How to bench test a used oil pump?

[boyt911sc]

There are a couple of used oil pumps from 911 SC engines that I would like to determine their conditions. One of the pumps was suspected to be flowing less oil during operation (?). I never tested one before. So any help or guidance how to test them would be highly appreciated. PP sells parts for oil pumps upgrade. Thanks.

Tony

[chris_seven]

You will need a variable speed drive with enough power to deal with the losses of the pump.

You will need an oil reservoir preferably with a heater and a flow measuring system or a calibrated container.

I would also think you will need a mounting plate.

Do you plan to measure the scavenge section as well as the pressure section?

What are the symptoms that makes you believe you have a flow problem?

[boyt911sc]

Originally Posted by chris_seven

You will need a variable speed drive with enough power to deal with the losses of the pump.

You will need an oil reservoir preferably with a heater and a flow measuring system or a calibrated container.

I would also think you will need a mounting plate.

Do you plan to measure the scavenge section as well as the pressure section?

What are the symptoms that makes you believe you have a flow problem?

Chris,

I am rebuilding an '83 SC motor for a fellow PP member that was delivered to me in boxes. Received a message from the owner that the oil pump was suspected to be delivering less oil pressure and caused a spun bearing. Knowing this information, I believe it would not be prudent to install this questionable oil pump without verifying its condition. Would you install this pump in your newly rebuild engine?

My intention is to find out some sort of procedure to bench test the oil pump and determine if I should install it or not. At this point, I am not confident to install it without knowing more about it. It would be unfair for the owner if I recklessly proceed with the assembly knowing now this was a suspected part.

Any advise, recommendation, etc. is highly sought. Thanks.

Tony

[cgarr]

I would take the pump apart and inspect the center and end plates for wear. The gears are a pretty close fit, if there is a lot of movement that would be an easy way to quickly tell it may have a problem. Now I have only taken apart a dozen pumps and don't claim to be an expert about pumps but some of this is just common sense.

[Walt Fricke]

A terminally damaged mag case might be the makings of a decent jig to clamp on each side of a pump. Fittings could easily be installed in the case, passages opened or blocked. Then you could alternatively test the pressure side, and then the scavenge. With pressure gauges installed to check - perhaps a vacuum gauge on the intake side of each as well.

Some sort of orifice on the outlet sides of each so pressure builds? Out there are some flow specs, but maybe pressure would be enough? Take known good pumps, see their specs, then compare with the one under test?

I've not heard of this being done. Craig's approach is what one hears about. But how much scoring on the end plates, for instance, is going to reduce pressure/delivery below what is needed? As long as the pressure setting valve piston can be moved down to its discharge position, it should be good?

[chris_seven]

Gear pumps are basically fixed displacement so it is quite difficult for the flow characteristics to change without modifying either the length of the gear or the tooth form.

As Craig says if there is increased clearance on the end plates or scoring of the rotor housing then the pump won't produce the correct pressure although the flow into an open tank may be unaffected. A relatively small change in axial float will have a relatively significant impact on pressure delivery so measuring this clearance would also help.

For a healthy pump pressure should be proportional to rpm until you reach the limit of the pressure regulator.

If you do decide to strip the pump then carefully mark the teeth that are engaged so you can re-assemble with the same teeth in engagement or you may find that the pump will wear quite quickly.

Henry Schmidt recently posted that he tested pumps with a modified lathe drive and maybe he already has the facility to test the pump?

We have just bought an inverter drive and have started to make a pump test stand but it is not finished and we are in the UK.

I would agree that you do need to check the pump but I would also look for other causes for the spun bearing.

[Walt Fricke]

How much would annular (if that is the right word) scoring of the end plates reduce flow, assuming the much larger areas between the scored circles is still flat, and at the same clearance from the gear?

Or is it radial scoring of the ends of the gear teeth which is especially important, since those channels would allow oil to slip backward directly?

[Henry Schmidt]

We had a fixture back in the day but that walked away some years ago. Now we just run the pump on the lathe and check for torsional load and vacuum created by turning the pump.

[Walt Fricke]

Henry: How do you check for torsional load? What load specs and vacuum levels are you looking for as go/no go levels?

And great snakes, who would walk away with such a special purpose gizmo like that? Nice Snap-on tools, computers, micrometers, that stuff must be the bane of a shop owner with employee's life over time. But a fixture?

[boyt911sc]

Originally Posted by Henry Schmidt

We had a fixture back in the day but that walked away some years ago. Now we just run the pump on the lathe and check for torsional load and vacuum created by turning the pump.

Henry,

I am delighted to see your post and got some direction to go for my oil pump investigation. What speed or RPM do you use for such test? Any suggestion or tip you could share to get this test done reliably and safely? Thanks.

Tony

[chris_seven]

If you make measurement I would think that end float should be a minimum of 0.002" and a maximum of 0.005"

The gap between the end of the rotor and the housing should be a maximum of 0.005"

[KTL]

Agreed taking apart the pump is a wise thing to do. Its not difficult and it gives you visual evidence of what's going on inside the pump. I know from personal experience that an oil pump can ingest bearing material, spit it out and still feel nice and smooth to the feel test of spinning it by hand.

Spun bearing sets off deafening alarm bells in my mind. No way i'd ever consider putting the same pump back in an engine that had a spun bearing, w/out at least taking it apart for inspection.

Agreed with Chris on disassembly. Keep the gerotors/impellers marked so you can reassemble exactly as they came apart. Reason being is that the same teeth are always meshing with each other. If you use a paint pen to put a dot on opposing teeth of each gerotor, you'll see the dots always come together as the pump rotates 360 degrees. Also don't mix up the gerotor stack.

Here's a picture inside of my now wall art 964 pump that gobbled up a bunch of bearing debris. Spun just fine by hand. Not even the smallest hint of dragging or roughness.



That said, an oil pump test rig with an old case sounds like a great way to do the flow testing. Still, it's not a bad idea to look inside and see why the flow may not be up to snuff.

[racing97]

Here is a guy that claims to be a pump "porter and polisher" and in his literature it states that he increases the flow by 20% and gains pressure.
You could inquire if he flows the pumps on premises and how or where?

Glenn Yee Motorsports - Porsche Racing | Service
regards

[Henry Schmidt]

Rebuilt, modified for maximum flow, tested and ready to ship.
Outright or exchange. 911 Vintage Parts 760-731-4911

[KTL]

candy store...........

[Trackrash]

What about "porting" the oil pump for better flow. I did this to my oil pump in my type I hot rod.

Is there a tread outlining it on a 911 pump?

[racing97]

Well porting an oil pump is beyond my pay grade but the flow is a function of the restriction in the system, which in this case would be a labrinth of passages mostly very small I can't quite rap my head around the importance of such an exercise. Maybe Chris will chime in?

[chris_seven]

Originally Posted by racing97

Well porting an oil pump is beyond my pay grade but the flow is a function of the restriction in the system, which in this case would be a labrinth of passages mostly very small I can't quite rap my head around the importance of such an exercise. Maybe Chris will chime in?

The main restriction is, of course, the pressure relief valve and pump that are constant displacement will produce pressure that is directly proportional to rotational speed up to the set point of the relief valve.

I can understand how reducing restrictions will help efficiency and we are currently building a test rig to make flow measurements.

[Henry Schmidt]

The original pumps were designed with a flaw that was corrected by race engine builders in the late 70s. This modification improved the flow and enhanced pressure at the 8000 RPMs we saw in those engines.
The modification we did to race pumps has translated to a simple improvement we do to all pumps we install. We started offering this modification to shops and now DIY builders that seek to improve their oil flow/pressure without investing in a new/later pump.
In the early 90s, the Porsche factory incorporated this modification to most of their production pumps.

[0396]

Originally Posted by racing97

Here is a guy that claims to be a pump "porter and polisher" and in his literature it states that he increases the flow by 20% and gains pressure.
You could inquire if he flows the pumps on premises and how or where?

Glenn Yee Motorsports - Porsche Racing | Service
regards

+ one on Glenn's work.