Chasing leak down after valve job

[Solamar]

Short back story - Bought a used 3.2 engine last fall with bad leak down numbers (Cylinders 1 & 4 were leaking at the head to cylinder mate, 45% and 22%). I had the heads rebuilt and re-ringed the Mahle Nikasil P&C's.



Engine ran great. After 1000 miles I changed oil and filter and adjusted the valves. At 4000 miles I decided to do a leak down test, mostly just so I had a baseline going forward. Engine still started, idled and ran great.

Cylinder 1 - 13%
Cylinder 2 - 2%
Cylinder 3 - 3%
Cylinder 4 - 2%
Cylinder 5 - 23%
Cylinder 6 - 3%

1 & 5 concerned me, both leaks could be heard well through the exhaust.

Pulled and stripped engine, all the exhaust ports looked new with the exception of #5, it was black.



Took the heads back to the machinist who agreed to disassemble all the heads and re-work as necessary.

While he had the heads I devised a plan to check the leak down of the cylinders without the heads. I was very curious how much of the leak down was air passing the rings (slight air sound from breather while performing the leakdown tests). I made a block-off plate with an air chuck nipple to mate the LD tester to, only problem was the piston domes prevented me from testing at TDC. No prob, I just rotated past TDC and locked the flywheel for the tests.



Without heads, just ring test -
Cylinder 1 - 2%
Cylinder 2 - 2%
Cylinder 3 - 2%
Cylinder 4 - 2%
Cylinder 5 - 2%
Cylinder 6 - 2%

I have nothing to compare those numbers to, but I'm certainly happy with the consistency!

Got the heads back, installed, torqued to 30 ftlbs and retested the LD

Cylinder 1 - 12%
Cylinder 2 - 2%
Cylinder 3 - 8%
Cylinder 4 - 7%
Cylinder 5 - 9%
Cylinder 6 - 11%

Using a hose stuck in my ear I can easily hear air rushing past several of the valves.

So I pull head #1 and fill the chamber with alcohol. After a minute I can see alcohol seeping past the intake valve.





(continued on next post)

[Solamar]

I pull the intake valve





Mark with Sharpie and gently lap to check the contact pattern





To my untrained eye they look pretty good..??

Clean, oil and re-install valve and re-check with alcohol. Still seeping past the INTAKE valve slightly. EXHAUST side is dry as a bone.

Here's where it gets weird... I reinstalled the head, torqued and checked leak down. Now LD is 7% (down from 12%), I cant hear ANY air escaping past the INTAKE valve but can clearly hear air going past the EXHAUST valve. WTF??

Does torquing down the head load the seats in a way that the seating can change?

Am I just chasing my tail here and should just put it back together and enjoy??

[Trackrash]

I always lap my valves to make sure they are perfectly seated. Looks like those were? The seat does look a little rough even so.

After it has run a while the valve will pound into the seat and probably seal completely.

[dannobee]

It's fine. By looking at the seat and valve, the valves were lapped in place. Like Gordon said, it'll seal up better once it's been run.

Given the initial leakdown rates and what I know, I wouldn't have torn it down in the first place. A quick tap on the suspect valve with a plastic mallet usually changes the leakdown reading for the better.

With respect to dyno work, I think a lot of us have wasted plenty of time trying to get leakdown rates near zero, and saw absolutely zero change on the dyno. It actually takes quite a bit of leakage (as measured with a "standard" leakdown tester) to change the dyno number even by a couple of hp. FWIW

[yelcab1]

In search of perfection for a 40 years old motor... never gonna happen. Did any work get done with the case surfaces? Did the machinist cut all the heads to the same height? Did the cylinders get decked as well? Too many variables on an engine with too many sealing surfaces for the old engine to turn out perfect. On a newly rebuilt valve job, I certainly would not worry if you have a bit of leakage, not enough to redo the engine anyway.

You need to drive it.

Great engineering skills for the leakage tester on the cylinders. Nice engine stand too.

[Solamar]

Many thanks for the feedback guys! Very helpful to hear from people that have 'been there done that'.

It does look like all the valves were lapped. I touched up a couple of the obvious leakers, improved things a bit.

I'm waiting on a couple gaskets, then Ill be back in action...

[Solamar]

Originally Posted by dannobee

With respect to dyno work, I think a lot of us have wasted plenty of time trying to get leakdown rates near zero, and saw absolutely zero change on the dyno. It actually takes quite a bit of leakage (as measured with a "standard" leakdown tester) to change the dyno number even by a couple of hp. FWIW

Appreciate that, glad I'm not the only one to fall down this rabbit hole

In regards to "standard leakdown tester", I'm playing with two different testers and the results are wildly different. They both identify leaky cylinders, but the % numbers are 2-3X higher on my old Snap On leak down tester. Also, the compressor pressure (which changes) and the point that you "Zero" the gauge are rather fluid causing several % points of change if your not careful.

If you look at my newer LD tester, 0-10% is the range to set your zero, 10%-40% is considered LOW leak down and 40%-70% is Moderate. Chasing 7-12% leak down was feeling rather foolish after a bit...

Originally Posted by dannobee

Given the initial leakdown rates and what I know, I wouldn't have torn it down in the first place. A quick tap on the suspect valve with a plastic mallet usually changes the leakdown reading for the better.

The black exhaust port (and 23% LD) on #5 was enough for me to justify the tear down. If I hadn't intervened I bet it would have burnt the valve. Fortunately these engines are quite simple to pull, just an enjoyable weekend playing with the Porsche...

[Solamar]

Quick follow up -

I put another 1000 miles on the engine, been running flawlessly. Yesterday I dropped the oil, adjusted the valves and performed another leak down test (cold).

All the plugs looked the same -





Leak down results -

Cylinder 1 - 2%
Cylinder 2 - 1%
Cylinder 3 - 2%
Cylinder 4 - 2%
Cylinder 5 - 2%
Cylinder 6 - 1%



Very happy with how the engine has turned out.

A trick I found that makes doing the LD test easier - cut a piece of 1" x .06 wall alum tubing at 5-3/8". Drop the tube down the spark plug hole as a guide for the floppy LD tube. Makes threading in the hose a piece of cake.

[Flat Six]

Originally Posted by Solamar

A trick I found that makes doing the LD test easier - cut a piece of 1" x .06 wall alum tubing at 5-3/8". Drop the tube down the spark plug hole as a guide for the floppy LD tube. Makes threading in the hose a piece of cake.

Nice tip, thanks!

[icarp]

The way I was taught to do a leak down test is to only use 80 lbs of air pressure . Using more psi is not fair to the rings , the are used to having a blast of pressure not a steady pressure . If your surface on the nicky surface was done to the proper spec , RA of 6 +1 -1 the rings should seat within the 1%-2% range . It would seem to me the head work man just did a routine job on your heads . Not the best he could have done . Your final numbers look very good to me

[gomezoneill]

Nice job on the photos.