Al Gietzen wrote:

As I look at ordering a custom oil cooler and sizing lines and fittings for
the oil cooling loop, I find some seemingly fundamental roadblocks to a
consistent design.  It is easy enough to compute the expected heat load to
the oil as a function of power level.  It is also pretty straight forward to
then compute the oil flow required to carry that much heat for different
temperature drops in the loop.  Hot oil has a fairly low specific heat (0.53
Btu/lb F), so it takes a lot of oil to move much heat.

What you find (Fred Moreno also gave us these numbers some time ago) is that
for a reasonably optimum design of the cooler with a 30 F inlet to outlet
temp drop, you'd need about 10 ¼ gpm per 100 hp.  Even at a 40 F drop you
need 7 ¾ gpm per 100 hp.  So designing for a modest sustained 200 hp
climbout with my 20B, I need somewhere between 16 and 20 gpm of oil flow.

OK so far; because some data I got from Dave Atkins suggests that the "high
output" pump he installed would deliver that much at rpms over about 5000;
that is, to a fixed pressure of about 60 psi.  First it becomes obvious that
these high flow rates require large lines and fittings to keep the fluid
velocities (and pressure drops) down - even a 3/4" line is going to have
flow velocities of close to 20 ft./sec.

Then you note that fittings on the engine, and the inlet and outlet passages
have much smaller diameter; like about 1 cm.  The stock 20B uses equally
small hoses to the oil coolers.  What gives?  In checking oil flow data from
dyno runs on a 13B I find that the oil flow is not linear with rpm, and
never gets above about 8 gpm.  And then I wonder whether the reason they
don't use bigger lines is because they know that you can't force flows
larger than that through the engine with reasonable pressure levels.  This
is OK for automotive use because you would never expect to see such high
power levels for sustained periods.

So what do you racing guys do?  It seems you either have to design for a
very large temp drop across the cooler (lower average temperature, much
larger cooler, which we don't want in an airplane) or have an auxiliary
cooling loop that does not flow through the engine.  What am I missing here?

Any thoughts, ideas, or further info appreciated.

Al Gietzen

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