I've used my acetylene torch on many bolts. Note, this is a plumbing acetylene torch, not a welding oxy-acetylene torch, doesn't get as hot.
My normal procedure is to heat the aluminum where the bolt goes in. Heat both parts that the bolt is holding together, and heat it as evenly as possible. With an acetylene torch, it will take quite a while because the aluminum spreads the heat very well. Next, I'll put the socket on the bolt and TIGHTEN it. You only want to give it just enough to turn the bolt head a bit (not even a 1/16th of a turn). Then go back and LOOSEN the bolt the same amount. Alternate tightening and loosening 4 or 5 times. There are a few indicators that may let you know if you're successful in breaking the bolt free:
1) If you get a solid cracking sound and a bit of movement, it's usually broken free and able to be loosened.
2) If you hear a squeaking or squealing sound, you should be able to loosen the bolt with some resistance. If it gets harder as you are removing the bolt, reapply heat and keep going.
3) Many times I'll get a puff of dust out around the bolt head. I've never had this happen once when the bolt has remained stuck. If it puffs smoke, you're good to go.
If none of these happen, I'll try to go back and heat the aluminum again. If it doesn't work your 2nd or 3rd time, it's unlikely that the bolt is coming out easily.
Next steps are to remove the bolt head as Pappy described. Once the part is removed, there should be a stud left. I'll normally go with an oversized nut and weld it onto the stud (sometimes if the stud is fully threaded you can use a standard sized nut. This gives me a good way to get a socket back on the stud. Now, I'll heat the stud&nut with the acetylene torch until it's cherry red. The aluminum doesn't melt, because the corrosion between the bolt and aluminum provides insulation, and any heat to the aluminum gets drawn away pretty quickly. Again, I've never melted any aluminum doing this with a plumber's torch. With the bolt cherry red, I'll hit it with some penetrating lubricant to cool it. (It'll smoke like you wouldn't believe) The objective is to expand the bolt by heating it and then contract it by cooling it again. You're trying to break up the corrosion. I'll repeat that a couple of times too.
Next, I let the bolt cool down, and then re-apply heat to the aluminum, and repeat the original process above. Normally this will get it out.
to date, I've only had a handful of bolts that I couldn't get off with these steps (out of a few hundred), and usually these are in a location were it is difficult to get adequate heat to the aluminum around the bolt. I've had reasonable luck with these few cases drilling and adding a Helicoil afterward. No lost parts yet.
Good luck!