Big Test Lady FWE arrives in Aero/Combustion Laboratory

[Dang911]

First and foremost, I would like to thank larry cottrill who provided me with a 90% complete FWE engine built with my supplied pressure taps for instrumentation. Upon receiving the engine, all that needed to be done was some light grinding, fitting the exhaust pipe, and running fuel/air line and all auxiliary instrumentation. There was some speculation on the length of the tail pipe and several options were suggested as means to tune the engine if it weren't to run right away. I decided since the exhaust tube fit well into the socket, to go ahead and just nicrome spot weld the tail pipe in three places to hold it in place. Nichrome can be easily chiseled off if I were to want to remove the exhaust for modification. I have a few pictures of what I have to work with.

Tomorrow the engine will possibly run without instrumentation in the cell and if all goes well I will begin to string it like a Christmas tree with type K TC's and pressure transducers and best of all one or two dynamic pressure transducers suited for combustion acoustics (http://www.kulite.com/products.asp). Originally I also wanted to place a few skin TC's to read the combustion chamber wall temperature, but I think the air temp will serve pretty close. Other good news is that I am now able to flow check the Simpl-Jector Larry supplied me. This will give me the effective area and a flow curve which should be VERY useful when setting fuel flow and pressures rates. It will also allow me to measure lean/rich blowout and optimal starting conditions. There is a chance I may be able to use a very expensive but "proper" pilot primary from a turbine for my fuel delivery, in which case I will have even more information on flow rates, cone angles...etc

I think that's enough for now....

[larry cottrill]

All right, Dang -- NOW you've got something that looks like an engine!

I find special amusement in the hardware required to hold the fuel pipe in place. I want to shake the hand of the guy that engineered that ;-) I think she looks really good with all those "calibrated leaks" carefully capped off, too.

I would never have understood what you meant by "nichrome spot welding" without your closeup of it. Elegantly simple, probably overkill in this case if the pipe fits snugly, since I think all there is acting on the pipe (in a rearward direction) is basically hot gas drag forces, probably measurable in ounces. But, it certainly looks like an effective way of making sure nothing loosens up from vibration!

Can you go into more detail about the kinds of analysis available? Is there any capability of exhaust gas analysis for presence of pollutant chemicals, etc.? ("They always want more." ;-) For the "dynamic" transducers, I think the most interesting places to start will be at the "zero percent" points of the chamber (out in the dome) and the intake (right where the intake pipe spouts into the transition cowl). Results there might suggest other interesting spots, like the nozzle point of the chamber or whatever. (Note that there are only a couple of stations where temps and pressures can actually be measured simultaneously; namely, the front dome and the nozzle point.)

Will you be adding any fittings to the tailpipe itself? It would be fun to see if the coming and going of the tail "piston" can be verified!

Great stuff, Greg -- fire that sucker up!

L Cottrill

[PyroJoe]

Shhhhweeet,
I agree with Larry, fire that puppy up!
Joe

[larry cottrill]

Heard from Greg a little while ago. They had to spend most of the day tearing down a government project test they just wrapped up; once they get that done, the next step is some basic flow measurements on the fuel pipe I sent along. Nobody's really sure if that pipe is big enough, so it's important to measure the characteristics and be ready to substitute something else if it doesn't seem adequate. Even if it works fine, flow measurements are an important first step in setting up the engine tests.

L Cottrill

[Dang911]

Hey now, we had to use a 1/2 fitting so the pipe bend radius would clear and then because it’s a compression fitting if you wanted to be able to take it apart it needed to be a male thread welded to the engine - so that's the most efficient way to step from a 1/2 to a 1/4. We do that all the time on rigs.

I think you mean burned up, instead of wrapped up, parts got melted, the test engineer pushed the rig too hard and ultimately the combustion acoustics (singing flames) were strong enough to break a pilot primary fuel line causing a temp spike in one of the cups and it burned through the liner and nozzle ~3500F.

Anyway, yes we have the capability to measure emissions that is a major focus for many of the tests I deal with. It requires a gas/liquid chromatography machine and emission sampling rakes, which are water-cooled. The sample emission must be then kept hot with steam while it travels through the building to the actual sampling machine were it will analyze the sample (the sample has to be kept a certain temperature to preserve compounds and avoid the sample changing compounds through bonding) It’s a lot of setup and requires a special crew to operate, they have their own console they operate from. So yes, it would be neat to know but it is not going to be the focus for this test.

Fuel flow is one of my focuses because if/when all the data is collected we will be able to correlate fuel flow rates with the acoustic performance of the engine. This may lead to better designs in the future or at least designs where a fuel flow rate can be calculated and performance estimated, I am not sure if anyone out there has been calculating fuel pressures and flow rates curves based on design and not experimentation.

The nichrome (light bulb filament) material is great for spot welding to steel. We use it to hold down sensors and small pipe runs and we often nichrome a bolt head down as an ultimate vibration proof safety for internal test hardware, which receives a lot of combustion acoustics.

I may add a temp sensor at the exhaust (via nichrome most likely) I will not need pressure because it will be venting to atmosphere (what ever the barometer reads).

Looks like the fire-up may have to wait until Monday, but tomorrow I am confident I can complete the flow test and finish setting up the rig in the cell.

[larry cottrill]

Just received an email from Greg; he has completed flow rate testing of the .003-inch Simpl-Jector I sent with the chamber. He provided the spreadsheet data, but the attached graph from the XLS file pretty much sums it up. The device chokes at a little under 60 PSIG, but the graph is pretty smooth, much of it is practically linear, and it looks like a lot of propane to me! Haven't heard any judgment from Greg or their Chief Engineer yet on whether they think this will actually work.

L Cottrill

[Dang911]

I sent Larry the data seconds after taking it and it wasn't until I looked at the data and thought about the experiment that I realized I DID NOT flow the injector all the way up to being choked. I did however max out the instrument air supply for that particular test. By looks of the curve it was getting close but hadn't flattened off enough for me to call it a full choke. Larry I also have to correct you about calling the graph quite linear (although I know you meant the latter half), a typical flow graph (we fall in as very typical) tends to follow the shape of y=sqrt(x). Regardless of where the choke lives, I believe I covered most if not all realistic flow adequately. I re-posted Larry's graph with the addition of propane flow rate in PPH.

After the flow test I setup a bench flame propagation test. I ran the injector lit, through the typical flow range and it looked ok as far as the flame goes, but it was not a uniform burn and the flame was only propagated a little less than 180 degrees on what was theoretically a full annular design. After the flame test I re-flowed the injector on several points to ensure no thermal expansion had changed the affected area of the injector.

[Irvine.J]

This is a great thread thank you dang for your efforts.
I bet larry is smiling ear to ear!

[Eric]

Neat setup. Is the tail adjustable or just really short? From the camera angle it looks like it could use another 30% length.

Eric

[larry cottrill]

This is a great thread thank you dang for your efforts.
I bet larry is smiling ear to ear!

I've smiled quite a few times in the last few days. Before that, I was more concerned about destroying my marriage and/or losing my day job while trying to get the sucker built. I'm really pleased with how the engine turned out.

I am less pleased with the Simpl-Jector I sent along. The flow curve is cool, I think, but the defective pattern is painfully obvious, and is about what I predicted in my own criticism of the finished product. I wish I had gone ahead and done a couple by spot welding, even though I'd have to hire it done. As things stand, neither the Chief Engineer of the lab nor Greg is really very impressed with it. The one-sided "half cone" pattern is why I showed it in the plan lying low in the intake and spraying upward into the mid-pipe flow. They should have something lying around in a drawer somewhere that will do better, and perhaps my sending that along is just adding an element of confusion to the test process.

Neat setup. Is the tail adjustable or just really short? From the camera angle it looks like it could use another 30% length.

The tailpipe seems unusually short because of the intake length, and because this is the farthest rearward intake position of any FWE I've done. The intake ID is proportionately a little larger than on the original FWE design, so the pipe needed more length to get the impedance up where it needs to be. I was thankful for that length, in terms of getting all those fittings worked in. UFLOW1D and my usual approach with the internal temps claim it should be right on the money. It does look strange, though -- we'll soon see, I guess. See my original 'Big Test Lady' thread for the original dimensions -- I did actually add a few cm to the tailpipe since then, to correct for a slight but important error in my original UFLOW run.

L Cottrill

[Dang911]

I guess its a little on the short side, it measures just about 26" in length from the combustion chamber. If you look on the other thread Larry has plans posted. I hope I can get it to start with this length tube. It is only nichrome tack welded in so I could change it to a longer pipe, but I would try wrapping sheet metal around it first to lengthen it.

[Dang911]

Well I wouldn't say that I was unimpressed, we don't know that it wont be adequate to start the engine, and lets face it I am sure that there are plenty of sloppy injectors out there that are working for people. Not everyone does a flame propagation test.

[larry cottrill]

Last night I spent a little while trying to gently re-bend the outer tube of a defective Simpl-Jector to alter the cone pattern. In other words, the idea was you could bend the tube just enough that you would get the screw head re-centered properly. What I found was that re-bending works, but it is IMPOSSIBLE to get "just enough" to correctly center it -- no matter how careful I tried to be, the pattern was always lopsided one way or the other. If you could observe the end view with a low-power microscope (the 3D type where you observe with both eyes) and had a fine screw-type device (e.g. micrometer head) to do the bending very gradually, I'm sure you could get it. My crude methods and tooling are simply inadequate for the delicacy of this task.

I'm sure that my suggested alternate method, resistance spot welding, would produce a good one. Also, it might help to refine the flare inner surface to make sure it is a true concentric cone before mounting the screw. An ordinary tubing flaring tool would probably work for this, or you could lathe spin each tube in about a minute. The screws themselves are practically as good as lathe-turned parts, unless you happen to pull a bent one out of the box.

L Cottrill

[Dang911]

I'll keep this short, but I got her fueled up and ready to go today. She never self sustained today but I got a lot of strong grumbling out of her. I have began remaking new injectors one will be more or less an open tube, the other will have 4 holes over 2 inches drilled and the same pattern offset 90 and by half a hole with a larger hole in the end pointing forward. Hopefully that should yield better results. I am also thinking about changing where the fuel is injected. I think that has a strong affect on tuning. Lastly I noticed with the old injector that the further I put the starting air to the injector the better performance I got, I think that tells me I need better fuel air mixing and hopefully the new injector will do just that. All tests were on vapor and I don't see why it can't run on vapor. New injectors will have more flow and I will give flow test results if/when I get a winning configuration.

I am willing to take any comments/suggestions at this point because I have 4 days left to get it running to collect the data we are all anticipating.

[larry cottrill]

Well, it's painfully obvious that the skepticism re the Simpl-Jector was warranted. I'm sorry for the hours that messing with this has cost you; it would appear to be quite inadequate for a chamber of this size.

It occurs to me that this engine has an unusually long path from the forward end of the intake tube to the dome. I think a fairly fast, forward-spouting fuel nozzle might be the answer to this, i.e. try to get the mixing action much farther forward. My thinking is that we may be getting ignition of the air/fuel charge before it can get into the front of the chamber where the explosion should be happening. If that's the case, a larger side-spouting one won't work much better. On the other hand, if this "early mixing" hypothesis is wrong (i.e. inadequate fuel flow is essentially the whole problem), that could be just the ticket. So, I'm glad you are trying two approaches.

With your two injector designs, don't get hung up on the first one that seems promising. Instead, go back and forth a couple of times and see if you can discern the characteristics of each, really thinking through what you're observing. Only then, make a choice and see how far you can take it. It is observing and then THINKING THROUGH what you see and hear that will get it to yield to you.

I think we haven't seen enough quite yet to conclude that the basic tuning is at fault. It may well be, of course, but I don't think we are very close to knowing that right now.

L Cottrill