Autopsy of a "good" IMS bearing
 

A stock 6204 single row deep groove ball bearing was cut apart for inspection. Before dis-assembly, the bearing seemed tight, and the side seals appeared intact. The bearing was taken from a used replacement IMS shaft with unknown history. There was oil in the IMS shaft tube when the bearing was removed. This picture shows the bearing upon removal from the IMS shaft. the seal looks good.


http://986forum.com/forums/uploads01...1385590171.jpg

Taking off the side seal, the interior was filled with what could best be described as brown oily goo. It appeared to be slightly thicker than motor oil.


http://986forum.com/forums/uploads01...1385590332.jpg

Cutting apart the bearing to inspect the inner races revealed a series of equally spaced scuffed lines running across the bearing race. These appeared to be discrete areas of fine pitting on the races. There were 8 lines on the inner race, and 7 lines on the outer race. This is the inner race:


http://986forum.com/forums/uploads01...1385590736.jpg

and this is the outer race - and a bottle of wine, think I'll have a glass after this post:


http://986forum.com/forums/uploads01...1385590827.jpg

This looks like the very beginning of the bearing failing. The 8 balls in the bearing looked like the finish was going and there was at least one noticeable scratch evident. My sense is this bearing is failing due to contamination and improper lubrication - low speed operation with too low of a viscosity lubricant.

Thanks for the detailed info! How many miles?

Unknown mileage. Came out of a shaft I bought on Ebay.

Nice post! I rarely read threads with IMS in the title but it was good to see this.

Thought it was interesting/sad that apparently the current part# for new IMS ends in 06 indicating 6th revision/improvement? :mad:

James....
As per BYprodriver said.

But without knowing the history, the results are too simplistic to draw any meaninfull conclusions, as the pictures only show a slightly worn bearing that could have been in operation for 10,000 miles or 110,000 miles.
If the mileage & service history was available, the resulting wear would be more meaningfull. And as you know, some high mileage engines won't have as much bearing wear as others even though they have had a similar maintenance regime.

I found it interesting that there are 8 balls in the bearing, and 8 equally spaced wear patches on the inner ring with 7 on the outer. I don't know what that means, if anything. I was hoping someone else might chime in.;)

Hi,
I have been playing with ball bearings, running them at high speed until they fail to see how and why. What I have observed is that as they wear, they produce a metallic dust that is incredibly fine. I think that the oily "goo" that so many people have mentioned is a mixture of oil and this metallic dust.

Interesting for sure. . But no conclusions of any type can really be drawn from this sample bearing.
Doesn't really look that bad to an untrained pair of eyes like mine.

^ What Steve said. That's why I was wondering about the mileage.

Glad to see this autopsy done. This is what we do with the bearings that are extracted, and its how we have gathered so much trend data.

When people have an IMSR carried out and they say "I felt the bearing and it was fine, I changed it for no reason", I just shake my head and ask them if they cut the bearing apart. The answer is always no.

If they did, they'd find something resembling what is detailed in this post. Judging a book by its cover has never been a good idea.

So, can someone draw any conclusions based on the photographic evidence? Approx mileage? How much longer would it have lasted?

Well not everyone says "no" :). However I will go back over it and look again with stronger magnification. Mine though was double row, very low failure rate.

That bearing was at a stage 3 failure point. I have caught a few of them like this, the single row IMSB is at stage 3 for a very short period of time compared to the dual row which will run this way for much, much longer.

Hi James,
I am curious about the inside diameter of the IM shaft, the tube, behind the bearing. Do you have means to measure that?
Thanks

As close as I could measure was 34.25 mm. I'd put +/- 1 mm on there for tolerance as the measurement was not straightforward.

Thanks James. I have been looking for a shaft on eBay.

Well now that is interesting. Too low of a viscosity lubricant. I've been wondering about running a heavier oil in my engine. Like the 10W40 that my original owner's manual indicates as opposed to the 0W40 that everyone talks about now. Is the tolerance on the Porsche engines from the 2.5l engines as tight as the new Hondas that indicate 5W20? I compare to a Honda engine because they run for multi 100k without coming apart. ???

No, and the engines lose a ton of internal clearance during expansion. With %/20 oil you'll be opening a can of worms that you won't like. The oil pump in these engines suffers greatly from internal losses due to clearances changing with thermal expansion, which makes it all worse.

I have developed another product to help with these losses and its going through the patent processes now.

5/20 oil might work fine in the M96 engine if operated in the Arctic, otherwise you'll have issues with valve lash, hydraulic tensioners and many other issues. This has been proven through direct experience and development on the street, track, dyno and otherwise with the M96 engines as our only focus. What the IMSB likes, isn't necessarily what the engine likes as a whole, that is unless you finally just suck it up and install the IMS Solution and allow your engine's oil to support the loads of your IMSB by taking that damn ball bearing out of the equation altogether.

oil viscosity
 

Thanks Jake, That's interesting how they built the engine around such strange tolerances. How does Subaru get around this with their boxer engines that crank out some incredible HP per displacement? (of course Subies have head gasket issues). How about all those motorcycle engines out there with roller bearings located all over the place in the engine? Or is that just because of how the ball bearings are located and consequent loading in line with the bearing or a thrust loading?

Anyway, I appreciated the experiential response! :)

Those other engines are not quite so impacted by the accounting department. Suby engines have 2X tighter running clearances than the M96, which is both a pro and a con, because tighter clearances lead to shear faster.

The M96 is the M96 and it isn't anything else. People refuse to listen when I say this, yet I have cars here from 30 states primarily because people treat them like something they are not.

how to treat the M96
 

So drive it regularly, get the RPM's up. pay attention to sounds, keep the oil changed with the right kind of oil, and enjoy the ride? Mine's a 2.5l:cheers:

I am speculating that it may be a shock that is sent through the cage when the balls change from being loaded to unloaded. A ball that is under load pushes the cage, and the balls that are not under load are pushed by the cage. Perhaps a shock is sent through the cage each time a ball transitions from loaded to unloaded. This causes the ball to slam from one side of its cell to the other side of the cell. Maybe this creates a shock that travels through the whole cage and affects all the balls at once. Maybe this shock causes the balls to skid for an instant as the cell wall crashes into its ball.

Looking at new, good used and failing bearing surfaces (races and balls) under an electron microscope really tells the story.

These photos are published in my book :-)

Here is some information I found. Information on this site seems easier to find than most.

Knowledge of Bearings

Rather than start a new thread here is some text and photos of my old IMS from a 105k miles 2004 car.

http://www.boxa.net/forum/index.php?showtopic=59044&page=3#entry709368

Just took the seals off my bearing. It looks quite different to the first post.

Your link does not work.

It does, you just have to be logged in as a member to see it. God knows why they want to restrict traffic like that, but they do.

Hmm.... Sure would be convenient if it were posted here...

Here's a cut and past of the relevant posts:

Well, I now have the old IMS bearing. First things first, I am no engineer despite working for an aerospace engineering company! My photos and commentary are merely observations.

For context, and repeated from earlier posts, the car is a 2004 986S with 105k miles. No paperwork but confirmed decent history from OPC with regular servicing. I don't know for sure but I believe the car was a daily commuter although the condition of the interior suggests a car with half the miles and a weekender. I believe the oil changes were all done with 0w40 Mobil 1 (oil change prior to sale was and this spec oil is what OPC use).

The original NSK bearing was replaced with an LN ceramic one as part of a schedule of works including a clutch change.

Russ commented that this was one of the most solid bearings seen. Holding the threaded shaft and wiggling the bearing shows no play at all. Holding the shaft and attempting to spin the bearing by finger then next to noting happens. You have to maintain contact with the bearing outer race and rotate it rather than spin it. Whewn turning it there is no notchiness, the movement is very consistent and smooth. There are no weired noises.

The seals seem completely intact, no pitting or any sign of damage. There is a small indentation on one part of the seal pictured below but I presume this was done at removal or in transit via post. The outer race has some light markings that go round the circumference. Running your nail over them you cannot feel them.

Old bearing:
http://i.imgur.com/JrdEdH7.jpg

Marking around circumference:
http://i.imgur.com/CdP2XhW.jpg


http://i.imgur.com/x0To8sg.jpg


Indentation on seal:
http://i.imgur.com/SPPOUtS.jpg



When I get the chance I will remove the seals and have a look inside.

Here is the next IMS instalment. The seals are now removed. Again, take what I say with a pinch of salt. I don't see many bearings in different states of ab/use.

To remove the seals was reasonably straight forward. The smallest flat blade bit pierces the seal. Trying to lever it out on the outer edge proved pretty hopeless.

Here is one side of the bearing opened up.

Nothing much untoward here:

http://i.imgur.com/a6Druiv.jpg


http://i.imgur.com/chsqFbi.jpg


Back of the seal was pretty clean, no real evidence of any debris:

http://i.imgur.com/H5aRSii.jpg


In this shot I wiped my finger over some of the ball bearings. You can see at the 3pm position the bearing is cleaned of grease/oil and is silver. The bearings are otherwise covered in a light film of grease/oil which has a silver (probably metal) tinge to it.

http://i.imgur.com/4076t8J.jpg


This shot shows the seals and various finger wipe marks to show the colour of the oil/grease inside the bearing. It seems evident that oil has got past the seal but is generally a golden brown colour and the same sort of colour as a good clean oil. The grey smears are what has come off the face of the ball bearings:

http://i.imgur.com/SeZMbYB.jpg

Here is a shot of the other side of the bearing. You can see at the 4pm position again the finger wiped ball bearing. Look at the grey film grease suspension that has been removed:

http://i.imgur.com/gI9JBSc.jpg



Next step would be to remove the ball bearings to look at the race. However, I am all out of tools for that aside from brute force.

FYI my now not so clean white board was a polystyrene bit of pizza packaging :D

Looks like there was no grease in the bearing, just oil. This looks similar to the Pelican Parts replacement kit. What are the numbers on the bearing's grease shield?

I'll have a look later to check for any part numbers. This isn't a pelican part as far as I know. It was taken out of my standard engine.

I have to retract my last post - The pelican parts is a single row bearing and this is a double. I was looking at the bolt and it looks beefier than the stock bolt out of my single row.

Mine is a 2004 car so a single row ...

Poland nsk 917 6204du17