Hi Floyd,
what if you have leakage on any tuned system?
I can well see that any engine that could run well with an open exhaust would not need a sealed system but surely a good seal is needed in a truly closed system.
You definitely need a sealed header joint on a tuned pipe system. It's much more critical on some engines than others, but they almost all care about leaks near the header. Some can tolerate small amounts of leakage towards the tail end, and some cannot. This used to be a big problem, and people would spend a half hour a night in the motel bathtub pressure-testing and sealing up their pipes. The original Bobby Hunt types were not too bad, the first Billy pipes were very prone to leaks. But the later Billy/Randy pipes only rarely leak, Eather's are always OK unless you completely carbonize it, and the current Randy types are also nearly bulletproof. I have been using the same 3 later manufacture Billy/Randy pipes for 15 years now, one has a small patch on it near the rear support, otherwise, no problems.
The effects of a leak tend to vary, but among the things that can happen is that is suddenly starts running like it is a Fox 35 on suction, speeding up throughout the flight. That's how I discovered that header leaks mattered - in my first big contest with a piped engine (40VF, 1989(?) Golden State meet the year we had it in Visalia and I had a blinding migraine all day Saturday) I was flying it, and all of a sudden it was picking up .3-.4 seconds a lap over the flight, and 4-2 breaking all over the place. I had been using the water pump gasket material but hadn't really paid any attention. I was complaining about it, and Ted tells me "oh yeah, Paul Walker said he seals his up with 5-minute epoxy, says it won't run right otherwise, let's try that". All of a sudden, ran perfect again.
For the header, do not use a conventional rubberized paper "water pump gasket" material, or aluminum tape, or anything like that. Face both sides of the interface (or all four on a RO-Jett) off with 240 grit sandpaper taped to a bit of glass or thick plexiglass until it is exactly flat on every mating surface. Clean everything up with acetone, and then use 3m High Temperature silicone sealer, from the auto parts store, intended for car header gaskets. Cover the entire mating surface with a very thin layer and remove any excess that gets inside. Push the screws through so you get some sealant on the threads. Then bolt it up. Clean up any excess with acetone right away, and clean all the tools, because it gets everywhere and makes a sticky mess
The instructions probably say to just snug it up, wait a few hours, then tighten it, and wait overnight. I screw it down tight immediately, and if necessary, put it in the airplane right then, and fly it 10 minutes later. I have never had it leak doing it that way, and I figure it cures itself in about 10 seconds of running. You end up with a microscopically thin layer and the exhaust will certainly not push it out of the bond line. It holds pretty good once cured, but you can get it off later.
The red header gasket silicone is almost certainly overkill. Regular old clear Silicone II bathtub or fish tank caulk is certainly adequate for the temperatures we get, and the usage. Lowly 5-minute epoxy turns brown and any on the outside can be easily scraped off after a run or two, but the adhesive qualities don't really matter, it just fills up any gaps that might be present, and if the header fits properly there's no way it is going to ooze out in operation.
I have run the weakly-tuned 20FP type mufflers with and without a seal, and I couldn't tell the difference for sure. It was close enough that the difference could have been my imagination. Sealing them up, however, certainly reduces the mess, particularly when the mating surfaces are inside the cowl. Same techniques as above, and I used clear silicone bathtub caulk on ST46s.
Brett