I have an identical problem, having been rinding the bike without seat clamp for more than two years now.
I suppose you have already read something about it, if not I recommend Sheldon Brown on the subject. But I must warn you that the ammonia theory didn't work for me, and I tried ammonia on some aluminum sample parts without any visible effect (corrosion, oxide removal, whatever).
The problems are:
- Aluminum oxide occupies more space than pure aluminum. This, oxidation of aluminum posts create EXTRA VOLUME inside the seat tube, and this can create huge forces;
- Aluminum and steel create a GALVANIC PAIR of metals, inducing electrolytic reaction between them, with dire consequences as you have already noticed.
I believe the solution, failing the chemical approach, must be oriented towards two possibilities:
- Dilation, a drastic one, but TO THE COLD! Since aluminum dilates more than steel, you should FREEZE the seatpost, by pouring some very very cold substance inside. Liquid nitrogen (frostbite danger), dry ice (perhaps not cold enough) and CO2 cartridges (doesn't remain inside for too long) are good candidates. Perhaps if you wrap a towel around the seat tube, pour boiling water on it, then pour the freezing stuff of your preference inside the seat tube, you'll crack the bond indeed, as you planned.
- Total mechanical disintegration. You'd look for some heavy machining tools (lathes and all) and eat the post from inside. Some damage to the inner surface of the seat tube is very probable.
I don't believe too much in brute force if the seatpost is frozen inside for too long (more than three months since last "movement"), but skilled shop owners have been asuring me that it is possible (I didn't go all the way with this for now because fortunately the current seat height is working and I don't want to risk being too many days without the bike).
Hope this helps, send some feedback if you get some result!