Peter Clouston

This structure looks very unsafe.  Several posters have already pointed out the need for extra studs in each corner to make an “L” section. The L-section needs to be full height. Not doing this is asking for the studs to buckle, it’s not just a simple matter of reducing sway. Stability against sway needs proper bracing, of the full height of the studs, not the half bracing you have here.  All four sides need to be braced.  Alternatively, 3 sides, and or even 2 adjacent sides, could be adequate but, in that case, there needs to be bracing of all four studs in the horizontal plane, at both floor and platform level.

The way the horizontal beams of the platform are attached to the studs is not normal timber frame practice.  Such beams should be supported on a solid base, either by lying on the top of the stud, as in a top plate in timber frame construction, checked at least 12 mm into the stud, or on top of a jack stud, as for a lintel. Skew-nailing or screwing is only suitable for stopping the horizontal beam slipping sideways, not supporting vertical loads.  The problem with your approach, and that of most of the commentators here is that you are loading the beams near their ends in ways that are tending to split the grain. This is the weakest aspect of wood.  

I am a retired civil engineer.  I have DIY structurally modified several timber frame houses and had my work approved by council building inspectors over the years.  That being said, I am not a carpenter.  I think you should talk to someone in your locality , either a framing carpenter or an inspector, who is familiar with you local framing timber and good practice in its use.

Ken, This is an old but interesting topic. If you are still watching this site, did you build it and, if so, how did it go?

I have been doing a bit more "research" on the web, from my hotel room. If anyone else is thinking of trying this before I get home and can experiment, be careful not to make the angle of widening more than 10 degrees. Basically the added diameter at the top of the riser should be no more than about 1/6th of the height of the riser. For example, if you start at 6 inch diameter at the bottom and the riser is 30 inches high, the maximum desirable diameter at the top is 11 inches. This still gets a substantial increase in volume, and time in the flue will depend on volume, I think.

If the angle is greater than about 10 degrees, you will get lots of turbulence, but not in a good way possibly. The drag forces will go up a lot. And the flow can be concentrated in a "core" that wobbles around. It is possible that if the wobbling is random and covers the entire cross section we mill get good mixing. But we don't really need that mixing by the time we get into the riser and we may instead get some gas stuck for a long time whilst other parts of the flow are hardly delayed at all.

Even a 10 degree taper introduces a substantial extra drag. That being said, standard rocket mass heaters work fine, even though some of them have an aerodynamically horrible detail at the transition between the riser and the down flow in the cooling annulus. Sharp corners, expansion, then contraction. The extra drag from the taper could be quite small compared to that from this "gas torture" section.

Hi Peter,

More good info. Thanks.

OOOPs. I feel stupid. When I calculated the flue gas temperature, I completely forgot to include the effect of the wood moisture content and the latent heat effect. I am not sure if the latent heat correction has to be included for the water of combustion as well as the free moisture, it looks as if it does ATM. Back to the drawing board! That means that to get the maximum temperature that you measured, you must have had very dry wood and also I think a lambda even lower than 1.5?

Peter

Hi Allen,

You are introducing a whole new area of technology here. As a Road Engineer, solution chemistry is way out of my field, combustion is as far out of it as I dare venture at my age and that may well be too far! I can see that getting the water into the salt is easy, but how are you going to get it out again? You have to recycle it somehow.

Actually, once you have condensed the steam, you have already got most of the heat out of it. After that, getting it out of the heated space before it can grab all the latent heat back by evaporating inside the room is easy and heat efficient. The heat left in slightly warm water isn't worth the problems involved in trying to capture it.

As I have said, in some climates you may actually prefer the humidity over the extra heat. However, in my climate I'd be doing my utmost to keep the condensate out of the cob. I'd be really obsessive about avoiding flue corrosion for example.

I'd say that there would be much more payback in energy saved per brains and muscle engaged by working on making the burning more efficient, which also helps cut pollution. Excess air above the minimum necessary to get full combustion is a complete waste. Not only is it heated air that goes out the chimney, it carries water with it that would have been condensed had the excess air not diluted the water vapour concentration in the flue gases.

Peter

Peter,

Thanks for the great link. I've saved the file. Do you have a link where it discusses the stove designers aiming at lambdas of 2-3? It's not covered in this particular pdf.

The higher lambda numbers explains why the vemiculite survives, temperatures are a lot lower in general than I though they'd be. But it increases my puzzlement at the long lives obtained from 200litre drums, as in many cases the oxygen content of the flue gases is not much below ambient. And there is plenty of water around in the form of steam.

Is 500 or even 800 deg C really hot enough to get a really clean burn? I thought that I read 100C was the minimum. I have to confess however that it might be only 1000F, curse these American units out of the dark ages. I get confused enough working between C and K as it is.

I enclose a graph of velocity profiles in a pipe. The paper I grabbed it from was the first I found on the web, it's got more partial differential equations than you can shake a stick at. I will have to pass on that until I have several days free to try and dredge up long buried learning, if it wasn't in those brain cells that age has destroyed that is. According to the paper n is generally 6-10 in pipe flow. I think that that would be for water and it would be higher for gas but that's only a guess.

Hi Peter,

Thanks for the numbers. A lambda of 1.5 seems pretty good. I calculated the theoretical max flue temperature from the published stoichiometric adiabatic combustion temperature of wood burning in air, 1980 degrees C, and assumed the excess 50% air in would be 0 to 20 degrees C and got 1327 C. That makes your 1170 degrees C correspond to a 10% lossof heat to the walls. That seems pretty right as even industrial sized furnaces lose 5%. This may have no practical application whatever but as a geek, I find it fun when theory meets reality so well.

What then causes the much lower temperatures that you are normally seeing. I'm assuming that it cannot be partial combustion, because your Testo? would tell you if it was. If the losses are constant at 10%, or even less, as I'd expect since the gases are cooler, the only explanation is that you are normally running at lambdas of around 2.2 to get 800C and 3.5 to get down to 500. Does this make sense?

Your model of layers of gas building up in speed from zero at the wall is correct, for laminar flow for the entire bore of the pipe. In turbulent flow, it only applies to the boundary layer, which is very thin. So in turbulent flow, to all practical purposes all the gas in the pipe is on average going at the same speed. I say "on average" because turbulent flow is just that turbulent, eddies within eddies within eddies, ad infinitum. Like the fleas's fleas. All flows in these systems are always turbulent. I can't find how to post a diagram, is it possible on this site?

I am in total agreement that the vast proportion of combustion occurs in the lower tube. If you said 98%, I'm not arguing. However, you never get to 100%. And going from 98 to even 99% is just as hard as the first stage from 0 to 98. The only way is to allow as much time as possible at as high a temperature as possible. That means slowing the flow by expanding the cross-section, as a very long tube at the same cross section would be practically impossible to insulate enough to keep the gases warm enough. If you search the web on furnace design you will find heaps of articles describing systems that do just that. And it is claimed to be one of the mechanisms for the excellent cleanliness of a masonry stove.

Thanks for the links. I actually still have the fluid mechanics textbook that the diagrams come from, must be one of the oldest editions still in existence though.

Thanks also for the info on the draft pressure difference. this shows that all the gases in the system are at very close to atmospheric pressure at all times. That's useful to know.

Peter

Hi Allen,

Thanks for that info on the life of drums. It's really quite remarkable and most reassuring. Especially as the gases will still contain about a third of the normal oxygen content when running at 50% excess air (lambda = 1.5), as Peter is running. I'd guess a lot of stoves are running considerably more excess air than that.

I wasn't seriously suggesting that the benches would turn into a boggy morass. They will of course tend to equilibrium with the humidity of the air around them, which is not enough to destroy a bench, even in a greenhouse. What I meant to emphasise was that the bench won't just soak up water indefinitely, if it could do that it really would have to bog down. Instead as it gets wetter, it compensates by evaporating more. In the long run, condensate in must equal evaporation out. This automatically cancels out any heat gain you get from the condensation in the first place!

I believe that if you want to profit from the condensation heat, it is better to drain the water somewhere where it doesn't rob heat from the room air as it evaporates. Ideally you would store it in a container where it cannot evaporate until it cooled, then chuck it out. It's more practical just to ensure the flue slopes so the water just runs outside, as you have mentioned.

Of course some people in freezing climates may prefer to have the extra humidity and trade a little heat for it. I still think there must be cleaner ways to do that.

In a Greenhouse, your plants love humidity. It's not supposed to be an environment for the comfort of people. In NZ or Seattle in winter, anything that reduces humidity is a plus, we can do without a humidifier thanks! It's not that the high humidity is terribly uncomfortable. The main trouble is that the walls, floor, ceiling and furniture are all trying to come to equilibrium with the humidity, so they soak up lots of water. Then when you try to heat the room, the relative humidity of the room drops and the water starts to evaporate. The evaporative cooling turns every absorbent surface in the room into one giant negative radiator! It takes a dickens of a lot of heat to compensate for that.

Peter

Hi Chris,

Thanks for your reply. However, I'm afraid I don't agree with your physics. It's been over 40 years since I last studied fluid mechanics but I'm reasonably sure that your explanation isn't right. I'm less sure that mine is .

There are very good reasons for building conventional external flues straight-sided. One is that it's easier that way and there is no good reason to build it any other way, given that combustion is not happening in the chimney. Two is that it makes it easier to put a rain cap on. Three is that you genuinely need speed and momentum to punch the air past any contrary wind gusts and through any low-lying inversion layer. Historically, number four was that getting the smoke and pollution as high as possible dispersed the crap over as wide an area as possible to avoid complaints from neighbours. Even so hyperboloid cooling towers, which are chimneys, do bell out.

Many modern industrial furnace systems deliberately slow down the combustion gases in order to get more complete combustion, often in a hot cyclone. Of course, they are using forced-draft.


Hi Allan,

Hmmm If the condensation soaks into the cob it must be evaporated out again, or the cob would eventually go soggy. The latent heat required for this will simply suck back out of the room air all the heat you have gained from the condensation in the first place! What you have is a humidifier in effect, not a heat recovery condenser. And you will also re-evaporate any condensible pollutants, unless these bind to the clay in the cob. After a few years, the binding sites inn the clay will become saturated, however, I guess it wouldn't be hard to rebuild the bench when it starts to pong! Although a humidifier is a good idea in many cold-climate winters, It wouldn't do here, or in Seattle, for that matter. Average relative humidities in winter in Seattle and year around here are over 80% and at night over 90. we need all the de-humidifying that we can get!

Peter

Hi Peter,

I've looked at Donkey's thread now. It's not what I was thinking of. He has the taper the other way around. It seems he wants to speed up the airflow, to improve mixing. They also discuss putting turbulators in the riser for the same reason. That seems to me to be back to front, you want mixing to happen right at the start, which is what the peter-channel and your trip achieve. Once you have got mixing, the slower the flow the better I would have thought. You say that most of the combustion happens in the horizontal flue, which tends to strengthen my first impression. After that, my theory is that the riser is essentially a polishing location, where the last little bit of tar and consumable ash is burnt out, needing time at temperature but no further mixing. Does this make sense to you?

I have been reading your threads but coming to them late can make it hard to follow some things. What lambda are you operating at? What sort of peak flue gas temperatures are you getting? The reason that I am interested in this is that I have read that you need at least 1000degrees C to fully combust wood. But many of these stoves are built with vermiculite or perlite, which melt or soften at 850-900. even pumice doesn't make 1000. What am I missing?

I notice that your own stoves use fire brick throughout. Is there an issue with starting from cold with such a lot of conductive thermal mass?

I'd also like to check another assumption of mine. The jet mass heater flue seems to me to be a gas syphon, driven by the differences in density between gas in the riser and the down-flow heat exchanger(the 55 US gallon, or 200litre drum). If that were the case flow velocity would not affect the draft, only the height of the riser and the effectiveness of cooling of the gas in the outer drum branch of the flue have any effect. Good velocity is thus a result of the draft, not its cause. Some on these forums seem to have the opposite opinion. Who is right? Am I missing something?

Sorry to ask so many questions. I am a slow worker compared to you and Donkey, so rather than just rip into it and experiment I would like to at least understand things a bit better before I start.

Peter

Hi Allen,

I'm sorry if I bamfoozled you. I have spent far too many hours on the web while I am stuck living in hotels, far from hearth and family. So I am full of half to a quarter-digested science and no practical experience at all. Can you help me with some practical experience? I was wondering about the performance of these stoves in practice. How long does it take to burn out a 55gallon drum?
If the flue outlet temperatures after the bench are as low as 90 degrees F, there must be an awful lot of condensation. Where does it go? What happens when the flues buried in the bench rust out? Doesn't the condensation then soak into the cob and thence into the house?

I think you must have posted as I was writing this this. Thanks a lot. I never knew about either site. Can't understand how I missed them. Will check them out.

Peter