Seized 930 Oil Pump

[jluetjen]

OK. I've finally got to the point were I'm starting to put a few things together on my race engine project, and I was trial-fitting the 930 oil-pump that I bought off of Ebay a few years back. I noticed that the wouldn't spin, even though the original (2.4 TK) oil pump that came with the engine core would spin freely by hand. Hmmmm....?

Background: As I said, it's a 930 oil pump that will be going into the Competition Engineering reworked case that I have. The pump was clean when I bought it, but I don't remember it spinning then either. I took it apart and it's clean on the inside, but it was dry. I'm guessing that it was run through a parts cleaner before it was sold, because there was no sign of any oil on anything. There's no sign of rust anywhere, nor do I see any obvious nicks or cuts that could be fouling the rotors. I lubed everything up with some Swepco (that I'm using as assembly lube) and I can get the individual pieces to move, but it won't move when it's assembled.

After a few more iterations of taking it apart, watching the parts spin and putting it back together, I've localized the problem to the suction section of the pump. Each of the rotors is actually a long rotor (I'm guessing from the standard non-turbo pump), and a second piece that provides the extra capacity. Both pairs have a keyway cut on the inside, but only the upper pair is keyed to the shaft. Once again, I've wet everything down with some Swepco, and reassembled. They seem to spin, and then foul on something, but I can't see what. I'm guessing that the lash may not be right.

Any thoughts?


PS: Anyone have a short "Turbo" shaft extension? If so, please PM me.

[davidppp]

Hello John.

I supsect its been slightly rewoked:

Normal method is to reduce the end clearance to zero by crossfiling or even running on a belt sander..

This means it will spin fine until the nuts are snuggedn down..

The remedy is a little crossfiling of the rotors, or if very slight, lapping paste on the ends..and run with a lathe or drill motor.

Kind regards
David

[jluetjen]

Thanks David;
Just to clarify, I've been able to get the rotors to sieze occasionally even when the pump is open. But that may just be due to slight alignment issues given the fact that the pump is open, I'm sure that the tolerances are tight. It does seem to be a sure thing when the pump is closed that it doesn't spin -- so you may have ID'd the problem.

[jluetjen]

OK, I took the pump apart a few more times and played with it further and I think that I spotted a problem.

Note the galling on this rotor:


The corresponding surface on the pump case has this galling or what I'm guessing is some micro-welded material.



(my apologies for the picture quality, my digital camera doesn't have macro capability.)

The question is what to do? I'm pretty comfortable with sanding or fileing the rotor because it's easy to get to the surface. But what about the case? I could take the trusty moto-tool, but I'm concerned about doing more damage then good. Any other ideas before I start cutting metal?

[jluetjen]

That was it.

For the case, I took one of the smallest (and longest) cylindrical cutting bits for the moto-tool, at a fairly slow speed and chipped that excess metal off of the case while being careful not to dig a hole in the case itself.

For the rotor, I very judicously hand filed the end of the rotor, as well as corners of the damaged vanes. I also found a few places where the middle of the vanes were siezing on the case, and lightly filed those areas down to. I really took off almost an imperceptable amount of metal, but it was enough.

I feel a lot better now that that issue is resolved. I think I'll next move to assembling the heads next.

[BURN-BROS]

Good job John!

[Grady Clay]

John,

Here is a product that everyone should have. It is great for
spotting high spots in the oil pump. This should be done even
with new OE Porsche pumps.
Dykem Steel Blue Layout Fluid.



These seem to last almost forever. I think I bought a case of
six many years ago and still have two full cans – the copyright
date on the can is 1967!


Measuring the radial clearances and gear clearances are very
difficult.
Plastigauge is a good and inexpensive technique.



Best,
Grady

[jluetjen]

Grady;
I don't have a turbo manual -- do you know what the clearance specs are for a Turbo pump?

[Grady Clay]

John,

I may be wrong but I don’t think Porsche ever published specifications for the oil pumps on any engines. That includes the oil pumps, Sportomatic pumps and Turbo pumps. If you get into 915, G50&64 and Tiptronic, there are many more.

Disassembling and visually inspecting an oil pump is part of any competent rebuild. If something is amiss, it is usually obvious. It was those situations where everything appears OK that we got into measuring actual running clearances. Somewhere I have a binder of all that but I haven’t seen it in 25 years.

Today, with the Pelican Engine Forum, we can share this information. Some shops are hesitant to release what they consider proprietary information. That is fine, we can just add to their database without any expectation of reciprocity. I think DIY only adds to the pro shops abilities. When I have my son’s race engines built (three 2.8 SCCA GT2 engines), they will be done in a professional shop and not at home.

It has only been recently that I have seen a new defective Factory OE pump. When the housing was cast, the interior part of the casting mold shifted. This left a “half-moon” shaped cavity along-side the radial gear surface. At maximum it extended about 25 mm from the blind end of the scavenge casting. The depth of this void was about 0.8 mm.

This was a Factory new GT3 Cup part and a race 2.8 was built around it. The symptom was the engine would quit scavenging (net between the pressure pump and the scavenge pump) after the oil got hot (3-5 laps). The good news is it didn’t cause any damage other than requiring a teardown. The bad news is it casts a pall over Factory parts. I will now measure and test every Factory part. Gone is that confidence.

A benefit here can be to develop an inspection procedure. There are a lot of dimensions a perfectly good oil pump must meet. There are also visual clues to a problem.

If I go around Denver, I suspect I can find many hundreds of 911 oil pumps in various conditions.

If we were to develop a procedure, what should it be? It should include visual inspection, measurement and actual performance pumping oil. What are the best procedures and standards?

Best,
Grady

BTW, Thanks to Chris Bennet (for giving a 930 manual) and others, I have somewhat reconstituted my Factory Workshop Manuals. Still much to go as I gave away too much in the ‘80s.
G.

[304065]

Reviving this thread.

So David, when you say the normal repair method is to reduce the end clearance to zero, I assume this is done by taking the pump case apart and filing, sanding or lapping the case halves so that the case becomes shorter relative to the length of the rotors, therefore reducing the end clearance between the rotor and the housing. This would prevent oil from escaping around the end of the rotor.

I suppose one could measure this using a depth micrometer on each half of the pump case pointing inward to where the rotor end rests, sum the two, and compare this to the overall length of the rotor. The case should be a fraction longer to allow the rotor to turn. Bill Fisher's How to Hotrod VW engines cites an end clearance of 0.0005 inches for the 69-71 VW pumps. Also says you want not more than 0.003 inch clearance between the gear teeth and the pump body. VW pumps use a gasket between halves so that's why they get away with ZERO end play.

Grady, where the dykem comes in is in lapping the pump case halves? Or would you coat the rotors with dykem and spin the pump with a drill motor, disassemble and note the wear pattern? Is it possible to move one rotor with respect to the other without welding up the pump case?

[HawgRyder]

You might also try indexing the gears to find the best running rotation.
Harley Davidson used to do this long ago, the gears were marked with a punched dot in the end.
Worth a try if there is a tight spot.
Bob