SNIP
        The seal was a National part.  Going from memory, I think the
    p/n was
        3500090, or something similar.  I would have to check on that to
be
        sure.  It was a nice tight fit.

        Mark St.
        Hi Mark,

        Can you check the seal P/N?  I'm working around a National P/N
    350909
        and before I get too carried away I'd like to make sure I've got
the
        correct seal. I'm embedding the seal in a new housing "half" to
    control
        the eccentricities better. This will also provide back-up to the
    seal so
        internal pressure forces tighter sealing rather than the tendency
to
        blow the seal out. ( IOW's I'm using housing material to act
    like your
        screwed in retainer ).      Thanx    Dave M

        I would not muck about with cheap shaft seals. I would use these.

        Paul Lamar



    Paul,
    Now you tell me!!!  There is no way that I'm pulling that PSRU again
    unless I have to for some other reason.  The seal I used is exactly
the
    same as what Tracy installed new.  In fact, I took the old plate with
me
    to get a match on a replacement.  The old one was a different make,
and
    possibly had a slightly smaller o.d. as it worked its way out of the
    hole.  The new one was a nice press fit.  It required a little
    "persuasion" to get properly seated.

    Dave, I'll have to check my receipt tonight and then I can let you
know
    for sure.

    Mark St.

    Wait awhile. It will leak again :)

    Part of the problem is the current plate and the PSRU are not
positively
    positioned concentrically using  dowel pins. Universal good practice
    for bolted gear boxes, bell housings and rotary engines for that
    matter.
    Larry's bell housing is going to address that problem
    as it will have a machined pilot bore for Tracy's PSRU.

    Furthermore the PSRU bore face will be positively parallel with the
    aft face
    of the engine within half a thou. Currently ANY variation in spacer
    length can
    cause this problem.
    --
    Paul Lamar ...No rotor no motor.


Paul,

I believe my name is on the list for a bellhousing.  Until then, I'll
have to make do with what I have.

BTW, the holes in my adapter plate were located and bored on a
Bridgeport with a digital readout.  Locator pins would have been a nice
addition, but I didn't think of it at the time.

Mark St.

If Larry agrees I may have some castings next week. Are you willing
to mount one, run the 3 rotor engine up to static RPM and have some people
move
the tail up and down (with stops of course) to simulate gyro loads?
One foot high anchor stands for the mains would help for more angle.
I'll pay for the casting. Larry may donate the machining cost.

That is all well and good but there are other things that can go
wrong. Slop in the 10 mmm bolt holes. Bolts that don't protrude
perpendicular to the end housing. Spacers of varying lengths.
etc. Did you check the exact position of the 10 mm bolt holes
relative to the center of the end housing with a DRO?

If you have a 3" to 4" micrometer measure the exact length of all the
spacers. Errors can be accumulative in a case like this.
It is sometimes called a "stack up" of tolerances.

Paul Lamar ...No rotor no motor.

Paul,
Bob D. did the machine work, and the calculations.  We got the locations
from the drawing on your web site.

I think the recent problem was due to too loose a fit between the hole and
the seal.  It didn't start to leak until the seal worked its way out of the
hole.  I think that was partially due to 100+ psi oil pressure and possibly
an undersized seal and that I didn't use any sealant on the mating
surfaces.  I'm watching the new one closely.  The new seal was a much
tighter fit than the old one.  I'm pretty confident that it will stay put.

Also, the old seal didn't leak until it moved in the hole.  The sealing
surface is very close to the splines on the input shaft.  When it moved, the
seal was now riding on the edge of the splines, which I feel did damage to
the sealing surface.  Moving it back into place didn't fix the leak as now
the sealing surface was damaged.  The new seal looks like it will work.  The
old seal may have been a cheap Chinese version of what Tracy uses.  It was
green instead of black.  I couldn't read the name as that is where it was
rubbing on the hub.

Mark St.

I finally found the specs on that National oil seal.
Here it is. A friend loaned me a Timken catalog on CD.
IMHO it is marginal for our applications.
The shaft is rotating up to 7500 RPM and the heat can
be well above 200 F which is the limit for nitrile.

I highly recommend a spring loaded graphite Teflon seal in this
application. Plus of course a truly concentric and parallel installation.
I am looking for one that will be a perfect replacement. No need for a bushing.

Paul Lamar ...No rotor no motor.

This is as close as I can find in the National brand.
It will require boring out the adapter plate or the new bell housing.