No amount of whacking, soaking or torching would budge my siezed Motobecane engine, but here is what worked: I put the head back on, filled the cylinder up with water, put the spark plug back in and took the torch to it and left it there to heat up the water. Knowing from thermodynamics that water at 1500 degrees and 1 m3/kg has a pressure in the range of megapascals, something in there is going to want to move. I hoped that it would either move the piston or force the water / steam in through the piston / cylinder seal where the siezure is. I ended up losing it through the head gasket seal and the decompression valve, but when I took it back apart and hammered it again, the piston moved, and I wasn't hititng nearly as hard as I had been earlier.
I tried soaking it in solvent overnight, but it didn't make a difference. I can see how it would break down corrosion in there but if there was cold welding between the piston and cylinder, it wouldn't be affected. I don't know if this was the case or not, but based on what the piston looks like now, I'm pretty sure it was corrosion of the aluminum piston. I don't see how torching and then whacking would help because the aluminum piston would want to expand more than the steel cylinder sleeve would with the same amount of heat because of its coefficient of expansion, so you would somehow have to get the cylinder a lot hotter than the piston. I do see how the uneven expansion would create stresses that would break down the bonds, so what I did the second time was heat it up really good and then quench it with water. The piston still wouldn't move, even while bringing the hammer down with both hands from over my head, as hard as I could. It just split the block of wood in half and sent one half of it flying.
For curiosity, I tried to find the pressure that would result from water heated to 2500 degrees farenheit while maintaining the same specific volume. My thermodynamic tables from school didn't go up that high, and nothing on the internet did either, and to interpolate from 500 degrees up to 2500 won't get you anywhere close to the accurate result. Does anybody know where I could find this information? I know there had to be a great deal of pressure in that cylinder to get the piston loose.