3.6 Case seam oil leak after rebuild?

[detroit]

So, a bit of input from a guy who designs engines at GM for a living...
As Dannobee states, modern engines as a general rule just don't leak. While it is obvious that flanges have to be really clean when the sealant is applied, the real key is in the design of the joint. A 911 case has 20mm of sealing land. The problem is that with all those heat cycles, the joint moves. A micron or two or less of sealant doesn't offer enough compliance to account for that movement, so eventually the joint leaks.
At GM we do the job with less than 9mm. The difference is that one of the two flanges has a 1mm step in it. The RTV fills in that gap, and creates a 1mm thick flexible seal. Bingo. No leaks. I'm frankly surprised that none of you professional builders, or shops like Ollies, haven't adopted the design. I think there's more than enough stock on a 911 case flange to accept it. A simple cutter path, and you'd have a much more robust seal.
I would unequivocally do it to my own case that I'm building, but hiring a shop to program the cutter path, fixture the part, etc. for a one-off is a little cost prohibitive for my wallet...

[Henry Schmidt]

The better mouse trap.....
Properly clean, inspect and gently massage the sealing surfaces [Scotch Brite by hand] and a non-silicone [snakes in the drain] sealant that is flexible and you're good to go. We have engines that have clocked in excess of 250K miles without case half leakage.
Our sealant of choice is ThreeBond 1184 for case halves. Is that a guarantee that this method will work? we give a 2 year unlimited mile warranty against leaking, so I would say we have confidence in this method. Even on twisted old mag case engines.

[PeteKz]

Henry, as always, thanks for sharing your hard-earned professional advice.

Detroit: Thanks for that insight too. I have noticed that the mating parts in a number of applications don't line up exactly. I always thought that was just manufacturing differences, but now I understand why and that it's deliberate.

I've done something like that by using a hand file to chamfer the inner edges of parts that have to contain fluids, to create a little V between the mating surfaces to hold squeezed-out sealant. Not even a millimeter. It seems to work. I did not do this on my last engine build, but I will next time! Since the bottom of the case springs the most leaks, maybe the flanges from the front oil seal to the rear oil seal are all that's needed.

[Old ITW]

I've put together cases for 40+ years and have gone thru a number of Permatex and Loctite products. Here is a new one not mentioned in any posts above which I might try on the engine I'm building now ... Gasgacinch 440C. The product has been around for decades, but I've never heard of it being used to seal cases. Has anyone else tried it?

[motomiata]

I’m planning to fix my leaking case seam using the vacuum infusion process described in post #2.
Is it viable to do this with the engine in situ? I would expect the vacuum to pull oil from the oil tank through the pump an in to the case. Or will it easily seal the 10 psi vacuum? Has anyone done this ?
Thanks

[Sboxin]

Originally Posted by motomiata

I’m planning to fix my leaking case seam using the vacuum infusion process described in post #2.
Is it viable to do this with the engine in situ? I would expect the vacuum to pull oil from the oil tank through the pump an in to the case. Or will it easily seal the 10 psi vacuum? Has anyone done this ?
Thanks

motomiata,

Our shop plans to do this procedure with the engine in the car.
Shop will add die to the oil and look for the leak with a black light.
All oil is removed from case and tank - seam needs to be dry and clean.
Our shop plans to do with 10 psi vacuum.
As reported in #2 this has been done several times.
Our car goes into the shop on Monday 7-10-23 . . . I'll report back on the results.


Regards,
Roy T

[motomiata]

Originally Posted by Sboxin

motomiata,

Our shop plans to do this procedure with the engine in the car.
Shop will add die to the oil and look for the leak with a black light.
All oil is removed from case and tank - seam needs to be dry and clean.
Our shop plans to do with 10 psi vacuum.
As reported in #2 this has been done several times.
Our car goes into the shop on Monday 7-10-23 . . . I'll report back on the results.


Regards,
Roy T

Thanks Roy,
Your feedback on the process once carried out will be highly appreciated.

Regards
Hans

[al lkosmal]

Originally Posted by michaelmount123

I've solved this before, so method is sound:

First, clean the area and dust some baby power in the area. The leaking spot will become very visible. Mark the spot.
Next beg or borrow a vacuum pump and connect it to the breather port in the engine. 10psi should be more than enough.
Spray some BraKleen on the area to clean remaining oil from the joint.
Apply some LocTite 574 (or similar) to the leaking area and watch it get sucked in. Remove the vacuum AS the sealant is being sucked in.
Wait a while, perhaps take a nap... Leak is gone without disassembling the engine!
You're welcome!

This may be a dumb question, but.....After marking the leak spot ,are you then draining the engine of oil, prior to applying vacuum and sealant?

[Sboxin]

Originally Posted by al lkosmal

This may be a dumb question, but.....After marking the leak spot ,are you then draining the engine of oil, prior to applying vacuum and sealant?

Please refer to post #27 . . . yes all oil is drained . . .

Regards,
Roy T

[Sboxin]

The results are in =

NOT a case seam leak!
NOT a case seam leak! OOOPS . . . yes, a very very small case leak just below the major leak in the intermediate shaft bearing O ring leak see my post on 7-21-2023. . . The shop will now do the vacuum procedure . . .

The Intermediate shaft bearing O Ring was leaking - why (?) because there was a lot of machine work done on the case including a line bore and shaving of the case - causing the IS Oring hole to be oval rather than a perfect circle. The Tech found it right away and removed the cover and redid the O Ring with more silicon - leak stopped - ran the engine up to temp (20 minutes) and let set to see if the leak has stopped - the shop will watch it for a few days to make sure it is fixed. The case half seal does not leak it was just the place the oil ran to by gravity . . . ha ha . . .The shop Tech Roland explained all of this very thoroughly to me - Thanks to Don Jackson's in Phoenix!

Photos from another post in 2022 (thanks to Uwon) that show the cover and O Ring - a different engine but the same 2 bolt cover (should be 3 like the older engines (?))





From Bruce on Uwon's post : "Cover for the intermediate shaft
There is a green rubber seal that goes on it to replace original red
Lube it and install without pinching
Bruce"
https://forums.pelicanparts.com/911-engine-rebuilding-forum/1114558-seniors-moment-leaking-cover.html

Thanks to michaelmount123, Bill Verburg, Neil Harvey, John Walker, Henry Schmidt, dannobee, detroit and others for contributing to my request!!

Regards,
Roy T

PS: The shop did this repair with the engine in the car - just had to remove some tin and the saddle carrier (?) to get to the cover plate.

[Turbo_pro]

Originally Posted by Sboxin

The Intermediate shaft bearing O Ring was leaking - why (?) because there was a lot of machine work done on the case including a line bore and shaving of the case - causing the IS Oring hole to be oval rather than a perfect circle. The Tech found it right away and removed the cover and redid the O Ring with more silicon - leak stopped - ran the engine up to temp (20 minutes) and let set to see if the leak has stopped -....................................

Not to high jack this thread but your post suggests that it was machined enough to affect the ovality of the sealing surface. After machining the case to return it to standard aline bore, how do you recover the distributor gear backlash?
I've seen this approach touted over the years and never really understood the recovery process.

[Turbo_pro]

Originally Posted by Sboxin

Sorry, I have no answer to your question - I am not an engine builder ( the last one I built was a356 in 1965 - very simple like a VW . . . ha ha) . . . But, if I can remember your question I will ask our engine builder ( I am very old and don't remember a lot of things )

Regards,
Roy T

The question was actually directed at the other engine builders on this thread.

[Sboxin]

NOT a case seam leak! OOOPS . . . yes, a very very small case leak just below the major leak in the intermediate shaft bearing O ring leak see my post on 7-21-2023. . . The shop will now do the vacuum procedure . . .


Thanks again for all of your help!
I'll update when we have some more results . . .

Regards,
RoyT

[Hokie6]

I found a similar leak in mine at the same location oddly enough. I plan to try the vacuum procedure and hope it will seal. Will be interested in your results on this -

[Sboxin]

Originally Posted by Hokie6

I found a similar leak in mine at the same location oddly enough. I plan to try the vacuum procedure and hope it will seal. Will be interested in your results on this -

The engine shop had some difficulty with the vacuum procedure and we still have a pin hole size leak that we will live with - the problem seems to be in the timing of the sealant - I didn't understand all of their explanation (because I'm not a "detail" person).

Of course the big leak was in the intermediate shaft O ring - which they fixed.

Wish you the best on your project (leak)

Regards,
Roy T

[Hokie6]

Sorry to hear you had problems with this. I guess I got lucky, in that my leak turned out to be from a torn cam gasket. Once I go that fixed, no more leak, at least not there. I have a small one now at the rear, but might be coming off the top of the motor. Almost too small to worry about at the moment. So glad I didn't have to do the vacuum procedure though.

[PeteKz]

I was suffering from insomnia one night, so I counted all the places oil can leak out of an air cooled 911 engine. I got to almost 100 before I fell asleep. I’m sure I missed a few.

[porschedude996]

10 psi vacuum? Maybe 10 inches of vacuum?