Lynn,

    Would you happen to have additional info (photos,etc) on the

 modifications to the stock 13B oil pump? This is one area that has

 not been  completely explored in this forum although you have

 covered mods to the pickup tube before. Are there some more of

 the magic tricks you would share with us?

 .....Inquiring minds would like to know!



 Best Regards,


 Kelly Troyer
 Dyke Delta/13B/RD1C/EC2

Here is a picture of a poorly done pump land on a 12A. Even this was
performing well before disassembly. The kidney shaped cavities in the iron are
intersected by drillings to form the suction supply from the sump on the left and
pressurized oil out on the right. The inside of the cavities should be much
smoother than those in the picture. Polished like chrome is the ideal.

The junctions with the drilled passages should show the largest possible
radius.
Mother Nature loves flowers and smooth polished surfaces.

I cut the last 3/8" off of a pump. Just the ring and the mounting flanges.
Stick in a ring, and a rotor and shaft assembly, and slowly turn the pump in the
correct direction. That would be clock wise. Just turn it slowly and watch
how the opening cavity traps a slug of oil on the land between the cavities at
the top of the pump land, and then squeezes it out over the pressure side.

Notice that the land area has sharp corners all around. Oil on the right is
at high pressure, and is always trying to get back into the low pressure left
side of the pump.

For racing, I have indicated two small locations that will make the face of
the land area agree more closely with the cavity shape generated by the
impeller and ring interface. This gets you better pressure at high speed, and far
worse pressure at low speed, and alarmingly low pressure at idle. Do not touch
the land area at the bottom of the picture. It should have sharp corners and
poor flow. It is just a seal surface between the high and low pressure sides.

Time is part of a flow equation. At speed there is a smaller time period for
oil to leak backwards through the pump. At idle there is a larger period of
time for that leak to function, and so it does. Just logical. Even the separator
between the sections and the cavities at the very front of the pump body have
a similar shape and also can be improved.

For just aircraft use, I would do the radius and polish job in the Kidney
cavities, and one other change.

The pumps tend to be (or were in the past), sloppy in end clearance. You need
about
.001" to be safe and there is sometimes up to .006"

Place a straight edge across the open face of the assembled pump. Trap
smaller and smaller feeler gauges under the straight edge and against the lobe and
rotor assembly, until one just slips out with a slight drag. That gage
thickness is the end clearance.

Lightly oil a pane of glass or your kitchen (Formica) table. Lay down a piece
of 400 wet or dry silicon carbide paper (I love that stuff) on the oil to
hold it still, and spray some more oil on the paper.
Hold just the pump body on the surface and make a figure "8" motion with the
pump body. This goes real fast, so stop, reassemble the pump and measure
often. This closes up the spaces inside the assembled pump where high pressure oil
leaks back through the pump to the low pressure side.

In operation, that clearance gets bigger, because the aluminum housing gets
longer with each unit of heat than do the steel (sort of) rings and rotors. The
pump is made from powdered metal. Note the lack of a single tooling mark on
any ferrous surface.

The pump from the engine pictured, had been shortened with a flat file, (gag)
and it still worked.

As you may remember the Swiss had cooling problems that they determined to be
loss of pressure after extended runs at operational RPM. There are a number
of items that were determined to have contributed the that picture.

The pump operates in a constant suction lift situation. Where the pump is
sitting some distance above the oil level. So we must rely on ambient air
pressure to force oil up the pickup tube when the pump reduces the suction side
pressure in that tube.

If the pump could produce a perfect vacuum in the tube (not possible) there
would be less than 14.7 pounds of local air pressure to push oil up the tube.
The higher you fly the less likely it is that this is going to work.

Here is where a remote sump mounted tank with the oil supply has an obvious
advantage.

The Swiss found that the bug screen was very restrictive. So they left it
off. They were going to add a windage try to help defoam the oil anyway, so part
of the plan was to add a large much coarser screen in the center of that tray.
I have not seen that yet. And still that was just to keep big things out of
the pan.

There is no worse flowing shape in nature than a sharp ended tube in suction.
This is a disaster. The pickup tube then functions as though it were only
half as big as it is, and it is on the small side in stock trim. When I still
used them, I removed the cover and screen and used a like piece from a Pontiac,
or some huge American car.
About 5" in diameter. You don't need a screen unless like me, you glue
engines together with silicone. If you do it right there is little to fear from a
few drops of Vaseline or Holomar.

You do need the cover the screen was mounted on. It helps prevent a suction
vortex from forming when the oil level gets close the suction inlet, as in long
climbs.
If you have one of those bug screens that are just bigger than a silver
dollar, scrap the cover too and make one from a flat sheet, and about 5" in
diameter. Then install the pan and cut the cover so that it has 1/4" clearance to the
pan in any direction.

Then there is the Vena Contracta formed at the tube end.

I braze on a thick washer ( 2 1/2" X 3/8") flush with the end of the tube,
using a lot of braze. The extra braze fillet allows me to radius the tube
opening into a trumpet bell shape that Mother really likes. The outer edge of the
very thick washer is rounded off top and bottom. Smooth and shiny.

There you go. Oil pressure up the kazoo and almost no foam to well over 7,000
RPM. And from parts you already own.

Lynn E. Hanover

 
The Rotary Engine NewsLetter.  Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/            
Copyright 1998-2004 All world wide rights reserved.