Uflow assisted PJ design

[Dave_G]

I have now tested my first Uflow-assisted design of a valveless pulsejet. It is an Escopetta variant, with an approximately 97mm diameter combustor and an OAL of approximately 1800mm. The basic design parameters are the result of some recent research and testing, and I used Uflow to examine the proposed design and verify that the engine wasn't somehow fatally flawed. Bill hinote has been of exceptional assistance in this regard, and has quite literally been mentoring me in Uflow. Bill, I am especially grateful! Bill has also been quite careful to educate me regarding the limitations of Uflow. Altogether, I have found it to be a helpful tool for design assistance.

This engine currently produces 24 lbs continuous thrust on propane vapor. Not great, but okay with me, all things considered. I expect this to go a bit higher when I can feed it more propane. I have yet to take video, but I will be doing additional testing through the coming weeks, hoping to drive the engine to rich extinction (inadequate propane flow is the current holdback). I have one other engine that I am working on that I need to finish first, and then I will return to this one and post video.
Once again, Bill, thank you very much for your help.

Dave

[hinote]

I have now tested my first Uflow-assisted design of a valveless pulsejet. It is an Escopetta variant, with an approximately 97mm diameter combustor and an OAL of approximately 1800mm.

Dave:

Please let me be the first to congratulate you, on your success!! We can hardly wait to see some images of the beast.

To the rest of the Forum:

I have been vaguely aware of this project, in my communication with Dave; his success is entirely the result of his own efforts, and should be recognized as such. This is quite an accomplishment, considering it's his first effort with UFlow.

[NanoSoft]

Very nice indeed. Can't wait for more details and of course some good pictures of it in action. Best of luck with further testing ;)

mikhail jones

[hinote]

I have now tested my first Uflow-assisted design of a valveless pulsejet. It is an Escopetta variant, with an approximately 97mm diameter combustor and an OAL of approximately 1800mm.

This engine currently produces 24 lbs continuous thrust on propane vapor.

Here's some initial designer's impressions, without additional data from Dave's effort:

1. The indicated thrust is quite excellent, and consistent with SNECMA's quote of 14 kgf from their 120 mm combustion chamber diameter.

2. IMO the stated OAL of 1800 mm may be too short; in combination with the first thrust results, this indicates a possible compromise in optimizing the various operating parameters of this particular design.

3. OTOH if Dave's motor can produce decent TSFC along with his thrust results, we may be seeing a breakthrough here.

I hope Dave doesn't mind me anticipating further results from his testing. This particular motor is my current specialty--and I'm excited to have another designer participate in its development.

I believe the Escopeta-derived VPJ motors will represent the forefront of PJ design developments in the near future.

P.S.: Dave: Thanks in advance for letting me invade your thread; I hope you can bask in the glow of your accomplishment. I promise to keep my d!@#$ed mouth shut, unless you specifically ask for my presence!

[NickC]

I've got a question about calculating the speed of the pressure waves so you can compare frequencies of different sized motors. Is the overall length of a valveless pulsejet equal to one half a wavelength that the frequency refers to?

Also, where can I download Uflow?

[Dave_G]

Bill,

Your comments are not at all an invasion! And your critiques are most helpful and appreciated. I could not have done this without all your help!

And you are right about the length. I made a mistake in my post; the engine OAL is 1.98 meters, not 1.8 meters (didn't have matls w/ me when posted). This is perhaps still slightly too short by your reckoning, but I wanted to correct my mistaken post. Thanks for bringing it up!

Nick,

To answer your question, yes, but remember that wave velocity is a function of temperature, and real temperature within an actual motor is partly dependent on design and timing. Also, effective acoustic length corrections are required when calculating overall lengths. I'm going to defer to the expertise of others to address these acoustic questions for two reasons: One, I don't want to inadvertently betray any confidences that may have been placed in me, and Two, there are other forum members that are far more competent than I to answer questions regarding acoustics.

You can get some basic acoustic formulas here (and forgive me if you've seen this already):

http://hyperphysics.phy-astr.gsu.edu/hb ... tml#soucon

And, I got Uflow here:

http://www.imst.upv.es/uflow.php

The English version page doesn't work, you'll have to use the Spanish version. Fill out the form and you can download Uflow for free. I don't speak Spanish (just enough to get in a fight), but the download request form is pretty simple.

Be prepared to have Uflow crash a lot and you won't be disappointed. The causes become apparent pretty quickly and you can learn to avoid them.

Dave

[Mike Everman]

Wonderful, Dave!
That is a very respectable thrust for that size duct! Really looking forward to video.

Nick,
If you have trouble with the uflow download, Al Belli posted the whole thing here:
http://www.pulse-jets.com/phpbb2/viewtopic.php?t=2406

[leo]

I like the animation function in Uflow here I try to post a screen record off it.

[Dave_G]

Here's the promised video of my new engine. I've designated it the R1.

The engine is being operated on propane vapor. My fueling setup uses a ball valve (wide open) in series with a needle valve to control flow. The ball valve is for safety so I can instantly shut off any motor under test if it flames out.

For the last six seconds of the run, the engine is running at maximum flow with the current fueling setup. The engine does not reach rich extinction. It stops running in the video when I rapidly shut off the ball valve after it has run a few seconds at full power. My next experiment will be to supply it with additional fuel to find the rich limit and the associated thrust value.

Maximum thrust is currently running at 26.5 lbs continuous. I filmed the thrust gage with a second camera during the same run with the intent of showing it also, but it's in a format I can't edit and the video is too long to post. I had to make a new thrust stand for testing this motor after it bounced over my old one. As can be seen from the video, I need to go back to a thrust stand that uses linear bearings or captured rollers as opposed to the oven-rack-on-wheels used for this test. The fact that the rolling elements are not captured allows the engine to fishtail and bounce around as it's powered up. At full power it settles back down.

I've made several compromises to performance with this engine. One of those compromises was a segmented U-bend, instead of a smooth formed one. It is shown in the attached photo. Also, the bustle is much smaller in diameter than it could have been; this was done to use an existing piece of tubing. The exhaust throat is not visible in the video, being shrouded by a combustor extension that also serves as a fuel preheat section. I hope to run this engine on liquid fuel soon.

The theoretical power limit for an engine this size is, I'm told, about 28 lbs. thrust. I'm looking forward to seeing how much more (if any) thrust I can wrangle out of this motor.

[Dave_G]

BTW, I've done some experimenting with the inlet of this motor, slightly altering inlet length as I experimented with different flares and bellmouths on the inlet. The Overall length as shown in the video is 1899 mm. I rechecked the combustor diameter and it is exactly 97.5 mm (previous numbers were listed as 'approximate'). I have not yet checked fuel flow rate -- I've been focusing on thrust and its' relation to inlet geometry -- but I will measure TSFC soon. I'm not expecting anything great -- honestly I wouldn't be surprised if the TSFC is around 4. I have no guess as to the actual consumption yet, however.

[Eric]

Unless you are inverting your tank or sticking it in a tub of very hot water and using quite large fuel lines you wont get enough fuel flow for a 26 lb engine to run at a TSFC of 4 for any length of time before fuel pressure drops, and thrust decreases greatly. Since your not getting flame outs, I would suspect that you are running with the tank upright.

From the looks of it, it is rather similar to the length proportions and cross sectional ratios of a "classical chinese" If you got a TSFC way above 2.0 I would be quite suprised.

Segmented pipes dont affect performance greatly, if you use a smooth pipe it would have slightly better thrust / tsfc performance but not a lot.

[Dave_G]

Tank's upright, vapor feed only. And yes, you're right, the tank would empty pretty quickly if the TSFC was 4. I really need to measure the flowrate.

Audio waveform display is attached. Upon looking at it, the operating frequency seems to be around 131 Hz. It appears that every other wave represents a 'cycle'.

The engine was running a little rough, as can be heard in the video file. I'm not going to complain about this, considering the thrust. I made two runs prior to the videotaped run, and they were all very consistently at 26.5 lbs steady thrust (not peak).

It's nice to know that segmented U-bends are a viable option. They're easy to make, and choice of available diameters is not a limitation as it is with purchased U-bends.

[hinote]

Audio waveform display is attached. Upon looking at it, the operating frequency seems to be around 131 Hz. It appears that every other wave represents a 'cycle'.

Dave:

I made some assumptions, and it looks like your acoustic temp is about 645-650K.

My guess is that you're going to see a TSFC of about 2.5 or so; injector design can improve on this a lot.

In fact--this is a good argument to consider: Are we better off improving max thrust density in the motor cofiguration, and then improving the TSFC with ancillaries like injectors and augmentors? It's an interesting thought--and considerable data acquisition would be necessary to create the graphed lines for the variables, which would then yield the exact point where the tradeoffs would be best optimized.

[Dave_G]

Thanks Bill!

I'll experiment with a few changes and see what I can do to improve it.