Subject: Intake air velocity was manifold weight
From: ACRE
Date: 6/7/2003, 6:43 PM


Chris Cosman wrote:

At what RPM would you be getting supersonic airflows?

Good question. Lets say the displacement is
80 cubic inches  per revolution.
Volumetric efficiency E = 100% Mazda rotaries can do better
than that by a bunch.

Snips of my original erroneous approach to the problem.

Taylor says on page 203 Vol. 1 inlets average .5 Mach for high VE
engines.

Another way of calculating it is to use the max speed of the rotor face
moving away from the port adjusted by the port area to rotor face
area ratio. It is really hard for me to figure the max rotor face
velocity
right at the moment :)  There are some charts for this in the book
"Rotary Engine" as I recall.

For a piston engine peak inlet air speed in fps is peak piston speed
in fps times piston diameter squared divided by port diameter squared.

Gas speed = PS X (D^2/d^2)

According to the book "Scientific Design of Intake & Exhaust Systems".
Page 78.

They don't say whether D and d is in feet or inches.
I suspect inches. I guess the ratio would be the same
so it may not matter.

You can figure the peak piston speed by knowing the stroke and RPM.

Peak piston speed inches per minute =  stroke X pi X RPM

Note also this is not the speed of the pressure wave which does indeed
move somewhere around Mach one depending on a couple of factors.

Paul Lamar

I think your approach is correct.

So... to properly tune intakes you'd need to account for the speed of
the pressure wave
minus the speed of the medium it's travelling through, is that correct?

Chris C.
microcosman.com/aircraft

I disagree. I think my original approach was and still is incorrect.
I think the approach they take in the book  Scientific Design of Intake &
Exhaust Systems IS correct.

Yes on the one hand pressure is increased in one direction and decreased in the
other. It becomes very complicated. I would have to summarize the book Scientific
Design of Intake & Exhaust Systems or scan in the whole book.
I suggest everybody just get a copy that is interested in this subject.

Paul Lamar

Chris brings up an interesting point and I wrote a little BASIC program
to explore the situation. The assumptions here are the intake port
just opened and the air flow velocity instantly became 400 FPS.
Not accurate by any stretch. A possible refinements would
be to assume that the air flow velocity was sinusoidal in nature.
Starts at zero rises to 400 FPS and ends at zero when the port closes.

 Intake situation at 6000 RPM. 100 RPS. Total time for one rev is .01 sec.
 Given Air velocity 400 FPS or 4800 IPS.
 Given Wave velocity 1000 FPS  or 12,000 IPS
------------------------------------------
Time   Angle  Air trav'ed  Wave trav'ed  Wave trav -   Wave trav +
Sec    E-shaft  Inches       Inches       Inches       Inches
0.0005     18      2           6           4           8
0.0010     36      5          12           7          17
0.0015     54      7          18          11          25
0.0020     72     10          24          14          34
0.0025     90     12          30          18          42
0.0030    108     14          36          22          50
0.0035    126     17          42          25          59
0.0040    144     19          48          29          67
0.0045    162     22          54          32          76
0.0050    180     24          60          36          84
0.0055    198     26          66          40          92
0.0060    216     29          72          43         101
0.0065    234     31          78          47         109
0.0070    252     34          84          50         118
0.0075    270     36          90          54         126
0.0080    288     38          96          58         134
0.0085    306     41         102          61         143
0.0090    324     43         108          65         151
0.0095    342     46         114          68         160
0.0100    360     48         120          72         168

Note that the wave traveled minus the air flow velocity
more or less 36 inches at the e-shaft angle of 180 degrees which
IMHO confirms the dynamic chamber theory more or less.
IOW the wave arrives at the other rotor intake port just as
it is opening.

This is a computer simulation which at this point in time is
not really an accurate view of what is going on in the manifold.
A lot of assumptions are made such as air flow velocity and
the speed of sound in the the intake manifold which can vary.

This is pretty much a work still in progress.

What we really need is some instrumentation with a high speed
pressure sensors and air flow velocity sensors inside a real
manifold on the dyno.

Paul Lamar

This one adds the SINE function to the intake air velocity and stops at 180 
degrees. It does not make much sense to continue on past that point
as the port is closed and the wave is going to get where ever
it is going by that time.  The SINE function was used as the
intake chamber will be nearly full and the pressure will
have built up in the chamber retarding additional flow into the chamber.
Air has inertia so this is not entirely accurate as the inertia
will tend to keep the air speed up for while. 

This suggest an interesting test. 
If a pressure tap into the chamber is used and recorded
the peak pressure reached will be an indication of the amount
of charge air sucked in.
 
 Intake situation at 6000 RPM. 100 RAPS. Total time .01 sec.
 Given peak air velocity 400 FPS 4800 IS. 
 Given wave velocity 1000 FPS  12,000 IS
------------------------------------------
Time   Angle    Air Vel  Air trav     Wave trav    Wave trav -   Wave trav +
Sec    E-shaft    FPS     Inches       Inches       Inches       Inches
0.0005     18     124       0.7          6.0          5.3          6.7
0.0010     36     235       2.2         12.0          9.8         14.2
0.0015     54     324       4.1         18.0         13.9         22.1
0.0020     72     380       6.4         24.0         17.6         30.4
0.0025     90     400       8.8         30.0         21.2         38.8
0.0030    108     380      11.1         36.0         24.9         47.1
0.0035    126     324      13.0         42.0         29.0         55.0
0.0040    144     235      14.4         48.0         33.6         62.4
0.0045    162     124      15.2         54.0         38.8         69.2
0.0050    180      -0      15.2         60.0         44.8         75.2

The big unknowns (and there are two
major ones) are actual speed of sound in the intake and the peak
air velocity. Lacking this information this is a wild ass guess
at best. On with the dyno.

BTW if you don't believe wave charging has anything to do with it
and velocity is everything then there is no point in using s 
dynamic chamber.

Paul Lamar
 
The AirCraft Rotary Engine NewsLetter.  Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/            
Copyright 1998-2003 All world wide rights reserved.