Bob, feel free to add your thoughts from today.
Cheers,
greg
As Greg described, we brought the cylinder over to a local engine rebuild shop, and the shop's "cylinder guy" worked with us.
First we tested the valve movement by setting the cylinder on the floor, and pushing with a wood stick on the top of the stem/spring. As Greg said, both valves moved, but the exhaust valve had a little stick to it as it first moved.
Then we removed the springs...an interesting task with no rocker arm shaft. The pic above was the second interation of adapt and overcome (the cylinder guy is a Marine). The intake valve dropped through its guide, and the exhaust valve did not. I thought it took more than a little tap...actually took a couple...but he did say it was not nearly as hard to pop out as many stuck valves he's had to pound out. So it wasn't stuck hard at the 45 or so degrees in the shop. Would it be much harder stuck at 400, 600 or 1000 degrees? Hard to say, but conventional wisdom says it would be much tighter at operational temperature.
After measuring the stem (.001 under, with a .002 under limit) I felt the stem, and it had a tiny burr in one spot, but we discussed this, and there is no way to tell if it was there before the failure, or was caused by the failure. Our AI bud thought it was too small to be significant, but I mention it, just FYI.
Then the cylinder mech used the "no-go tool" on the guide. It looks like a valve stem, and one end should fit in the guide, while the other should not. The big end did not fit in either end of the guide, so the guide was not worn past limits (makes sense).
When testing with the small end, from the combustion chamber side, he was barely able to put it in the guide, I thought maybe 1/4 inch..should have gone a couple inches in...like Greg said. On the rocker arm side, the no-go tool penetrated over an inch and a half.
From the article linked earlier and our discussion today, guides normally show wear just opposite of what was seen here...the combustion chamber side gets larger, and thats how corrosion and deposits form in there, leading to a sticky valve. There are some deposits in the valve seat area (the pocket the valve head fits in) as you can see from the pic, and some grunge around the end of the valve guide, but that grunge came off when I wiped the end of the guide with my finger. Couldn't tell if there were visible deposits in the guide itself (not sure once could visibly see that, since the tolerances are so tight...I just don't know what to look for).
The cylinder mech said he thought the guide might be tapered incorrectly, but again, is that how it was before the failure, or did the failure tweak it? Or is it just deposits on the hot side of the guide that tightened up the space? Lots of questions remain.
The post-exam convesation focused on whether the valve would have been stuck hard enough at temperature to cause the bent push rod and the failed bosses. We also wonder why the push rod didn't bend further, and the bosses failed instead.
Some opinions only: The engine shop owner and the AI both feel it was a bad casting. However, the AI has talked with friends at the FSDO, NTSB, and at several engine shops, and none of those that he talked with had heard of this failure on these particular cylinders. All are aware that many are watching this thread, and no one wants to bad-mouth the cylinders or cause a fear that this is in any way systemic. Again...important to note...opinions only.
Greg is faced with the decision to replace the other 3 cylinders, or fly these. Knowing that these cylinders have close serial numbers adds to his concern, but the fact that this brand of cylinder has a very large sample group with a very large number of hours on them, and this is a rare failure (perhaps the first for this particular cylinder) perhaps mitigates that concern somewhat.
We've all discussed further inspections, and would like to get a better look at the cam lobe and lifter from that valve, as we didn't get a good look at it in Jean that day. But I know Greg would like to do the due diligence on this, as he's a mountain flyer, and a careful guy. I saw the concern in his face as we talked this out, so I know he'd like to hear further discussion and feedback.
And Dan, in discussing the leaning process that day, it sounds like (my interpretation) that when the engine coughed, he thought he was perhaps on the rich side, so he leaned a bit to see if it smoothed out...no major pulls on the red knob. From the graph, after that, it looked like he may have pulled, then added a little throttle, and that cylinder did a quick dead cat bounce, then gave up the ghost (red EGT line to zero). All FWIW, and just one possible interpretation of the graphs.
Still think Greg did a helluva job that day, in decision making and flying!!
Cheers,
Bob