You may be looking at the difference between thermal mass and insulation, not the steel as such.
There are insulative castable materials you can use to make a heat riser that will handle the high temperatures required for complete combustion.
We have seen a lot of steel and sheet metal warp or burn out over time in both fireboxes and heat risers. The temperatures required for clean combustion are in the 1000 to 2000 F range. With insulating materials, we have seen perlite melt out which indicates it's hotter than 2300 to 2400 F. It is very easy to get these fireboxes hotter than the max. performance specs for mild steel, or even to reach forge-welding or smelting temperatures with very good insulation.
Incidentally, even with good steel and thick walls, all-metal firebox interiors don't work well for a J-type firebox design. They will conduct heat to where it's not wanted, making it hard to keep the system drafting in the right direction. And they tend to burn out, warp, or crack over time. Even if there's
enough steel there to last a while, the process of warping and corroding can release chunky sheets of oxidized rust, which will change and possibly block your flame path.
The attached picture is a thin galvanized steel duct piece that was deliberately used as sacrificial formwork, but we didn't anticipate the amount of warping it would do as it burned out. A messy process, and unnecessary when you can get
cardboard tubes for formwork that burn out cleanly. We've also seen severe warping in triple-wall metalbestos pipe (in one prototype that we ran for only a few sessions, the inner wall expanded and crumpled inward, blocking part of the flame channel and allowing the flames instead to access the less-durable insulation between the walls; most of the insulation that
should have been in the damaged area was gone by the time we pulled it apart to inspect the damage.)
The thicker steel tends to do more flaking and corroding before it is actually gone, but it can also crack surrounding masonry parts with its thermal expansion, or even "walk" due to repeated expanding and contracting cycles with each fire. You can solve each of these problems individually, except for the problem of the metal itself oxidizing and wasting away.
If you want better performance and durability, I'd look at insulating kiln brick (very light weight, like it's made of foam, you can cut it with a
wood saw), or some of the castable refractories with high insulation values, or ceramic-fiber refractory boards and blankets.
We've probably added to the confusion on this point over the years by using spare stovepipe in our
workshop demonstrations, for a "quickie" demonstration of how the heat riser changes the draft of the firebox compared to no heat riser. This is an easy thing to pull out of the scrap
yard for a bonfire demonstration, but we never meant to imply they were the best material for long-term use.
For durable heat risers, we tend to favor half-width firebrick plus 1" of refractory insulation blanket, or 2" of perlite, caged with wire mesh or sheet metal
on the outside. These versions do take a little longer to heat up, but also retain heat for safer cool-down in the dying-embers phase, and make it easier starting the next fire 12 to 20 hours later.
For the quickest-heating fireboxes, we have had good results with ceramic-fiber castables, and Matt Walker has shown good results with some other insulating castable refractories. These are not quite as durable; the one-piece castings tend to crack, so an experienced designer tends to think in terms of either multi-part castings for expansion joint stress relief, and/or build a non-combustible surround that will stabilize the firebox and prevent air leaks after a few cracks start to happen.
The insulating kiln bricks seem like the best of both worlds, but they are still a little more fragile than the dense bricks, so tend to wear sooner in the
feed area.
Yours,
Erica W