50 replies to this topic

[kcrucible]

Well, my turn to ask for opinions and assistance in diagnosing.

Full page here,
http://kcrucible.wor...-up-the-burner/



But the summary for convenience of the reader... the burner looks pretty good, though it seems like I might need to open up for just a little more air as I get some tongues of orange when I get to the high end.

Low pressure:


High pressure:




But, I realized that the flame was combusting inside the burner tube, not (solely) post-flare.

Low pressure:


High pressure:




So, what's the cause? Backpressure from the flare at the end of the pipe slowing down the gas too much so that the flame can crawl back down the tube? I haven't tried a non-flare firing yet.

[Fe-Wood]

its hard to tell without seeing the whole burner. It looks to me that there is not enough air intake and maybe some sort of turbulance is causing the flame to clime back up the tube.

[kcrucible]

Thanks. Here's a couple of pics of the burner in case it's helpful.








I'm suspecting that the flare is slowing things down too much.


I may try a run this morning without it, and if the behavior changes signifigantly concentrate on that aspect... narrow the flare port to be closer to inline, make sure there's a smooth transition, etc.

[Phil Krankowski]

Can you adjust the position of the injector tip? It looks like moving it back will draw more air, increasing the velocity in the burner tube and leaning the mixture improving your performance.
Phil

[kcrucible]

Yeah, I can move back the tip a bit by heating up the silver solder and sliding it backwards... good idea for increasing air intake without needing to increase the slot size. Definitely easier!


I don't think it's likely to change the flame location though. I just had another idea... the slot channels are rounded on the interior. Maybe that's not making the flow better as I had hoped, instead introducing vortices right at the end of the air intakes, hurting flow?

[Phil Krankowski]

Just change one thing at a time.

You may want to unsolder the connection so it can move free and set up a temporary clamp for the burner tube till it acts nice then re-solder.

On the intake I would expect that smooth round edges are desirable, on the injector you need a clean sharp edge so the gas parts away from the nozzle. While you have it apart, look at it and see if there is a burr on the end of the contact tip. Use a torch tip cleaner to fix the problem if it exists, they are about $10 at many places and may last a lifetime.

Phil

[kcrucible]

A note about the edges... though the picture doesn't show it well, the flames on High pressure are scrolled at that point. They swirl around and combine, from the outside in... like someone sticking their finger in dough. I'd read another post about the injector (yours probably) and thought a similar effect my be happening with the air-intake, causing all of the air to push against each other slowing things down.

But if the air goes inline with the tube sooner, then it may not be a factor.

Step 1: Slide the tube back a little
Step 2: Widen the slots, especially towards the rear (so that pressure is normalized by the time it gets to the top end of the slot and doesn't need to draw much/any air in at that point.)


Is it possible that the jet has expanded TOO much by that point, and the gas is hitting the edges of the slot? In that case pushing it forward would probably be the solution. The reason I'm linking about this is the T-Rex burner...



Notice how the tip is pushed far into the tube so that the fuel can't catch the edges (I assume that's the reason anyway)? I could always make my slots bigger to accomodate.





Thanks for your suggestions, Phil.

[ThomasPowers]

The terminal flare should *NOT* create back pressure. It's function is to drop the pressure to where the flame propagation speed just equals the gas speed at some point in it---why it's generally tapered so there is a range of pressure/speeds in it.

[kcrucible]

Ah... so as long as there's a taper, the flare shouldn't be a part of this problem at all since the speed adjustment is all post-pipe? I was always curious about the point of the taper, but what you say makes sense. Ok good, I can ignore that and concentrate on the air intake then.

Thanks Thomas.

[Fe-Wood]

The other issue I see is the coupler in the middle of the tube, that should be a smooth section. The turbulence caused by that transition could be enough to cause the back pressure and pre-ignition you are getting.
I can't remember where but there is a formula for length of tube based on diameter. When I made my burners, I smoothed out the inside of the tubes to increase flow and reduce turbulence, they work great!

[kcrucible]

Fe-Wood, on 12 August 2010 - 11:19 AM, said:

The other issue I see is the coupler in the middle of the tube, that should be a smooth section. The turbulence caused by that transition could be enough to cause the back pressure and pre-ignition you are getting.
I can't remember where but there is a formula for length of tube based on diameter. When I made my burners, I smoothed out the inside of the tubes to increase flow and reduce turbulence, they work great!

Actually, that's not a coupler, it's a choke that slides along the outside MADE of a coupler (ground out the threads.)


Yes, I found the formula you're talking about at Ron Reil's site... 9:1 ratio of length to diameter from the tip. So, for a 1 inch diamter tube you'd have 9" of pipe to mix in. That's a good part of why the jet is placed where it is. I can probably afford to sacrifice an inch though . Apparently other people have managed somewhat smaller without toooo much impact. At 6-7" burner performance gets worse though. Maybe 8"-8.5" is acceptible.

You do bring up a point about making sure that the interior of the tube is clean of ridges. I'll have to check that out. Thanks.

[Sweany]

From the pic it looks like the Mig tip is too far back.
Where the tip is now I don't beleive it will draw enough air...

[Fe-Wood]

kcrucible, on 12 August 2010 - 07:52 AM, said:

Yeah, I can move back the tip a bit by heating up the silver solder and sliding it backwards... good idea for increasing air intake without needing to increase the slot size. Definitely easier!


I don't think it's likely to change the flame location though. I just had another idea... the slot channels are rounded on the interior. Maybe that's not making the flow better as I had hoped, instead introducing vortices right at the end of the air intakes, hurting flow?

A note on silver solder-
I used silver solder for my first set of jets. When I shut down the forge after the first long run time, I didn't close the air intake ports and the escaping heat up the pipe caused the solder to melt. I have since gone to threaded Brass.

If you look at ron's T-Rex burner above, his air intake slots appear to be almost twice what yours are as I beleeve he has 5 or 6 slots. When I made mine I used 6 slots with the inlarge bell design shown on Ron's burner.

[kcrucible]

Experiment #1:

Adjusting the tip forward doesn't change the "combusting inside the tube" scenario, it just moves the point of combustion a couple of inches further up the tube and sputters like crazy. So, not simply a matter of the jet causing turbulance by brushing against the air intake. It does, however, totally kill burner performance.


More experiments tomorrow.

Quote

If you look at ron's T-Rex burner above, his air intake slots appear to be almost twice what yours are as I beleeve he has 5 or 6 slots. When I made mine I used 6 slots with the inlarge bell design shown on Ron's burner.

That particular burner is the 3/4" burner, which he flares out to 1" (I think) for the air intake. That particular one only has 4 slots as near as I can tell (can't see metal on the other side of the slot... so the slot is 4way symetrical). His larger burners have more slots. It looks pretty comparable to mine I think. He just uses the bigger jet in the smaller tube by expanding the air intake?

In any event, at the original position of the jet needed just a little bit more air, but that's fixable. I can easily widen the slots without compromising anything I think. This premature combusting thing is the first order of business though. Whether there is enough or not enough oxygen, it shouldn't even be trying to fire up till after the pipe.

[Robert Simmons]

From my experimentation I am pretty sure that it is the lack of the constriction that is your problem You want the gas to be blown into a conical constriction. This would take advantage of the venturi principal and speed up the gas flowing through the tube into the burner. Right now your mix is taking place early and the burn is taking place immediately on the end of the supplied mix. So your flame front velocity exceeds your gas velocity and that means it will burn back until it runs out of air to mix with which is justi n front of your injector.

You will need to have the gas get injected before a restriction and then you can asdjust how far from the constriction. You can accomplish this as I did with a reducing pipe or you can use castable refractory and two cones to cast the venturi right into the burner.

[Sweany]

How far forward did you go with the tip?
Note on the Trex that the tip is past the air slots. I think you have plenty of air.

You might try some a wrap of duct tape on the slots, to get your burner more into the T-rex design. this should tell you if the tip is to far from the tube.

In my experience adjustability is vey good in this area. Make sure the tip in centered in the tube also, from the looks of the flanme though, it seems centered.

[Robert Simmons]

Sweany, on 13 August 2010 - 07:13 AM, said:

How far forward did you go with the tip?
Note on the Trex that the tip is past the air slots. I think you have plenty of air.

You might try some a wrap of duct tape on the slots, to get your burner more into the T-rex design. this should tell you if the tip is to far from the tube.

In my experience adjustability is vey good in this area. Make sure the tip in centered in the tube also, from the looks of the flanme though, it seems centered.

The T-Rex has a machined internal venturi. Without a venturi, it wont matter as the flame front velocity will outrun he gas velocity. Look at the pic above and you see the constriction on the burner tube? There is your venturi.

[kcrucible]

Quote

How far forward did you go with the tip?

The tip was a little more than .5 inch from the end of the slot, which was as far forward as I can go with the nipple that's currently in place.

Robert Simmons, on 12 August 2010 - 11:25 PM, said:

From my experimentation I am pretty sure that it is the lack of the constriction that is your problem You want the gas to be blown into a conical constriction. This would take advantage of the venturi principal and speed up the gas flowing through the tube into the burner. Right now your mix is taking place early and the burn is taking place immediately on the end of the supplied mix. So your flame front velocity exceeds your gas velocity and that means it will burn back until it runs out of air to mix with which is justi n front of your injector.

You will need to have the gas get injected before a restriction and then you can asdjust how far from the constriction. You can accomplish this as I did with a reducing pipe or you can use castable refractory and two cones to cast the venturi right into the burner.

I do believe that you're correct regarding gas velocity. I came to the same conclusion last night.


I cleaned up the tube (the interior was ROUGH... like 50 grit sandpaper) and cleared out the jet a bit (a small bur inside.) Still the same basic behavior. Then I removed the flare... the gas stayed outside the tube for a couple of seconds that time then..... WHOMP, the flame popped back up the tube.

So, I'm left with fuel/gas speed being insufficient to push the burn outside the tube.


How to accomodate that? Here's what I'd come up with:


1) Smaller diameter tip. The M. Porter book lists 2 sizes for a 1" tube. I had chosen the bigger size (45) as it was available locally, but the bigger size means lower velocity. The other size (35?) would give a faster jet of propane, which would also pull in more air. It's hard to read, anyone make it out better or have personal experiance in size choices?

14T-35 Tweco Contact Tips (1140-1302)Tapered .035" contact tip for all Tweco® #2, #3, #4 style guns.

14T-40 Tweco Contact Tips (1140-1303)Tapered .040" contact tip for all Tweco® #2, #3, #4 style guns.

14T-45 Tweco Contact Tips (1140-1304)Tapered .045" contact tip for all Tweco® #2, #3, #4 style guns.



2) Bigger slots. If the slots aren't allowing the jet to pull air in as easily as it would like, then the suction that results would be a drag on the air that does go down the pipe, causing the velocity to be slow.


3) There is an implication in the M. Porter book that the 1/8" pipe nipple is an "accelerator." That you want this to be as long and unrestricted as possible so that the pressure will accellerate the propane to maximum speed before it gets constricted and ejected from the tip. I blew off this observation, noting that the original reil burners, with an effective length of 0 worked. Still, maybe something to it. My pipe nipple is 2", though I've got another 2" or so in various adapers before that... they may not count.

http://www.abana.org...bteaser_mig.pdf

Quote

the most important section for acceleration is the last few inches
of the accelerator assembly. The pressurized gas is also gaining
momentum in the pipe portion of the accelerator assembly. It
takes between three and four inches of pipe length for the gas to
reach full velocity before it encounters the contact tip. A short
pipe on the accelerator will ruin burner performance (the
advanced accelerator is a partial exception to the rule).

4) One other bit of info I gleaned from the Amazon version of the book is that square slots function better (less drag) than rounded slot ends, so that's another easy-to-test thing.




You suggestion about the cone really speaks to #2, essentially. Easier airflow. Ultimately, no more air is going to go through the tube than a given amount of suction/venturi/speed can pull into the burner tube. The cone is mainly about more air draw surface as far as I can tell.

I'd seen a reference to a venturi insert once, but would need to research more before going that route to see if it's even possible to retrofit in this burner. Basically just put the venturi at the end of the tube (instead-of/in-addition-to the flare?) so that the flame can't come back inside due to the speed at the point of contriction?



Does anyone have any comments about 1 or 3, specifically? Comments about 2 are welcome too.

[Robert Simmons]

kcrucible, on 13 August 2010 - 12:44 PM, said:

You suggestion about the cone really speaks to #2, essentially. Easier airflow. Ultimately, no more air is going to go through the tube than a given amount of suction/venturi/speed can pull into the burner tube. The cone is mainly about more air draw surface as far as I can tell.

Not really just easier airflow or the larger the venturi the better the torch would perform and that isnt always true. I have tried all sorts of combos. By the Venturi principal mathematics, the velocity is porportional to the constriction ratio. Think of it this way, in the large size of the cone there is X area capable of pulling some volume of air per second. When you introduce the venturi, that volume of air must still be pulled through the pipe but in order to do so it MUST speed up because air flow is a function of velocity times area. The gas pipe is firing propane gas into the tube and accelerating, generating a low pressure area in the pipe. The surrounding air must compensate but the pressure is all forwards so air must come in the back. The air rushes in from the back to fill the low pressure and then must be accelerated to maintain the same flow per second as it had when it was in a large tube. When the acceleration of the air equals the flame front velocity, you have a stable burner.

The real problem you will run into is when you stick that into the forge. The air pressure drops initially as it should and air is accelerated into the chamber but the problem is most of the air in our atmosphere is inert and doesnt burn. Only the oxygen burns leaving 79.05% or so of nitrogen and other parts blowing at high speed into the forge. This air has nowhere to go except out the door. So in the forge the air pressure will rise and eventuallt start to counteract the higher airpressure on the other side of the venturi. The back pressure eventually cancells out the venturi pressure and the flame burns without enough oxygen (rich) or goes out altogether.

So then the race is on to find a way to inject more oxygen while keeping the back pressure from cancelling out the venturi. If you can put enough oxygen in to get a complete burn of the propane then back pressure wont be an issue because the propane wont be able to slow down as it is consumed so fast, furthermore the high temperature in the forge will reduce the pressure of the air.

DO NOT TRY TO INJECT AIR FROM A WELDER'S O2 BOTTLE! If you do, we will be reading about your house exploding on the evening news. You do not want to go down that route.

My solution was to try to go to a blown burner to counteract the back pressure and feed the flame more o2 at this high altitude. There were numerous problems with flame front velocity not being fast enough to counteract the blown force. After thinking about it, it doesnt surprise me because again 79.05% of the gas is inert. What I needed was a way to introduce more air without pushing the flame front off the end of the torch. Hence the idea to introduce the air directly into the chamber.

kcrucible, on 13 August 2010 - 12:44 PM, said:

I'd seen a reference to a venturi insert once, but would need to research more before going that route to see if it's even possible to retrofit in this burner. Basically just put the venturi at the end of the tube so that the flame can't come back inside due to the speed at the point of contriction?

You would get a venturi there but you would find the gas doesnt have enough time to accelerate and mix with the air. Also remember you want that jet shooting into the cone so the gas is sped up as well. The pressure from the bottle is the driver here and the gas will have already slowed down considerably before it hits the venturi on the end. In addition you will need something like a flame stabilizer on the end. Flares are all well and good until they go red and then yellow in the forge. A flame stabilizer will negate the need for the flame holder and then it will slow down the mix by means of turbulence on the side while allowing the center mix to get through unobstructed. The flame will be stabilized because the gas will want to push the flame front off the end in the middle but will be too slow and want to burn back where the turbulence is. The balance stabilizes the flame just a fraction of an inch off the end of the tube.

P.S. When you put the burner in the forge, everything will change and leave you thinking a lot more.

[kcrucible]

Robert Simmons, on 13 August 2010 - 01:30 PM, said:

Not really just easier airflow or the larger the venturi the better the torch would perform and that isnt always true. I have tried all sorts of combos. By the Venturi principal mathematics, the velocity is porportional to the constriction ratio. Think of it this way, in the large size of the cone there is X area capable of pulling some volume of air per second. When you introduce the venturi, that volume of air must still be pulled through the pipe but in order to do so it MUST speed up because air flow is a function of velocity times area. The gas pipe is firing propane gas into the tube and accelerating, generating a low pressure area in the pipe. The surrounding air must compensate but the pressure is all forwards so air must come in the back. The air rushes in from the back to fill the low pressure and then must be accelerated to maintain the same flow per second as it had when it was in a large tube. When the acceleration of the air equals the flame front velocity, you have a stable burner.

But it can still only speed up to match the propane jet. The suction that the jet creates dictates the speed, and thus the volume, of air in the chamber. Whether that volume of air originated in a 3" chamber or a 1" chamber would seem to be rather immaterial? The cone feed might make it easier to get up to matching the jet speed, causing less vacume-drag. That's the only advantage I could possibly see.

Quote

The real problem you will run into is when you stick that into the forge. The air pressure drops initially as it should and air is accelerated into the chamber but the problem is most of the air in our atmosphere is inert and doesnt burn. Only the oxygen burns leaving 79.05% or so of nitrogen and other parts blowing at high speed into the forge. This air has nowhere to go except out the door. So in the forge the air pressure will rise and eventuallt start to counteract the higher airpressure on the other side of the venturi. The back pressure eventually cancells out the venturi pressure and the flame burns without enough oxygen (rich) or goes out altogether.

One difference is that I'm actually building a foundry furnace, not a horizontal forge, so I can take advantage of the chimney effect. The hot gas will rise and go out the exhaust naturally, pulling the lower gasses along after it. I wouldn't expect a large increase in air pressure after lighting.

Quote

You would get a venturi there but you would find the gas doesnt have enough time to accelerate and mix with the air. Also remember you want that jet shooting into the cone so the gas is sped up as well. The pressure from the bottle is the driver here and the gas will have already slowed down considerably before it hits the venturi on the end.

But if you shrink down, then come back to normal size (say, 3" total venturi tube length), the velocity will be reduced again to what it was before. Sucking enough air volume isn't really my main problem... it's speed of the mix. The conclusion is that you stay at the smaller size until you're ready to exit the tube so that the burn-back doesn't happen?