Using Engine Monitor to Detect Problems by Peter Cassidy
AirPlay Second Quarter 2012
One of the great benefits of an engine monitor is its ability to store engine data for later analysis. If you monitor your engine's performance, you can usually spot problems like burned valves and failed spark plugs before they become serious. I make it a practice to download and review my JPI engine monitor data every 10 hours. Here's an example of how this practice pays off.
The graph shows a portion of the EGT traces from a 2 hour flight on Jan. 24, 2012. Notice how the top, light blue line (EGT #6 cylinder) is jumping around. There is clearly a problem even though the engine was running smooth and the JPI display did not indicate anything was abnormal. When I saw this trace, I immediately suspected a spark plug problem.
Pulling the top plug on cylinder #6 it was easy to see the problem. Look at the black gunk on the ignition lead and similar buildup inside the spark plug. The black is carbon tracks that are shorting out the plug probably through a crack in the ceramic. Cleaning with MEK fixed the problem, at least for nw. I fully expect to replace the plug before long. I did look at the JPI data and plug connection after another 8 hours flying and all was well. If I had suspected a problem in the air, I would have done an in-flight mag check. That would have pinpointed the problem as the top plug on #6.
It's worth noting that this plug was new 62 hours ago. It is a Champion fine wire type that should last over 1000 hours. At $85 each, I think twice about tossing them. In recent years there have been reports of cracked ceramic in Champion fine wire plugs so I'm not surprised. Another area where cracks in the ceramic show up is in the firing tip. In this case part of the ceramic can break off. I had this happen in 2009. It was found during an inspection.
Pay attention to the data stored in your engine monitor. I picked 10 hours for a regular review for two reasons. First, it is frequent enough to catch serious problems like burnt valves before they get destructive. Second, at a 2 second sample rate, 10 hours is comfortably under my JPI data storage capacity. If I change the sample rate to 6 sec., I can go 25 hours before the data overflows. But I prefer the finer data rate and 10 hours is when I should be looking at the data anyway. Also, save the data from previous downloads. It's helpful to look at earlier traces to see what normal is.
It's not surprising there was no indication of this problem either during the ground runup or in-flight. During a mag check on the ground, it takes very little to fire a plug compared to in-flight. It is not uncommon to have a good mag check on the ground but fail an in-flight check. In-flight, where we have significantly higher cylinder pressures and leaner fuel-air mixture, is significantly more demanding of the ignition system. In addition, we have two plugs in each cylinder and the engine will run just fine on one. That's redundancy doing its job. One plug failed is not an emergency. However, it does need to be corrected should the second plug fail in which case we will have a very rough running engine and be looking for a place to land as soon as possible. If you are running LOP (Lean of Peak) as I almost always do, switching to ROP (Rich of Peak) will often bring a marginal plug back online.
There are a number of ways to monitor our engine health. Reviewing the engine monitor data is just one. Included in the list should be, compression checks, borescope inspections, oil analysis, oil filter inspections, in-flight mag checks, and GAMI lean tests. If we do these on a regular basis, we can dramatically reduce the likelihood of a catastrophic engine failure.