Hi Paul,

As promised, here are some pictures of the oil pump test rig.  We

create

back pressure by progressively closing the valve on the right hand

side.

We measure pressure with a 0-10bar pressure gauge, just out of the

iron.

We measure flow with an ultrasound digital flowmeter (not installed

when

these pics were taken), and we measure RPM on the pump pulley with a
stroboscope.

We tested both the '89-'92 13B single-turbo pump and the '93-'95
twin-turbo pump.  The one you see on the rig is the latter one.  We

feed

it oil two ways:  through the regular pickup tube attached to the

bottom

of the iron (not visible on the picture) and through a SECOND PICKUP
TUBE (the copper one visible on the picture) inserted in a 5/8'' hole
drilled through the side of the pump body and feeding the oil to the
side passageway Mazda put in its later model pumps.

Best regards,
Francois Badoux

OK I get it now. Good idea to use a front end housing. We got those
lying around too. Is that a synchronous electric motor there? RPM?

If I were doing it I would add a valve in the outlet pipe to
simulate the oil pressure drop across the block.

Paul Lamar

 Paul, will you pay attention?! There IS a valve on the outlet pipe to
simulate pressure drop across the block!  Look on the right hand side of
the rig's picture. I also mentioned it in my previous email above.

Yes, it's a synchronous electric motor, RPM 2780.  We vary the pump RPM
by changing pulleys and cross check with a stroboscope to make sure
there is no slippage (which happens if we are not very careful at
dropping no oil on the transmission belt). All flows mentioned below are
measured at 3630 pump RPM, which corresponds on this engine to 5450
shaft RPM.

We made some new tests and we now have a pretty complete picture of what
happens:

1.  We changed oil from AeroShell 15W50 to a mineral single-grade
SAE-50.  It improves the situation somewhat but doesn't solve the
cavitation problem.  The difference in the rheology of the two oil
qualities seems to have only a second degree order of influence.

2.  We varied back pressure at iron's outlet between 3.5 bar (50 psi)
and 7 bar (100 psi).  Flow varies as a function of back pressure but
only marginally:  20.7 gal/mn at 3.5 bar vs 19.0 gal/mn at 7 bar.

3.  We varied oil temperature and therefore viscosity:  no measurable
influence on the oil flow.

4.  The REAL DIFFERENCE comes from a) changing the pump body to the
'93-'95 twin-turbo pump, and b) decreasing substantially the headloss on
the suction side to feed BOTH pump rotors properly (thanks Lynn, your
suggestions worked... and we added one of our own to complete the job).

Conclusion:  we missed being at SNF with our Arrow because we needed 16
gal/mn of oil flow across the oil cooler and only had 9.  Now, with a
few relatively simple mods to the suction side of the pump, we have
flows in excess of 20 gal/mn.  Let's implement these changes on the
plane and go flying!

Thanks to all who contributed suggestions.  This newsletter is great
value indeed.

Francois badoux

Sorry! I saw something blue on the outlet pipe but it was not clear what that
was. It is not like yours are the only messages I have to read :)

Let me get this straight. Your oil pan is deeper than the stock oil
pan so you were losing suction with the 89-'92 13B single-turbo pump
but not the 93-'95 twin-turbo pump?

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.