George,

Your comment about change in water flowcharteristics whenP-ports are
inserted in a large water passage isapertient observation. I
calculated that hole area is over 4.5 cm2.Waterflow in the coolant
system would have to reduce and/or thesystem backpressure increase to
  maintain stock water flow when  blocked.  In  affect it is like shuttinga
downstreamvalve.I would think we wanthigher than stock flow through
  the  block as we are working the engine at  higher power densities.Anyone
with access to the Mazda Racing housingscould confirm thesize of the
  holes in the water passages to see if there  is a  difference betweenthem
and stock housings.


Doug inJapan


With the sealing block system water can still flow aroundtheSS p-port
  tube. You don't have to pot the area with epoxy if you  are  concernedwith
the
water flow in that area. If you do pot ityou may beable to drill
a 1/2 inch dia. or larger water passage.Another possibilityis a small
  external return line from the rear  rotor housing back to the  input tothe
rad.
This is commonpractice on Chev V8 racingengines.

I agree that robust waterflow is a good thing.

Theblockage is not that bad. You have tofigure the percentage of the
totalarea blocked and if you look notat the rotor housing but at the
centerhousing you will find thecenter housing side ports block a lot
of thereturn waterflow.

The bottom line is the metal temperature adjacent toboththe wear
surface
and the spark plugs explained in my book How to CoolYour Wankel.

Mark Steitle installed thermal couples in thatlocationand found
the temperatures a couple of hundred F below themetaltemperatures
in a Continentalliquid cooled aircraftpiston engine. Ihave tons of
data on this subject.

IMHO whatis far more importantis apex seal temperature which is
affected
byrotor cooling and apex sealoil supply. Hot apex seals wear faster
and
the apex seal slot itself wearsfaster. In fact the early RX8with what
turned out to be inadequate oilsupply to cool the apexseals resulted
in less than 100,000 mile life.Mazda addressed theproblem by adding
additional oil injectors about 2007.I still add a1/4 ounce of 2 cycle
oil to each gallon of gas in my 2004 RX8alongwith the regular
metering pump.


PaulLamar

I justchecked some factory race Periphery Port housings. At the top of
the
port cavity is a 25/64" [.39"] hole for coolant flow. Justone.

Lynn E. Hanover


Lynn

Very good. Thanks.Iasked because my O-ringed insert PP tube
effectively blocks the waterpassageway.There is however enough room
near the outer top of thepassageway to remove some metal and allow
water to flow through.Itake it the other cavities holes on the
return side are about the same asstock?

Doug in Japan


I will check on that in the AM and get back to you.
Lynn E. Hanover


Paul, I had a notion on testing my waterflow once I make some parts. I am
 going to bolt the parallel flow center housing to a rotor housing ondplumb
 a garden hose to first the inlet and then the outlet and run the waterinto
 a 5 gallon bucket to time how long  each side takes to fill the bucketwith
 flow through the rotor housing.I want to have slight back pressure fromthe
 return water to even out flow distribution on the inlet side water. I came
 up with a way to bypass/transfer the side housings so as not to restrict
 flow through the rotor housings where seal coolingmight be improved.Look
 on the rotor face side of the housings on the cast iron wear plate. The
 solid models shoe a 1/8" thick blockage but that is removed with the
 machining. Also all the part faces include a 3/16 inch machining allowance
 which is taken off with the lathework. all the holes are cncmilling
 machined as are the ports.It seems lie a drilled hole through the potting
 mix is good insurance ifonly to purge air out of the cooling system or
 trapped steam from the exhaust area below .

 george grimes

Back in '02 I did actual measurements on the flow of the Mazda (13-B) water
pump as a function of pressure and flow rate. I used a 55 gallon reservoir,
and then timed how long to fill a 5 gallon bucket. Did it at several
different RPM, also measured the pressure drop across an evaporator core as
a function of flow rate. The results are attached.I actually did a short
write-up on this, but cannot find the file. A couple of pertinent points: 1)
Max pressure of a centrifugal pump is generated at zero flow, 2) Max flow at
zero pressure.The intersection of the resistance curve of the plumbing
circuit and the pump curve establishes where it is going to operate.


Bill Schertz
KIS Cruiser #4045
N343BS
Phase I testing

I probably have the write up Bill. I'll just have to search for it.
Can you offer any key words?

Paul Lamar