Irene Kightley's rocket mass heater - unfinished but working well

I've not finished the cobbing yet as we're lambing and busy on other projects but already we're getting a good idea of how much heat it gives out for the amount of wood we use.

The studio where the built the stove is 13x5 metres and the east wall isn't insulated yet. The rest of the walls have 30cms of stone, 20 of terra cotta brick and 40cms of compressed straw. The windows and doors are double glazed. Here's a video of the stove so far :

https://picasaweb.google.com/101842447547367164524/FilmsAndSeriesOfPhotos#5713155398613403762

I'll get better results once the cob is finished and has dried out properly but I'm experimenting with it a lot before I finish the beast. We're still getting condensation dripping out of the pipes in places so I'll have to make a collector to incorporate under the cob.

I've started taking notes of temperatures at certain times and I'll take photos of how much wood we use. Once I've enough data, I'll publish the results.

Irene Kightley wrote: We're still getting condensation dripping out of the pipes in places so I'll have to make a collector to incorporate under the cob.

I found water a problem too. I was hoping that putting insulation under and mass around would keep more heat in and evaporate any water. I haven't get that far though. So the cob may help with this problem I would like to know how it goes... I asked about condensate but heard nothing back.

Putting insulation around the pipes seems to me to defeat the purpose of them, that is, to release the heat into the cob but you can try it out and see how it goes.

We had loads of water dripping when it was first fired which seems normal. Our cob is still drying out (It's winter here) but the cobbing around the bell and at the foot of the barrel seems to be completely dry which isn't surprising as we've lit the heater about thirty times now. At the beginning I used really dry wood and then I started experimenting with branches which weren't absolutely dry so the water kept dripping. Recently I've been using very dry wood and there seems to be a bit less water but it may not always be us who use the stove and you can't rely on a perfect user.

I'm not talking about a lot of water - so far I've collected just three quarters of a bucket and I know cob is forgiving and can soak up and re-release the humidity but I'd prefer to be on the safe side and make a collector which I can slide out and empty every so often.

At this stage in the process, before I've completely finished cobbing and plastering, I'm taking my time to interact with the heater as much as possible using different kinds of wood, lighting in different weather conditions, changing shapes and the size of the wood feed and direction and size of the outside air feed and as well as getting to know the heater, I'm trying to get it as right as is possible.

Irene Kightley wrote:Putting insulation around the pipes seems to me to defeat the purpose of them, that is, to release the heat into the cob but you can try it out and see how it goes.

I have been less clear than I should My bench is significantly different from the standard RMH, though quite common these days in the masonry heater world. They would call it a "bell". It relies of stratification of gases by temperature so the input is at the top and the exhaust at the bottom. The bench is quite short due to availability of materials. It is about 4 feet long but will be closer to 6 when closed in. Here are some pictures from before it is finished... I need to upload some later ones as I have tested it since and it works really well for storing heat.

https://permies.com/t/10653/stoves/Yet-another-portable-RMH#104236

There is quite a large difference in temperature from intake to exhaust and the mass is easy to adjust to store however much heat is needed... I could even adjust mass by season. Anyway, my tests have been outside on cold concrete so the bottom of my container is busy radiating whatever heat is there to that cold concrete. Because of that, the bottom surface never reaches a high enough temperature to re-evaporate and so water collects there. My thought is that putting insulation under the container would allow this bottom surface to get hot enough (212F ) to get rid of that water. I have two stories of flue anyway and so the gas temperature needs to be high enough to get up there anyway... on a good day the rocket may generate enough draft on it's own, but I want it to always work even if it needs to be somewhat less efficient to do it. So what I am suggesting is not to insulate the whole bench (though there may in some designs be reason for that, such as a small mass at a high temperature to slow down the speed of heat released) but rather underneath. This may be reasonable even in a stock RMH design to keep the heat from leaking into the ground. It may also solve a condensation problem. The difference in the two systems (steel vs. cob) is the type of insulation used. because my container (bell) is held off the ground by a stand, I can use a bat kind of insulation and will be using Roxul which has as high a heat rating as brick or cob. With a cob bench it would have to be strong enough to hold up the bench... like a cob mix high in sawdust, straw or vermiculite (or whatever. The alternate is to build a hollow platform with bricks and patio slabs under and use a softer insulation... but that does not make sense to me.

Having said all that... I kind of wish I had cut the barrel down the center and made it 7 feet long by welding the two halves together and found another way to hold it up. Maybe next time. With this one I need to find out how long of a burn (with what weight and moisture content) I need to get the bench up to temperature with various amounts of mass. The amount of mass will be indicated by how long it stays at a usable temperature. I need 8 to 10 hours where we are now, but where I am planning to move to (vaguely), I would require more mass and a longer time of heat... that may be the time for a longer bench or maybe two of them anyway. It may also be time to move to an 8inch system instead of a 6inch. So far I have been very happy with the performance of my system. It may be a bit costlier for the average builder than the cob model, but as I already have a welder and not much yard to dig up it has worked well for me. When we can move, it is hoped that we will have more land and so may make more use of cob in that design.

I'm very intrigued by the clay collar that connects the drum to the flu, I've never seen one of those before. Is that something you had made, something unique to your part of the world or should they be available pretty much anywhere?

If your system is leaking water in the positive side would that not mean you'd be leaking CO 2 as well? ?

no its just the water from the system being wet. you are driving moisture out of the cob and the rick and the insulation and the wood. it condenses and runs out. collect it in a pan and toss it out. Or check its acid and water plants with it. it soon stops and you will not have to worry about it again.

Sorry not have read all your posts and seen your photos before Len, I see exactly what you mean now and yes, perhaps insulation would help in this case.

I'm very intrigued by the clay collar that connects the drum to the flu, I've never seen one of those before. Is that something you had made, something unique to your part of the world or should they be available pretty much anywhere?

Andy, the collar is simply a fired clay chimney pot which are readily available here in France and in the UK in all sorts of shapes and sizes. As we're building a house, neighbours have given us a lot of materials and this pot just happened to be exactly the size I needed to do the job of wrapping round the barrel instead of using cob.

Dan, this is my concern too.

Almost everyone I know who has built a RMH has problems with water leakage but it's a subject that's rarely mentioned. I imagine that the gasses are lighter than water and are taken out of the flue whereas the liquid condenses and falls to the bottom of the pipes - but perhaps that's just wishful thinking.

Our posts just crossed Ernie, I think the issue is that if a liquid can get out of the pipes despite extreme care in finishing the joints properly, then perhaps a gas can too ?

in this case i would doubt it; it takes the liquid a long time to soak the cob enough to escape. and the weight of gas to liquid is different.
well let me amend that; I have not observed this to happen so far. these things still surprise me on occasion so i cannot rule it out. get one of those cheapy CO sniffers and see.

So I just wanted to chime in on this condensate topic as I was alarmed at first. On my first RMH I was shocked by the condensate that ran out my cleanouts at the end of the horizontal flue run (which I pitched to the cleanouts so I saw it all). After a few weeks of firing it stopped. I was only burning very dry wood and it concerned me but I let it go when it stopped. Then, three years later I decided it was time to really ream the flues with a stack brush and low and behold the condensate happened again! I theorize that there is something about the buildup in the flues that helps to mitigate the condensate issue. I have a few thoughts but it is just guesses, the bottom line is that there is some kind of "seasoning" of the bowels that needs to happen to releive the RMH of condensate running. I have since seen this happen on all the other ones and it is a mess that one should expect and plan for dealing with the juices with catchment of some sort that can be easily removed & emptied without spilling the black staining stuff. Canyon

Thank you very much for that information.

I can appreciate that as the pipes clog up slightly they'll "soak up the juice".

I'll make some collectors.

From my understanding of combustion of hydrocarbons, one of the products is CO2, and another big one is H2O, there's no getting away from it. If the flue is above 212ºF, then the water won't condense. If it is less than 212ºF, concensation will occur, and if the surfaces are smooth the water will run downhill. If the surface is sufficiently dirty, the water gets spread out by capillary attraction/absorbtion or whatever, so you won't notice it.

If it is less than 212ºF, concensation will occur

That's not always true. If you are burning really wet wood then probably, but if you are burning really dry wood the amount of h2o produced in combustion might not saturate the air at lets say 150F to 90F. In that case when the temp of the gas drops from 500F to these temps you may not get condensation if your gas is not at the saturation level with h2o. It all depends on the amount of h2o vapor AND temperature.

Daniel Truax wrote:

If it is less than 212ºF, condensation will occur

That's not always true. If you are burning really wet wood then probably, but if you are burning really dry wood the amount of h2o produced in combustion might not saturate the air at lets say 150F to 90F. In that case when the temp of the gas drops from 500F to these temps you may not get condensation if your gas is not at the saturation level with h2o. It all depends on the amount of h2o vapor AND temperature.

This one could go on forever

We want as high of a saturation level as we can get and still run. Ideally, we want all the oxygen going in to get used up and become either water or carbon dioxide. Any extra air going in is just taking warm air and exhausting it from our home... and it is also making the burn cooler. These extra gases are called ballast gases and are unavoidable... however, keeping them as small as possible will get the most heat from fuel to cob to butt.

Having said that, I don't know what the water saturation would be for perfectly dry wood (as a starting point). I also don't know what temperature that saturation starts to condense at. So I'll start with what I do know.... There is just under 21% oxygen in the air... maybe a bit less in a house that is trying to stay warm... lets use 21% for easy math. Most of the rest is nitrogen, but there is some CO2 and water already too (the air is wetter here during heating season in the PNW but may be drier where it is colder).

Wood is a hydrocarbon. I don't know the exact amounts of carbon a hydrogen locked up in it, but most hydrocarbons are made of a long string of carbons with two hydrogens per carbon. There will be two extra hydrogen atoms at the ends, but I am going to ignore them because any numbers I come up with are going to be drier than reality anyway... and besides a 2:1 ratio is easy to work with

So each two hydrogen atoms require one oxygen atom for one water molecule. Each carbon atom requires two oxygen atoms for one CO2 molecule. So 1/3 of the oxygen becomes water and 2/3 becomes CO2. That is the burned gases have about 7% water in them (actually that 21% oxygen is O2 so that would be 14%)... in a gas I think that would be by volume... remember, this is for perfectly dry wood... 0% moisture content... it doesn't exist in nature.

really dry wood is about 10% moisture content, most of the stuff we will be burning will be higher. That moisture content will be by weight. Carbon has a mol. weight of 12 and oxygen 16, so that 10% moisture is less than 10% of the generated gas molecules... only about 3/4, say about 7.5%

So of that 21% oxygen we end up with 28% CO2, 14% H2O and another 1.5% extra water from moisture. That doesn't seem like much... maybe someone else has better math? Anyway even really moist wood (50%) would only add 7.5% to that 14% from burning. So most of the water in the exhaust is from burning it looks like ...

Now does some else know what the temperature to moisture content for condensation is?

I have learned from others that for general efficient burning conditions w/ avg wood that under around 140 F is the point which condensation is a concern. I try and get a fine tuned rmh to run an exit temp from 160 to 200 F to avoid the issue. There is not a great amount of btu's easily available below that to warrant dealing with the condensate in my and many other masonry stove builders opinions. Perhaps there is someone with more detailed info?

I'm gonna put a jag in the conversation..
@Len Ovens,
A proper Bell has the input AND the output at floor level. If the input is at the top (easier to do with a rocket stove) it will tend to "pump" the heat out of the bell or at least stir the contents so that it won't stratify properly.
Stratification of heat is the name of the game in bells! Very important detail..

Kirk Mobert wrote:I'm gonna put a jag in the conversation..
@Len Ovens,
A proper Bell has the input AND the output at floor level. If the input is at the top (easier to do with a rocket stove) it will tend to "pump" the heat out of the bell or at least stir the contents so that it won't stratify properly.
Stratification of heat is the name of the game in bells! Very important detail..

Lots of bells around, Some have the entrance at the bottom, some don't. Even those that do, often have the input in such a way that the flow is pointed upwards. In my case, because of the shape of my container, I could not put both input and exhaust on the bottom unless I ran one of them around some corners and back. The main idea with the RMH is to reap the benefits of a masonry heater without the cost of materials... that is using found or free stuff as much as possible. Another thing you may have noticed about most normal bells is that they are quite narrow. Even a square one is full of channels. I am not sure why as I would think that would make for more gas movement and less stratification, The three reasons I can think of are: To provide support for the top, to keep the input from being too close to the exhaust, to provide more mass to soak up the heat. I am not sure which of these are important... and as I am not building with brick I am not too worried I looked at the container I had and thought about the best way to use it to extract heat from the flue gas. The first thing is that without the bell/bench the flue gas is moving, but it is not a torrent... not like fan forced air would be, it is moving slowly. At the point it comes out of the main barrel it has cooled enough that I can block half of the exhaust without having smoke back. That is, with a 6 inch riser, I can get away with a 4 inch exhaust even before the bench. So the flue gas as it enters the bell is slowing down even more and spreads out across the flat top transferring it's heat to the mass on top. As the gas gives up it's heat it falls... the bell is not meant to have static flue gas during a burn, but constantly replace cool gas with hot gas so there will be some stirring no matter what the physical configuration is. I put the input pipe half way through the chamber to keep as much separation between the intake and exhaust.

In the end, whatever a "real" bell is supposed to be like (and really, I have looked at the build process of as many as I could... they are all different) this one works very well so far as testing has shown. The pipe goes to the center close to the top, but the masonry just above the pipe does not get as warm as the rest where the gas rises to meet the mass. There does not seem to be (other than right over the intake pipe) hot or cold spots as would be expected if the stratification was much disturbed or there was a direct flue gas path. I don't know for sure, but I think the gas goes in a horizontal circle with the cooler gases falling faster than the warmer ones.

Having said all that, all of my trials have been with a cold system outside. It will be interesting to see how it does starting warm and getting hot enough to evaporate the water. So far with three hours of brisk burning it has done really well, maybe too well. I may have to go to an 8 inch system to really get the best out of it.

Len Ovens wrote:That is, with a 6 inch riser, I can get away with a 4 inch exhaust even before the bench.

Are you talking about a RMH, or a different type of stove? I thought the exit at the bottom of the barrel in a RMH had to be bigger than the riser.

Hi Len,
I have been wanting to build one of these Rocket stoves for some time also and I am also interested in the Bell type. I have been building woodgas stoves for about 5 years now just for a hobby and I found some math that might interest you.
http://mb-soft.com/public3/woodburn.html
Here they talk about how much carbon, hydrogen, Oxygen is in the wood.

ken

Roy Clarke wrote:

Len Ovens wrote:That is, with a 6 inch riser, I can get away with a 4 inch exhaust even before the bench.

Are you talking about a RMH, or a different type of stove? I thought the exit at the bottom of the barrel in a RMH had to be bigger than the riser.

Yes I am talking about a RMH. Yes I know the RMH mantra is "constant CSA throughout the system". Yes I agree with Ernie that robust trumps that last little bit of efficiency. I was also talking to "Donky" (not sure if I spelled it right, sorry) who also tests a lot of these things, and he noted that the volume of the flue gas changes dramatically through the different stages of the RMH. He has used a CSA at the end of the bench with as little as half riser CSA he says. I don't think any of us would recommend deviating from the constant CSA method though.

In my case the exit pipe from the base of the barrel is 6 inches same as the riser. So constant csa. In testing, I tried putting a block across the exit pipe to see what would happen, and it still worked... end of experiment and the flue goes into the bench at full CSA.

RE, most RMH builds... the barrel cob bench interface seems to be the trickiest part to make big enough to get riser CSA. I did the math... and realised that the bottom of the exit pipe is shadowed by the floor and often the sides around the exit pipe are shadowed too. Then the riser base is normally quite big being made of firebrick and 4.5 inches thick. So for an 8 inch riser, the gap between the exit pipe and the riser is only 4 inches (assuming a 24 inch barrel). This is why it appears the that the exit is so scooped out so much. To get 50sqin at 4 inches means it has to be 12inches wide at least. wider is better because there is also a 90degree bend at this point. In my case I have only an 18 inch barrel, but, my outside of the riser at the exit point is only 8 inches. This means I have 5 inches of space and because of my design only the bottom is shadowed not the sides. So there is lots of room for the gas to move and reorient itself to the new direction of travel.

ken smith wrote:Hi Len,
I have been wanting to build one of these Rocket stoves for some time also and I am also interested in the Bell type. I have been building woodgas stoves for about 5 years now just for a hobby and I found some math that might interest you.
http://mb-soft.com/public3/woodburn.html
Here they talk about how much carbon, hydrogen, Oxygen is in the wood.

That was great! I knew there was oxygen in the wood as most organic CHs do, but I am used to alcohols and veg oil where there is one oxygen per molecule. In this case really we have carbon and water and would only have to add enough air to burn the carbon. I could redo my calculations ... but I still didn't come up with a relative humidity of the exhaust and I wouldn't be any closer from this info. It would still appear that most of the water comes from combustion with reasonably dry wood... but I still don't know how to calculate what temperature that would stop condensing at.

One can see from that page how just heating the wood up with no air would produce charcoal though. Very interesting.

Len Ovens wrote:
Lots of bells around, Some have the entrance at the bottom, some don't. Even those that do, often have the input in such a way that the flow is pointed upwards. In my case, because of the shape of my container, I could not put both input and exhaust on the bottom unless I ran one of them around some corners and back. The main idea with the RMH is to reap the benefits of a masonry heater without the cost of materials... that is using found or free stuff as much as possible. Another thing you may have noticed about most normal bells is that they are quite narrow. Even a square one is full of channels. I am not sure why as I would think that would make for more gas movement and less stratification, The three reasons I can think of are: To provide support for the top, to keep the input from being too close to the exhaust, to provide more mass to soak up the heat. I am not sure which of these are important... and as I am not building with brick I am not too worried I looked at the container I had and thought about the best way to use it to extract heat from the flue gas. The first thing is that without the bell/bench the flue gas is moving, but it is not a torrent... not like fan forced air would be, it is moving slowly. At the point it comes out of the main barrel it has cooled enough that I can block half of the exhaust without having smoke back. That is, with a 6 inch riser, I can get away with a 4 inch exhaust even before the bench. So the flue gas as it enters the bell is slowing down even more and spreads out across the flat top transferring it's heat to the mass on top. As the gas gives up it's heat it falls... the bell is not meant to have static flue gas during a burn, but constantly replace cool gas with hot gas so there will be some stirring no matter what the physical configuration is. I put the input pipe half way through the chamber to keep as much separation between the intake and exhaust.

In the end, whatever a "real" bell is supposed to be like (and really, I have looked at the build process of as many as I could... they are all different) this one works very well so far as testing has shown. The pipe goes to the center close to the top, but the masonry just above the pipe does not get as warm as the rest where the gas rises to meet the mass. There does not seem to be (other than right over the intake pipe) hot or cold spots as would be expected if the stratification was much disturbed or there was a direct flue gas path. I don't know for sure, but I think the gas goes in a horizontal circle with the cooler gases falling faster than the warmer ones.

Having said all that, all of my trials have been with a cold system outside. It will be interesting to see how it does starting warm and getting hot enough to evaporate the water. So far with three hours of brisk burning it has done really well, maybe too well. I may have to go to an 8 inch system to really get the best out of it.

The channels do help hold up the top, but more importantly they create surface area. Surface area is the main limit to heat transfer in a bell (really, any heating/ed device).
As to mixing... Bells rely on the principles of "free gas movement" to do their thing, the less you swirl the mix, the better it will stratify, etc. Some stove builders count from the bottom of the firebox of the first bell, some say that isn't a "true" bell.. I tend to fall with the latter, though any large chamber, regardless of it's flue arrangements will act a little like a bell. 'Course, you've got to take it all with a grain of salt. There are new combustion products coming in at a given rate, the old stuff has to leave at the same rate or the stove will back up.. Call it what you will, there's pumping and stirring going on regardless.
I'm not convinced that you NEED to separate the flows much. Seems to me that if the in and outs were side by side at floor level,(in a bell system) there would be no problem, no tendency for flue products to short circuit the bell. If they are placed both at the top, well that's obviously a bust, though a certain amount of heat would still circulate the bell. The physics of bells dictate not just stratification, but dispersion. When a particle (of anything) moves from a pipe to a larger chamber, back to a pipe again, that particle tends to bounce around and touch every part of the chamber before leaving. It's in the nature of things to act this way, and no matter how poor the setup for it is, some of this will be going on no matter what. 'Course, it hurts NOTHING to be sure. There is no harm to making the pipes "blind" to each other, probably helps somewhat.. Just as it's natures way to be chaotic, it's also natures way to NOT follow the neat rules we set for it..

Kirk Mobert wrote:

The channels do help hold up the top, but more importantly they create surface area. Surface area is the main limit to heat transfer in a bell (really, any heating/ed device).

That makes sense, The hold up the top thing is more tradition I think as firebrick (or castable refractory) comes in bigger sizes now... just take a look at how many "white ovens" in heaters are made. I may be able to improve my bench with a row of brick in the center, but I am not so sure... right now it cools off the flue gas enough for condensate to be a problem. I have seen experiments with a long, tall, skinny bell. One brick wide, four foot or so high and four foot or so long. While it did have a lot of surface area to absorb heat, the temperature reading over the surface showed a lot of flow... enough that while the top was quite warm, the center was cooler than the bottom. It did seem a bit better when the flue gas entered part way up the one end of the bell. I would really like to see the same kind of test done with cube shaped bell with no baffles. It would require somehow holding temperature sensors in a 3d grid inside the cube without disturbing the gas flow.... because...

As to mixing... Bells rely on the principles of "free gas movement" to do their thing, the less you swirl the mix, the better it will stratify, etc.

That is why I would think that a wider more cubic bell would work better and why I left my box open with no baffles. I may try reducing the length of my intake pipe (I have three to play with ) after what you have said here and to shrink the cool spot right over the intake pipe. I think narrow channels may be working against the bell effect.

Some stove builders count from the bottom of the firebox of the first bell, some say that isn't a "true" bell.. I tend to fall with the latter, though any large chamber, regardless of it's flue arrangements will act a little like a bell. 'Course, you've got to take it all with a grain of salt. There are new combustion products coming in at a given rate, the old stuff has to leave at the same rate or the stove will back up.. Call it what you will, there's pumping and stirring going on regardless.

It would depend on the volume of the bell area if there was room for the gases to stratify. Most I have seen the channels are too small for the flue gas to slow enough for much bell effect.... so I would tend to agree that most "double bells" are really contraflow with a bell on top. Some people have referred to the barrel of a RMH as a bell (I may have even) but I think the flue flow at the top of the riser messes up things too much... though there is a lot of temperature difference from top to bottom of the barrel. Some of that may be that the flue gas is forced against the top and then flows past a thin layer of gas next to the barrel sides without transferring as much heat there. Lots of people have made surface measurements, but I am not aware of anyone who has made center of flow measurements.

I'm not convinced that you NEED to separate the flows much. Seems to me that if the in and outs were side by side at floor level,(in a bell system) there would be no problem, no tendency for flue products to short circuit the bell. If they are placed both at the top, well that's obviously a bust, though a certain amount of heat would still circulate the bell. The physics of bells dictate not just stratification, but dispersion. When a particle (of anything) moves from a pipe to a larger chamber, back to a pipe again, that particle tends to bounce around and touch every part of the chamber before leaving. It's in the nature of things to act this way, and no matter how poor the setup for it is, some of this will be going on no matter what. 'Course, it hurts NOTHING to be sure. There is no harm to making the pipes "blind" to each other, probably helps somewhat.. Just as it's natures way to be chaotic, it's also natures way to NOT follow the neat rules we set for it..

Yes, in the end what works, works. I have done a number of things a little different than the rules and yet it draws fine and the flue gas at the end is as cool as I would want it to get... maybe too cool. My bench is nice to sit on and the exhaust is much cooler than the input. I do know I can't explain all of it... even though I think I know why... the reality is that I don't really because I haven't made enough measurements. The nice thing is that the flue gas is sealed from the mass and so I don't have to seal the mass, I can just set more on top or at the side and see what happens.

Len Ovens wrote:Some people have referred to the barrel of a RMH as a bell (I may have even) but I think the flue flow at the top of the riser messes up things too much... though there is a lot of temperature difference from top to bottom of the barrel. Some of that may be that the flue gas is forced against the top and then flows past a thin layer of gas next to the barrel sides without transferring as much heat there. Lots of people have made surface measurements, but I am not aware of anyone who has made center of flow measurements.

While I don't consider the barrel to be a proper bell, I HAVE noticed that if you gap the top and sides extra wide, they can be made to act a bit like bells..

Kirk Mobert wrote:

Len Ovens wrote:Some people have referred to the barrel of a RMH as a bell (I may have even) but I think the flue flow at the top of the riser messes up things too much... though there is a lot of temperature difference from top to bottom of the barrel. Some of that may be that the flue gas is forced against the top and then flows past a thin layer of gas next to the barrel sides without transferring as much heat there. Lots of people have made surface measurements, but I am not aware of anyone who has made center of flow measurements.

While I don't consider the barrel to be a proper bell, I HAVE noticed that if you gap the top and sides extra wide, they can be made to act a bit like bells..

The think that with the RMH barrel, a wide gap just slows the gas flow from top to bottom so that there is more time for heat to transfer to the barrel surface... I don't think there is any part of the top surface where the gas movement isn't faster than any other part of the the heater... besides inside the riser.

As an aside... we seem to have hijacked this thread. My apologies Irene. Anyone with further questions or comments relating to my bench or RMH should probably redirect them to:

https://permies.com/t/10653/stoves/Yet-another-portable-RMH

I think there was also a thread about general discussion of how things work for some of this too, but I couldn't find it just now.