Hi Mark,
From what I have read and heard, a well
designed exhaust system for a PP is more of a challenge than for any
other porting scheme. As Lynn has reported, the PP is even more effected
by any back pressure in the exhaust system than our regular rotary
engines.
That said, I dug out a pretty good book on intakes and
exhausts (yes, they understand and address Finite Amplitude Waves in the
exhaust system). Anyhow I thought I would extract what
appears to be some interesting insights for exhausts in general. There
was formulas for getting a rough estimate of header lengths and diameters -
now these were for piston engines but it would appear the dynamics of the
pulse turning would be the same for the rotary.
I put the calculations into a spreadsheet (attached) and
the results do not seem unreasonable - one thing I don't know the effect
of is the generally higher exhaust temp of the rotary and what effect it might
have on the numbers. Also I don't know your PP port timing - so the one
used is the factory Mazda PP Exhaust port closing from Paul Yaw's web site -
73 Deg BBDC.
There appear to be two primary power
enhancing benefits from a well designed exhaust system.
1. Minimum restriction to exhaust gas flow -
apparently velocity is critical so diameter of pipes is important - apparently
optimum velocity is between Mach .45 and Mach .5 (got a Mach meter handy
{:>))
2. Scavenging effect of the reflected exhaust FAW
to aid in pulling more mixture into the combustion chamber.
The well design system appear to maximize both
attributes but generally for only a limited rpm range.
Engine displacement, RPM and intake/exhaust over lap
appear to be the major engine factors in determining the desired exhaust
parameters.
Ok, here are some extracted information - none of it
necessarily new, but perhaps useful to review (even thought I know you know it
already)
Exhaust tuning works by using the proper header pipe
length to time the arrival of the "suction" (expansion) wave reflected back
from the end of the header to arrive at the engine exhaust port during the
overlap period - between Intake Value Opening and Exhaust Valve Closing.
If the scavenging wave is wide enough (time wise) and
the overlap (intake/exhaust) is short enough (probably not your case), the
exhaust system will remain in tune throughout a wide rpm range.
High out put, High speed engines with long overlaps
periods require much more precise tubing lengths to optimize power. As
power levels increase, the margin for error decreases.
Collectors: (Length and Diameter - see spreadsheet
for some examples)
1. Longer collectors force the torque peak at
lower rpm because at higher rpm the reflected wave from the end of
the longer collector does not get back in time to help scavenge the combustion
chamber and vice versa for short collectors. A "reverse funnel" opening
into the collector appears to broaden the reflected wave and broadens the
effective rpm range - but reduces the amplitude of the wave.
2. The Amplitude of the reflected wave is
dependent on the difference in cross sectional area between the pipe and
collector. A smaller dia collector represents less area change and
therefore reflects back a lower amplitude wave and vice versa for a large
diameter collector
One interesting thing I read was the notion that in some
cases, perhaps the headers should NOT be of equal length -
IF you want a broader power range. The theory is that
with slightly different lengths the scavenging effect would be spread
across a broader range and be less "peaky". I had never hear
of potential benefit of unequal headers before, but it would seem to make
sense if you want a less peaky power point. But, if you want the bigger
boost near one rpm range then equal length headers would appear
to apply
So there, that's what you get for waking me
up.
Ed
Sent: Saturday, November 20, 2010 6:32 AM
Subject: [FlyRotary] Tangential Muffler
Guys,
It has been way too quiet on the list lately, so I thought I
would post this report. Yes, another muffler experiment. Seems I
can't get beyond the exhaust system.
I did an experiment yesterday... with interesting results. While I
am very pleased with the muffling qualities of the tangential muffler, I have
been suspicious of it being responsible for the less-than-expected performance
of my p-port 20b. So, I removed the tangential muffler and replaced it
with the previous 3-into-1 exhaust w/DNA muffler that I had been running for
about 100 hrs.
First the bad news. The 3-into-1 system is much louder, to the
point that it almost overwhelms my Zulu. I haven't shattered any
windows, but little children and small animals have been reported running for
for their mothers. FWIW, I'm running the stock exhaust splitters that
came in the 20B housings.
The positive news is that I picked up 10-12 mph on top speed! So,
it confirms my suspicions that the tangential muffler is too restrictive for
my engine. While definitely not optimal, the 3-into-1 design is clearly
better than the tangential muffler when it comes to making power.
So, my plan is to build another exhaust from 625 inconel, making all
three runners 31.5". I haven't decided what I will use for a muffler.
I'll report back when I have more data.
Mark