|
Post by pinhead on Dec 18, 2014 9:20:51 GMT -8
I've been experimenting with my bell bench for a few weeks now and after adding a bell dividing wall, I have come to troubling conclusion: The second bell is being wasted. I haven't yet built the rocket stove portion of the project yet, though, so luckily I am able to experiment with different divider configurations in order to get the most out of the bells. As it is currently, I have two half-barrels divided by a brick wall with a "slot" running along the floor of the wall. It's the classic "double bell" as demonstrated here: energy.concord.org/energy2d/double-bell2.htmlAs you can see from the demo, as much heat is drafted straight out of the chimney as makes its way into the second bell. This is undesirable in my situation, though; with the amount of heat a 7" Batch Box with S-Portal produces, I shouldn't have any trouble with low stack temperatures. In other words, I'll be losing more energy out the chimney when using the standard double-bell configuration than needed to draft. I've been modifying the simulations and running ideas through my head and I believe I know what I'm going to try to do. Luckily it's really pretty simple. Behind the dividing wall - between it and the chimney - I'll build a brick wall covering the bottom 2/3 or so of the bell. This will force the gasses up towards the top of the bell where they can then "stream" over to the rest of the bell. Hopefully there will be enough open volume there to allow the gasses to stratify and only exhaust the lowest temperature gasses through the chimney. I'll try to get some pics when I work on it next.
|
|
|
Post by Daryl on Dec 18, 2014 9:50:48 GMT -8
Can't wait for the replies and feedback. My little experience has led me to favor channels for gas flow but what do I know.
|
|
|
Post by Robert on Dec 18, 2014 12:36:35 GMT -8
I built one bell cookstove rocket. One batch of wood and 24 hours the bell is warm. I have no doubts in bell at all... But we need to know what we are doing thats for sure...
|
|
|
Post by pinhead on Dec 18, 2014 12:57:12 GMT -8
I suppose I should have reworded my complaints.
I don't doubt the efficacy of bells in general; I still believe they're superior to tubes.
My theory is, it would be better to allow the gasses to enter towards the top and exit at the bottom vs. having both inlet and outlet at floor level.
In my bell, the surface temperatures of the second bell is warmer at floor level just in front of the chimney than it is on top, just aft of the dividing wall.
In other words, it looks like the simulation linked above is fairly accurate, with the hottest gasses colliding with the chimney wall and a large portion never "rising" in the bell.
I also recognize that lower draft/slower velocity will greatly enhance the stratification of gasses and the bells' heat exchange capacity.
|
|
gjh42
Junior Member
Posts: 61
|
Post by gjh42 on Dec 18, 2014 14:06:46 GMT -8
Playing with the simulator, I got a drastic change in behavior (with the divider removed and acting as one bell) when I added a small lip just above the chimney entrance. The (hottest) gases coming down the far bell wall rolled back to the center floor rather than going straight out the chimney.
With a double bell as you have it set up in the simulator, a divider at the floor certainly does make a beneficial change in flow patterns.
I also played with making it act as a dead-end bell (exhaust below entrance), which looks amazingly effective at making even heat distribution.
|
|
|
Post by DCish on Dec 18, 2014 14:23:27 GMT -8
I believe i recall (was it Peter?) mention that a straight-through path should never be facilitated... that is, entrance should be above exit (but no higher if possible) if they are across from one another. On the Russian bell site they show a vertical entrance at the bottom with gasses streaming straight up, and the exit as far away as possible at the bottom. If you put entrance at top you might create high mixing that disrupts stratification.
I like the wall idea. If you are able, a nice test might be to start with a lower wall, close to the inlet, and only as high as the inlet stream, interrupting the straight flow and perhaps encouraging stratification without forcing cooler gasses to climb a tall wall, effectively making it a channel rather than a bell.
In the bell I am (slowly!) building the gas inlet and outlet are side by side on one end of a narrow (14" wide, 6' long) bench. The outlet draws from floor level, the inlet is directed out into the bell as low as possible but above the level of the outlet. Fingers crossed!
[Edit: Didn't see gjh42's post while I was writing my own... I'm basically trying to say the same thing, but he has better data...]
|
|
|
Post by PNW Dave on Dec 18, 2014 16:09:59 GMT -8
I made a couple changes to your model, attempting to create a "steady state" instead of the constant turbulence in your original. It becomes stable very quickly, with very little apparent motion while the simulation runs. The image below is after leaving it to run for a few minutes. This simulator is a ton of fun to play with, and I have lots of thoughts, but I must be off to work now. I would really like to see how my bench design works in this simulator (just the upper drawings portion)-
|
|
gjh42
Junior Member
Posts: 61
|
Post by gjh42 on Dec 18, 2014 18:38:31 GMT -8
One of the things I did while playing with the simulator was like your dead-end bell... it quickly got white (hot) throughout the bell, with little heat leaving.
It also quickly settled into a smooth flow pattern, out at the top and back at the bottom, with the hot gases going all the way to the end.
|
|
|
Post by patamos on Dec 18, 2014 22:26:19 GMT -8
Pinhead, I couldn't view your images (my plugins get blocked from time to time...), so i may be way off in what i am picturing. But i would be inclined to imagine the sucking/vacuum forces of the chimney and ways to alter the trajectory of the gasses as they flow into the bell. Height of inlet port will matter, but i think direction of flow too.
If inlet and exhaust are at the same end, you could consider running a flue within the bell most way to the far end and then let gasses stratify on the way back.
hope this is more than pure gibberish.
|
|
|
Post by satamax on Dec 18, 2014 23:47:54 GMT -8
Pinhead, i'm no expert. But you have forgoten something, first bell should always be bigger than second bell. I can't remember where i've read this, might be on stoves.ru. www.stove.ru/index.php?lng=1&rs=16By the way, is that energy 2D a program or what? Thanks. Max. Ney worries, i checked and downloaded it.
|
|
|
Post by PNW Dave on Dec 19, 2014 0:33:34 GMT -8
One of the things I did while playing with the simulator was like your dead-end bell... it quickly got white (hot) throughout the bell, with little heat leaving. It also quickly settled into a smooth flow pattern, out at the top and back at the bottom, with the hot gases going all the way to the end. Awesome, I'm going to give it a try too.
|
|
|
Post by Daryl on Dec 19, 2014 2:27:10 GMT -8
I made a couple changes to your model, attempting to create a "steady state" instead of the constant turbulence in your original. It becomes stable very quickly, with very little apparent motion while the simulation runs. The image below is after leaving it to run for a few minutes. This simulator is a ton of fun to play with, and I have lots of thoughts, but I must be off to work now. I would really like to see how my bench design works in this simulator (just the upper drawings portion)-
When the gases cool down, they slow down. What is the ideal temperature for gases exiting a flue? Would condensation be a factor if all the heat is left in the bell?
|
|
|
Post by peterberg on Dec 19, 2014 5:19:06 GMT -8
What is the ideal temperature for gases exiting a flue? Would condensation be a factor if all the heat is left in the bell? This is an important point. Running the simulation, it appears the distribution of heat can be changed a great deal. But... at the same time the temperature at the end of the chimney need to be well above 40 C within about 15 minutes after a cold start. Otherwise, chances are on the horizon a chimney stall is on its way. Also, this is a 2D simulation, the likelyhood of the exact same behaviour in a three-dimensional heater is marginal, to say the least. In short, there need to be a significantly temperature difference between the room and the chimney end. If not, the condensation point will come down and the draw will stop altogether.
|
|
|
Post by patamos on Dec 19, 2014 17:37:21 GMT -8
Which is what makes a bypass flue so helpful for start up and tweaking on the fly.
Combines the best of double flue German Grundofens and Russian Bells.
|
|
|
Post by wrekinwanderer on Dec 21, 2014 14:51:00 GMT -8
Satamax, thanks for the link to Kuznetsov's stoves, interesting stuff that confirms the principles we like playing with. I couldn't find in the article . Anyone have anything to say on this? Certainly on most rocket stoves the first bell, in the barrel, is smaller than the second bell, the bench...
Patamos, with a is this something you would manually close with a damper when the stove was up to speed? After which the heat would then flow into any subsequent bells....
|
|