|
Post by racket on Jul 23, 2011 19:09:51 GMT -5
Hi Wolf
You shouldn't be having combustion problems when using propane , its the most "user friendly" fuel possible .
The "muffler" could still be the problem , also there are good reasons for having the fuel injector centred in the flametube ........................I get the feeling that perhaps the combustor is a bit too "radical" :-(
A very long ducting is going to produce lots of losses , ...................but will give time for flames to "subside" :-)
Cheers John
|
|
wolfdragon
Senior Member
Joined: April 2011
Posts: 287
|
Post by wolfdragon on Jul 23, 2011 20:43:42 GMT -5
Well after some welding, I put the finished ducting on the transition cone: The Cone has a length of approx 6", the first leg of the transition duct is also about 6" and the flanged leg is 8". I forget what the turn radius is. The video below is of the high end of the firing test, the combustor will "idle" at 750 deg F with a 145 deg F inlet temperature. I plan on insulating that ducting to help minimize losses, but it seems like it did the trick... you can see light from the flames, but there is no fire coming out until 1600 degrees F or so (which I know is well above redline) John, why do you want to fuel injection directly center of the tube, even with the simple drilled hex bolt won't it still be relatively easy for the fuel to just stay in the center and not spread out and burn? Since I wanted to disperse the fuel as much as I can, I put the injector radially away from the center as possible so it has to catch air and get blown to the center (which it does). Am I missing a finer point? Does anyone know of a good high temperature spark plug boot? The one that came with my Mallory wires is not fairing too well...
|
|
|
Post by racket on Jul 24, 2011 17:41:40 GMT -5
Hi Wolf
We don't want the fuel just burning in the centre, we want it to burn across the whole cross sectional area of the primary zone to "burn" all the primary oxygen , a centrally positioned fuel injector will/should provide a uniform distribution of fuel across the cross section , probably the reason why its "the standard" for commercially produced engines where symmetry is a positive .
Your combustor has a lot of induced swirl from the positioning of the delivery duct at a reasonably large radius , the delivery tube is of modest dimensions with reasonably high air velocity................ ( have you worked out its velocity??) ........................assuming it is roughly the same velocity as exits the comp diffuser , for that air to travel back to the centre of your flametube will require a large drop in static pressure with a very large increase in velocity ( free vortex laws , think tornado ) .....................the reverse of what happens in your comp diffuser .............any wonder the fuel isn't burning until late in the flametube .
I'm not a great fan of tangential delivery tubes , they increase losses from the scrubbing of air against metal , ..................swirl in the flametube has its benefits , but with our generously proportioned flametubes its not as necessary to have huge amounts of swirl , some "bending" of primary zone holes to induce a bit of swirl is probably all thats required , ....................I've found that even with no delivery tube swirl angle or primary hole orientation , that the flametube "automatically" induces a certain amount of swirl possibly from the pressure drop across the FT holes and the acceleration of the air/gases as they travel down the FT and are expanded by the burning fuel .
Too much swirl can be a negative , we want enough for good combustion ( not a problem with propane) but not too much that it produces a "hollow" combustion pattern , I've seen Pics of flametubes with so much swirl that the flame is only taking up a quarter of the crossection but the full length of the FT because its been impossible for the grossly swirling incoming air to mix with the fuel because of that "vortex theory".
The fact the swirling incoming air will have to increase velocity(pressure drop) to travel into the centre of the flametube will induce losses , having to "straighten out" that swirling will induce more losses , all that high speed air/gases scrubbing against metal will induce more losses , the losses can become so great that the engine produces little power , the aim of "tuning" a gas turbine is to reduce losses as we can't increase power any other way as we are already spinning it as fast as possible and running as high a temperature as possible
Hope this explains things :-)
Chers John
|
|
wolfdragon
Senior Member
Joined: April 2011
Posts: 287
|
Post by wolfdragon on Jul 24, 2011 18:46:46 GMT -5
Seeing that I have the same diameter tube from the compressor exit all the way to the combustor inlet, the velocity should be not too different than the compressor exit velocity...
I see what you are saying about how the fuel and air mix needs to be, but there is one bit that doesn't quite follow with what I have seen. After going back over all the pictures and video I have to the various testing phases of the combustor, the beginning of the flame tube and the primary zone has always been the root of the flame, which also explains why the whole front end of the combustor assembly gets bastardly hot...
With the initial placement of the injector at roughly an inch out from center, the flame only takes up about 1/2 - 2/3 of the total tube diameter with the propane bottle wide open, with a heck of a tail rushing out of it. After I moved the injector up to about 5.5 inches out from center, the entire 6" ID of the flametube is full of fire from the face of the flametube all the way to somewhere between the primary zone ends (~30% length) to where the secondary zone begins (50% length) and it took much less than wide open throttle to get it there.
In both instances, by the time to flame hits the tertiary zone at 75% flametube length it has been crammed down into a 2-4" cone that further tapers out nothing over the next few inches depending on the burn rate. The geometry of the flame pattern is what was making me lean toward putting a diffuser cone of approx 4" between the secondary and tertiary zone to force the flame back out and closer to the cooling air. So I am fairly certain that the burn is occurring in the right place and not farther down the tube than it should, the swirl just seems to be carrying the last bit of flame out the back of the tube instead of leaving it inside to cool off.
None of the pictures, video, or myself and my help ever saw a tornado type void in the center, so I am assuming that all is well in that respect.
One thing that I have noticed is that in the progression of just the open ended combustor, adding the transition cone, and finally the ducting is that it is requiring the cheap needle valve I have to be further and further closed to achieve roughly the same amount of fire and sound (I don't have temperature data for all cases to make the quantitative judgement so only the qualitative one will do). I also have noticed that the garage smells a lot less like a propane leak while the fire is on. Is it possible that the backpressure induced by the inherently restricting elements is forcing the fuel to burn more completely? I just remember the first time I tested it with the giant 6" hole in the back that I stopped after a minute or so to look for a gas leak, this was not the case with the 30 minute burn I did with the finished ducting installed.
I think if I have to redo the combustor I will still try a tangential inlet duct but add some space for what would basically be straightening vanes to use the radial flow to setup even distribution and then the vanes to make it axial flow. I might also try short lengths (1/2" max) of Stainless or Nickel tubing for the primary zone to put zero swirl but say a 25 degree backflow towards the beginning of the flametube, to force a toroidal type flame geometry. Then use the secondary and tertiary zone to induce a bit of swirl. Just trying to think of ways to improve this one in the event of another (which will probably happen... much to my wife's chagrin...)
I think that this combustor is all set for progression to the next step of the build process, but I would like to know what others think now that it is not belching flames like Bender at a frat party...
|
|
|
Post by racket on Jul 25, 2011 1:09:47 GMT -5
Hi Wolf
The reason that the flame takes up such a small percentage of the flametube crossection is because its in "the eye" of the air tornado and only burning at the interface of propane and air
Another reason why its best not to have so much swirl is that your FT hole area is severely reduced because the air isn't going in radially its going tangentially , and those small primary holes wouldn't allow the air jets to penetrate into the propane on such a large flametube .
May I suggest you make a conventional fuel injector with 6 small radial holes and position the injector back near the centre of the FT , the high speed propane jets will extend out into the incoming air compared to your current "muffler" which "injects??" a ball of propane which won't extend radially very far .
We don't need swirl in the secondary or tertiary zones , swirl is only required in the primary zone to produce faster mixing of fuel and air to improve combustion speeds .
Cheers John
|
|
wolfdragon
Senior Member
Joined: April 2011
Posts: 287
|
Post by wolfdragon on Jul 25, 2011 4:53:21 GMT -5
What do you think the hole size in the hex flats should be?
I have a few 1/32 bits as well as a standard 115 piece drill index.
I should be able to get a brass plug from the hardware store, hopefully I can find a hex plug and not just a square one.
|
|
|
Post by racket on Jul 25, 2011 16:14:56 GMT -5
Hi Wolf
6 X 1/32" should do the job :-))
It'll provide a "normal??" injection that you can compare with your current one , hopefully it will produce a short "blue" flame
Cheers John
|
|
wolfdragon
Senior Member
Joined: April 2011
Posts: 287
|
Post by wolfdragon on Jul 25, 2011 18:01:08 GMT -5
The flame length will be hard to see without cutting the rear plenum off the combustor, but I will be able to see either orange or blue in the reflection in the bend.
The big box home improvement store only had square head pipe plugs, so I guess I will give 4x 1/32" holes a shot first, doing holes on the corners to make an 8 pattern would be tough with that small of a bit...
Will see if I can get to that over the weekend, looks like it its going to be busy with non garage related activities thought...
|
|