|
|
Post by shilo on Nov 27, 2014 23:49:59 GMT -8
good people
|
|
|
|
Post by matthewwalker on Dec 8, 2014 16:34:30 GMT -8
So, I've been quiet for a bit, but it's time for an update. Terry (Griz) was kind enough to send two different tube configurations for testing. I ran both of them for a couple days, seeing how it was starting and feeding them. Then I did a series of runs on each. The first one I tested I'll call v6. It looks like this:  3.14"sq. inlet size, and I think 2.7"sq. or so outlet size. Terry, is that right? I didn't measure, just going off of what I remember from our conversations. So, it was kinda a bear to warm up. It seemed like it wasn't flowing that freely at the outlet side, so it didn't start the secondary burn until things were going fairly well. I needed to give it more primary during warm up, and for longer than with v5. It did still deliver some nice charts, although minimum CO was higher than v5. Here's three runs in a row, from a cold start to a warm system...    Pretty darn good, although V5 was more predictable, faster to warm up, and overall ran with lower CO. Still, this was interesting. I can see through the top glass to the port, and this one had some moments when things were really raging back there. It looked cool, and made a heck of a roar. The next one I'm calling v7. It looked like this:  Same CSA numbers as far as I know. Again, sorta slow to warm up, although better than v6. This one performed better overall in daily use, and the charts, while close, I think show this one doing better out into the end of the burn. CO was also lower here, riding around 20ppm for most of the burn. Again, not quite as stable and low as v5, but pretty dang good. I'm not posting three runs of this one, it struggled a bit on my middle load and I ended up running the thing wide open to get it up to temp. I have that chart, it looks more like a normal rocket chart with the whole thing burning down in 20 minutes or so. I think this one shows some promise, here's the start up chart, and then the warm run. Not bad...   Overall, it would appear that once the stove is warm there isn't a whole lot of difference being made by the delivery shape. v5 flows really well at lower draft, like start up and warm outdoor conditions. I think part of that is the open top of the tube, which means that inlet size is smaller than available outlet size. For some reason, I think that is helping. It also is doing better at mixing, with average low CO under 5ppm for a long time in the middle of the burn. Neither v6 or v7 ever got down to 10ppm, with v6 staying up by 50ppm for most of the run. Not bad, but indicative in my mind of a less optimal mixing pattern. Maybe. Anyway, very cool to play with these. Terry, I'm so, so pleased that you stepped in to participate like this, and I'm down to continue trying stuff if you or anyone else would like to keep plugging away at it. So, I'm back to v5 for now, and am quite pleased with it in terms of living with it. It's warm out now, 50°F or so, and will stay that way for most of this week, along with 10-20 inches of rain. Yep, 20 inches! Yikes. Anyway, when it gets cold again, and not so windy, I'll switch back to standard configuration for a bit and see how I feel about all this. So far, I feel like I'm using less wood, messing with the stove less, and staying warmer. I also think I'm having longer moments of dirtier burning with the long coaling stage, so it's not all roses. I still need to do some long tests of both the standard configuration and this one to see exactly what's happening from start to all the way out, but overall I think this is an interesting and worthwhile configuration option for sure. |
|
|
|
Post by grizbach on Dec 9, 2014 1:51:49 GMT -8
Matt,
I've been waiting with bated breath. It's great to see the results!
My thought process for these were to maximize the curtian of air scenario to try to get as much secondary air mixing before the mixture hit the tunnel. Both configurations have the same slotted area(2.78sq").
I had the most hope for V6 with it's sandwich mixture of flame,air, flame, air, flame.
Because the V7 flows more at low draft, I would sumerize that with quicker mixing and burning of V6, there is more pressure before and in the tunnel causing the slower flow. It is also possible that by splitting up the flame three ways, caused too much restriction.
Matt, with the slower flow you noticed I realize we don't have a measurement for this. Just like with my rocket at home with a more restrictive system, I needed to tweak my intakes to my flow. As I wrote about draft in the library, flow is directly related to oxygen to the fire.
It is very interesting that as soon as the extra air goes below 10%, the CO almost instantly spikes.
There could also be too much restriction with the thin slots. When you are ready for testing again I would try to open the slots by inserting a screwdriver and twisting them open. If this doesn't make a difference then there IS too much pressure being generated by quicker burning before the riser.
Also, with your V5, I am curious how it will run with a lid on the top.
Thanks for testing Matt, Every bit of info helps the big picture!
Terry
|
|
|
|
Post by pinhead on Dec 9, 2014 7:02:33 GMT -8
Good work and excellent tests!
But I'm going to armchair quarterback a bit.
It seems like the tubes are too small to get even flow through the ports.
For instance, it has been found that - at least for the S-Portal - 2" square tubing is required to flow enough secondary air for a 7" system.
That's with 12.5% primary air inlet and 12.5% secondary air inlet.
Throwing a big 2" square tube in front of the port would be rather restrictive, though (either in overall airflow or it would simply take too much room in the firebox).
|
|
|
|
Post by grizbach on Dec 9, 2014 8:28:46 GMT -8
pinhead,
It may not be clear in the pics, but V6 and V7 have the same 3.88sq" supply tube. With this in Matt's existing door opening he is left with 11% primary.
V6 has a manifold that splits into duals right at the base of the business end. These smaller tubes have 1.49sq" CSA each.
Group, Does anyone have any ideas how to get lower excess O2 in the flue?
|
|
|
|
Post by matthewwalker on Dec 9, 2014 11:16:24 GMT -8
Lower excess air is quite simple, even with these designs. I can run the primary wide open, which speeds up pyrolization and delivers a huge huff of wood gas to the after burner in a short time. All of these designs actually are fairly stable down to about 8% O2, it's below that that things get dicey. The way they are tuned now, is that they have enough air to be able to handle those big off gas events that happen during a burn, but still be relatively stable and not too high in O2 during the body of the burn. So, in other words, we could tune them to run at low O2 when shut down and gasifying, but they would be fuel rich and blowing smoke during warm up and loading and fuel drops and such.
To improve upon it, to my mind, means lots of air adjustments during the burn to adjust for the different fuel rates. At DC last year, I know the winning stove did use a spring driven air control that automatically did some of this adjusting. I think it's possible, but I'm going for a simple two setting stove. One starting setting, and one running setting.
I do not find that the Testo shows a great Eff. advantage between 8% O2 and 11%O2. I've payed very close attention to it during all phases, and while lower O2 is a good goal, I don't think it's going to change the end figure by very much. There is a big discrepancy in Peter's and my Testo with regards to efficiency, same O2/CO/T numbers yield almost 5% higher eff. in his readings. We don't know why.
Oh, and Griz, I get a constant draft reading in IN./H20 from the Testo, so I can see how much pull we are getting at least. All of these ran way better when draft was higher than -.06 in./h20. Startup was around -.004 and a hot stove on a cold day was around -.1 I think, I need to pay more attention there.
|
|
|
|
Post by grizbach on Dec 10, 2014 1:32:32 GMT -8
Matt,
With these floor based secondary models, there are two channels on either side of the supply tube that might give a more consistent supply of primary air to the coals. Do you think this might give a more even heat to the fuelbox. Also, if you had the standard chamfered floor, this could burn the coals farther into the run.
With my stove, I always like when the wood doesn't drop down flat onto the floor. It allows the primary to go deep instead of around the logs.
T
|
|
|
|
Post by matthewwalker on Dec 10, 2014 10:43:34 GMT -8
Terry, I do think that it would see some improvement in the charts if it had Peter's sloping floor to minimize the coaling stage. In a final configuration I would recess the secondary inlet tube into the floor, and keep Peter's dimensions with the flat strip of the bottom of the firebox being the top of the secondary delivery tube.
As for the primary down low, I'm actually trying to minimize that, and once the thing is hot my goal is to send all the primary up over the fuel to minimize off gassing(to better control fuel delivery to the after burner) and to minimize ash transportation into the gas stream in an effort to reduce particulate.
At some point I'll make a stove with the recessed channel in the floor and the sloped floor, but for now this one is a decent test bed that is relatively flexible. I will probably create that shape with cob here before too long as a test in this stove, once I'm settled in on the configuration.
|
|
|
|
Post by DCish on Dec 12, 2014 8:52:48 GMT -8
So here's a crazy idea to maybe try... PvdB says that introducing air before turbulence is a good thing. What if this idea is taken to the extreme and the slot in the vertical tube is put facing the door instead of the port? Or if you basically spin the vertical pipe from version 7 (without the top slot) around to face the door to ensure a stream of air on each side of the vertical pipe? The counterintuitive part of this is that it is injecting air into an area that is of higher pressure than immediately after the tube. On the other hand, you're keeping primary air so low that the slots would probably have no trouble injecting air. Perhaps this earlier introduction will allow greater mixing immediately behind the tube, rather than having air injected into, and perhaps simply filling, a pocket of low pressure behind the tube, which filling may actually mute the amount of turbulence rather than taking full advantage of it. Could be a crazy idea, buy since you're already testing, it might be a fun test to do, even just as a rule-out. Successive approximations of this might also be tried, such as slots that are perpendicular to the flow (sort of the inverse of sjang's two pipes flanking the throat with slots on the flat of the square pipe... Slots coming directly out the sides toward the stove walls). Perhaps also a square vertical pipe with side slots could be tried.
LOVE watching this thread develop, thank you so much for sharing!
[EDIT: Of course a significant down side to this is that it would be relatively much easier for fuel to obstruct the secondary air supply]
|
|
|
|
Post by matthewwalker on Dec 12, 2014 9:44:41 GMT -8
I like it DC, I had kinda had some similar thoughts, especially since when I got the goose egg on CO I was injecting the air before the tube. You could see the way it helped mixing as it streamed around the cylinder. I agree, it should have no problem being drawn in regardless of the orientation. I'll try it, although it may be a bit. I'm sorta settled into v5 for the time being. But you never know, these long dark days get me into experiment mode for sure.
|
|
|
|
Post by Donkey on Jan 1, 2015 12:46:33 GMT -8
This is some GREAT stuff here Matt!
Some things come to mind, Re. versions 6 and 7, in particular V6 (I was cheering for V6, though the simplicity of V7 is more attractive). Those smooth shapes seem less likely to create the kinds of draw/turbulence required for consistent airflow.
I'd like to see these re-done with a sharper drop-off to the air ports, more of a bluff-body instead of a smooth transition. It might be possible to (carefully!) re-work them with a hammer on an anvil.. It would be really easy to screw them up entirely, but maybe just maybe sharper lines could be tapped in. I would think that V7 would be a better candidate for this treatment.
??!?
Happy New Year!!
|
|
|
|
Post by matthewwalker on Jan 1, 2015 19:45:49 GMT -8
Thanks Kirk! Yeah, I do think delivery shape is important, and I had good results with earlier versions made from square stock. When I get back into experimental mode I will for sure try that, and DC's idea of the reversed delivery slot.
I have just been living with this thing for the last few weeks, trying not to mess with it too much and just get a feel for it. I'm currently pretty enamored of the design for every day use, but it does have it's quirks. Really, really requires good fuel and a knowledgeable operator. Either one of those factors can make this thing into a smoke dragon, which is one of the beauties of the J. It's really hard to blow smoke with that one unless your fuel is terrible. Still, I'm loving loading this thing up with bigger format fuel, to the brim, shutting it down and heading to town and coming back to it still going strong. And to be fair, the chimney is just as smokeless as it was with the J, and the Testo shows it's burning incredibly cleanly. It just has the potential for abuse, same as any stove with a door and air controls.
Right back at'cha with the Happy New Year my friend!
|
|
|
|
Post by jliebler on Jan 4, 2015 16:10:04 GMT -8
Matthew,
I've just found this thread, VERY IMPRESSIVE! What is the distance from the V5 vertical tube to the firebox side that contains the port? Also how tall is the vertical steel tube and how tall is the port?
Thanks
Jerry
|
|
|
|
Post by matthewwalker on Jan 4, 2015 17:03:06 GMT -8
Hi Jerry, the thing is about 2" back from the port, which is 9" tall. Tube has an overall height of 9".
|
|
|
|
Post by jliebler on Jan 5, 2015 14:42:53 GMT -8
FWIW my observations. Version 6 looks very much like what has been described as the "S portal", but version 5 is very different, although the ratio of secondary port inlet area to system size is very close to the "S portal. Version 5's slotted vertical tube is right in front of the port and changes the flow into the port, dividing it into two streams with the secondary air centered into the port. With such outstanding results it's hard to see room for much improvement but I think a vertical tube with square cross section oriented diagonally with the slot on the corner facing the port could be positioned closer to the port, maybe even against that side of the firebox, making the flow through the port even more turbulent. Another possibility with the square tube is two smaller slots on the sides of the tube closer to the port side corner. Thinner walled tubing will warm up faster but might not last as long. Placing the horizontal tube at the top of the firebox is certainly possible, it should "warm up" faster but are there disadvantages?
|
|
