more of my own valuable time left in the end.

While it took me a while to learn to listen to them, the old hands around here usually do what they do the way they do it for a reason. (Sometimes they'll even tell us why.

[Richard Hull]

Chris is right......

Sage wisdom often comes late, but it comes. Such wisdom is rarely cheap but can cost the pocket book and, or, the spirit via untold aggravations and failures that may not cost money but cost time and effort to be cast aside as waste. Something must be billed for such wisdom.

It is tough to experiment in the truest sense with a sealed fusor chamber. If one is not a true experimenter, but instead, a single pass, "git 'er done" builder, then a locked down, spherical ball of most any metal with ports added via brazing is just fine.

In the end, a highly versatile fusioning fusor that is meant for experimenting, must be a device with two, 304 stainless steel hemispheres with vacuum rated, seperable, SS flanges which are attached via the TIG welding process as are all view ports, gas inlets, and vacuum ports. These ports are welded from the inside, only. This may not save money, but it will get you there with the least time and avoid a myriad of headaches associated with vacuum and mechanical, chamber body failures.

Few of us are into building ships in a bottle.

Richard Hull

[Nanos]

Why welding from the inside only ?

Whilst I'm at it, why not MIG welding if one hasn't a TIG one for example ?

[Chris Trent]

I think I finally figured this out myself so I'll venture an answer and someone else can tell me if I'm right, or just full of it.

The welding from the inside prevents cavities from being formed by the joints inside the fusor that can trap gas and act as annoying virtual leaks. Ok, it doesn't prevent the cavities it just keeps them from being on the vacuum side.

MIG welding generally creates more porous welds than TIG welding. This can cause both virtual leaks and some subtle real ones. In High Vacuum weld quality counts. Compare a good MIG weld with a Vacuum quality TIG weld and the difference is readily apparent.

[Nanos]

My next question would be, what kind of spec should I go for on a TIG welder ?

Baring in mind I would rather buy something I can use later on (I plan to build my own home, so something dual purpose is always good.) any reason to choose a particular brand over any other, or just go for a cheap model ?

I'm in UK, a quick look on Ebay I notice a new one can be had from £200+ (At the moment my budget range is more like £200) and I could take my time learning how to use it. (I had wanted to do a welding course but there simply are no courses in London!)

[Nanos]

I was wondering, in the very old days everything was bolts or rivets, could one rather than welding the flanges to the sphere halves, instead bolt them on though the sphere edges ? using perhaps coach bolts with added washers shaped as to contour the shape of the sphere and prevent the bolt head from digging its edges into the thin spheres.

[Richard Hull]

Chris is correct on all points. MIG welding is fine if you have the exact SS wire type for the work loaded in the welder. It is just not used in Vacuum system welding, normally.

Bolting sphere halves is just not done, mainly due to the massive linear seal lengths involved. Conflat seals are just as large, but the knife edges and gaskets are designed for a flawless bolt down seal against an inflexible 1" or thicker Stainless flange.

The key TIG welder buy point is the current it supports. I bought one that will handle 150 amps, but most work rarely needs over 100 amps.

Richard Hull