Advice wanted re bell design for use instead of a mass

I am considering replacing the mass of my RMH with a masonry "bell" or "bells". I've read lots of articles on the construction of a series of "bells" which allow the hottest flue gases to rise and shed their heat to the walls of the bell and thus only the coolest gases fall to the bottom of the bell to exit to the next bell. Thus only the coolest gases are drawn off to the chimney flue.

Most designs show a series of 2 or three bells arranged either horizontally (beside each other) or vertically (one on top of the other).

My question is this....are there advantages to either of these over simply constructing one larger chamber where all the hot flue gases can rise, shed their heat evenly through the whole space at the top of the one chamber and fall gradually as they cool and exit the coolest gases through the chimney flue at the very bottom of that one chamber?   My "gut feeling" would be that a single large chamber would likely shed and store more heat as the entire top area would be equally hot. The gradation of temps would be from top to bottom of that one chamber rather than from one hot chamber to successively cooler ones. Has anyone here actually built and tested a single large chamber in contrast to multiple smaller chambers which run at successively cooler temps?   Basically I'm asking, if one used the same amount of masonry material in one, two or three chamber systems, or constructed a one chamber system equal in volume to a two or three chamber system... which would be most efficient  i.e. store the most heat and shed the coolest exhaust to the chimney?

I'd rather not try to "re-invent the wheel" if others have already done the leg work and have the answers to the above questions!

What have YOU learned?

BTW, my RMH already has an integral mass of sand around the barrel and the barrel exit temps never get above 475F. So my proposed bell doesn't need to harvest all the heat, only the remaining heat. This means that I don't need refractory brick or mortar as  normal brick and mortar (I've read) will withstand temps of 600-800F. So I'm thinking of building of brick and capped with concrete patio slabs. Maybe even including some large  rocks  and steel pails of water within the bell to absorb and store heat.  No need to worry, the water won't boil!  I can't even get a pan of water on the top of my barrel to boil when the barrel top runs at 550 or 600F!  Any suggestions re 1,2 or 3 chamber designs?

Bruce, the Kuznetsov type with bell above bell

http://www.stove.ru/index.php?lng=1&rs=16

are meant to store heat evenly at all levels into the room. Because heat is radiated from the bell, it is better to do so at human height, than below the ceiling.

More than two bell, in my opinion, is to be kept for very big systems. And a barrel, should be followed only by a bell, again, according to my own highly valuable opinion

What you need to know, is the heat you have left after the covered barrel. To assume what you need in term of bell.

Or do it by ISA.   For example, a 6 batch can cope with 5 to 6m² of ISA. (internal surface area, which extracts the heat)

From the few experiments with J tubes, that some guys have done. You can say 4m² for a 6"J  (that's about 40 sqft)

I think 6m² has been done with an 8 J This is the total of the 2 bells.

In your case, you have a barrel at first, so this is 1.89m² to retract from the total.

My 8" (maybe 7 1/2") J-tube with brick bell mostly covered with cob has been operating for a few weeks now in winter conditions, so I can add a data point.

The bell has about 50 square ft of ISA, including about 10 square ft of steel cooktop insert and an access panel. After the fire has been going for an hour or so, the temporary stovepipe leading from the base of the bell gets too hot to touch, 200F at least, and a bit hotter with a long hard burn. The single thickness of red brick on edge at the bottom of the back wall where the future chimney will connect gets comfortably hot.

This tells me that there is enough heat left over for the cob bench bell I plan to insert next to the main bell, at about 20 square ft ISA, while still having good draft.

My system has a slight draft stone cold, and a fierce draft within 30 seconds of starting the kindling, before hot air could fill the bell and reach the chimney, so I might have a more favorable situation than average. When I replace the 6" stovepipe through the wall (12' total rise) with the permanent masonry internal chimney (23' rise), I will probably get even better draft.

Thanks so much Satamax. That will give me some basis for calculating how large my system should be.

Thank you too Glenn. I looked over the pics of your build a few weeks ago and was quite impressed.

Satamax, Glenn and any others with experience with bells used with their RMH's...

In addition to the brick of the bell, is there any rule of thumb re how much more thickness should be added (brick, cob or whatever)to the outside of the brick?  

Would adding loose, crisscrossed bricks or sealed pails of water INSIDE on the floor of the bell be advantageous or a hindrance to overall heat absorption?

I'm thinking that such could add to the total heat absorption of the system but at the same time would decrease the space and thus speed up the rate at which the flue gases would have to pass through. (This rapid movement of gases through the flue of most horizontal masses is one drawback with such (IMHO).

Any thoughts in this regard?

Hollow brick, 20cm thick, filled with concrete, takes 6 hours to let the heat through. I nearly have 2 days, between a day with 4 fires (about 200/240kw  mass at 45C° ) and stone cold. I have 113m² and about 635m3 to heat. 1/4 metal radiator (what others call barrel, but which is not, in my case.) 3/4 mass. About 4 metric tons mass.

In an insulated house, i would advise less "barrel" and more mass than what i have in the workshop. Where i needed more quick heat.

My bell with red brick and firebrick on edge (2 1/2" thickness) starts to get hot through within a half hour at the (firebrick) top. The 6-7" thick cob sections take 4 to 6 hours of steady burning to start to feel warm on the surface (and are then warm for 12-24 hours after the fire is out). Note that the cob is added to the brick thickness, for a total of about 9". Corners hardly ever feel more than faintly warm. I think 4 or 5 inches of cob would be a better general purpose thickness. Also, outside corners can be very rounded not just for looks but for evenness of heat radiating. Sharper corners lose heat too fast and will never feel as warm as the flat surfaces.

Satamax & Glenn,
Thanks again for your advice as I plan my bell system. Glenn, I have often wondered about corners and your post makes me think I'll build a more rounded bell.