Several areas of greatest concern / challenge come to my mind:
Isolation of the extreme heat generated in the fire box (
feed tube, burn chamber, and fire riser; which I am collectively calling the "fire box" so I don't have to type all that out).
Controlling heat transmission to ceiling above barrel / drum.
Proper draft from the exhaust/chimney.
Heat Isolation of Fire Box
I have a test build going right now. I am not yet certain what I think of the heat isolation, and I monitor it very closely. In fact I bought a digital thermometer that accepts k-type plugs and high temperature probes to go along with it (these are quite affordable: the digital reader/display itself was only $20 or so, and the thermocouples range in price from a few dollars on up; I believe the 700 F cost about $5 and the 1400 F cost about $35).
I am testing this above a wooden floor. I used 3.5 inches of perlite, stabilized with fire clay, as the insulation under the fire box. I am measuring the temperature about 3-inches under the spot where the coals get piled up, just inside the entrance of the burn chamber. From my outdoor tests, I know this area reaches it's maximum temperatures several hours after the fire burns out. (Also, a close temperature relationship exists, between this maximum and the maximum measured during the burn, immediately behind the fire brick on the far wall, where the burn chamber and fire riser make their 90-degree turn - this is where the flames hit against the back brick, so it makes sense it gets hottest burning the active burn; likewise, it makes sense that the area under the coals gets hottest some time after the coals have sat there, completing their burn cycle).
So anyway, so far I have measured as high as 170 F to 175 F 3-inches below where the coals are piled up to finish the full burn cycle. And that happens about 3 or 4 hours after the flames die off.
I also note that it takes longer than 24-hours for this area (170-175 F) to return to room temperature (which in my case is on the order of 55 F). Thus my concern that if I burn twice a day, I may over-heat this area, which I do not really want to rise above 200 F. I also have noted, that from the point where I decide I no longer want to add
wood, and just let what is in the feed tube burn up, I see about an additional 40 F rise in this area 3-inches under the coals (in the perlite-clay insulating mixture). These numbers and rates of change I would expect to vary from one built to the next, but I would also expect there to be some normal range of temperatures observed, as the heat from the fire box moves into the insulating material surrounding it.
Thus, my suggestion would be to use 6-inches of perlite-clay insulation under the fire box -and a foot or so beyond this to all sides- so as to maintain safer temperature build up.
Below this, of
course, there is an air gap of a minimum of 1-inch (mine is about 1.5-inches), because any space of 1-inch or more readily allows air movement, which in turn carries away any heat transferred to the surface exposed to the open air space.
I need to keep weight low due to the floor, so I did not build the decking and support with masonry materials, but rather 2x6s. This is my concern and attention to monitoring the heat so closely. 170 F seems to be quite safe for stud walls, floors, ceilings, etc. So that would be my ideal high temperature limit. But I'm willing to run that up as high as 200 F, with the information I've seen so far (I'm open to changing my mind on this point).
Just so you have the build details, which I am not recommending, I'm just saying this is what I have done (and plan to change, increasing to the aforementioned six inches of perlite-clay):
floor
deck of 2x6s laid
side by side
2x4 laid across floor deck, at 90-degrees, to form air space channels (roughly spaced 12 to 18 inches apart, as best I recall)
another deck of 2x6s
layer of aluminum foil, shinny side upwards (facing the heat of the fire box; also facing the inside and top of the 2x4s in the next step)
2x4 screwed on top of deck (to help hold the fresh, damp perlite-clay mix, and give a level guide; a simple "fix" would have been to use 2x6 or 1x6 for these sides, and fill that with 5.5 inches of the perlite-clay insulation mixture)
Then I filled this in with the perlite-clay mix.
Then I double checked the level, paying extra attention to the area upon which the fire box would sit.
Then a layer of aluminum foil below where the fire box is to be built.
For a test, I then laid down three 12x12-inch ceramic tiles, 1/4" thick. I'm curious if they are going to shatter. They
should be able to take the heat, but I'm not sure they can take the rate of change.
Then another layer of aluminum foil, under the fire box area.
Then about 1/2-inch of perlite clay, mostly for leveling the fire brick floor (I could have used sand too; but the perlite clay mix was lighter in weight).
Then a full thickness fire brick floor, and build the rest of the fire box on that.
So that's what I've done.
I'd suggest two potential changes. I'm considering both modifications, but since mine is built, I will closely monitor the heat building up under the fire box and decide if I need to do this now, on this build, or if I can wait to make the modifications on the permanent build I plan next summer/fall.
One I've already mentioned, which is increasing the perlite clay insulation under the fire box from 3.5-inches to 6-inches.
The other, is inserting pipe through the perlite-clay insultion layer, where it meets the wooden deck upon which it rests. These would be hollow tubes, and sized so that I could use a dowel or electrical grounding rod (or other properly sized stiff rod) to clear out any perlite-clay that may get into the hollow pipe as I drive it through the perlite clay (temporarily capping the pipe might be a smart idea, heheh).
Once the pipe has been inserted between the perlite-clay and wooden deck, I'd make sure it was uncapped and clear, so that air would move through it. To encourage this, one end could be angled down toward the floor and the other end angled upwards toward the ceiling, although I'm not certain that would be required, I suspect it would encourage air movement. This would carry away heat build up in the insulation.
The other observation I would make about the idea of adding pipes under the fire box to carry away built up heat, is that if they are located a minimum of 3 or 4 inches away from the fire box, I think the temperatures there will be low enough, and reach that depth in the insulation late enough, that removing heat build up at this depth will *not* inhibit the needed high temperatures inside the fire box. One might even be able to locate them a little closer, shallower, but that would require some more measuring of temperature movement through the insulation mixture to determine.