From the EAA online rag:
...The exhaust system is notably unremarkable. Tom started out with set of long, equal length, stainless steel headers feeding a four-into-one tailpipe that collected under the engine and exited between the landing gear (under the pilot’s seat). This more elaborate exhaust system was never actually installed because the firewall size advantage gained with the dry sump would be negated if these pipes were used. Dyno testing at Lycon revealed that the four-into-one netted an 8 hp gain at wide open throttle, but Tom wasn’t convinced that the extra drag would be worth it...
http://www.eaa.org/experimenter/articles/2009-07_phantom.asp
How about that... 8 HP, just like the CAFE study indicates. I guess they (and every other gearhead, racer, engine builder, etc in the world) were right - Headers flat out work.
BTW, the "code" to break on the CAFE article is the 38-42 inch primaries, not the megaphone exhaust tip.
I just looked at the video of Tom's plane as it was being built and then readied for its first flight in August of 03 and at that time the airplane was unpainted and was in really rough external condition. It had the long exhaust pipes sticking down from each cylinder with a slight backward curve at the end. I guess what you described must have happened a long time before that.
I've been privy to a lot of LyCon's data sheets for different engines and I've seen claims for lots of incredible power, but they usually base their power estimates on FF. Would you believe an IO-360 putting out 280 HP at 2700 rpm? No? Neither do I! One sheet I saw last summer on an O-200 gave lots more power based on FF, but the sheet also contained both rpm and torque, and if you have those two you know exactly what the true power is, and it was quite a bit less than their FF number.
As far as the megaphone end is concerned I must disagree. If you look at the pressure curves that go with it you'll see that it gave a low pressure in the cylinder during valve overlap to help draw in fresh charge. That was one of their biggest recommendations, so why isn't anyone doing it.
So since the resonant length for 2700 rpm is about 76", how do you account for the four-into-one exhaust boosting power 8HP whenits much less length than that? Do you have a well-reasoned explanation, or are you just passing on what you read or heard? I would accept these numbers if I saw some definitive before-after test data. One of the things that was mentioned is that its necessary to change the cam timing, overlap, and duration for a resonant sytem to work.
As far as gearheads and racers are concerned, they can have the longer pipes that are necessary for resonant operation on their cars. Also they can ruun at much higher rpm where shorter length is necessary. You can hardly make a valid comparison of cars and planes for that, among other reasons. But keep in mind that they must also get past the region where an engine will almost die when the rpm is at the anti-resonance for the pipe.
But keep in mind that my original response had to do with optional means of mounting exhaust pipes that would be much simpler, lighter, and less expensive than the multi-pipe sytems, and they could also be integrated in with a better method of dumping the cooling air overboard to reduce cooling drag. Sometimes people are more interested in getting into quibbling with dubious arguments than actually looking into how something like this might be implemented.
Think for a moment about the things I've done on my plane. I built a carbon fiber pan with the carb on the back side, rather than below, feeding forward through fiberglass ducts. I have separate inlet carbon and glass plenums that feed the cylinders on each side between the push-rod tubes. My cylinders are wrapped with carbon fiber from the push-rod tubes to the bottom of the cylinders. There in a box that is fed engine exhaust through a special Coanda nozzle I designed and built, the cooling air and exhaust are mixed together and pass through a rectangular duct and the combination exits below the firewall through a variable aperture exit. The Coanda nozzle creates a low pressure that helps draw air through the cylinder, and then after the two mix, the higher energy exhaust helps get the cooling air up to freestream velocity at the exit. I designed and built my own air-oil separator, and I use the heated air from my oil-cooler for cabin heat with no carbon monoxide worries. I also designed and built my voltage regulator with over-voltage protection, but it doesn't rely on popping a circuit breaker to work. It just shuts off the regulation until the fault is removed. I also designed my VOR, Glideslope, Comm, and transponder antennae and they are all contained within the structure. Also I have two Plasma I ignitions, which I also designed for Klaus. Those are my credentials.
Crickey, mites! I left out one of the best things on my plane, that three-blade, high-efficiency, noise-free fan on the front!