Friday, November 20, 2020

When Cooler Heads Prevailed

Date Aircraft Route of Flight Time (hrs) Total (hrs)
20 Nov 2020 N21481 SDC (Sodus, NY) - local 0.6 2183.7

Photographed 14 November 2020

On August 13, Warrior 481 experienced an overtemperature alarm for the #2 cylinder head temperature (CHT) while Kristy, The Bear, and I climbed away from Wiscasset, Maine. I immediately pushed the nose over, pulled the throttle back, and executed an emergency landing back at Wiscasset. Though we were able to get the airplane home that day and I have continued to fly while being mindful of the temperature, Ray and I have spent the last few months in a frustrating cycle of hypothesis, experiment, and disappointment.

For reference, Lycoming considers 500°F to be the redline for their cylinders. However, there is significant literature available from Lycoming and others showing that prolonged operation above 400°F is deleterious to engine longevity. 

Downloading the engine data out of my ship's JPI EDM-700 engine monitor was a critical troubleshooting tool that allowed objective review of the temperature data at home without trying to monitor the indications while simultaneously flying an airplane. 

CHT data from 20 July 2020

As described previously, high CHTs are a hallmark of the engine break-in process. Within the first 25 hours since major overhaul (SMOH) of the engine, CHTs came down into a range comparable to what I would have expected from Warrior 481's previous powerplant. Cylinders #1 and #2, located at the front of the engine, ran the coolest whereas #3 and #4 ran the hottest. At this point, I was reasonably certain that the break-in was mostly complete.

What controls CHT? It basically comes down to three things:
  • Too lean of a mixture while demanding high power
  • Ignition timing that is too advanced
  • Inadequate air cooling of the engine
When the engine monitor alarmed leaving Wiscasset, it was cylinder #2 that tripped the alarm, tying that cylinder to the problem in my mind for many months. We immediately began troubleshooting possible causes while asking the key question "what changed?" Penn Yan, the engine overhaul shop, was initially unconcerned because my CHTs were below Lycoming's redline, but conceded that such a drastic change was worrying.

In troubleshooting, Ray did the following:
  • Verified that the JPI CHT probes were reading accurately by corroborating CHTs with an IR thermometer
  • Verified that the timing was correct (the SureFly electronic ignition module was an early suspect but was quickly exonerated)
  • Checked for causes of a lean mixture:
    • Verified that there was adequate fuel pressure supplying the carburetor
    • Penn Yan bench-tested the carburetor to ensure adequate fuel flow; a minor tweak was made but nothing significant was found
    • Checked for air leaks at the intakes, the carburetor, and the primer nozzles
    • Verified that the mixture control went all the way to the forward stop
  • Inspected and made tweaks to the baffles to ensure that there were no leaks
None of these activities seemed to significantly move the needle and the engine continued to run hotter than normal. Cylinder #2 frequently demonstrated 20-30°F higher temperatures than the other cylinders. What I did not realize at the time, however, was that this did not happen consistently until after we worked on the baffles.

For an air cooled aircraft engine, cooling air enters the cowling near the front of the engine and largely flows above the horizontally-opposed cylinders. Baffles crowning the engine pressurize that air, creating a high pressure region above the engine that forces the air downward through the cylinder cooling fins. Once beneath the engine, the air exits the engine compartment at the lower aft end of the cowling. The cowling is specifically designed so that the slipstream draws air out of the engine compartment.

In an effort to increase airflow over the top of the engine, we added some RTV to the leading edge of cylinder #2 to ensure that no air was escaping below the cylinder. This was done immediately prior to our 05 September 2020 flight through the New York Bravo.

In consultation with Penn Yan, their best suggestion was to replace the carburetor. Despite the prior flow check, there was concern that the carburetor was providing inadequate fuel flow or too lean of a mixture. A new carburetor arrived for installation in November.

CHT data from 08 November 2020

The plot of CHTs shown above was the return flight from Sky Acres on 08 November 2020. I ran at 65% power with a full rich mixture to manage the temperature and, even so, struggled to keep CHT #2 near 400°F while the other cylinders were closer to 380°F.

Photographed 14 November 2020

Shortly thereafter, Ray swapped the carburetors and I took Warrior 481 for a test flight on 14 November 2020. In flight, it was clear that cylinder #2 was still running hotter than the other cylinders (below).

CHT data from 14 November 2020

I was discouraged by these data because CHT #2 was still running disproportionately high and I walked away from the carburetor swap thinking that we failed to solve the problem. On further analysis, this was not actually the case. In fact, ignoring cylinder #2 for a moment, all of the other cylinders came down an impressive 40°F due to the carburetor swap! This strongly suggests that the primary problem was the carburetor all along, though we did not realize it at the time.

I described my tale of high temperature woe to the denizens of the Piper Forum and received an intriguing reply from Domenick, another Warrior owner with whom I have had intermittent correspondence over the years. It described a fix for high #2 CHT on Lycoming engines by opening up the forward baffle slightly to allow more cooling air to pass around/beneath cylinder #2. This was actually the opposite of what we did with the RTV! So, Ray removed the RTV and opened up the spacing between the forward metal baffle and the cooling fins on the leading edge of cylinder #2.

Forward baffle on cylinder #2 after RTV removal and removal of portion of metal baffle (photo by Ray)

CHT data from 20 November 2020

After Ray worked on the #2 baffle, the CHT dropped right into step with the other cylinders (above). Warrior 481's previous engine never ran with CHTs clustered this tightly together.

After going through all of the CHT data in detail, it became evident to me that we had two problems. The first and primary problem was that all cylinders began running hot sometime before our vacation to Maine. Swapping the carburetor out made the greatest difference. I am pleased with Penn Yan for providing the new carb under warranty.

Photographed 14 November 2020

The second problem was one that we created ourselves. In an effort to force more air over the top of the engine, we did not allow enough cooling air around the front face of cylinder #2. This is corroborated by the captured engine monitor data showing that #2 did not start running consistently hotter until after the early September baffle work. But because the whole adventure began with #2 getting too hot in Wiscasset, I did not recognize that #2 was not always the hottest cylinder; to my mind, it was problematic all along. After the baffle work, I became so focused on the differential between the #2 CHT and the rest of the CHTs that I came to regard it as the primary of symptom of the general problem. As a result, I did not initially recognize how significantly we improved the engine operating temperature with the new carburetor. Hindsight is 20/20.

After a multi-month odyssey, from the initial impulse coupling failure at Dansville, to the Warrior being stranded in Dansville for four months, to high temperature concerns with the new engine, I have had my fill of maintenance woes for the year. I am looking forward to just being able to fly again. It is a shame that there are not many places that I can realistically go due to the pandemic.

2 comments:

  1. It amazes me what subtle changes can do for cooling. Even slag on the cooling fins can alter temps. Great job digging deeper and trying different fixes. Good to hear 481 is squared away.

    ReplyDelete
    Replies
    1. Fingers are crossed, but she's been running reliably cool ever since. I have some concern about what next summer looks like when the ambient air temperature is higher.

      Delete