Back a while ago we were talking about methods of butt welding and the idea of using a gap came up, of cooling came up, hammer welding, there are many different methods to join two pieces of sheetmetal and most were mentioned.
I got to thinking how could you really test them, head to head? Without some sort of measuring method, how could you honestly say one is better than another? Sure you could feel it and know that one is better than another but how could you show the difference. I came up with this idea, to get a piece of 18 gauge mild steel bent into a 90 degree angle, cut a slice into it and butt weld it. Then check it with a straight edge and see how much your weld altered it.
Hmmm, this could work. It worked pretty well on this one little example.
I drilled a hole right at the bend in the metal so I could cut all the say to that hole at the bend. Using tin snips to cut the perfect tight butt weld that I believe in couldn't be done without putting that hole so I just put the hole in both of them so they would be the same. The other piece I cut with a 1/32" cut off wheel.
Both pieces were perfectly straight when I started, checked with an aluminum yard stick. I cleaned the metal on both sides as well so they would be exactly the same outside of the cut method.
I did a nice hot tack on my butt welded piece then let it cool naturally, no cooling what so ever with water or air or anything. Just a nice fast tack with the welder set hotter than recommended for sheetmetal. My thoughts are if you weld it fast you are creating less heat because you don't have to heat the metal with the weld to get it to penetrate. WHAM it's done this way, very little heat.
The other one I welded short tacks then cooled the weld with water. I did let it set between welds so it wasn't like a bead, just tack, cool then set it down and to the other one and even walk away for a bit. There was no rush, just weld and cool and let it set for a while.
You can barely see the results (it was hard to capture it in a photo) but the one with the tight gap and letting the welds cool naturally has zero shrinkage by the test. The piece remained perfectly straight. The other piece is noticeably bent with the yardstick resting on the center there is about an 1/8" inch gap on both sides!
Now, this is far from a scientific test, but I kept everything as close to perfect as I could with the metal being cut from the same piece and exactly the same length. I planned on doing a bunch of them with cooling and a gap and not cooling and a gap and so on but I only had the two pieces to use and thought I would do the both extremes and leave it at that for now.