Case Shuffle
 

Does anyone know the reason that 911 cases shuffle.

Is it crank bending, vibration of poor design?

Which cases/crank combinations are affected.

Is 'shuffle pinning' the best way to solve this problem or are there other remedies

Does your case show signs of fretting? I have never seen this with a 911 case but have with aircraft engines: I machine the case at 10 of the mains thru bolts with sleeves just like the #8 bearing for people who like it done, usually on mag cases.

http://i30.photobucket.com/albums/c3.../casebore3.jpg

http://i30.photobucket.com/albums/c3.../casebore2.jpg

Okay, I'll bite.

Fretting, in and of itself, is interesting to learn about. Basic premise:

Fretting - Wikipedia, the free encyclopedia

Nanometers of movement can cause the condition. Vibration, thermal expansion - you name it. When the engine gets hot and the case (aluminum or magnesium) expands greater than the pin does, does the pin still work when we're talking about nanometers of movement? Probably not. How about a sprinkle of glue between two faces (like loctite)? Probably not... if there is a present condition then I think that it will dominate the glue.

That said - the fretting may be a result of harsh operation before the case is brought up to temperature when things are loose. Perhaps then, the pins function just so to prevent dancing between surfaces.

It doesn't surprise me that the main saddle surfaces have this condition. Think about how the case is secured. It's pinned at the flywheel end with locating sleeves and it's bolted and bushed at the pulley end with studs and the #8 bearing housing. In between, it's secured with the looooooong thru bolts. While the thru bolts provide the clamping force, they also rely on the grip/friction between the case halves to resist shuffling of the case here, do they not?

I'm betting its vibration related. The 911 crank is pretty stout with 8 main bearings. I've seen the fretting on SC cases. The SC 70.4mm crank is considered a fairly good crank. But it's also a victim of the harmonic issue that plagues the earlier 6 bolt 70.4mm cranks. I recall it's in the 7000-8000 rpm range where this occurs?

I think I heard that the harmonics are worst in the 2.7L engines. Maybe even worse than with the 3.0 due to different jounal sizes? And either way, no doubt made worse due to the magnesium cases on the 2.7. I'm a fan of shuffle pinning even the aluminum cases for best stability, though some would say it's overkill. I had my sandcast aluminum 2.0L pinned when I built my 2.7 race motor.

Scott

A discussion of the causes and modelling of fretting could be interesting but wasn't the exact reason for my question.

I would think that fretting caused by slip displacements in range of nanometres is a very rare condition and very material dependent.

Fretting due to displacements in the order of 5 to 100 microns is much more common particularly when considering soft materials.

Most fretting in cast magnesium alloys such as AZ91 seems to be due to adhesive wear and the mathematical model of this process is very complex and involves both the level of the axial displacement and the normal stresses.

It looks as if increased clamping loads could potentially make matters worse rather than better and there appears to be critical combination where damage is maximised.

The main reason for my interest is that in the UK 'shuffle pinning' is a process that has almost become de rigueur.

Looking at the methods being used it seems as if there are a number of different approaches some involve line boring and expensive bearings and others claim this isn't needed.

I have also seem a few engines which have had the saddles machined and then re-bored to standard but I am, not sure if this is a good idea.

I was interested to know the real cause and if is specific to a few engine configurations or if all engines are affected.

I have recently heard of 964 and 993 engine cases being pinned.

I have never seen cases that seem to have fretted badly.

There is also a new cold spray technique being used to repair magnesium helicopter gearbox cases and once the repair has been carried out the resulting face is said to less prone to fretting.

I've been told that the pinning technique like cgarr mentioned (pinning at the oil galleys) is less likely to require a line bore. I'm not sure why though.

As for machining and line boring back to standard, chris_seven you mentioned machining the bearing saddles. This is true, but I believe the entire case mating surfaces (including the perimter) needs to be machined. You may have meant this, but just mentioning in case. I had this done to my sandcast alu case by Competition Engineering here in the US a bunch of years ago.

Scott

Thank You - we do some wet liner engine work now and commonly machine cylinder block perimeters but I appreciate the reminder.

I do realise that the complete case would need to be machined and this would narrow the engine, which I don't think is a great idea.

I am trying to work out the increase in stiffness due to the dowels and the clearance becomes an issue.

I have been told that it is possible to position the holes accurately enough to avoid line boring using a jig bore (expensive machine!!) but the typical cost of this, in the UK would be between $150 to $180 per hour.

In general I would expect a dowel fitted into a soft alloy to use an interference of around 25 microns.

I have been told that for shuffle pinning the company that used the jig bore allowed this interference on one half of the case but a clearance of between5 and 10 microns on the other side.

This seems to me to be sufficient to still allow fretting especially as this clearance will increase with temperature.

I would also worry about the clearance holes opening up due to fretting.

Is it possible that the e 'hollow' dowel may be a little more flexible and this may help ?

Hi Chris,

We are starting to see cracking in the mag cases in odd spots, like stud bosses, chain ramp bolt bosses, etc. There have been reports of cracking due to shuffle pins although I have not come across it myself. The general consensus from colleagues is to no longer shuffle pin them.

Aaron,

I am coming to the conclusion that this technique is risky and am not too keen on some of the issues - particularly constraint.

case bolts also seem to have been made with 'lumps' in the centre for use on the 917, RSR and 934/5

I have seen articles which say that these lumps are used to 'pin' the cases and others which say that they were a loose fit in the oil galleries and were just designed to break up resonances.

Any thoughts?

I have no experience with the bolts. I like the idea, but one would check the accuracy of the thru bolt holes. both location and diameter. I am not sure the factory required the holes to be that accurate.


As a side note. The GT3 cup engines experience very little fretting. Very light fretting on every other main saddle. This may be due to increased rigidity of the case, the design of the cylinder bank and head(both are gang of threes) along with the reduction of reciprocating weight. Piston,ring package and pin are under 500 grams....difficult to do on bigger air cooled engines with 90cc combustion chambers.

I wonder if expanded-diameter through bolts could be used to pin the cases ... doing so would block oil from passing (I think at the through holes are oil galleys)?

They are that's why you can't block those passages, another reason you use o-rings under the washers. The thru holes are not machined so a tight fit would be impossible even if you were to do it and the outside case holes are smaller than the holes at the mains so you would never get the thru bolt to fit either without some changes to the case.


Sent from me

That's why they have a flat machined on them and the case does need to be modified.

http://i197.photobucket.com/albums/a...psr1xlb922.jpg

Factory Through Bolts were a little different an used spherical seats but I am not sure this is essential as the bolt is long enough to accommodate slight misalignment by bending.

911 RSR / 935 Crankcase Bolts Photo Gallery by 914-6 GT at pbase.com

Case Shuffle
 

How do they route oil? That would not work in a stock 911 case without oil galley mods unless that flat spot is to allow oil past in which case you loose a bit of alignment if the case is ever under any stress.


Sent from me

The oil routes along the side of the flat.

I believe that as there is a full diameter and only a small chord section is cut away there will be sufficient alignment to prevent fretting.

They seemed to work on the RSR, 934 and 935.

The question I started with was what caused shuffling and what forces were involved.

If it is only fretting then the forces may be very small and the deformations in the order of a few microns.

If this is the case then it seems reasonable that bolts of this type will work as it is unlikely that sufficient force will be available to make the oil hole go out of round.

A similar bolting arrangement was, I believe was also used on the 917.

I don't know what the internal webbing of those cases look like or how they were designed. I was trying to figure out if there was a way to work in a stick 911 case, always open to improvements and easier ways to do things.


Sent from me

I think that the 3.0RSR is identical to the Carrera 3 but don't know if there were any internal mods.

I can ask about a 935 motor which could be interesting.

A poster in an almost identical thread on Early 911S Registry has some of the RSR bolts and says they are not a tight fit, therefore they won't limit shuffle.

Seems we can set an upper limit on shuffle displacement based on what the bearings/oil wedge can withstand without spinning (I have not heard of engines dying because of shuffle, and the fretted cases seem to be from motors that ran fine). I imagine it's not much, so we are in the micron range and anything that doesn't have an interference in both halves of the case won't make a difference.

Max,

Are you assuming that there is sufficient radial load being generated to cause the case halves to separate and overcome clamping force as this seems very unlikely.

Fretting can occur when the deformation of the case is entirely within the elastic region and increasing the clamping force could potentially make matters worse.

I have some experience of fretting in steel to steel joints and the solution is to always use a surface active lubricant at the joint interface rather than to increase clamping force.

I am not sure how to achieve this with a Magnesium Alloy but there should be a suitable method and I have asked some colleagues that design Helicopter Gearbox cases for some advice.

I agree with your comments that any clearance will still allow shuffle to occur and certainly in the UK it seems common practice to allow between 5 and 10 microns on one set of the dowel holes.

This would still need a reasonable force for assembly of the case. The clearance would of course increase at running temperature.

The bolt in the photograph I posted has been used for some years and is an interference in the oil gallery which is line honed for an m6 fit.

They seem to work and are relatively easy to install.

I have always thought that factory bolts had a clearance but I have never found out how much.


Again a few microns would still need them to be pulled into place with a reasonable force.

I do believe that the 'lump' is to prevent the bolt from losing clamping at high rpm due to 'ringing'

Interesting, I thought there had to be sliding for fretting to happen.

Max,

You need to define sliding in this case. I agree that there needs to relative motion but sliding can imply low levels of friction.

Fretting is much more of a stick/slip problem and is likely to occur at the peak of a 'friction hill' rather than when sliding is really occurring and friction levels have fallen.

I believe soft metals fret due to adhesive wear and this can occur when movement is very limited.

It would be interesting to have some idea of the amplitude of case vibrations to try to determine what is happening and maybe it is possible to install an accelerometer and look at the g levels.

Next time we run an engine on a dyno I will see what we can learn.

I favor the bushing approach shown in Craig's photos.

You can do this by clamping a case together, and boring down through the throughbolt holes to, and then a certain distance beyond, the parting line. This insures that the bushing holes are perfectly aligned. Then you can clearance the near side a little based on case to bushing fit, press the bushing into the other side so it butts up to the end of the drilling (all the while being careful of oil squirter passages), and you are nearly set. The trick for this is then to press narrowing inserts into the outside of the throughbolt holes where you went in to bring their ID back to standard.

Much easier than jigs, and I think more precise than standard pinning.

Whether or not this really prevents some harm I don't know. I have had this done on all my cases, mag or aluminum, just because.

I've got no idea what was done to factory racing cases so the througbolt bores lined up well enough at the parting line to use the flat sided bulge as a parting line stiffener. Are there instructions for assembly of these motors which specify an orientation for the center part?

I've wondered about the need for those self-aligning washers, but if a set fell into my hands I'd use them for the cool factor.

Pumping 500 or more HP from air cooled VW 4 cylinder cases might be where shuffle pinning got its start? Or their 356 progeny? With the 911, didn't Ferry Porsche say that if they thought the original case could be used to produce the 300 HP of the 2.8 RSR, or just the 210 of the 2.7RS, they would have made a lighter (and less expensive?) case.

I know this doesn't answer Chris' question. And it is always good to hear from someone who has helicopter transmission engineers for buddies, and shows that lubrication, rather than increased clamping, may be better to prevent fretting. And explains stick/slip by analogy to the difference between sliding and static friction, which most of us have heard of even if we can't quantify it.

Walt,

I agree with most of your comments and all of your sentiments. I am not opposed to trying to pin cases but I am not confident in some of the methods used.

The problem that I have is when faced with detailed questions the machine shops that I have spoken to just tell me that they can do the job and not to worry.

Unhappily I do like to worry and have always believed that unless you either leave a clearance around one set of pins you would have to line bore and I repeatedly told that the job is quite straightforward.

That's where I start to come unstuck - none of the machine shops I have spoken to can provide any information on 'fits and clearances'. None of them can show any drawings or provide either inspection plans or inspection reports on cases that have been pinned.

I don't expect them to hand over copies for me to keep but just sight of the relevant details would help.

The money isn't the issue but from an engineering perspective why carry out a process that doesn't do much and suffer the attendant risk.

The appeal of the case bolts with the 'lump' is the fact that they can be installed with the case bolted up and we have made a centreless ground bar to line up the bearing housings before carrying out the mods.

Opening up one side of the case and then reaming the oil holes is not too difficult.

The bolts we have now are marked with 'FLAT' laser etched onto the head so they are aligned in a specific position.

We are planning to install them with an m6 interference which should deal adequately with expansion.

I would agree that the hollow dowels used by Craig and installed in a similar manner are a much better idea than trying to line up two separately machined halves.

I have looked at Spirol Hollow Dowels and their design data states that fits and clearances should be identical to those used for solid dowels.

I am not sure of the force needed to press fit a single dowel with a typical m6 interference but is you multiply this by the number of dowels the load required is not insignificant.

There is good data available for the clearance for a dowel to be a 'sliding fit' and for a 10mm dowel this would mean 0.002" on diameter or 37.5 microns in a single direction. More than enough for fretting to still occur.

if after pinning the two cases do they just slide together? If so then maybe they will still fret.

If you have to press them together, how do they come apart?

The fitted bolt concept which is a design feature we used commonly to align the high speed titanium couplings we used to manufacture for Military Helicopter Engine Test Rigs - Typically for T-700, Makila and RTM 322 engines- and they do work well. These engine run at around 21000rpm and will not suffer any significant vibration on the shaft bearings so balance and smooth running is essential.

It is quite easy to maker a puller to install the bolts and as they are fitted one at a time it may give the case an easier time.

We are just doing a layout to convert the bolt to a 'Tie Bar' so it can more easily fitted and removed.

We have drawn it up with 2 x flats to give the oil circulation even more chance and then use self-aligning washers on both ends.

With regard to lubrication on the surfaces I have a telephone conversation booked later today and should find more information.

We need a compound that will 'adsorb' onto the surface of the magnesium so it won't be squeezed out.

We are also looking at repairing badly fretted cases without having to narrow the engine using a 'Cold-Spray' that is already being used for helicopter gearbox repair.