Jerry Hey wrote:

I finished the swirl pot yesterday.   All that remains for the

cooling

system is the overflow tank which I think I will make out of

welded

aluminum.  Is there any problem with having it fill and empty

from

the

bottom?    Jerry

I use the original Mazda system from the RX-2 on the race car. The

overflow tank has the pressure regulating cap. My swirl pot has a filler

cap

to seal it but no guts in the cap to regulate pressure. The overflow

bottle

acts as an accumulator. The outlet of the overflow bottle could exit th

cowl

in front of the pilots side windscreen. This would allert you to a

problem.

Lynn E. Hanover

Cliff Manzke wrote:

Jerry, hate to tell you now, but the on-axis, bottom outlet of your swirl
pot is bad news.  Here's why:  When you first fill the system with water

you

will also fill it with some air.  When you start the engine, the swirl pot
starts to do its thing by acting as a centrifuge, forcing the water

outward

and at the same time forcing the air inward, forming a column of air in

the

center of the swirl pot.  At the same time, a strong whirlpool will be
formed inside the swirl pot by the outlet.  So, when enough time has

passed

(several seconds) and enough air gets separated out, the column of air

will

reach the bottom of the tank, at which point the whirlpool will instantly
suck it down and back into the system, starting the process all over

again.

It may sound crazy, but these are words of experience that Jeff Spitzer

can

confirm.  We (at General Atomics) pulled our hair for a few days over this
one.

The solution is either a much taller swirl pot (heavy) or one with a
tangential outlet as shown in the attached picture. With this design, the
tank can be quite small and still function.  Something like 6 inches tall
and 3 inched in diameter ought to do.

Also, the swirl pot does not need to be at the highest point in the system
to do its job, although filling the system with water is impossible if it
isn't.  A separate filler plug at the highest point will solve this.  In

any

case, small bleed lines from all the high points back to the swirl pot are

a

must.  Your swirl pot may look like a porcupine when you are done.  If you
don't get ALL the air out of the cooling system your just shooting

yourself

in the foot!

It is also my opinion that the "radiator" cap (and swirl pot) should be on
the inlet side of the water pump so that the cap controls the pressure on
the inlet of the pump.  Pump cavitation is a big no-no.  Having a known
positive pressure (cap pressure, something like 15 psi)) at the pump inlet
will help prevent (or at least reduce) cavitation.  Another name for a

swirl

pot is "header tank" which implies "keeps a constant head (pressure) on

the

pump inlet."  Entertain yourself and install a pressure gauge (or for

some,

a vacuum gauge) on the inlet side of the water pump and see what it reads

at

full engine speed.  You may be surprised!

Another good idea is to install a clear sight tube (permanent or

temporary)

in the water system.  When you have air entrained in the flowing water it
will be obvious.  When you get ALL the air out, the sight tube will be

clear

and you'll swear your pump stopped working.

Happy flying!
Cliff Manzke

Hi Cliff, how is the robot fighter plane business going?

Yes on the pressure taps. I am getting ready to instrument Robin's
RX7 with pressure and temp gauges on both sides of the rad.
Thanks for the great tips.

Paul Lamar

You'll have to ask Jeff.  I've been out of the unmanned aircraft game (and
General Atomics) for two and a half years now.  Lately I've been having all
kinds of fun using high speed data acq equipment (AVL) to look at real time
cylinder pressure and fuel injection events on diesel engines.......the kind
of equipment you'll need for your intake manifold tuning
experiments....$$$$$

Cliff Manzke

So you think it is going to cost a lot? How about just an oscilloscope and a 
high speed pressure sensor?

Paul Lamar
 
The AirCraft Rotary Engine NewsLetter.              Powered by Linux.
ACRE NL web site.
http://home.earthlink.net/~rotaryeng/            http://www.linux.org