Ask the Expert: Auto-Ignition – History and Usage

If you were to gather a group of King Air pilots together and ask them their understanding of and usage of the Engine Auto-Ignition system, I will wager that you would receive a wide variety of responses. Some will arm it on every takeoff and leave it armed until after landing, while others will use it only when in icing conditions. Some will say it prevents an engine flameout, while others will say its purpose is to provide a relight after the flameout has already occurred. Let me see if I can add some historical context to this system and describe its purpose in detail.

The very first King Air models did not have an auto-ignition system. In fact, they did not have an ice vane system! “But, Tom, last month you wrote how important ice vane usage is for engine ice protection, and now you are telling me inertial separators weren’t even installed on the King Air initially?! How did they fly in ice?!”

As surprising as it is today, after ice vanes have existed for so long, the first model 65-90s (also known as “Straight 90s”) used alcohol spray nozzles in the cowling. Similar to alcohol windshield or prop anti-icing, a pump forced the alcohol mixture from a storage tank out through the ejectors. The spray very effectively eliminated the ability for ice to form on the engine intake screen. The system worked quite well … until the alcohol tank went dry. It could sometimes be quite a hassle to find an FBO that could readily refill the tank so complaints were received by Beech’s Customer Service department that this alcohol method certainly had its drawbacks.

Before continuing with our main discussion, allow me to add a couple of interesting tidbits. If you look closely at a picture of the prototype King Air or one of its first progeny, you will notice that the cowling is different. It is missing the oil cooler housing, or scoop, on the bottom. Due to the absence of the scoop, the cowling is noticeably cleaner in design. So where is the oil cooler and where is it getting its airflow? It was located near the back of the cowling, below and behind the location where the inlet air turned the corner to reach the engine’s inlet plenum. (Its location, in fact, was quite similar to how the oil cooler is housed in the entire 200-series.) Some of the air entering the cowling inlet continues aft to flow across the fins of the cooler instead of being ingested by the engine.

The second interesting tidbit is to point out that it was a good thing that the early King Airs did not use engine bleed air as the pressurization air source! Since that air would be mixed with alcohol when the anti-icing system was in use, the pilots and passengers might be getting a bit higher than the airplane as the alcohol affected their brains! Instead of bleed air, the pressurization source was from a roots blower type of supercharger driven by the left engine. Its intake was located in a place such that no alcohol found its way in.

Back to our main story: In response to the complaints about the inconvenience of refilling the alcohol tank, the Beech engineers came up with an inertial separator system, the first ice vane design. Instead of the well-known T-handles that operate the vanes mechanically via a push-pull cable, the first system deployed the vanes via an air-operated piston/cylinder arrangement. Yes, this air was engine bleed air. All King Airs, even those with superchargers, still utilize “Little P3” bleed air for things like deice boots and, in this case, the ice vane actuators. By the way, the engineers were smart enough to design the system such that if all electric power were lost, the pneumatic actuator would default to the extended position, thus protecting the engine in the event that icing conditions were encountered.

In 1966, the second edition of the King Air series, the A90, replaced the “straight 90,” and one of its many large and desirable improvements was the incorporation of the ice vane system replacing the alcohol spray system. But then, a big problem raised its ugly head: Some of the A90s were experiencing loss of engine power due to ice ingestion! In fact, one of the first cases involved a Beech demonstrator A90 with the vice president of sales on board, at night over the Rockies!

“What’s wrong?! How can this be?! We tested the system thoroughly and the FAA certified the design! What’s going on?!” said the incredulous design engineers. Beech and Pratt & Whitney immediately started a new series of icing flight tests, trying to find why the problem was manifesting itself at this time. It was now the winter of 1966 – 1967. Closed circuit TV cameras were installed in the cowling to try to see what was actually taking place. Time and again Beech sought out icing conditions, flew in them extensively, and yet the airplane came through just fine. It was observed that all the engine flameouts related to them by the operators took place at 16,000 feet or above, so those conditions were emphasized. Weeks elapsed without any problems being discovered.

Then it happened. The Beech test airplane had a double engine flameout due to ice ingestion. The crew got the engines running at a lower altitude – although with compressor damage – and landed safely. The investigation revealed the culprit. With the clarity of 20-20 hindsight it is amazing that no one thought of the problem before, but here it is: The deflected ice particles came in contact with the oil cooler face, the warm oil melted the ice, and the subsequent water blew out the cowling from the aft side of the cooler. What was being overlooked, however, is the little device called the vernatherm valve, the gadget that regulates the flow of oil through the cooler to maintain the desired temperature. Under very cold OAT conditions, all of the oil is bypassing the cooler, so now the deflected ice particles coat the face of the cooler leaving absolutely no other path than directly into the engine intake. Damn! That explains why the problem was only showing up at 16,000 feet and above … due to the cold winter OATs up that high causing the oil to bypass the cooler.

Back to the drawing boards the engineers went – with the FAA in close oversight, since King Airs losing engines in flight had definitely gotten their attention, causing a temporary emergency Airworthiness Directive (AD) to be issued to prohibit icing flight – and two changes were made. First, the oil cooler was relocated to the scoop that was attached to the bottom of the nacelle, so that if and when it was bypassed and iced up there would be no effect on engine airflow. Second, engine auto-ignition was designed and installed.

Prior to this, King Airs had left and right ignition switches so that the ignitors (glow plugs, back then) could be activated without the starters being energized. This was for use while doing windmilling airstarts. But it appeared that too many early King Air pilots – who likely had almost no previous turbine experience – forgot that these switches had to be turned on to relight engines that had suffered a flameout due to ice. (“Heck, it didn’t work that way in the P-38 or C-47!”) So, both Beech and the FAA wanted a system that would automatically turn on the ignitors when power was lost. The system is comprised of a simple electrical relay that activates the ignitors whenever torque is below about 400 ft.-lbs. when the switches are in the “arm” position. To state the obvious: When an engine flames out in flight, torque immediately goes to zero (actually, a negative value), well below 400 ft.-lbs.

There were so few King Airs in existence at that time – probably less than 150 – Beech set up a modification line at their Salina, Kansas, facility and the airplanes were flown back “home” for the cowling modification and the addition of the auto-ignition switches and torque sensors. I believe that most, if not all, of the straight 90s were also converted on the line to the new style of system … at Beech’s expense. The Pilot’s Operating Manuals (POMs) were revised to include the requirement to arm the auto-ignition switches in icing conditions, as well as at night above 14,000 feet. The clouds were harder to see at night, of course, and a 2,000-foot buffer zone was subtracted from 16,000 feet, the lowest altitude where engine ice ingestion had been problematic.

With the corrected design of the ice vane system, a pilot who uses the vanes at all times when flying in visible moisture with the OAT at 5° C or below will never experience an engine flameout due to ice ingestion. Thus, his arming of the auto-ignition switches provides no benefit. Peace of mind? Sure. Staying in compliance with the airplane’s POM? Of course. So we will go ahead and arm the switches. But the need for the ignitors to re-ignite the fuel/air mixture the windmilling engine is still providing following a flameout is nil if the ice vanes are properly used.

Glow plugs were replaced with spark ignitors beginning with the introduction of the 200 model in 1974. Thinking that these new types of ignitors would have an almost infinite lifetime as compared to glow plugs, the decision was made – a bad decision, in my mind – to have the 200 operators arm the auto-ignition switches for all flights, all of the time. It did not take long for Beech to realize that the spark ignitors were in fact life-limited. Too many operators reported that they were replacing these new ignitors nearly as often as the old style. Of course, the reason they were failing is that, with auto-ignition armed at all times, the plugs were actually sparking – and wearing the electrodes – whenever torque was low. Silly as it now seems, the first model 200 checklists had the switches armed from soon after start to right before shutdown, so the ignitors were sparking for most ground operation. A POM revision (actually a POH revision, Pilots’ Operating Handbook, since the name and format had been changed by that time) was issued that moved the arming of the switches to the Runway Lineup procedure and the disarming came in the After Landing section. This direction was carried over into the F90 POH upon that model’s appearance in 1978.

Finally, with the appearance of the 300 model in 1984, the Beech checklist writers moved the arming and disarming of auto-ignition back to what it had been in the 90-series: Use for icing conditions and at night when icing conditions may be entered unknowingly.

I would venture to say that most King Air pilots arm auto-ignition when taking the runway on every flight, even when it is severe clear and warm. I know that some King Air training providers advocate this. If that makes the pilots happy, so be it. Nothing is being harmed by doing so except perhaps slightly more plug wear. To replace a few more spark ignitors during thousands of hours of operation makes nary a ripple in the overall cost of operation. But please realize that the system is useless unless a flameout takes place. Although there have been a handful of reports of engine flameouts caused by something other than ice ingestion – a condition lever cable rigged too close to fuel cutoff, fuel starvation due to mismanagement – these types of situations are extremely rare. I believe that the propensity to arm auto-ignition on every takeoff comes partly from pilots with jet experience, in which the ignitors can provide a relight following bird ingestion into the engine. Make no mistake, however, bird ingestion cannot and does not happen in a PT6 turboprop.

In summary, I will never state that a pilot is in the wrong if he or she arms auto-ignition for every takeoff and throughout the entire flight. But I hope they will accept that their colleagues are also not in the wrong if they reserve auto-ignition usage for icing flight.

Do you recall the windmilling airstart “envelope?” The POMs/POHs state that airspeed must be above 140 KIAS and altitude below 20,000 feet when conducting this procedure. During my years of conducting inflight training, I can verify that the lower the altitude and the higher the speed, the cooler that starting ITT will be, due to more air entering the engine. Although, as I have written here, use of auto-ignition is a rather moot point if the ice vanes are used properly, nonetheless I have pondered the ramifications of having auto-ignition provide a relight following ice ingestion when cruising at high altitudes. I hope none of us have the experience, but my belief is that there’d be an excellent chance of overtemping the engine if auto-ignition provided an automatic relight up above FL200.

Conclusion? Make sure auto-ignition is armed for icing flight but realize that the proper use of ice vanes nullifies the need for the relight that the ignition would provide.