Guest Post: TransAsia 235 Crash Timeline

[Editor’s note: IG kingpin Michael Exner has delved into the recent crash of TransAsia Flight 235 with characteristic vigor, and recently sent around some images he generated from ADS-B data that go a long way toward establishing what happened. I asked him if I could reprint the material here and he agreed. — JW]

I reviewed the flight paths for flight GE235 for the previous week. In every case, the path was nearly the same. The plane makes a departure to the east and turns to the right about 160 degrees shortly after takeoff. [See image above.] This is what happened on the day of the crash too. However, in the case of the crash day, the airspeed and altitude just seconds into the flight, at the start of the first right turn, was already low and slow. For all flights in the previous week, the plane was between 2000-3000 feet at the first turn (point furthest east), and the speed was between 130 and 170 kts. For the day of the crash, the plane was at 1250 feet and 83 kts, which is below stall speed for 15 degrees flap setting. [See image below, after the jump; click to enlarge.] This suggests that the left engine was not producing full power from a point very early in the flight. Indeed, the ADS-B data indicates that the airspeed at liftoff was normal (136 knots at 100 feet), but began falling immediately after liftoff. This suggests that the left engine flameout occurred immediately after liftoff, near the east end of the runway.

[Editor’s note: If a twin turboprop loses power in one engine while climbing away from the runway, the situation is dangerous for a number of reasons. For one thing, the plane now obviously has much less thrust, which means it might struggle to climb clear of obstructions while at the same time maintaining necessary airspeed. If it doesn’t maintain enough airspeed, then it could experience an aerodyanamic stall without having enough altitude to recover. (Mike points out that the ground speed when the plane was at 200 feet was 85 knots; with a headwind of 10 knots, this means the plane was just below the stall speed (with 15 degrees flaps) of 96 knots.) Without airspeed, the rudder might not generate enough force to overcome the asymmetric thrust caused by the fact that one engine is running and the other is not. This asymmetry of thrust is even more dangerous if the blades of the inoperative propellor are not ‘feathered,’ that is to say, turned sideways to the airstream to reduce drag; feathering is sometimes done automatically and sometimes not, but in the case of TransAsia 235 the non-turning prop appeared not to be feathered. An inexperienced pilot can make the asymmetry of thrust worse still by adding power to the engine that is running. This might seem a tempting prospect at first glance, if the plane is struggling to gain speed and altitude, but it might create a turning moment too high for the rudder to counteract and cause the plane to roll over and crash.

Given all that, it might be that AirAsia had initiated its planned turn to the right just after takeoff when it lost power in its port (left) engine. Thrust from the healthy starboard engine caused it then to veer to the left.

To give another perspective on this dynamic, here’s a fascinating video reconstruction of such a crash that occurred in 2010. — JW]