If you look through your logbook, there’s probably an expression with the word “seat” to describe many of your flights. A ringside seat. The hot seat. Left seat. Right seat. The catbird seat. Best seat in the house. A nosebleed seat. Ok, I’ll stop. Because engine data analysis is more interested in the seats in front of the firewall – the valve seats. Thanks in part to Savvy’s borescope initiative, we’re seeing more and better pictures of valve heads, stems and seats. And we’re seeing something that threatens engine longevity – waterfall erosion. More about that when we look at the pictures.
But first a shout out to my Savvy colleagues Dave Pasquale and Jess Nordlander. Dave’s leading the charge on the erosion (also called guttering) phenomenon much like he did earlier with standardized borescope inspections and with valve lapping. Jess is our lead borescope analyst having logged thousands of hours reviewing piston domes, valve trains and cylinder walls.
This month we’ll look at data from a Cirrus SR22T, a Carbon Cub, an RV-10 and a Cessna Turbo 206.
First up is a Cirrus SR22T powered by a Continental TSIO–550 with data from a Garmin perspective with a one second sample rate. The owner didn’t report trouble – just asked us to take a look. Here’s EGTs, CHTs and FF for a typical flight.
My colleague Ray noticed that FF rises during the climb. More during the first climb but the second climb at 18:20 was an enroute climb from FL100 to 140. Neither increase was pilot-commanded. Ray wondered if MAP was climbing along with the FF. Here’s that data.
So the answer is yes but not as much. The airplane went into annual and during the inspection they discovered this hole.
The ram’s horn flange broke off when the clamp was removed. It may have been weakened but it was still attached. Our theory is that the exhaust leak resulted in lower pressure and the turbo controller compensated by dialing up MAP and FF. We would have bet that a hole of that size would heat up the air enough to show up as higher CHTs, but – going back to the first chart – we’re not seeing that.
Next up is a Carbon Cub powered by an Aerosport 340 with data from a Garmin G3X with a one second sample rate. Here are EGTs, CHTs and FF for a typical flight. Cursor is off to the right out of the way.
The owner was performing his annual condition inspection and reported lower compression on cylinder 2 with leakage audible from the carb side (vs the exhaust side.) Looking at the data, 1 and 2 are a little different from 3 and 4, but 1 and 2 are pretty similar. Not seeing anyplace where 2 goes rogue. btw, analysis kudos if you noticed that the Garmin is logging (non-existent) TIT for this engine. I hid it using the Appearance menu.
The owner submitted a set of borescope pictures and these two from the cylinder 2 intake valve caught our eye.
Waterfall erosion — also called guttering – is the name for this phenomenon. The metal on the seat wears away and allows air past the closed valve. We mostly see it on exhaust valves, but here it’s on an intake valve. The photo on the right shows erosion on the valve face, too. Initial observations are that ECI cylinders tend to show more wear on the valve face while Continentals and Lycomings tend to show more wear on the seat. More data should be able to support or refute that initial observation.
This stage of erosion resulted in about a 10-15 psi drop in compression. So 70’s to around 60. At this stage, we’d expect lapping the valve to allow the valve to remain in service. With late-stage erosion we’d expect to see more leakage, and compression readings closer to the master office, and lapping would probably not save the valve. Thanks again to Dave and Jess for their pioneering work on this, and we’ll update you as we discover more.
Next up is an RV-10 powered by a Lycoming IO-540 with data from a Garmin G900X. The owner reported rough running and uncommanded RPM changes of 50-70 RPMs in cruise. He submitted a series of mixture sweeps and mag checks. Here are EGTs, CHTs and FF from that flight.
Mixture sweeps showed good distribution – unchanged from previous sweeps – and good spark all around. Before submitting the data, the owner had sent both Emag P200 6XLs out for board and firmware updates. That didn’t fix the rough running. He reamed the valve guides. That improved the rough running a little, but it didn’t go away.
He noticed a mark on the cylinder 1 rocker arm – looked like it had impacted the spring. He replaced the rocker and the lifter, bled the new lifter and set the dry tappet clearance. Better, but still rough. He decided to IRAN the governor. During the I, they discovered a cracked shaft.
A new governor was cheaper than overhauling the old one. Once it was installed the random RPM changes stopped.
The owner posted — It looks more like installation damage on the arm spline (by the original builder), not a true “cracked shaft” that would create free-play in the RPM control.
We’ll finish this month with a Cessna Turbo 206 powered by a Lycoming TIO-540 with data from a Garmin G3X with a one second sample rate. Here are EGTs, CHTs and FF for a recent flight. Cursor is off to the left out of the way.
Fortunately, this cylinder failure happened near an airport and the pilot was able to get on the ground safely. Valves get a lot of blame, but this one happened lower in the case. Here’s the bottom of the cylinder 6 piston.
The natural question after an event like this is “is there something in the data that would have alerted me to this upcoming event”? Let’s zoom into the event and add oil pressure on the FF rank in dark blue.
CHT 6 starts a slow rise a little more than a minute before EGT drops off. Oil pressure rises a degree or two but doesn’t really convey urgency. It’s a little less than 5 minutes from the failure until landing. Let’s hope we all have a fortunate outcome when it’s our turn in the hot seat.
