Electric valve

[Hondaen]

Hi

I have been working with pulsejets for several years now.

I have a idea for valve design, and I wonder if this has been done before, and if it will work.

My plan is to build a head, with valves from a four stroke engine. The valves is operated from a electric magnet.

The electric magnet is operated from a mikrokontroller. A pressure sensor, senses the pressure in the combustion chamber. A "timing chart" in the mikrokontroller wil operate the valves, so I can controll the thrust and resonance on the engine.

If this works, I will make my own valves. If the valves are machende like a "egg", there will be less turbulense behind the valve, and maybe rice the effectivines of the engine.

My concern is the electic valve, it may not be able to operate at hundereds of Hz, but I`m sure that problem can be solved.

Any thougts people ? :-)

Thanks

[Anders Troberg]

My concern is the electic valve, it may not be able to operate at hundereds of Hz, but I`m sure that problem can be solved.

I think that can be solved. I suspect the problem is that a microcontroller will have difficulties adapting to the dynamic nature of a pulse jet. The sensors that has to work in the CC will probably have big problems surviving.

My idea on this theme, which has never failed (since it has not been tested...) is to have one small valve that operates like the usual mechanic valves. Then, add some big, heavy duty, electrically operated valves, which are slaved to the small one so that they will operate in sync. This way, the engine can find its own operation parameters, but you still get the nice, big, electric valves.

If you have problems with speed (it will take some time to open and close the large valves, both due to inertia and to the time it takes to saturate a solenoid), you can delay them one cycle and advance the timing slightly. It's a cyclic sequence, so that will work just fine.

As you can put the sensor for the small master valve outside the CC, it will have a much nicer environment to work in.

You could probably also throttle it and adjust how it works by putting some transforms on the wave (ie the time plot of the position) that the master valve sends to the slave valves. Amplify it a bit and they will open more, make the positive pulses shorter to open the valve a shorter time, filter out high frequencies to focus more on the base working frequency, advance or retard the timing to compensate for slowness by shifting the phase, detect edges or zero passes and insert a spike at them for faster operation of the valves and so on.

Not only that, if you could log and save that curve someway, I think there are some people here who would almost be willing to sell their soul for that graph...

As I said, it's untested, so no guarantees, but perhaps it can give you some ideas. You indicate that you were going to use a microcontroller, so I guess that this construction is well within your abilities and that the theory will make some sense to you. My gut feeling say that this will be a simpler solution, since it will all regulate itself.

[Jim Berquist]

Some place way back when I posted about a rotor valve system. It seems the problem to overcome is the inherent resident tuned frequency of the P.J. A rotor valve could be acoustically coupled via feedback to synchronize with the same. A phase lock loop could be set to adjust the motor speed to resident freq. The use of 4 stroke valve head would be more difficult I would think....

Jim

[MB]

Remember the resonant frequency of a PJ is determined by its length - much like the tone of a church-organ pipe. So once determined, you needn't expect it to wander around the spectrum. And you certainly shouldn't expect to be able to alter it electronically.

I'd say sustaining the engine running seems practical, but your challenge will be in actually starting the engine. Like calculus, the devil is on the boundaries.

[Jonny69]

The frequency does change on throttling though, especially when you are pumping large amounts of propane gas in the frequency goes down significantly and you can hear it.

For a fixed throttle position I can't see why a poppet valve wouldn't work as it moves that fast in an IC engine at high revs. It'll just be a matter of getting enough power to whatever is driving the valve(s)

You should be able to hear if the frequency is out as it'll cause the sound pressure to cycle up and down as it somes in and out of phase. You can either adjust the length or alter the frequency of the valve. As said above it couldn't be too hard to use a PIC in a comparator circuit to 'listen' to the exhaust frequency and compare it to the valve frequency and regulate the two.

Interesting project if you decide to go ahead. I'll be watching with interest.

[MB]

Did I say that? What a dummy I am. Total length is the largest factor in the resonant frequency of a PJ, but the other factor is the Speed of Sound, which is the rate at which the shock wave travels. According to Wikipedia.com, "At sea level, at a temperature of 15 °C (59 °F) and under normal atmospheric conditions, the speed of sound is 344 m.s-1 (1225 km.h-1 or 761 mph)." It then goes on to say, "The speed varies depending on atmospheric conditions; the most important factor is the temperature." So indeed, as Jonny69 correctly pointed out, as the fuel flow changes, so does the combustion activity, and hence the temperature, and the speed the shock wave will travel -- and hence the resonant frequency.

All that said, the subtle nuances of measuring conditions to predict the timing will surely be less reliable than simply measuring it directly; which was suggested above.

[Jim Berquist]

You would just have to get it running then tweak the rpm of valve to match the HTZ of the pj............