3.6 cylinder base leaks

[Bill Verburg]

'95 993 engine built on a 964 case
lots of leaks before rebuild but none at the cylinder bases
rebuilt once w/ Victor Reinz green base gaskets, leaked
rebuilt again w/ Wrightwood racing black base gaskets leaks worse

any ideas?

[Henry Schmidt]

Originally Posted by Bill Verburg

'95 993 engine built on a 964 case
lots of leaks before rebuild but none at the cylinder bases
rebuilt once w/ Victor Reinz green base gaskets, leaked
rebuilt again w/ Wrightwood racing black base gaskets leaks worse

any ideas?

ThreeBond 1211.
What head studs are you running and what torque?

[Bill Verburg]

Originally Posted by Henry Schmidt

ThreeBond 1211.
What head studs are you running and what torque?

headstuds are ARP, torqued to spec

[PeteKz]

I ditched the copper case gaskets on my 3.2 build and used a thin bead of Permatex Ultra Black (max oil resistance) applied to the cylinder base "step." 6000 miles, no leaks yet. I try to make sure it squishes down between the cylinder and the bore of the case. I did the same on my last 3.0 build with Permatex Blue, and it didn't leak either.

[mikedsilva]

Originally Posted by PeteKz

I ditched the copper case gaskets on my 3.2 build and used a thin bead of Permatex Ultra Black (max oil resistance) applied to the cylinder base "step." 6000 miles, no leaks yet. I try to make sure it squishes down between the cylinder and the bore of the case. I did the same on my last 3.0 build with Permatex Blue, and it didn't leak either.

How did you achieve deck height and in turn, compression ratio without a gasket?

[PeteKz]

Mike:

The copper gaskets in the 3.0 engine are about .004"/0.1mm IIRC. I did several test assemblies of the parts dry to measure the clearance between the top of the piston dome and the head, and find the point of minimum clearance (it's on the outside of the dome near the exhaust valve depression, BTW). I had about .030-040"/0.75-1.0mm of clearance at the minimum points. That's a fairly wide range of error, but that was due to the difficulty of measuring how thick a pice of clay is after it's been squished, and how much further a piston will go before contact by measuring that with a dial indicator. Within my measurement error, the "deck" clearance was more than .030, but not more than .040. That's good enough for me. As noted in my other postings about the M1 cam, I ended up pretty tight on the valve-piston clearance, with a minimum clearance of .050" at TDC. That's less than the .060"/1.5mm I would have liked, but I decided to stay with it. For me, this is a street car, not a race car, and I don't like to spin my street engine to 7000+ RPM.

I used the term "deck" in quotation marks, because the pistons I used are the Mahle Sports for the 3.0 and 3.2 engines, which have an offset dome like the OE pistons, and they don't have a real deck, like an American 2-valve V8, or the 914 and VW engines. In American V8's, the generally accepted minimum deck clearance is .030" to allow for rod stretch at high RPM and for some rocking of the piston in the cylinder, without touching the heads (I've read that in multiple sources, but I'll refer to Smokey Yunick's book on engine building as authoritative). But, the tighter a builder can make the deck clearance, the better the squish and combustion chamber turbulence and mixing. If you are not spinning to 7000+ RPM, you can go tighter than that. I built a couple 914-2.0 engines for track and street use that I went as tight as .020" without slapping the heads, but being a pushrod engine with stock valve train components, I limited RPM to 6500.

Compression ratios: In the calculation of CR, changing the deck height by a few thou makes very little difference. Say you have a chamber volume of 60cc, and the diameter is 95mm, giving an area of 22.5 sq cm. Decreasing the "deck" by .01cm (not mm) decreases the chamber volume by 0.25cm, reducing the nominal 60cc to 59.75cc, or a decrease in volume and CR of about 0.4%. Extrapolating, a change of .25mm/.010" amounts to a 1% change in chamber volume and CR. Pretty small potatoes.

FYI, Running the calculation for the same amount of change in the bore shows why changing the bore by a given amount makes a bigger change in displacement than changing the stroke by the same amount.

[Old H2S]

I just did a 3.0 to 3.2 bump up with Nickies and CP pistons and ditched the cylinder gaskets too, and brought the CR up to 9.75.. I coated the base with Permatex Right Stuff black wihich was the only sealer I found to stop my Mustang race motor from leaking after I built it 17 times. Now I see Henry says that Threebond 1121? works so maybe something new to try.

[PeteKz]

I think any of the name brand silicone gasket maker/sealers will work, what’s more important is good clean surfaces for it to adhere to.

[Bill Verburg]

3.6 cylinder base seals are very different form 3.0 or 3.2 don't know why you are posting about them here


here's my 3.8 RS which actually has a groove for the o-ring, the base 3.6s don't have that groove but still use an o ring


another shot of the 3.8 case grooves



these are the available 3.6 base seals Victor Reinz are too small and guaranteed to leek, the oe Porsche apparently only sold by Wrightwood racing are the ones that are supposed to work

[dannobee]

Henry's right, Three Bond on the base (where the baseline 964 o-ring goes) forgives a lot of sins. But be expeditious when you assemble everything because it'll set up before you know it. Have everything ready to apply some pressure to the cylinders (studs in, hold down nut tools ready, etc). Make sure that it oozes out without any gaps, then clean up with paper towels and q-tips.

I'm convinced after over three decades of doing this stuff at dealership level that Three Bond (in its various iterations) is THE reason that modern cars don't leak oil.

[stownsen914]

I wonder if Curil T on the oring would help it seal

[Bill Verburg]

Originally Posted by stownsen914

I wonder if Curil T on the oring would help it seal

Seems like it ought to, a number of others have offered that advice

Unlike the 3.8s the 3.6 o-ring groove is at the base of the cylinder

the green Reinz gaskets are just not thick enough

the oe Porsche and Wrightwood are, it has been suggested that possibly the case through bolts are leaking but after very careful inspection they are not

[Mixed76]

O-rings and their grooves are engineered systems, if any of the measurements are off they don't seal. Or if the o-ring material changes (by the manufacturer) and they stretch or shrink or swell when exposed to heat, oil, etc.

Hard to say what is going on here, most likely the o-ring is different but if the cylinder base diameter was changed, the surface roughness increased, or the o-ring groove cleaned out aggressively there could now be a dimensional problem.

If you can verify OD of the cylinder, ID of the spigot, and ID of the o-ring groove and compare that to the o-ring installed thickness it might tell you something. The o-ring should be squished in the groove by 10-20%. Not under much pressure and not moving so groove axial length shouldn't matter much.

ID of the o-ring groove has to be clean and smooth, as does the surface closest to the spigot flange.

Dunno if that helps.

Sent from my Pixel 7 using Tapatalk

[john walker's workshop]

Black Wrightwood o-rings, and Permatex Aviation brown camel snot and being that you're right there, just do the through bolt o-rings with Wrightwood green viton. Too much work to deal with them later.

[Neil Harvey]

You should not need any glue.

Are the decks flat and parallel? Are the Cylinders the same height? The O ring will seal with very little crush. Typically you will see approx 0.018" - 0.020" crush. Check the groove depth against the O ring cross section. If the grooves are correct depth, either ring will seal.

Its easy to get caught up in the issue. Remember this seal system works. Focus on what each component does to seal and check their integrity. The problem I will bet, is not the O ring.

[scootergt3]

i'm thinking the heights are off and when the cam housings are torqued they are lifting.

[MoreGAS]

There are atleast 2 different O rings for the 3.8 (109 mm) cylinder sealing arrangement pictured, and likely you ad a stock 3.6 case machined correct? . We make sure to have the O rings in-hand before machining the case as it is The way to get the proper engineered fit. I don't recall the clearance we settled on but it depend son several factors. This link below can provide some insight into a lot of what is considered, material being one important consideration also.

https://www.specialistsealingproducts.co.uk/optimum-squeezing-force-for-an-o-ring/

I can also share that the Red O rings are what we always use and when installed properly one damn near has to use a plastic hammer to get the cylinder into the case, and for insurance we use a Bomber red silicone for the bases, I recall it is called Curil T( there are several PM me for a pic if needed as it is hard to find ) and it remains pliable at very high temps (beyond 500 F from memory) , and it is a bear to get the cylinder back out after use. Call this a backstop if the O ring leaks(maybe from poor machining of O ring groove) . But better than having to do it again for free.. as we did when Client wanted to take a short cut and not machine the O- rings into his case when we refreshed his 3.8. Remember those who do it cheap do it often".

As someone suggested make sure the deck is flat from the start, then the that cylinders are the same Hts, if the cylinder have alot of time on them you will find they have collapsed on the exhaust sides, maybe .002-.003" on some, clearly this is not going to help the compression sealing either.

Last unrelated thing, stock copper gaskets for 95, 98 90 MM bases of cyls are .010" and .25 MM, think folks above got their units mixed up. If it all possible I would run them for sealing, even though some Brands don't use them, I never feel lucky and don't want to do 25-30 hrs of labor for free when they leak on a new engine.

Kevin
GAS Motorsport

[Bill Verburg]

Thanks to all that responded, it helps a lot

[Jeff Alton]

Bill,

We have an original true 993 3.8RS motor being torn down right now. Once we get the P/C set off I can take any measurements you may need and supply you with pics of what we found.

Cheers

[Jeff Alton]

Bill,

Are you using the 3.6 base seals in the 3.8 block location? I got that sense based off the picture you posted of the 2 o-rings side by side. If so, that may explain. The seal Porsche produces for the 3.8 case is not round in profile, and they are available from Porsche. 999 705 023 41 is the part. Apologies if I miss understood the pic.


Cheers