Since the question has been asked - here are some cobbled together extracts from other postings on the subject that contain a lot of technical information about the problem and our solution. Because I DON'T HAVE TIME TO EDIT THE CONTENT IT MAY REPEAT DETAILS A BIT SORRY!
The positioning of the intermediate shaft long ways is maintained by the bearing press fit. Although it has a small wire circlip in a groove but this can be overcome to remove the bearing by shock loading and it is mainly the interference fit that holds it in place (the clip later changed to a circlip).
New bearings arrived for assembly with grease inside and seals on both outsides. There is no way for oil to pass in or out of the bearing initially. While grease is OK for lubrication at the loads I have calculated and the revs - a heavy interference fit on both the inside and the outside of the bearing results in the balls inside grinding some minute particles of metal away as they "run in". I think everyone knows that when running in you usually have to change oil to remove those initial wear particles while an engine settles in - but there is no way this happens in this bearing - which is perfectly adequate for the loads applied to it - given sensible lubrication so IMHO a more expensive bearing is not neccessary (just the right spec and clearance) and we have had none fail in years of modifying and replacing them and to three figure numbers.
We do not fit seals because with fresh oil being thrown in the direction of this bearing by the two chains nearby running in an oil bath in the sump area - at up to 40mph - there is plenty of potential lubrication available and as the ball tracks are hollow there would always be sufficient oil sat in the bottom of the groove to accommodate initial start up – and now fresh clean oil will be present.
Some we have replaced early have a mixture of grease and oil and what looks like grinding paste inside the bearing and with two ball tracks in the earlier bearing the amount is - suprisingly large. The single row bearing specs up slightly better than the double row because the double row has less deep ball tracks – so the replacement is actually probably better and of course half the metallic particles to wear away.
There is also no cooling since no oil flows in or out.
It seems to me that some are tighter than others (normal mix of manufacturing tolerances will allow a small number to have the tightest fit on both inside and outside while most are less tight) and of course some will run hotter than others while manufacturing tolerances will result in some actual bearings needing more running in (and creating more metal particles) than others.
Most survive until the seal wears and allows some oil in which mixes with the grease forming a type of grinding paste and later on as the seal edge suffers it washes some grease and particles out and eventually works more like the seal was not there in the first place. How soon this happens is again a function of the tolerances and use.
Therefore some - once they get past say 50K or so – generally improve and therefore last a long time (the loads are low - the other end of the shaft which has one chain running against it s just a steel diameter in an alloy hole).
We have seen some in which the ball cage has cracked and although the balls rotate in the same direction – when they crowd together the face of the balls runs in the opposite direction to the next ball so the surface of the balls run against each other at twice the speed that the balls are rotating at and wear rates increase dramatically increasing wear particles. Others have actually cracked the outer bearing housing through metal fatigue (clearly marked) implying that they simply ran too tight for too long or the lack of a cage and the balls crowding over one side of the track overloads the outer race.
Because the spindle is such a basically poor design with a very weak centre - this then often snaps.
By removing the seals - most bearings should be OK since the grease is quickly replaced by the oil and the wear particles can escape and be collected by the magnet or filter. However while removing the bearing in situ is possible (and we have a tool to do so safely) replacing it strains the weak original spindle and could cause a fatigue crack initiation - so a stronger design of spindle should be used to reassemble it. It is also necessary to take off both chain tensioners to avoid the chain load forcing the IMS over too much to one side and expecting the spindle and housing to force it back against the hydraulic tensioners. This obviously takes longer. We also change oil and filter at the same time and usually the RMS.
In my view there was nothing wrong with the idea of an IMS - only the basic engineering thinking was flawed - but then only enough to fail in very small numbers statistically - so not an absolute disaster.
What amazes me more than anything else however is how simple the IMS problem was to analyse and fix and how long it remained a weak spot despite several re-designs - none of which addressed the obvious cause - the need for a tight fit and the lack of proper lubrication resulting from the seals.
By repairing them for those that cannot afford a new engine - we actually IMHO help preserve the confidence of owners and the reputation and value of the cars.
We mainly deal with cars over 4 years old and so have to deal with owners with generally smaller financial resources – than those buying brand new cars and changing every three years.
To do that better than anyone else we introduced our Lifetime Maintenance Plan 10 years ago (well ahead of the market) and invested in engineering solutions, equipment, space and staff - to enable less well off owners to afford to keep their cars - even if and when they become one of the small in number whose experience is devastated by a failure that could have so easily been avoided.
There is no other UK business offering anything like as comprehensive a warranty, nor anything like the quality and variety of repair solutions or the ability and willingness to adjust rebuilds to suit the pocket and future intentions of the owner.
I did contact the manufacturer to buy replacement bearings – which were out of production but in the course of the conversation they confirmed that they too do not understand why Porsche sealed the bearing and recommended they did not. Furthermore- any engineer who is familiar with engines will know that a grease filled bearing inside an engine will gradually lose its grease and then run on the oil. The debris trapped inside and mixed with the grease initially and with old oil later on – prematurely wears the seal edges to allow the grease/sludge to seep out and new oil to seep in and then sit there.
We have seen a failure of a later M97 engine IMS but it is rare – however a larger bearing will resist contracting as much under an interference fit and has more space for more grease and for more contaminants to spread. Similarly we have seen the gearbox bearings (which are also sealed) fail.
I do not expect every IMS bearing to fail and I don't think that this failure is related to use or even servicing and I believe it is entirely random for the good engineering reasons I previously stated about the quality of the bearing and the build- up of tolerances and have even explained why after say 50K they become more reliable again..
The bearing is a sealed for life bearing but they are not sealed for life when running in hot oil and that is the mistake IMHO. Indeed when you remove the bearing some very black old oil mixed with metal particles slowly drips out of the IMS tube in which it was trapped on the inside of the seal and bearing.
On inspection of those within the danger zone – some lower mileage ones have a mixture of old oil and grease and metallic particles like a grinding paste inside, older ones just have old oil and particles and very old ones have just old oil left.
Clearly oil slowly ingresses and mixes first with the grease and particles and later washes the grease away just remaining as old oil mixed with particles.
A heavy interference fit will inevitably vary from part to part as will the fit and quality of the bearings – so some have a build up of poor quality and tight fit while others may be of better quality and a loser fit – most somewhere in the middle. Some will be run in slowly and others more aggressively and oil changes may be more or less frequent. We simply don't know if this has any influence or not but out if it all some fail and most do not. It is a marginal design and removing the seal and allowing fresh oil to wash out the running in particles and keep it cool and properly lubricated – is the best improvement.
I have wondered if the seal is to protect bearings and pre-assembled components lining up for mass production – but it should have been removed on assembly.
Numbers are however minute – it is not such a big technical failing – so therefore – to come back to exactly where I started from – my advice is just get some cover that you can afford to protect you from the cost of a failure – just like you do for house
Insurance and car theft etc.
Eventually all the grease has mixed with the oil and become entrapped in the tube leaving very old oil mixed with bits of metal trying to lubricate the now prematurely worn bearing with minute spots of new fresh oil gradually seeping in.
Once a new bearing has been fitted without a seal (of whatever type and quality) I would always promote the best oil you can afford and more importantly changing it often. My only argument is that I don't see changing oil quality making any difference if the original bearing and seals are still in place because the problem mix that is already trapped within the bearing as it came from the factory is still in there and if you are going to take it to bits to clean it you may as well change the bearing and leave out the seal - after which I believe std oil types are perfectly OK..
A better quality product is always better than a lower quality one but more expensive and if it is unnecessary the extra cost is wasted. The question is always - is it necessary to fulfil the function. In my view and engineering experience and the evidence of those we have repaired – the std replacement bearing of the right grade fitted with the stronger spindle is perfectly up to the job, has proven to be OK and we guarantee it anyway. The problem was never the specification of the bearing but was the insistence of fitting it with seals that appear to have actually caused the problem in the first place.
All parts are made to minutely different sizes and inevitably sometimes you get the largest pin going in the smallest bearing hole in which the inner track is big and the outer track small with the biggest balls in it and fitting into a housing that is the smallest. The odds against it all working one way are huge and therefore the number of "tight" fits is low - but it does happen and likely to present more of a running in problem and more debris to contend with. The amount of grease in there also probably varies - all leading to a small number of failures simply because the design is IMHO marginal.
Bearings are available in a variety of fits or tightnesses and when an interference fit is in use the designer should chose the right one for the application. We have made our choice and we find it works OK.
I will try and attach some photos but if they do not appear and someone sends me an E-mail address who can display them I will send them on to them.
Baz