Conducted 12/30/2012 -
The morning of the 26th,
John's parents wanted to get an early start to their 13 hour car ride home (the
Cozy will have us there in a little over 2 hours eventually). John wanted to get
an early start on the crank seal replacement anyway (Christine had to work
unfortunately), so all was working out well. John picked up the seal and
adhesive, and paid a visit to our EAA Tech Counselor for the nifty little
Lycoming tool for stretching the seal around the crank flange. He then went out
to the hangar to get started. Fortunately we're able to heat the hangar when we
need to work, but for curing things we can’t obviously leave the torpedo heater
running. For fiberglass work we have found 60 watt light bulbs at close range
will keep surface temps elevated enough. For this we were going to need a bit
more though, so we picked up a couple of 300watt heat lamps to point at the seal
after it was installed.
The seal install went well, and the heat lamps were able to keep the area of the seal around 85-90deg F according to our little IR thermometer. The next couple of days were clouds/snow, so that gave the adhesive ample time to cure. Saturday (the 29th) we reinstalled the flywheel, prop extension, prop, and cowlings. We cleaned up the hangar and ballasted the plane for a hopeful Sunday morning flight.
Sunday morning the sky was clear, but the car reminded us all the way to the airport that it was cold…alternating between the “ICE” warning and the Outside Temp display of -8F. All John could think about was how cold it was going to be at 10k feet. There was little wind though, so it really didn’t feel THAT bad outside. We still elected to do the preflight in the hangar with the heater on though. John's flight report to the Cozy list is below:
"I taxied out and did my run up. It was at that point that I realized I hadn’t ground run the engine since replacing the seal. “How the heck did I forget THAT?” I thought. Obviously I wasn’t expecting any change in it’s ability to RUN, just that I should have checked it for leaks again. What I should have done was head back to the hangar right then, let it warm up a little longer, then at least pull the top cowl and check it over. Instead I somehow rationalized that even if it was still leaking, it wouldn’t leak that much. It had only lost at most ½ qt in 1.7 hours on the last flight, and I had a new seal installed now.
The run up went fine, and I was quickly cleared for take-off. The acceleration was as brisk as the morning air, and in what appears to be about 11 seconds in the log, the wheels broke ground. One second later, with an altitude of 19 feet AGL, the engine speed dropped 300 RPM. To me it sounded like the engine died. I was already at 100Kts, and I wasn’t sure I’d be able to stop the plane before rolling off the end of the runway if I tried to set it back down. One second later, before I could finish my panicked analysis of the situation, the engine sped back up. I started wondering if I had just had some sort of ignition hiccup. I quickly tell the controller I’m going to remain in the pattern and may request a landing if things don’t look good on downwind. At that point I thought I noticed the hiccup again, but I can find no evidence in the log of it now. Maybe I was just overly sensitive.
As I turned downwind, I checked the key switch, fuel selector valve, and mixture. I noticed the mixture lever was pulled back slightly, and started to wonder if I had simply starved the engine for enough fuel in this cold air. It just didn’t seem like it should have been enough to cause such a drastic response from the engine. I elected to climb up to 5k’ and just circle the field while I checked the engine monitor for clues to what happened. I could find nothing. No elevated CHTs or EGTs. Both ignition systems were happily reporting their timing advance information. Fuel pressure and fuel flow were both good. At this point the engine was running perfectly. The slightly out of place mixture control was my only clue. “What could have caused that?” I thought. In hind sight the most important question was “Why on earth was I trying to solve the mystery while in the air?!?!”
I spent the next 30-40 minutes leaning the engine out at various throttle settings trying to see if it would do it again. It didn’t. At that point I figured I really needed to see the engine monitor log to know where to look, so I was going to have to call it a day. There was a pile of traffic in the pattern by this point, so it was a few minutes later before I dropped back down and made an uneventful landing. I had another hour in the book, and was finally past the ½ way point.
I examined the log but could find no indication of sudden loss of fuel or spark. I could see the RPM drop from the first ‘hiccup’, but could find no trace of what I thought was a second. I sent the data to a friend who spent some time looking at it, and called me last night with an interesting revelation. “Hey, did you notice the drop in oil pressure from 89 PSI to 79 PSI JUST before the RPM drop?” he asked. He had no idea how it could possibly be related, but thought it was odd. Even more odd was that the RPM starts dropping just as the pressure bottoms at 79 PSI, but as the RPM falls, the pressure climbs back up. Then it hits me! The prop hub had no oil in it when I started the engine. The governor I’m using requires about 2000RPM to make enough pressure to actuate the prop. Given the cold temps and lack of ground running, the drop in RPM could have just been the governor system purging air.
I called Whirlwind and confirmed my suspicions. The purging of air would create some pitch fluctuations that would have resulted in a noticeable RPM drop as governor finally pumped oil into the hub. Whirlwind asked “Why didn’t you cycle the prop during your run up to get the oil out to the hub?” I told them I tried, but that I can’t hold the plane at 2000 RPM with my brakes. They informed me that there IS a different spring I can use in the governor I have to get the actuation point lower. I guess I should have mentioned this to them earlier.
So the mystery is solved, and I’m over the 20 hour mark now, but I’m not pleased with the decisions I made on that last flight. I should have gone back and ground run the engine. It might not have mattered, but then again maybe it would have gotten enough oil into the hub to prevent the hiccup that had me contemplating an off-field landing scenario. I also should have just set the plane down on the first circuit around the pattern. Yeah, it turned out to be something that was a one-time, non issue. But running it on the ground for half hour would have been a safer place to be reviewing the possibilities. I would have felt better about taking it back up and actually continuing on with the planned test after that too."