I checked my Statics and Strength of Materials to see if there was a formula for bending sheet metal around a curve. The best that was there were the parabolic deflection beam formulas. I don't think they apply. However, with a distributed load on a straight beam, the apex of the parabola does have a computable radius after the material gets bent.
So let's say the Young's modulus of your stainless is known. Use the Young's modulus calculator, the bend radius, and the sheet thickness in:
to get the force required to hold the material in a 16 inch diameter circle. Apply an adjustable strap, slowly working up to that much tension in the strap, more or less, and your desired bend radius. The strap should be right down the middle. Use one of those endless hose clamps, available in lengths of up to 10 feet, if you don't have a two ladder lock buckles and a 5 foot length of nylon webbing. Use your judgement.
Then apply something a little stronger and resistant to heat, like two pieces of 3 mil aluminum tape 2 inches wide, and at least a foot long, along the edges, over the joint. The extra length is so the adhesive at the ends remains strong while the brazing or welding is done. Remove the center strap, if you used nylon, or leave the endless hose clamp in place. Either way, your choice.
Then, tag weld the seam, be it lapped or butted, one half the way from the center to the edge. Two weld nuggets about 1/8 inch diameter should do. I don't weld much at all. I am probably not as good a welder as you.
If having the seam lapped and then having the upper layer stubbornly remain, say, 1/16 inch from the lower, preventing a sound weld, is a problem, then clamps must be applied to a bar over the seam, leaving the weld area free, and using a more or less identical bar diametrically opposite. For instance, I'd think two pieces of 1x1 inch steel 18 inches long with a 1/4 inch hole through at each end 3/4 inch from the end would do, in combination with two pieces of
1/4-20 threaded rod about 18 inches long, and four matching nuts.
Once two tag welds are made, you can make several more to hold that edge down if it is a lap joint, or a few more if it is a butt joing, and then grind them all smooth. After that, my choice would be to braze rather than weld. Handy & Harman will provide a small sample of stainless compatible braze filler, about
3/16 inch diameter, which is an excellent color match and very strong. You'd use a hell of a lot of flux for stainless and try to work quickly, because the chromium in the steel becomes chromium oxide, a ceramic, at brazing temperatures, and even a large amount of, say Anti-Borax brand flux can only dissolve so much chromium oxide before turning black and forming an insoluble glass. This is an indication that the flux has been "burned". The material is very tough.
If you finish the joint and the flux is insoluble, you can add lots more flux and reheat the work to dissolve the chromium oxide. Then, soaking the seam in a tub of hot water overnight will dissolve all the flux. If necessary, set up a tray with vinegar and water or a solution of hydrochloric acid. This will dissolve the remaining flux. Be careful with HCl, you already know it's dangerous shit.
Now to finish this, you can use a sanding belt to smooth the braze and remove oxides from the metal that got hot but wasn't protected by the flux, and you're done.
I switched to decaf all day yesterday, at well, did a little work, and felt OK. I'd still like to know if when you read my first post in reply to your original post, whether you got the impression I had had too much coffee?
Doug Goncz (at aol dot com) Replikon Research
Replikon Research researches replikons, which are self-reproducing configurations of non-living matter in environments that support replication, analogous to organisms living in ecologies.