Paul, I am running the water in the center housings just below the
lower spark plug to an internal manifold. From there it divides and
feeds half through the front rotor housing and half through the back.
The end
housing internal flow manifold distributes the return to the center
housing and is used to heat the cool side to reduce temperature
variations in the housings side to side. The water exits the top on the
intermediate housing to carry steam away.There is an obvious difference
in flow passage capacity in the stock housings shown in your
attachments.
There is about three times as much area on the hot side as
the cold side. As seen in the attachment only two thruholes go directly
into a rotor housing opening while 9 line up with the hot side passages,
On the face side which is against the rotor housing there are recesses
cast in the stock rotor housing which allow flow around the intake port
area to additional rotor housing coolant passages. I have expanded on
the recesses and added more area thru the housing to increase flow back
through the cool side of the housing. This will not change the actual
flow
through the rotor housing but will allow future increases. Several
of the
cooling passages thru the stock rotor housings have bulkheads with
smaller thru passages to control flow rates and add turbulence. In
future I may drill out a percentage of blocked areas until the area is
equal in both directions. At that point additional end housing flow may
be useful.
I am using a balanced piston flow divider on the exit side
of the water flow where the two engine modules water combines to go to
the radiator. Any difference in flow pushes the piston towards the low
pressure side to increase area on the higher pressure side until the
flow is close to equal. There are no thermostats used.
I get the idea of the thermocouples to equalize cooling which
will be useful on a test stand. Trying to implement something to bias
the balanced piston flow control is doable if necessary. An adjustable
spring on each side of the balanced piston could be added and set at
maximum cooling load.
george grimes
Paul--Very interesting note from George regarding his plans for
cooling. Past engines show that basic flow path plan will work. I did
not see any photographs of his parts but his approach sounds somewhat
similar to our "center in, fore/aft and around" flow path for the
SuperLight engine that circled around the ends and returned to the
center housing on the other side to exit the engine. We did it in an
attempt to even out the temperature gradient around the rotor housings
and improve the ability of the housings to handle the additional heat
resulting from sustained high power settings. Prior to that time we
were using standard 1987 Mazda housings with no cooling modifications.
They worked very well on the original SuperLight prototype until
we worked up to sustained high power WOT output onvery long flight legs.
To help make sense of the photo for folks, we designed the engine from
the git-go indexed 90 degrees when compared to a standard Mazda auto
engine, with sparkplugs on top and peripheral intake and exhaust on
bottom. Cool water was introduced to engine first at the center of the
cooling manifold. The coolant flow path from the spark plugs/center
housing was by passages in the aluminum castings outward, toward each
end then around the corner, and back again to the center housing,
exiting on the other side.
In addition to its function as a coolant
manifold, the part also strengthened to some degree the overall engine
assembly with castings incorporating provisions for fasteners. In this
regard the part was similar to the function of the top and bottom
reinforcements featured on the four-rotor winning at LeMans in 1991. It
appears that our little engine shared the same basic coolant flow path
they used on that engine. Unknown to us at the time, we were organizing
final-machined pieces in the assembly room the day the 1991
LeMans started June 22. Our first dyno start was Sept. 21, 1991, the
day before my birthday.
It all worked very well when we got everything sorted out, but was
damned time-consuming and expensive. When the original intended market
went away, the engine was simply too complex and expensive for the kit
aircraft market. But it did have some nice features that we
incorporated on the Mazda conversion engine.
Steve Beckham
--
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
Youtube key word PaulLamar2
Copyright 1998-2010 All world wide rights reserved.