From Russia without Valve

[AlexanderAndrosov]

Hello Everyone, there will be a log of my and my comrades' attemts to build a valveless engines.
My name is Alexander, I am a postgraduate student of turbine engineering. Maybe the pulsejet will be a subject of my Ph.D.

FIrst of all: I collect a huge collection of documets/articles/photos/textbooks/blueprints/3D-models. The main part is from this site, but many are from different sources. I will share with you with an up-to-date version. Keep in mind that you won't find any order there, it's just a pile of everything I found. I'll organize there...never, most likely. Anyway. see the up-to-date version here: https://drive.google.com/file/d/1JxaNa3 ... share_link
P.S. My models and blueprints are most-likely .m3d and .cdw. I use KOMPAS-3D(russian AutoCad).

Secondly, what we already did...see the attachments! There is a batch of "Enics-type" VlPJ.
Thirdly, what I need. The blueprints of large PJ of course! I'm interested in big toys, 15kg(33lb)+.
And also I need your help in numerical modeling of PJ. I use ANSYS CFX/Fluent, but my level is brainless repetition of hindus from youtube.

You can always contact me here, by e-mail AlexanderAndrosov1998@mail.ru and in Telegram https://t.me/AlexanderAndrosov1998.
See you soon!

[Mike Everman]

Welcome, Alexander!

[tufty]

Hello Alexander. Nice looking jet, good work.

Numerical modelling of pulsejets is something that's been tried quite a few times, with varying levels of success. It's enormously complex and I'm not sure it will ever be cracked 100% - I suspect most "design" is a lot of copying wa's already been done, with a good dose of "that looks about right", and then the usual cut / weld / test cycle.

[VicHanby]

Hi, A good few years ago I had a very talented research assistant who I had to let go due to lack of funding to keep her on. She was Gang Li and went to work for Ricardo on the south coast of England. Before she left I had her construct a 2-D model of a Schmidt-tube (using axial coordinates). The model had an inlet valve with a quadratic resistance when open. She used CFX-4. The key to getting it to pulsate was to replace the Damkohler number to initiate combustion (the default) with another ratio which I'll have to delve into my notes to refresh my unreliable memory (at 81).
I've a pretty powerful iMac upstairs which I don't use any more as it runs very hot and sometimes cuts out. i'll happily fire it up again and see if I've got any of the colour pics and pressure traces from the output. I didn't publish the work as I was to preoccupied with other things to do any additional work - I was head of a large engineering department at Loughborough university at the time.

I'll get back on here if I can find and output, meanwhile keep going. I believe that a CFD model of either a valved or valveless pulse jet is a noble objective. I'd really like to be informed of your progress.

[Mike Everman]

Hi Vic, I would very much like to see that. Maybe start a thread in the Valved fora?
I enjoyed my time in Loughborough. Do you live thereabouts? I did some fun work for the Int'l Cricket Council at the National Cricket Performance Centre there.

[AlexanderAndrosov]

Hi, A good few years ago I had a very talented research assistant who I had to let go due to lack of funding to keep her on. She was Gang Li and went to work for Ricardo on the south coast of England. Before she left I had her construct a 2-D model of a Schmidt-tube (using axial coordinates). The model had an inlet valve with a quadratic resistance when open. She used CFX-4. The key to getting it to pulsate was to replace the Damkohler number to initiate combustion (the default) with another ratio which I'll have to delve into my notes to refresh my unreliable memory (at 81).
I've a pretty powerful iMac upstairs which I don't use any more as it runs very hot and sometimes cuts out. i'll happily fire it up again and see if I've got any of the colour pics and pressure traces from the output. I didn't publish the work as I was to preoccupied with other things to do any additional work - I was head of a large engineering department at Loughborough university at the time.

I'll get back on here if I can find and output, meanwhile keep going. I believe that a CFD model of either a valved or valveless pulse jet is a noble objective. I'd really like to be informed of your progress.

Thank you for your reply, Vic. It seems to me that I am not your talented researcher, because almost everything you have told me is a gray area. But I'm as interested as Mike is.

I started cheerfully and was silent for a long time. Although, probably, this is normal for this forum. The boss suspended our "das raketenprojekt" in favor of more work on new orders. Western countries have suspended the supply of spare parts for turbines, so our plant is loaded with work for once. And I'm busy "very interesting" designing filters and a ladder on wheels for an expanded warehouse.

In vacations between my very important and interesting work I tried to model pulsation non-premixed combustion in Ansys, using video tutorials from youtube, but there wasn't a PJ-type combustion tutorials, only final results. So my attempts are still unsuccessful.

But what I did. See the attachments.
I tested in Ansys this Telsa-Aerovalve construction. I just set some boundary conditions, get the first mass flow, swap the inlet and outlet and get the second mass flow.
The return flow was 39% less than the direct one. There is a flow pattern in the images. As for me, such an air supply gives a good annular turbulence of the flow, which is useful for the formation of a fuel mixture.

By the way!
I couldn't replace the archive with my collection, so a fresh version of this monster is here: https://drive.google.com/file/d/1t5ePnu ... drive_link

[VicHanby]

Hi Vic, I would very much like to see that. Maybe start a thread in the Valved fora?
I enjoyed my time in Loughborough. Do you live thereabouts? I did some fun work for the Int'l Cricket Council at the National Cricket Performance Centre there.

Hi Mike, During my time at Loughborough I lived in Barrow-upon-Soar, then Thrussington, both handy commutes by bicycle and motorcycle. Two years ago we moved down south to Tring in Hertfordshire. I've had no success in finding the output from the CFX simulations, I've a feeling they were on a Sun workstation and have just got lost. The Cricket centre came after I defected to De Montfort University in 2000 I think, did you enjoy your time there?

[Mike Everman]

I did. Mostly just work, though. I made a robot with a swingarm that would get a ball up to 100mph so we could get a controlled tick against an instrumented bat.
The final verdict was, if the microphones there don’t hear the tick, it didn’t happen, but if they hear it, it did not necessarily happen.
Within a millimeter made a similar sound that required finer descrimination.

[AlexanderAndrosov]

Guys, what do you think about palabolic reflectors for a dead-end parts of Enics-Type I've posted and another PJ's ?
See the attachments and links:
https://www.youtube.com/watch?v=BEjLEHTWaFk
https://www.geogebra.org/m/QEcTJmfE

Will it be enough just to make a backwall this way or for more efficiency it will be good to place a fuel injector/sparkplug into a focus of parabola as the center of detonation ?
I guess it's okay to weld a segmented parabolic wall , but is it worth it ?

[tufty]

There was a time when I would have said yes. These days, I'm more of the opinion that the inherent randomness and turbulent mixing in a pulsejet will mean that, at best, it makes no difference.

"Theories look great on paper, until reality comes along and scribbles all over the page"

[AlexanderAndrosov]

Hi Tufty.
Yes, this is an age-old conflict of interests between theory and practice. I realized this, first working as a worker in factories, and then as a design engineer, when a technologist or a worker beat me with my own pack of drawings when the product is not assembled, because I did not provide for mounting gaps at all.
Anyways, the first detonation looks nice. I used this resource: https://wavesimulator.starfree.app
It's not advanced CFD-model, but the first pre-CFD prediction.
Than deeper is parabola and its focus, then better reflection efficiency. So the engine becomes egg-head.
I hope it may be possible to achieve a quasi-stable cycle. When we'll do the next prototype, I'll order a paraboloid scan. We cut out the parts on a laser machine anyway. But it seems to me that'll be next year.

[Mike Everman]

Looking good so far.

As to the parabolic or other shape to the closed end,
I think a flatter wavefront coming off of some convex shape would be beneficial, and I've seen so many tries at that, with little evidence that it mattered. I think every designer goes there at some point.

For my own work, I would really want that flat wavefront after reflecting, so transitions further down the duct that are there to reinforce the lower harmonics will be hit sharply, not smearing over them with the slope of a convex wavefront, softening the effect of the transition.
Let's just think about the rarefaction wave reflecting back from the opening, My mind's eye can see different wavefronts later encountering this closed end, based on how long the pulse has traveled from whatever event down the duct. A rarefaction wavefront close to the opening is flat, but the longer it propagates, the more it becomes hemispherical. I think we can concern ourselves more with pressure waves that travel further.
At times I believe the proper shape will be something that flattens a hemispherical wavefront when it reflects, but again, the longer that wave propagates after reflecting, the more convex the wavefront becomes, until it is hemispherical again.
We see this in all simulations, and so, what do we gain by briefly flattening the wavefront? Maybe we tailor this shape to send the reflection out convex, and have it become flat further down the duct, at a transition that amplifies one of our lowest harmonics.

We've also theorized around here what the geometry might be like if that end of the duct came to a conical point, changing the sex of the acoustic wave that hits it (but not the pressure wave, I imagine). When I say acoustic wave, I mean maybe wrongly the velocity wave, and the pressure wave I liken to the massflow, and they are 90 degrees out of phase with each other as I recall. I am also likely to be all wet with this "understanding".

[Mike Everman]

I think the best shape is elliptical, specifically a 45deg ellipse whose major axis is the CC diameter. At least, if the goal is making a flat wavefront according to my last message. Admittedly a very simplistic point of view.

[AlexanderAndrosov]

We've also theorized around here what the geometry might be like if that end of the duct came to a conical point, changing the sex of the acoustic wave that hits it (but not the pressure wave, I imagine). When I say acoustic wave, I mean maybe wrongly the velocity wave, and the pressure wave I liken to the massflow, and they are 90 degrees out of phase with each other as I recall. I am also likely to be all wet with this "understanding".

When you are wet with this understanding

(Although I may misunderstand your idioms)

Yes, a noticeable disadvantage of the paraboloid, if you do not make a rounding transition, is that it will never be tangent to the walls of the combustion chamber.

What about elliptical in common and spherical as a special case of the first.
If to use an ellipsoid/sphere just as reflector, then even flat wave will become spherical:
https://www.awesomescreenshot.com/video ... 1037776635
If to place a detonation center into an ellipse's focus, then it will transfer a wave to the second focus:
https://www.youtube.com/watch?v=vlwkpXrH_Jc
It's okay if you want to burn ants using a lamp, but whats useful for PJ's gas dynamics?
Spherical case doesn't have a single focus, but the entire focus line:


Maybe it's even useful when detonation center unpredictable happens somewhere in combustion chamber.

[AlexanderAndrosov]

Just cone end is also cool. But parabola is still the best if to force someway a detonation center to be in it's focus.
I'll think about how to achieve this using some geometric manipulations, something like in the attached .pdf

What a pity that you have to do work at work! (by the way, if you need filters for turbines and large-capacity machines, you're welcome to contact )