Rotary Valve Pulse-Jets

Moderator: Mike Everman

SR71Fan
Posts: 22
Joined: Sat Oct 17, 2009 1:32 am
Antipspambot question: 125

Rotary Valve Pulse-Jets

Post by SR71Fan » Sat Oct 17, 2009 1:36 am

Hello All:

An idea I have considered is that of a rotary valve pulse-jet.

In its simplest form, a set of 4 to 7 rotary "butterfly" valves
(like engine throttle plates) are in CONTINUOUS ROTATING motion
and synchronized so that they are all at the same position at any
one time. The ganged valves are driven by a servo motor. The
individual butterfly valves and the supporting bearing structure
would have to be heavy enough to withstand the pressure of the
power pulse.

The butterfly valve clearance with the respective manifold barrel
must be sufficient to ensure no direct metal contact, even as the
valve components heat up. Cooling is accomplished by the air
induction portion of the cycle.

The rotation rate is set to synchronize with the pulse rate of the
engine. Initially, the rotation rate is set [empirically] to half
of the expected start-up pulse rate. The first spark firing for
ignition occurs then all the valves are closed. If all is well,
the valves will then open to allow the next charge of air as the
vacuum is forming in the firing chamber. Then, as the reflected
exhaust wave flows back into the chamber and the next firing event
occurs, the valves will have rotated back to the closed position
again.

I envision that the actual pulse rate is monitored by a
microcontroller, and the valve rate is "locked" in synch with the
pulse rate as it changes according to flight speed. The
microcontroller could be programmed to adjust the time lag in the
valve positioning to maximize thrust.

The butterfly valve is not the only, or perhaps even the best way
to impliment such an engine. An alternate form of valve consists
of two conical shaped valves containing flow ports on perhaps three
sides. One of the conical members is fixed, and the other is nested
inside the first, and rotates at the prescribed rate to open and
close the ports. The advantage of the conical valve is that the
ports can have sizes that provide a "dwell" which extends the
"closed" time period of the ports.

Although I am an aeronautical engineer by education, my real
expertise in such a project is the microcontroller and the software
algorithms that may be needed to make it work.

Having described the concept of this proposed pulse-jet, I am
wondering if you have ever seen such a concept during the course
of your research? I would expect that the feasibility of such
an engine would require the availability of the microcontroller
technology now so prevalent in automobiles.

I look forward to hearing anyone's thoughts on this concept.

SR71Fan

yourmom003
Posts: 7
Joined: Mon Oct 12, 2009 9:49 pm
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by yourmom003 » Tue Oct 20, 2009 6:54 pm

i dont know a whole lot about this, but it sounds like a really exspensive and complex way to do things, but if it is made it seems like it would be great!

SR71Fan
Posts: 22
Joined: Sat Oct 17, 2009 1:32 am
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by SR71Fan » Wed Oct 21, 2009 11:19 pm

Yes, the concept of a microprocessor-controlled pulse-jet is somewhat
complex, but I believe it opens the door to capabilities that are
outside the realm of more conventional implimentations.

To gain some insight as to WHY the complexity of microprocessor
control for a pulse-jet engine could be justified, I need to
first explain how the use of a rotary valve could provides some
tremendous advantages for a pulse-jet engine.

According to some investigators, the rotary valve is the
design of choice for pulse detonation engines. Some of the
reasons for this can applied to conventional pulse-jets.
So, for the monemt, I want to leave the subject of electronic
control, and focus on the characteristics of the rotary valve.


First of all, a rotary valve provides what I call "positive
valve action". By this I mean that when the valve is open,
it's positively open, and when its closed, it's positively
closed. There are a couple of significant consequences
to this.

a) The rotary valve is not held closed by spring pressure.
Therefore there is no airspeed limitation imposed by the
inability of springs to close the valve before the next
combustion pulse. Indeed, the inlet pressure ahead of
the rotary valve are limited only by the mechanical
strength of the valve assembly. With even conventional
steels, this can be made extremely strong. This fact
allows the rotary valve pulse-jet to operate at very
high flight speeds, which would generally not be possible
with spring-loaded valves.

b) When the rotary valve is open, it places no restrictions
on the incoming air flow. With conventional spring-loaded
valves, the airfow is always inhibited by spring pressure
on the valve petals or vanes. With little restriction
on the incoming air flow, the combustion chamber recharge
time can be made significantly shorter than with a spring
actuated valve.


The second, and perhaps more significant feature of a rotary
valve pulse-jet is the potential for synchronizing a group of
pulse-jets that operate together in a cluster. If the rotary
valves of a pair of matched pulse-jets are geared together, but
driven at 180 degree phase angle from each other, they will fire
alternately, and they will remain in that phase relationship.
The microprocessor control, in the process of keeping the one
of pulse-jet's rotary valve in the proper position for optimal
thrust, will in fact be keeping both of them at their optimal
thrust settings.

This technique can be extended to four, six, or even eight
matched engines running in a cluster. The only stipulation is
that the rotary valve phase relationships must be spaced for
equal time intervals between individual combustion events.

The gains of this design are at least two-fold:

a) the vibration effects of a single large engine can be
reduced by spreading the pulse events into a multitude
of smaller pulse events generating the same total thrust.

b) Engine scaling becomes possible; for example, a cluster of
six pulse-jets each generating 50 pounds of thrust now
becomes a single pulse-jet engine generating 300 pounds
of thrust.

c) Due to the steadier thrust profile, several methods for
thrust augmentation may become viable.

Notice that a single microprocessor can manage the entire engine
cluster, so long as it has sufficient computational horsepower.
for the number of individual pulse-jets being controlled. Also
note that a degree of throttling is possible by shutting down one
of more of the individual pulse-jets in a cluster. Nonetheless,
the microprocessor control also allows a limited degree of
throttling of individual pulse-jet engines by means of control
of the fuel injection timing.


With the potential gains in performance and functionality
outlined above, some justification can be seen for the complexity
of a digital "engine control unit" (ECU). All of the technology
and control mechanisms, except for the speed control required for
the electric motor driving the rotary valves, are current employed
in automotive ECU's, and have been for well over a decade.
A major part of my professionsl work experience has centered
around real-time microprocessor hardware and software design.
So I know full well what is possible.

I hope this lengthy explaination is not overly confusing, and I
hope the readers can see the potential that incorporating a
digital ECU. My goal in this research was to find a way to build
a powerful jet engine at a fraction of the cost of a conventional
turbojet, and without using exotic materials.

GRIM
Posts: 474
Joined: Mon Jul 30, 2007 2:43 am
Antipspambot question: 0
Location: British living in Chile

Re: Rotary Valve Pulse-Jets

Post by GRIM » Wed Oct 21, 2009 11:35 pm

Hi SR71,
I meant to reply to your first post , been busy , Welcome in any case , :D

I like your idea , and have had some dealings with butterfly valves , in kiln blower control ,
a couple o things come to mind ,
Butterfly valves are very Non linear , the "closed" part of a full rotation is only a few degrees , and the transition from closed to almost full flow is a little more than 45 degrees , just a thought , dont know how this would affect the timing ,

Another potential issue is the speed the valve might have to rotate , at 150hz for example ( a realistic frequency for a small / medium size engine ) the valve would have to rotate at 4500rpm according to my calculations , i guess this is in the realms of modern servo motors ,

well just a couple of thoughts , Interesting concept , and probably could be put to the test without breaking the bank if you can source some elecronics cheaply ,

Thanks for posting
G

SR71Fan
Posts: 22
Joined: Sat Oct 17, 2009 1:32 am
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by SR71Fan » Thu Oct 22, 2009 3:09 am

Hello Grim:

Thanks for your input.

You bring up a good point regarding the "closed" angle range being
only a few degrees, but I don't think that is going to be a major
problem. First, by shaping the edges of the aperture that the butterfly
rotates thru, you can increase the so-called "dwell" angle, and thus
the total time the valve is "closed". Realistically, thought, the
valve is never really "closed", as a sufficient clearance between
the valve and the aperture would have to be provided to account for
metal expansion and bearing wear. in any case, the hot gas blow-by
during the power pulse would not be sufficient to damage the valve,
and the valve would be efficiently cooled during the air intake
portion of the cycle. The valve itself should be made of a stainless
steel that is resistant to hot gas erosion.

Your estimate of 150 Hz pulse rate would be for a 42 inch long tube,
if I'm using the right formula. One of my test case examples used
a 6 foot tube. which would drop the pulse frequency and valve RPM
almost in half. While this engine concept should run on a 42 inch
pulse-jet, I had assumed a largee one. In any case, 4500 RPM is no
problem for the motor.

In terms of putting the concept to the test, the electronics is
actually the least of my concerns. The sheet metal, valve body
fabrication, and fuel feed system are where the real expense lies.
Remember that microcontrollers are a major aspect of my occupation.

yourmom003
Posts: 7
Joined: Mon Oct 12, 2009 9:49 pm
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by yourmom003 » Thu Oct 22, 2009 9:35 pm

i was thinking of how you guys said the butterfly valve onlyhas an open angle of the valve was small, but what if you just kept the valves rotating on the axis that they normaly rotated, just 360 degrees rotation timed out right. it seems to sound like it would work out better than rotating back and forth that angle. that is after you work out some bugs such as making the sizes of the butterfly and the whole match up withen a few thousands of in inch, so very little presure escapes. it would seem to be simpiler if that isnt what was already being sugested. (i think you where talking about it moving back and forth.)

SR71Fan
Posts: 22
Joined: Sat Oct 17, 2009 1:32 am
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by SR71Fan » Fri Oct 23, 2009 12:02 am

mom003:

I'm sorry I wasn't clear about this.

You are right, the butterfly actually rotates continuously, it does *not* move
back and forth. Moving the butterfly back and forth would take a lot of energy,
and it is not necessary as long as the butterfly has a symetrical shape.

Actually, you do *not* want to use very close dimensions to get a tight fit.
A small amount of gas escape is OK, and a small amount of clearance will
prevent parts from interfering because of expansion.

SR71

Rocket Man
Posts: 144
Joined: Sat Sep 26, 2009 9:59 pm
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by Rocket Man » Fri Oct 23, 2009 3:18 am

It is a lot easier than that but it will work only on very large engines with a slow pulse rate. Do you know how a full automatic machine gun works? Build a pulse jet along the same design as a machine gun. Each time the engine fires the gas pressure opens a set of normally closed valves so it can take in a new supply of air/fuel. Problem is the mass of the moving parts has a resonance frequency of its own. The engine must have a cycle rate slower or equal to the cycle rate of the moving valve parts.

SR71Fan
Posts: 22
Joined: Sat Oct 17, 2009 1:32 am
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by SR71Fan » Fri Oct 23, 2009 4:15 am

I don't know if the pressure-actuated idea has ever been tried, but I'm not knowledgable
about all pulse-jet variants. It is quite likely that a pulse-jet of the type you propose
would indeed be large, perhaps the size of the Argus V1 engine, or larger. And that
would merely exaggerate the problems that all large conventional pulse-jets have,and
that is the large reciprocating movements of the valve mechanism. The larger the engine,
the worse it becomes.

One of the major gains of the rotary valve engine is the elimination of all reciprocating
motion, and replacing them with pure rotation. No more valves beating themselves to
death, with the resulting short engine life.

SR71

Sebastian LFT
Posts: 57
Joined: Mon Aug 25, 2008 5:50 pm

Re: Rotary Valve Pulse-Jets

Post by Sebastian LFT » Fri Oct 23, 2009 12:13 pm

Just a quicky suggestion. Putting Rocket Man's and your ideas together one could design a ratcheting annular rotational valve. A stutter valve if you will. Imagine many radial port slits instead of circular or triangularly holes...
You'll never know 'till you try it.

SR71Fan
Posts: 22
Joined: Sat Oct 17, 2009 1:32 am
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by SR71Fan » Fri Oct 23, 2009 5:23 pm

As I've stated before, I consider reciprocating motion to be "anathema" [to be avoided]. Any way you actuate the valves,
if oscillatory motion is involved, it is still inefficient and prone to valve wear and breakdown.
This is particularly true of larger engines.

SR71

Sebastian LFT
Posts: 57
Joined: Mon Aug 25, 2008 5:50 pm

Re: Rotary Valve Pulse-Jets

Post by Sebastian LFT » Fri Oct 23, 2009 8:38 pm

I do look forward to the fruition of your work. :D
You'll never know 'till you try it.

ace_fedde
Posts: 421
Joined: Mon Feb 16, 2009 9:26 pm
Antipspambot question: 0
Location: The Netherlands

Re: Rotary Valve Pulse-Jets

Post by ace_fedde » Fri Oct 23, 2009 9:39 pm

SR71,
Keep an eye on this thread:
viewtopic.php?f=3&t=5535
I will post some ideas there that might be very interesting for your (valved) idea as well :D
Fedde
Your scepticism is fuel for my brain.

SR71Fan
Posts: 22
Joined: Sat Oct 17, 2009 1:32 am
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by SR71Fan » Sat Oct 24, 2009 12:03 am

I agree that discontinuous (or pulsed) fuel injection is the way to go,
for several reasons. However, I would recommend a fuel pressure
much higher than 3 bar (42 psi in the U.S.). I was actually thinking
of about 7 bar in order to achieve better atomization of the fuel spray.

SR71

Rocket Man
Posts: 144
Joined: Sat Sep 26, 2009 9:59 pm
Antipspambot question: 125

Re: Rotary Valve Pulse-Jets

Post by Rocket Man » Sat Oct 24, 2009 5:20 pm

I have a paper that was written by someone that tried to build a rotary valve pulse jet using a variable speed DC motor. They built a copy of a Dyna jet engine and got it running using the head with the reed valves from the original dyna jet engine. The tested it to find out how many pulses the engine was producing. The rotary valve was designed so it would open at the same rate as the reed valves. The DC motor was turned on and the air was blown into the intake over the flowjector. The engine tired to start it made a lot of popping sounds but it would not run. They adjusted the speed of the DC motor a little plus and minus a few pulses and it still only popped and would not run. The information does not show how the rotary valve was built it only says the valve opens and closes exactly at the same speed as the reed valves. It says the engine runs on the reed valves but not on the rotary valve.

This blows my idea. I was thinking of building an engine with valves that open on the same order as a full automatic 30 caliber machine gun. I was thinking of using the pressure in the exhaust pipe to power a piston that opens and closes the valves.

Post Reply