Low Leakdown Test on One Cylinder

[vigilant104]

I'm having the first Condition Inspection done since I've owned my plane, and during the differential leakdown test today, #3 (right rear) produced a reading of 59/80. All readings: 1: 69/80. 2: 69/80. 3: 59/80. 4: 75/80. The engine was cold, and the rocker arm assemblies were both off the engine (so I know that a valve wasn't being pushed open).
The Aerovee manual says anything below 60 or a large variation between individual cylinders requires additional investigation". The engine seemed to be running fine, has 215 hours on it, and I wasn't getting high oil temps like I'd expect if I had a lot of blow-by. For various reasons (a torrential downpour/hailstorm on the roof of my tin hangar) it wasn't possible to listen and tell where the air was leaking to.
Opinions?
I'm tempted to put everything back together and run the engine, then repeat the test when it is warm (and when it's quiet). Any experiences with cold vs hot leakdown test results in VW-based engines? Other ideas?

[eppre]

I am not a full a&p but have worked around our local FBO. We alway check the compression hot.

[fastj22]

On my jab3300, I had one read 63/80 cold. The rest mid 70s/80. Retested warm and that cylinder read 80/80. I've been told there are many variables like ring alignment and to use leak down values as a trend line.

[daleandee]

I'm having the first Condition Inspection done since I've owned my plane, and during the differential leakdown test today, #3 (right rear) produced a reading of 59/80. All readings: 1: 69/80. 2: 69/80. 3: 59/80. 4: 75/80. The engine was cold, and the rocker arm assemblies were both off the engine (so I know that a valve wasn't being pushed open).

Hi Mark,

It is said that a leak down test should be done on a warm engine but many of us don't like pulling steel plugs out of aluminum heads until the engine is stone cold. You could try "staking the valves" i.e. tapping the end of the valve with a plastic mallet to pop it open and closed quickly and perhaps dislodge a piece of carbon that may have gotten caught under the valve and is holding it open. In my experience this is almost never the case but rather is helping the valve to sit correctly when the guide is worn but the number will be better. :)

You could reassemble it and fly it for a a few more hours as sometimes the rings will align and cause the reading to be low on a particular cylinder. Again this is not the likely cause but is worth a try.

Here is a link to some of the best information I've seen lately on cylinder testing:

http://www.avweb.com/news/savvyaviator/188758-1.html?redirected=1

Having said all of that you also need to realize that the VW engine (especially the heads) work very hard in this application and sometimes need some attention after a few hundred hours. I've noticed lately that there have been nearly a dozen VW based engines on Sonex aircraft show up on Barnstormers with rebuilds or top end work reported with less than 300 hours. Of course there are various reasons for this but the fact that the heads are the weak link in the chain is well known. The great news is that lapping a leaking valve is pretty quick and nearly painless. Even in the extreme a set of replacement heads are, in the aviation world, considered to be dirt cheap.

Hope this helps,

Dale Williams
N319WF @ 6J2
Myunn - "daughter of Cleanex"
120 HP - 3.0 Corvair
Tail Wheel - Center Stick
63.0 hours / Status - Flying
http://kitplanes2.com/blog/2013/12/ownerbuilder-2/
http://www.youtube.com/channel/UC49h6Qijc17_Ebfz0CbRFtg

[daleandee]

*** Deleted Duplicate ***

[Sonerai13]

For various reasons (a torrential downpour/hailstorm on the roof of my tin hangar) it wasn't possible to listen and tell where the air was leaking to.
Opinions?

A differential compression test doesn't really tell you much until you determine where the leak is. I'd redo the test on that cylinder when you can listen for the leak. If it's rings (air coming out of the crankcase vent) I wouldn't worry about it much, as it is probably fine when the engine is hot. You can recheck it in a few hours after doing a lot of flying and it will probably show an increase. If the leak is through a valve, it's important to know which one. Most often it's an exhaust valve, but on a rare occasion an intake valve will leak.

Someone else mentioned "staking" the leaking valve. I've had this help exactly once in my entire career, but there's always that chance. In order to do it right you need two people and your differential compression tester. Hook up your tester, then have someone hold the prop at a point where the piston is about halfway up on the compression stroke. Then put about 20 lbs of air pressure into the cylinder with your compression tester. While the pressure is on and your partner is holding the prop, give the valve a couple of good raps with a soft mallet. You'll hear a nice "pop" as the valve opens and snaps shut. If there is a piece of foreign material on the valve seat, this should blow it out. Like I said, not real likely, but doesn't take a lot of effort so worth the try.

Also, if you or someone you know has a borescope with a 90 degree mirror attachment, and you or they know what to look for, you can borescope the cylinder and look at the valves without pulling the head. This will only show something if it's a very obvious problem though, and with your level of leakage I doubt you'll see anything too untoward.

Again, it's important to determine for sure what's leaking and how much before going forward with any repair strategy (if deemed necessary).

[vigilant104]

Thanks much for the helpful suggestions. Dale, Joe: I'm going to put the engine back together and run it for a bit, then repeat the test and give a good listen to learn where he leak may be. It will be a few days before I can get back on this project, but I'll check in with a report. If it turns out the ring gaps just happened to be lined up or there was a bit of lead on a valve seat and everything then checks good, I'll go buy a lottery ticket before the luck runs out.
Thanks again.

Mark

[vigilant104]

I'm back on the project.
I hooked up my new homebrew leakdown tester to the problem cylinder and got a reading close to the one from the A&P's real tester. The leak is definitely coming from the exhaust valve (can clearly hear the hissing at the exhaust pipe, nothing heard at the carb or oil breather. I "staked" that valve repeatedly and did see some small improvement (now the cylinder reads about 63/80, it was 55/80 with my homebrew tester when I started.) Dale, I suspect your explanation is correct: As I banged and retested, the readings got better in a somewhat erratic pattern. I don't know if I succeeded in dislodging anything or if the valve just worked itself into a more favorable spot.
I'm not optimistic that running the engine will change things much (i.e. the leak isn't coming from a chance alignment of the ring gaps, etc.)
So, again my leakdown numbers are now (Cyl 1 - 4): 69/80, 69/80, 63/80, 75/80. There were no problems with the way the engine was running, the leakdown test was just part of the annual condition inspection
I'm inclined to put everything back together and then watch things closely to catch a burned valve or excessive recession of the valve seat (e.g. do a leakdown check and monitor any changes in exhaust valve clearance every 10 hours to check for trends). Imprudent? Any votes for a new valve and seat right now? Other comments?

Thanks.

[daleandee]

I'm inclined to put everything back together and then watch things closely to catch a burned valve or excessive recession of the valve seat (e.g. do a leakdown check and monitor any changes in exhaust valve clearance every 10 hours to check for trends). Imprudent? Any votes for a new valve and seat right now? Other comments?

Hi Mark,

When I had a valve leaking I would set the head on a workbench and lap the valve using a hand tool and some "valve lapping compound." You will almost immediately be able to see where the valve isn't sealing by looking at it after a bit of lapping. To check for leaking while the head was off I'd use 100LL and add a bit of Marvel Mystery Oil to it for color. With the valve in place I would pour this mixture into the combustion part of the head and see how much it leaked out of the exhaust port. This will confirm when you have lapped the valve enough to get a complete seal.

Continuing to run a valve that is having hot exhaust gases blowing by it will cause it to warp or burn even more. It won't get any better. If you feel uncomfortable doing the work or if the leak is too large to take out with lapping, then you can always take the heads to a reputable repair station for a valve job.

Be careful on who you allow to do valve work on your heads as it ain't difficult but some shops are much better than others when it comes to air cooled VW heads. Perhaps there are Aerovee owners near you that can make a recommendation as to who is doing the good work in your neck of the woods.

But, at the end of the day 63/80 isn't a horrible number and I'd probably be inclined to do what you have in mind ... reassemble it a fly it and see what happens over the next few hours.

Hope this helps,

Dale
N319WF
64.0 Hours

[MichaelFarley56]

My vote Mark would be to put it back together and not worry about it! I think you'll be just fine! Get that engine back together so we can get some formation flying in here shortly!