Rotary valve pulsejet?

[dominik]

Fan and shaft on one axis would be perfect for making timings (at first sight) as easy as possible, however the appearance of such an arrangement would be different from usual pulse jets. The fan needs to be very large in comparison to the combustion chamber, because one rotation of the fan (even less than one) must push out the used gas and bring in a fresh load. If fan and valve should be of similar size, maybe a gearbox will do the trick to avoid an uC for controlling.
Personally this arrangement is somewhat out of my knowledge of fluid calculations, so I would prefer the control by uC, the program is the easiest part and can be done with a small Attiny (or similar). And it allows to play with fan / valve speed and ignition point.
PWM generation can be done by hardware timer to get a constant signal for a brushless fan (I already bought a 64mm one with ~800VA)).
So the air buffer tube will be also this diameter. The combustion chamber will be slightly smaller (60,3 stainless steel tubing) and be conical reduced like other pulse jets.

I have some more points that are not finally considered by myself::
Which kind of one directional valve can be used for the fuel line to be fast enough but also strong to survive the detonations? Small ball valve made with a small bearing ball?
(I would like to start with propane)

How to translate the 5V ignition signal with minimum delay? I’m not used with the new electronic igniters for cars, can they be of use for it?

And: what will be the maximum frequency of such a device to avoid self ignition? Maybe cooling the combustion chamber?

Will it produce „usable“ thrust?

[racket]

Have you checked out this Site for info www.pulse-jets.com/

[dominik]

Have checked the link/the Forum and found the "rotary valve not working" thread.
The unstable frequency will be difficult to solve if running on resonance - ok!
So I have to discard the additional operational mode (with self ignition)


In the meantime I purchased a dual shaft nema11 for my project box, quite bulky but
I believe a nema8 will be too weak. And I have an encoder from my cnc-mill lying around
that has nearly the same size as the nema11 and will peferctly fit on the other end.

I also measured the EDF and found that it will fit into a 76,1 pipe. Actually considering
to dismiss the 60,3 pipe and go for 76,1 for the air-tube and the combustion chamber as well.
Reduction from 76,1 to 42,4 on the exhaust side.

So the general concept and dimensions are more clear for me. Now I need some assistance:

The airtube will be around 200mm length and will hold the fan, the encoder and the stepper, I plan
to fix a hexagonal (or squared) axis to the stepper to drive the rotor of the valve and have some
play in the direction of the pressure bursts. The airtube will end with the first stator of the valve,
which will also be the mount of the stepper. This stator will be welded to the tube and evened later on the lathe,
given some outer ring to bolt the engine together. After this first stator the rotor will follow, then a
secondary stator which is welded to the combustion chamber. I would give the rotor a clearance of some hundreds
of a mm.

Now comes the tricky part:
1.) Which dimensions (length/volume) would you suggest for the combustion chamber and where would you mount the spark plug?

2.) How should the rotary valve be designed? High changes of the stepper speed (if any) will not work.
So the open and closed duration (ratio) must be archieved by the valve design itself. Due to the general design the
rotor will be easily exchangeable but the stator parts must be a direkt hit to do not have to rework the engine
all the time and I want to avoid extra parts (maybe exchangeable stator on inlet side).

Just a guess: The open time must be significiantly larger because the fan will need some time to fill the combustion chamber
and the closed time can be kept as short as possible because combustion, pressure rise and fall will be quite short.
If this time is too long there will be a performance loss (low frequency/backflow), if its too short it maybe possible to destroy the fan because
chamber is still pressurized.

3.) I have the feeling that a constant flow of fuel is not optimal, because it can lead to self ignition or "bad" air to fuel ratios.
I also believe to add an on/off function (for the fuel) to the rotary valve will be very difficult.
Is there any kind of fast valve that is not piloted (we have some very fast valves for pneumatics at work, but they are all piloted)?

regards

[enginewhisperer]

You could use an automotive fuel injector (or several)

People use them as remote valves by mounting them in a small "manifold" but it might be possible to mount the injector at the end of a short tube firing into the combustion chamber and shield it enough to survive.

[monty]

Dec 24, 2018 15:03:47 GMT -5 enginewhisperer said:

You could use an automotive fuel injector (or several)

People use them as remote valves by mounting them in a small "manifold" but it might be possible to mount the injector at the end of a short tube firing into the combustion chamber and shield it enough to survive.

Use a direct injection type....they will survive.

Monty

[dominik]

Thanks for the advices. I ordered two different ones.
I will check if one of them fits behind the rotary valve,
that should be enough shielding and cooling.
One was declared as direct injection but it does not look if it would survive large heat. I will see when they have arrived.

[dominik]

The project box for the poor mans PDE ist filling slowly: 



Valves and laser parts are still missing...

Would like to start the evaluations soon but after my actual turbine project I first have to take care on a new bathroom - damn.

Regards

[dominik]

The fuel injection valves have also arrived. So nearly everything available to start but no time, yet.
To have comparable results to an "old fashioned" pulse jet I plan to build the combustion
chamber to allow the attachment of different heads/valves. The evaluation is planned in the following
order with thrust measurements on all steps:
1.) Head with shutter valve (normal valved pulse jet)
2.) Head with shutter valve and boosted by fan (charged pulse jet)
3.) Head with rotary valve, boosted by fan and µC controlled ignition.
At the moment I´m unsure if propane or gasoline will be the fuel of choice. Maybe both options will be
tested..
I will start a separate build thread when I have time to manufacture first parts.

Regards

[racket]

Hi Dominik

My PJ experimental engine www.youtube.com/watch?v=uqMn0A2aUdo had a detachable front held in place with a large V Band clamp from the comp housing of a large turbo , it made for very easy removal for modifications to the valving , I used 2 stroke engine reed valve sets

Cheers
John

[dominik]

Hi John,
I´ve seen this video when I looked for the videos of your bike and cart.
It´s the first time that I see a pulse jet in use to drive a turbine.
Often read about it, but never seen one. Did you reach a point with the
experimental engine that it produced useful power output and also runs
smooth?

The electrical valves I just received are from a modern 4 stroke engine and have some
dozens of milliseconds switching time. I hope that I can synchronize the
fuel injection with the rotary valve, but this is remote future.

EDIT: For the steps 1 and 2 I plan to use customized reedvalves of 0,15mm stainless springsteel (1.4310) that are
watercut between to "sacrifice" steel plates.

Regards
Dominik

[racket]

Hi Dominik

Theres a sort of build thread on the PJ Site www.pulse-jets.com/ , do a Search in the valved section for " Turboprop PJ Experiment Update" , it contains a lot more info than I can remember, its been ~13 years since I finished playing with the engine .

Cheers
John

[dominik]

Hi,
found it, will take some time to read the hole thread.

Regards
Dominik

[dominik]

Hi, 
I read through the hole thread and it seem´s that you quit the project due to thermal issues and that, if I recognized correctly, the output power was not
that awesome.
The idea that arose near the end of the thread, to use a turbocharger as a freepower is quite nice, however I suppose this would only work well if the pressure
of the pulse jet is completely used (closed system). This of course will lead to other problems that the PJ can most probably no more designed to self sustain, 
because the only one I ever had, shut down immediately if you brought something near the end of the exhaust an back pressure built up.
However in the self sustaining mode the exhaust temperature would have been much too high for the turbine and I got your points that a constant flow turbine
normally differs from a pulsed one.
(But I had many loud bangs during my initial turbine testes that drove the turbine quite well - at least for a moment)

Maybe there´s a bridge to the actual idea of the charged pulse jet with rotary valve.

One possibility could be to use a complete Turbocharger and replace the normal combustion chamber with a valved one. 
The combustion cycles will be ignited electronically (maybe fuel injection also controlled).
The rotational mass of the compressor and turbine should keep the flow somewhat steady (after initial spool up).
A Reed valve should do for some first tests but I guess a well designed rotary valve would be more perfect to control the
cycles and have the engine last longer. I suppose it´ll be a bit difficult to start such an engine.

Or the PJ is charged by a fan and the turbine is used as a "pure" freepower, (maybe in addition with an alternator to get electricity
for the fan). This system could maybe designed to start on a push button with some clever µC.

Regarding the original idea (without freepower) I made some thoughts visible by CAD, here is some idea about the rotary valve that would have a fairly large
area with short "closed" and long "opened" times (as considered earlier to keep rotational speed as constant as possible).
At this time it´s just a sketch an the area is not calculated or validated against the fan´s volume flow:






Any ideas on that?

Regarding the connection between charger and PJ I would prefer some outer flange for the initial prototype like you suggested John a V-Band or maybe some screws in axial direction.
However this will produce a lot of drag if the prototype should be mounted on an RC-Plane for testing (if ever). Better would be a large inner thread, but this would reduce valve area
and is not that secure against getting loose. Still bemoaning this.

EDIT: LOL - a few minutes ago a colleague gave me half a canister Sidewinder 16 (methanol with 16% nitro) fuel that he does not need any more, seems to be a distinct hint to bring the poor man´s PDE
to life ;-)

Regards
Dominik

[dominik]

Little Update:
I had to order some laser parts today and included the rotary valve parts in the scrap areas:



Thea area of the smaller windows is 140mm² by 6 times ~840mm².
The exhaust pipe is made from a regular SS-pipe with an internal diameter of 38.4mm resulting in an area of 1158mm²

The IN-OUT ratio is slightly larger than it would have been chosen with normal reed valves, however a fast shoot from the hip
and considered as a possible starting point.
Due to fact, that this PJ will be operating very different from a normel PJ (in the 3rd evaluation stage) I
suppose it would make no sense to stay to the common reed valve design rules (Hopefully).

Material used is Ferrotherm, which is perfect for high temp application, here it makes not really sense because the
motor would die very fast, but as mentioned above I put the cuts into the remaining scrap of other parts.
I´ve never welded that stuff before, so maybe it´ll will not work as I hope.

Should be finished by the end of January, I will report how they turned out.

Best regards
Dominik

[racket]

Hi Dominik

The "interesting" thing with my PJ was the fact that the PJ operation wasn't affected by the downstream freepower even when direct coupled , I think the "reverse" flow of cold air slipped through the freepower without any "negative" consequenses .

The PJ was just a bit of a break from making gas turbines , I needed something different to play with for a while, it was fun , but very noisy...................I never did get into the intricasies of pressure waves etc etc , thas a "minefield" of complication ...........I was more interested in the power takeoff stage.

Cheers
John