Another 964 Through Bolt Question - O Rings

[mikedsilva]

It's been a while since I did a 964 motor.

I have the case apart and looking at the o ring situation and through bolts etc. When I removed the through bolts, there was only 1 O-Ring fitted to each side.
Also, the engine number I have does not fall into the category of the TSB that says they need modified washers and double o-rings.

So I measured the depth of the recess in the case with the end of my vernier.. and it came to approx 6.25mm.
When I compare the vernier to the through bolt with washer and o ring, it seems that the o ring will NOT be in compression at all.

If that is the case, then how does the through bolt seal?

Has anyone else measured the depth of the recess in their 3.6 case?

[burgermeister]

I cannot speak for 964 motors, but I've rebuilt a 3.2 as well as a 993.

The 3.2 seemed to smash the o-rings into oblivion - the chamfered area (not a counterbore) they fit in seemed to have insufficient volume to contain the o-rings. Upon disassembly, they are all mangled and useless.

The 3.6 seems to have some axial space for the o-ring, relying on seal between the ID (case bolt) and OD (case bolt counterbore). Upon disassembly, all the O-rings were intact.

The 993 version is to me the proper way to use an O-ring. The 993 motor had fewer leaks at the through bolts than the 3.2 did, and at least judging by wear of the rod bearings, the 993 motor had more mileage.

[mikedsilva]

Updated first post to include pics.

[burgermeister]

I never took pictures of mine, but that looks just like I remember my 993.

The O-ring seals because the ID of the case and the OD of the through bolt have less clearance between them than the thickness of the O-ring, so it gets flattened a bit when the bolt is pushed into the case. There is supposed to be a small amount of axial clearance for the O-ring.

If the required seal was between the washer flat and the bottom of the case counterbore, you would need axial 'interference' to squish the O-ring in that direction. But that would not stop oil from leaking out between the through bolt and the washer!

[detroit]

I'm no 964 expert, but I suspect Bugermeister is correct. The seal as designed is between the fastener and bore of the hole. In this case, it's the right way to do it. A face seal between the washer and counterbore won't work as oil will just leak through the joint between the bolt and washer. The problem as I see it is that with tolerance stacks, those bolts are probably almost never centered in their bores. Without enough compression on the o-rings... they're going to leak.

[mikedsilva]

Originally Posted by detroit

Without enough compression on the o-rings... they're going to leak.

Well, that's what I was thinking...

[Henry Schmidt]

Originally Posted by mikedsilva

Well, that's what I was thinking...

Mike, it looks like you have all the answers except one: why? Why is often a question not worth pursuing.
The o-ring must be compressed to create the designed sealing effect.
If the measured depth is X and the combined measurement of the washer protrusion and the o-ring is less than X, the stack is incorrect.
Just fix it.
Use either two o-rings if that will create the desired crush, find a different case washer with a deeper protrusion or find a bigger o-ring.
You might even be good with a small copper washer between the case and the o-ring.

[burgermeister]

Originally Posted by detroit

The problem as I see it is that with tolerance stacks, those bolts are probably almost never centered in their bores. Without enough compression on the o-rings... they're going to leak.

Wouldn't the washer, which sits in the bore with the O-ring, keep the case bolt centered fairly accurately?

The engine presumably was running for some time. How many of the case bolts appeared to be leaking, and how badly, when the engine was disassembled?

If you decide to add more o-rings or small washers to the stackup, do keep in mind that the O-ring will expand in the axial (along the case bolt) direction as it is compressed radially, and there needs to be sufficient room for this to occur or the O-ring will get mangled just like the older 911 O-rings do. I expect a mangled O-ring will seal less effectively than an intact one, but I could of course be wrong (and I often am)

Might be interesting to make a small test piece with the case bore's ID, install an O-ring and case bolt in it, and then see what the O-ring's axial dimension is when it is in the 'installed' state, rather than the 'free' state.

[mikedsilva]

Has anyone else ever measured the depth of the recess in a 3.6 case?
and compared it to the thickness of the washer and the o ring?
Surely I am not the first.

I would have thought as Henry states above, that the o ring must be compressed to create the seal, but I'm wondering if it has something to do with that instruction in the 964 assembly manual about not sliding the o ring all the way up the bolt?

[mikedsilva]

In the photo of the bolt with the washer and o ring attached, it's clear to see that the diameter of the o ring is greater than the washer protrusion... when it is inserted in the case there is interference between the o ring and the case.. does this compression of the o ring around its circumference, then make the rubber expand sideways creating a slight (ever so slight) compression also against the case?

[targa72e]

I believe the more important dimension for the o-ring to work correctly is the clearance of thru bolt to case bore. The o-ring should be slightly compressed in this direction. The oil pressure should push the o-ring against the washer to seal.

How o-rings work from people who make o-rings.

https://promo.parker.com/promotionsite/oring-ehandbook/us/ehome/ci.How-an-O%E2%80%93Ring-Works,EN.EN

john

[mikedsilva]

Originally Posted by targa72e

I believe the more important dimension for the o-ring to work correctly is the clearance of thru bolt to case bore. The o-ring should be slightly compressed in this direction. The oil pressure should push the o-ring against the washer to seal.

How o-rings work from people who make o-rings.

https://promo.parker.com/promotionsite/oring-ehandbook/us/ehome/ci.How-an-O%E2%80%93Ring-Works,EN.EN

john

Hi John
thankyou for posting this. It suggests to me that perhaps my o rings and washers are hopefully correct.

[Henry Schmidt]

Originally Posted by mikedsilva

Has anyone else ever measured the depth of the recess in a 3.6 case?
and compared it to the thickness of the washer and the o ring?
Surely I am not the first.

I would have thought as Henry states above, that the o ring must be compressed to create the seal, but I'm wondering if it has something to do with that instruction in the 964 assembly manual about not sliding the o ring all the way up the bolt?

Now this has piqued my interest.
It seems we have conflicting information.
I assume that even though the o-ring is dangled on a the minor diameter of the bolt, when inserted, the o-ring ends up on the expanded portion.
Why not slide it up by hand to "know" it has seated properly and that , when "stretched", the o-ring hasn't broken?

[mikedsilva]

Originally Posted by Henry Schmidt

Now this has piqued my interest.
It seems we have conflicting information.
I assume that even though the o-ring is dangled on a the minor diameter of the bolt, when inserted, the o-ring ends up on the expanded portion.
Why not slide it up by hand to "know" it has seated properly and that , when "stretched", the o-ring hasn't broken?

Exactly. I assembled a 964 motor (my first) 4 years ago and it never leaked and to this day is leak free, and I slid the o rings all the way up to the washer. I do not understand the reasoning for not sliding the o ring all the way.

I guess, this is one of those time where perhaps I should not question why but just do as I am told.

[mikedsilva]

Originally Posted by targa72e

How o-rings work from people who make o-rings.

https://promo.parker.com/promotionsite/oring-ehandbook/us/ehome/ci.How-an-O%E2%80%93Ring-Works,EN.EN

screen shots from the above website. Hoping this explains (to me) how the o rings seal in the 964 case.

[burgermeister]

Originally Posted by mikedsilva

I guess, this is one of those time where perhaps I should not question why but just do as I am told.

I disagree strongly. If your brain thinks, you ought to use it, as often as possible. If directions don't make any sense, you should probably do what makes sense instead. Everyone makes mistakes, even Porsche tech manual writers.

Regardless of the reasoning, it seems the nut side O-ring has to end up on the 10mm collar of the bolt - I fail to see what would be different on the bolt head side.