Exhaust and muffler design in the rotary installation is one of the more complex of all the installation issues. There are, and have been so many variations among the various installations that there is little statistical proof of anything. I'll add a few comments and opinions that may be relevant, and describe what I did. It may trigger some ideas for you to think about. My system on the 20B is now approaching 100 hours – not a long term proof – but it is still solid.
The exhaust temps out of the port are very high, typically in the neighborhood of 1600F, and sometimes maybe 1700F. Couple this with pressure pulses and vibrational loads, and corrosive environment, and you have a very demanding situation. When you look at material properties to handle this, things narrow down pretty rapidly, particularly if you also want light weight as we do in aviation. Stainless steels, like 321, can handle the temps and be a workable exhaust – but, design for low stress levels then becomes a must, because SS are subject to 'stress corrosion' at these temps. Combine the high temperatures and vibrational stress and you get inter-granular corrosion which weakens the material and it eventually falls apart.
On way to alleviate that is to use inconel. It gives you higher temperature capability and corrosion resistance. And it gives you higher cost. But is it worth it to reduce your risk a forced landing in an unfriendly place? Compared to the total cost of your airplane it's a small amount. Maybe cut cost somewhere where it is less critical to safety.
Another thing to consider is that the more quickly you can expand the exhaust gas, the more quickly you can deal with lower temperatures. Charles Law – temp (degree K) goes down in direct proportion to increased volume. This becomes more complex in an exhaust system because of other factors, but it still works in your favor. The gas will expand down a constant diameter pipe, but expanding into a BIG pipe can make a significant drop.
That can be one of the advantages of the tangential muffler/manifold, or the design that Neil presented. The amount of the temp drop of course depends on the pressure in that bigger can. These designs have their own possible failure modes associated with welded joints and thermal stresses, but at least there is nothing there that is going to plug up the flow downstream. The skill of the welder and the post-weld heat treatment are important factors.
These units are generally bolted directly to the engine via the short header pipes, so vibration loads are a factor. Ideally you'd like to have stress (and thermal expansion) de-coupling between the engine and the muffler/manifold, but since the engine can move relative to it's mount you either have to accommodate significant movement, or support it to the engine by some other means then the header pipes.
And then there is the matter of the exit pipe(s) and secondary mufflers. Those have to be supported as well – an unsupported length of pipe extending from the muffler is an ideal candidate for some vibrational resonance which will fail the system somewhere. And the further away from the engine centerline, the greater the loads.
My exhaust system is shown in the first attached photo. This is in a pusher configuration. It is an inconel tangential manifold/muffler supported to the engine by short inconel header pipes which are welded to a heavy RB steel flange. It has a convex 'head' at the front, and a conical outlet to the exit pipe. It has internal vanes welded at an angle on the inside surface opposite the exit from the headers (you can see the welds on the outside) to help break up the pulses and direct the exhaust toward the exit. They also prevent possible "swirl-flow choking" which could increase back pressure. There are 'straightening' vanes in the conical exit section.
The exit pipe is clamped (custom heavy SS clamp) to the inlet pipe of the secondary muffler (I'll call it a resonator). The resonator is also of my design and is made of 321 SS. It is basically a straight through 2 ¾" pipe that is drilled full of ¼" holes (about 100), contained within outer 5" dia. pipe. The inner pipe has an orifice plate at the center which has a 1 5/8" opening. This orifice produces some restriction to the flow through the resonator to force some of it outward through the holes, and back through the holes to exit. The purpose of the resonator is to knock down the pressure peaks a bit more. Measurements on the dyno showed that resonator knocked another 8 db off the sound level and had no noticeable effect on the HP.
The plug in the resonator closes a port originally intended for the O2 sensor. But it didn't work well in that location because the temperature was too low (interesting, huh). I had to move it to the inlet pipe.
Last but not least, there is a SS support at the end which clamps solidly to the redrive. The clamp is designed to be rigid laterally, but to also be an effective heat choke. This supports the resonator, and reduces the likelihood of any resonance vibration in the system.
I originally thought that the resonator internals may not last more than 50 hours, but at 95 hours they are still solid. Which brings up another point. It is easily inspected. I can see those internals from the exit end, and I can stick a screwdriver or ratchet handle or whatever; in there and bang around to be sure things are sound. I inspect all the welds in the exhaust system every time I remove the cowl, or at least every 10 hours or so. Make your system inspectable, and keep an eye on it.
I wouldn't call it "quiet", but I've had people say they like the way it sounds. Time will tell its reliability.
Best,
Al Gietzen
If you go through the archives, you'll find lots of examples of
failed muffler designs. Many by your's truly. I think I've tried every
concoction known to man and the Swiss. They all worked... for a while.
My best overall design (see attached) is a 2" tube, full of holes
inside a 5" tube. All made of 16ga SS, all welded together. Needless to
say, the flange is more like 3/16" - 1/4" SS. The inside end of the 2" tube
is welded to the end cap of the 5" tube. That blocks off the one end of the
2" tube and secures it from movement. The exhaust end of the 2" tube is
welded through a 2" hole in the other 5" end cap. Rather than drilling the
2" tube full of round holes, we cut slots with a saw. Then take a big flat
blade screwdriver, stick it in the slot and bend it over. This creates an
oblong hole. (Much easier than drilling into SS. This is what will go on
the Volmer.
The sound is quite acceptable, it fits inside the cowl and Jim M.'s
version lasted the life of the aircraft... 600+ hours.
Neil
PS: Are you considering Rough River?
-----Original Message-----
From: Rotary motors in aircraft [mailto:flyrotary@lancaironline.net] On
Behalf Of Al Wick
Sent: Saturday, June 07, 2008 4:57 PM
To: Rotary motors in aircraft
Subject: [FlyRotary] Re: Mistral Crash Analysis
C'mon guys. You do this every time there's a crash. Instantly go into
rationalization mode. It's unhealthy. Greatly increases risk builders won't
take action. Increases risk you won't research it thoroughly.
A healthy response would be:" Here's another example of how our engines
produce unusually destructive exhaust temperature and pulses. We have a rich
history of broken exhaust components. We need to be very thorough when
designing and building exhaust."
I designed my own muffler. It had two inlets, two outlets. So if (when) my
muffler failed, it could never block both pipes. I also put loose safety
wire around my pipes, because on a pusher loosing pipe wipes out prop. So
basically, I assume stuff will fail, then design it to control the way it
fails. I've heard of rotary guys doing same type of thing. This is a good
time to share those key items.
On your car, they deliberately design products to fail a certain way. They
will make a component weak, so it fails first. They do that with wheels and
hubs. So when the muffler fails, little pieces come apart, not big sections?
You guys do a great job of sharing successes, design and construction
details. This is another opportunity.
-al wick
<No doubt you are on the money, Rusty. When folks are already predisposed
to bad mouth the rotary - this will only be more ammunition. "See! even
with umpteen million dollars you can't get one to fly" {:>). But, I
serious doubt it will effect many who have researched the rotary and come to
understand its benefits - as for the rest, who cares {:>)
>
I'm certain it was a relief to Mistral that the culprit was not one of
their engine components.
Whew! a close one for sure.
Hi Ed,
Unfortunately, I bet the majority of people will only hear "Mistral
rotary", "lost power", and "crash" :-(
Rusty (RV-3 taking forever.)
____________________________________________________________
Orlando Vacations - Click Here!
http://thirdpartyoffers.juno.com/TGL2141/fc/Ioyw6i3nK7u59LkagfZ5IcEgGqYcvPRC
wA85GgQITbFK9vE7wPGKFP/
--
Homepage: http://www.flyrotary.com/
Archive and UnSub:
http://mail.lancaironline.net:81/lists/flyrotary/List.html