A: First, you’ll find a sterling base much easier to solder to. Sterling has a much higher thermal conductivity than does brass, so it’s much easier to heat it evenly, which helps greatly with the warping. Second, in soldering sterling to brass, be VERY careful not to overheat, and only use easy or extra easy solder. You make silver solder by mixing brass and silver, in essence, so the contact point between the silver and the brass will tend to dissolve the two metals into each other. The result is that very often, when soldering these two metals together, the one will unexpectedly just slump into the other, ruining everything. This can happen at later stages of assembly as well, even after the bezel is soldered, such as when you put findings on or whatever. Like I said, be careful.

Better than brass would be a commercial red bronze (which is actually still a red brass, being alloyed with zinc, not tin), or nu-gold, or the like. These high copper alloys, with less zinc, will be much less likely to give you these catastrophic accidents. I’m not, of course, saying you cannot solder brass and silver together; you can. It’s just you have to be very careful of the temperatures. Silver on jewelry scale items is fairly cheap. This silver, even with large bezels, should still only be a few dollars worth, and will result in a more valuable piece of jewelry. If you can afford it, I’d recommend using all silver. Obviously, you know your finances better than I, but still, what’s your time and effort worth?

To minimize the warping, either with brass or silver, follow these steps: After you’ve made, soldered together, and shaped your bezel, carefully anneal the bezel again, and anneal the base sheet as well. You will then probably have to slightly adjust the bezel again to get if flat and properly shaped, but will still now be stress-relieved, and won’t automatically distort on heating—and the same with the base sheet. Uneven heating will still distort either or both, however, causing the warping and buckling you noticed.

It can be internal stress in the sheet, if you’ve not annealed it and it’s stressed, but most of the time it’s simply that heating is uneven, what with the nature of torches and all, uneven thermal expansion from heating expands some areas more than others, thus causing buckling and warping. It’s not as likely to be difference between one side and the other of the sheet, unless you’re using very thick metal. With most sheet metal, the differences in temperature between the side the torch is hitting and the other will be not very great. Much greater are, for example, the differences between the outside edges and the center of the sheet, or one edge closest to the torch flame vs the far edge you didn’t realize you’re ignoring.

You can combat the warping of the base sheet by slightly doming it, just a hair. You do this by using a mallet on the center of the sheet over a slightly domed surface, or even a flat surface. Only the center is worked, until a very slight, only barely perceptible, dome shape is developed. The sheet is now not flat, but it is uniform. The convex surface is set on the upper side, and the bezel centered on the dome. Because both are symmetrical, the bezel will still fit, and the dome isn’t enough to bother the setting.

What the domed shape does is provide a preset direction for thermal expansion when you heat the sheet. While it will still expand, this time it won’t try to twist or warp so much. Instead, the dome will tend to either flatten or rise, depending on where the heat is applied, and the bezel will tend to still fit as this happens. On cooling, things relax again, and once you’ve trimmed the excess metal off the bezel, you usually can no longer can find that domed shape any more. While this technique is usually not required for bezels, being more commonly something people doing larger hollow construction, such as boxes and the like, have to contend with, it should work well for your large bezels as well.