Pipe Physics

(with counterpoint by Hoyt McKagen)

I was looking through some old emails and ran across this gem from Chris BeHanna (behanna@syl.nj.nec.com):

Subject: Pipes and Performance (Was: Re: SuperTrapp Staggered Duals)

Jane Caban  writes:
>Can someone explain what performance means in terms of pipes? I have always
>had the Supertrapps on my bike so I don't know how performance would be
>affected if I had another kind.

Couple of things:

1) Flow: you want to be able to flow as much exhaust gas as is practical. This means low-restriction.

2) Scavenging: you want the flow to have as high a velocity as possible so that it properly scavenges the exhaust gases out of the head when the exhaust valve opens. This places a practical limit on the inner diameter of the pipe that can be used with a given engine.

3) Reversion: when the pressure wave from the exhaust pulse reaches the end of the pipe, it will be reflected. At certain engine speeds at which the whole system (exhaust AND induction) are in the proper resonance, this pressure wave (reflected as a pulse of rarefied air), will reach the head at the precise moment of overlap in which both the exhaust valve and the intake valve are open, producing a partial vacuum that helps to draw air into the cylinder. In general, this will only happen over a narrow rpm range. For a Harley, there will only be one such range on the safe side of redline. At all other engine speeds, a compression pulse blows through to the carburetor and is carbureted once on the way out, and again on the way in, producing a rich stumble and a resultant loss of power at that rpm range. In general, the "sweet spot" where the reversion pulse helps will be well over 5000rpm, so pure straight pipes do NO GOOD AT ALL on street machines.

Generally, you will want some mechanical device (usually baffles of some sort) to break up the reversion pulse so that the rich stumble goes away. You trade off the high end power to gain low end and midrange since you do not want multiple carburetion of the exhaust pulse when it blows through the carb and gets sucked back in. The way to prevent this is by using baffles of some kind. This "breaking up of the reversion pulse" is often confused with "backpressure."

Repeat after me, folks: BACKPRESSURE HAS NOTHING TO DO WITH ANYTHING. We are dealing with an acoustic phenomenon here, as well as pure wave mechanics. The purpose of the baffles, besides keeping the bike quiet, is to break up the reversion pulse so that it doesn't screw up your carburetion. Period.

In general, flow and scavenging should be matched to the capabilities of the induction end of the system, as you have wave mechanical difficulties if one end of the system can flow more than the other. Generally, you should pick the pipe AFTER you've picked the carb, cam, and headwork. Just slapping on a pipe because you like the sound will hurt overall performance rather than help it.

Ideally, you'd have laminar flow through the system, except for the turbulence you need in the combustion cycle to get a complete burn. By matching the capabilities of induction, head, cam, and exhaust, you will most closely approximate laminar flow (laminar meaning "ordered". As an example, when people line up, they can get through a doorway much faster and more easily than when a crowd tries to push through the door. The lined up crowd is laminar flow; the pushy crowd is turbulent flow).

My $.02,

Chris BeHanna  DoD# 114          1975 CB360T - Baby Bike
behanna@syl.nj.nec.com           1991 ZX-11 - needs a name
Disclaimer:  Now why would NEC  
agree with any of this anyway?    I was raised by a pack of wild corn dogs.

Some performance data:

From Seth Zirin:
Here is some pipe data I got from Dennis at Bub Enterprises on 6/2/92. I forget the year bike but think it was a 1990 FXR. It had a stock carb, stock ignition with rev limiter disabled (Screaming Eagle), and an Andrews EV-3 cam.

They set the bike up on a DynoJet dyno and tested the following sets of pipes under the exact same conditions. The pipes tested were:

HD Stock			The stock HD exhaust system with the
				crossover as it left the factory.
HD Off Road			HD "slit fit" off-road mufflers on
				the stock header pipes with the
Python I			The original Python Pipes.
Super Trapp 2->1 (18 disks)	A SuperTrapp 2 into 1 system with
				18 disks installed.  He didn't mention
				the type of end cap.
Rich 2 into 1			The Rich Products ThunderHeader that
				has been the topic of recent discussion.
Bub Bad Dog			Bub Enterprises Bad Dog pipes.  This is
				their product.  I have these on my bike.

Rear Wheel HP at RPM 2500 3000 4000 5000 6000 ------------------------------------ HD Stock 17 26 40 42 41 HD Off Road 19 33 49 57 54 Python I 20 32 46 55 38 Super Trapp 2->1 (18 disks) 19 34 46 52 56 Rich 2 into 1 18 30 50 58 60 Bub Bad Dog 21 35 50 60 71

During their testing, they found that the engine produced 2 HP more at 6000 RPM with a full tank than with a half tank. They compared results with S&S and Jerry Branch and they also have seen this. A fuller tank probably keeps the float bowl at the full line a larger percentage of the time.

I don't remember how the following noise measurements were taken. The actual method isn't really important since all measurements were taken under the exact same conditions and are meant to be compared, not judged as absolute values.

				Noise (dB) at RPM
			2500	3000	4000	5000	6000
HD Stock		80	80	81	82	84
HD Off Road		86	88	92	94	96
Python I		87	89	91	94	96
Super Trapp 2->1
(18 disks)		82	82	87	96	93
Rich 2 into 1		88	90	92	96	99
Bub Bad Dog		83	84	90	93	93

Thanks to Seth Zirin (shz@mare.att.com) for this info.

Response to the above from Hoyt McKagen as posted on the yam650@micapeak.com discussion group.

There's no doubt that the pulses line up the wrong way and kill portions of the powerband but it's as much or more from stifling the flow as it is double carbing. There's not much backflow in modern motors and that's because it's a problem. Hence it's designed around. The usual cam timingshave small valve area exposed on overlap. And ex systems can kill backflow; all muffler types do it by construction, for example. Racing systems have used reverse meggas for decades for same reason. I've seen this site before BTW and think he's a bit extreme. The reverse ex flow problem may get more obvious with longer duration cams. But double carbing isn't a big issue with four-strokes.

Best wishes,

Hoyt McKagen
Belfab CNC - http://www.freeyellow.com/members/belfab/belfab.html
Best MC Repair -  http://www.freeyellow.com/members/batwings/best.html
Camping/Caving - http://www.freeyellow.com/members/batwings/caving.html
   When wise man argues with fool, people cannot tell the difference!

Back to Motorcycles