rocker arm ball misalignment

[jond]

Hi all,
I looked around for a similar post but found nothing.
I have a turbo Aerovee and have adjusted the valves to the specs recommended by Sonex: 014 for exhaust and .008 in intake. I have noticed that after the last few flights that after adjusting the valves that the flat face of the swivel rotates around so that the flat face in now on the backside. This causes the gap to decrease and leaves the exhaust ports continuously open. This has happened twice to the same exhaust ports today. Observers were present to verify that I was not misconfigured, both times.
Has anyone had this experience before? And if so, what was the cause and solution? The last time I adjusted the valves, I reduced the gap on the exhaust side from .014 to .013 and I will fly it tomorrow to see if it repeats the problem. I will update tomorrow.

[lgsievila]

jond,

I have noticed that after the last few flights that after adjusting the valves that the flat face of the swivel rotates around so that the flat face in now on the backside.

It happened to me on several rocker arms and is just one of the many reasons I gave up on the AeroVee.

[rizzz]

... 014 for exhaust ...

Really?
I just have a normal aspirated VW but I can see why you would experience the problem you're describing...

I'm no expert by any means but I have been wondering around the many VW forums on the net for a few years now (both aviation and automotive) and I've seen many of the 'experts' debate whether the gap should be 0.004 or 0.006 but I have never come across anything like this.
If you want to increase the duration the valve is closed, would the proper solution not involve a differently ground camshaft?

[wlarson861]

the gaps are different to compensate for the extra heat produced in the exhaust with the turbo. The exhaust valve expands more from the heat than in an aspirated engine so the gap is set larger. As for the flat spot rotating it is common when the valve train is removed or the gap increased away from the valve ,enough to allow it rotate. When the valve lash is set correctly the ball can not rotate away from the valve stem as the ball part takes more clearance than the flat. whenever I remove the rockers such as for a head torque, i make sure to rotate the flat portion towards the valve stems before tightening down the rocker assembly. If one rotates so all you see is the ball then rotating it back in place can be done with a gloved hand or dry paper towel.
I would not decrease the gar when setting the lash for fear of the closer tolerance not allowing the valve to fully close after the engine is good and hot.

[jond]

Thanks for the reply Bill. I hear you regarding the ball swiveling away when the valve train is removed and reinstalled. In my case, I was only performing a valve adjustment after noticing that a couple of of the cylinders appeared to have low compression when spinning the cold engine by the prop. So no, I did not disturb the valve train, other than to adjust to correct tolerances. I always check when using the gauge to make sure that the ball has flat side correct. And then I flew the plane to bring it up to temps and test the tune. I noticed a couple of hot exhausts, so after lunch checked the valves again. The swivel balls had rotated backwards again so I regapped to .013 on the offending exhaust ports. I did a few more circuits and suspect that they may have repeated the rotation again since I saw the two offending cylinders (1&4) looking hotter than the others on the egt. I haven't checked the gaps since I ran out of time, but will likely be at the airport tomorrow to continue the tuning process. Fyi, nothing seems unusual with the whole rocker assembly. On another note, I changed the swivelball adjustment arms last fall due to excessive play (1/32") of the swivel ball in its housing. When I inspected the new adjustment screws prior to installation I noticed nothing unusual.
I know this is not rocket science, but a mystery so far!

[lgsievila]

To clarify my AeroVee was normally aspirated and valves were set according to manual. I just had bad components in that two of the balls in the elephant's feet froze up and would not move causing the engine to run rough. When I took valve covers off I found a round spot on top of the valve stem on one cylinder. I replaced that bad one and when it happened again I replaced the other 7 adjusters.

[gammaxy]

Your post got me curious, so I estimated how much clearance you need before the ball is free to rotate all the way around at ~0.030". This is my best estimate from zoomed-in pictures so might be a little off. I wouldn't expect .014" to allow the ball to spin around, but I really have no idea what happens to this clearance on a running engine. Valve float might cause it to increase; valve expansion to decrease; pushrod expansion to decrease, etc.

.014" for the exhaust valve with the turbo is a mystery to me. Has Sonex officially explained the reason for this? The obvious explanation is temperature, but it's difficult for me to accept that the engine is so much hotter that it needs 75% more clearance than normally aspirated. They did run the engine in a test cell, so if anyone has measurements showing the necessity of that setting, it should be them.

I'd double-check the obvious things like the torque on the cylinder head nuts and the alignment of the rocker arms relative to the valve stems (is it possible they can move sideways off the face of the valve stem a little?) I don't know anything about you or your observers, but if you haven't set valve clearance before on a similar engine, you should probably watch this video Sonex made: http://www.eaavideo.org/video.aspx?v=23173638001

[rizzz]

.014" for the exhaust valve with the turbo is a mystery to me. Has Sonex officially explained the reason for this? The obvious explanation is temperature, but it's difficult for me to accept that the engine is so much hotter that it needs 75% more clearance than normally aspirated.

My thoughts exactly but obviously I'm not an expert. One thing I do know that thermal expansion of metals is linear (just asked one of the metallurgists here at work).
What EGT's are expected on a normal AeroVee vs. a Turbo?

That said, I don't think the 75% (actually it's 130%) number has much significance either though. Let me explain.
Yes the that is how much bigger the gap on a turbo engine is relative to the gap on an aspirated engine,
but,
it is the size of the valve/length of the valve stem that this needs to be relative to because it would be the thing expanding.

The idea of having any gap at all is because the valve expands with heat and the goal is to have zero gap on a hot engine, I get this,
So apart from just asking Sonex, one way to confirm if the 0.014 gap on the Turbo is still intended to close completely is simply measure the gap on a hot engine.

[Onex107]

What you have described can only happen when the clearance under the rocker ball is bigger than what you set. Have you considered that the problem valve is sticking at some point and giving the ball enough room to turn? It can happen with a sticking valve that is not seating all the way or by a weak valve spring that allows the valve to float. I had a stuck valve in a Cont. 0-200 that was caused by the use of 100 low lead. They called it morning sickness, the valve would be stuck when the engine is started and it would run rough for a minute then break loose and run normally. It was only stuck when the engine was cold. They reamed the valve guides to correct the problem. In that case they didn't remove the cylinders. They pushed the valves into the cylinder, reamed the guide, and then fished the valve back into the head. A piece of rope in the cylinder held the valve up while the spring was re-installed. It's a lot easier to remove the Aerovee head than a Continental cylinder.

[vwglenn]

20 years I've been adjusting valves on VWs. Never have I heard of such a large gap but I'm not a racer and never ran a turbo. If that's what is required (don't understand why) the "swivel foot" adjusters are probably going to rotate like that because they're going to lose contact with the valves. Stock adjusters may be a consideration if the gap needs to be that big.

My valves on my street VWs have always been adjusted to .006 in the intake and exhaust. It's always been my understanding that the gap was there to account for the pushrods expanding. Unless Sonex uses some sort of exotic pushrods, I don't see the point of an adjustment that large. I built a dunebuggy which had chromoly pushrods. Those were adjusted to .001 which was a tighter gap because those rods didn't expand as much as stock rods. Conventional wisdom and my method of adjusting those particular valves was to simply ensure the pushrods would spin and not worry too much about the paper thin gap.

Are you sure you're not missing a zero in there? .0013? If the turbo has chromoly pushrods, that would make sense to me. Perhaps a simple misprint (left out a zero) on the assembly manual?