The Need To Be Tactically Tactile
The expression “show ’em a clean set of heels” is often used by runners, when seeking to leave the competition behind. In aviation, without a clean set of wings, you’re just not in the race.
Many of the medium size and smaller jets have just that – a clean set of wings. Apart from the obligatory flap, ailerons, speedbrakes and spoilers behind the midspan chord, the leading edges of Fokker F28’s, F70’s, F100’s, CRJ’s 100 and 200, Challengers, and similar aircraft all rely on an unadorned leading edge.
Around that leading edge, we might have some internalized anti-ice heating but that’s all – and there’s the rub. Without a good tactile “rub”, it’s often hard to tell whether the wing’s all important upper surface is contaminated by frost or not.
The type of supercritical wing section used on these jets suffers greatly from an increased coefficient of drag (and a much decreased coefficient of lift), if hoarfrost is there for takeoff. Icing takeoff accidents have also occurred in the Yak 40, Citation, Falcon, etc.
However, the Citation’s wing section is nowhere near as vulnerable as the other types. They appear to have an icing sympathetic wing-section for thinner distributions of ice and frost.
This type of wing-spoiled accident is more to do with the type, location and area of icing and less to do with the weighty accumulation of ice that can occur airborne in SLD (supercooled liquid droplets or freezing rain), for instance.
That type of icing is of more concern to Cessna Caravan pilots, although they too need to ensure that their main and tailplane upper surfaces are ice-free before launching.
The “Fokker Ice Plan” was put in place in the early 90’s to address a number of ground icing accidents. These included an F-28 in LaGuardia in 1992 and a Fokker 100 in ’93 in Skopje. The plan included:
a. A black stripe on the outboard part of the wing to improve visibility of small ice particles against a grey wing (for ground use)
b. An improved Flight Director
c. An “on the ground” Wing Leading Edge Heating System
d. An icing awareness program consisting of a video (“Look Twice for Ice”)
* Presentations to operators
* Provision of ice data for simulator use
* Updates to Flight Manuals
One enhancement was cancelled: the over-wing ice inspection light. Testing showed that plain illumination was nowhere near as effective as a physical “hands on” tactile inspection. That lesson has been re-learned many times since the first such recorded takeoff accident in 1972 (see the table listing them at tinyurl.com/2ugffq ).
As you can see from the many accidents listed, a clean wing is not necessarily a safe wing when non-apparent icing and high angles of attack are in play.
After the accident to the USAir F28 at LaGuardia in 1992, the FAA sought to sort the wing ground icing problem with enhanced de-icing procedures before the next icing season. The approach was to educate pilots in keeping the wings clean before takeoff.
The NTSB was also keen to have Fokker’s test-pilots investigate how takeoff performance and stall margin would be affected by using a lower initial target pitch attitude on rotation. This was a precautionary measure designed to combat undetected wing upper surface icing – or icing that accumulated unseen during taxiing and pre-takeoff holding.
Simulations with different target pitch attitudes showed reductions in peak angles of attack (AoA) of the order of 2.3� to 2.6�— whenever the target Vr pitch angle was reduced from 18� to 10�. Wind tunnel tests showed that a fully contaminated wing simulating hoar frost via a particulate density of one per sq cm (of 1mm high crystals) stalls some 6� angle of attack earlier than an uncontaminated wing. That’s once again, as Shakespeare would have said, “the rub”.
The lift loss amounts to a “staggering” 26 percent. Because elevator and horizontal stabilizer effectiveness is relatively unaffected, having a different and inverted airfoil section, the loss of wing-lift is lethal. Why so? Aircraft can easily be rotated into a wing-stalled condition where wing drop (aka “roll-off”) is highly likely and any attempt to “pick up” a stalled wing with aileron just compounds that wing drop.
That’s the lethal accident scenario that’s to be seen time and time again in the listing referred to above. It’s apparently now happened again in Moscow on Feb. 13 to a Challenger 850 crew. Turbo-props don’t have the same takeoff problem. Much of their wing-lift is induced by the props and their wing sections tend not to come in a supercritical profile.
Fokker test pilots conducted a series of simulator takeoffs with normal and contaminated wings, using four different takeoff techniques. Contaminated techniques trialed were:
* Normal technique (resulted in a stalled condition on the ground)
* Variations in rotational pitch-rates varying from 2 to 8 degs/sec
* Reduction in final pitch target (nose-low take-offs)
* Increased rotation speeds (Vr)
They then tried without the Flight Director but by using a fixed initial pitch target. The object was to reduce the peak angle of attack (i.e., the danger of inadvertent stall upon rotate/unstick) but without unduly affecting ground-roll distances. In order of decreasing effectiveness they found that:
a a two-step rotation using an initial pitch target of 18� reducing to 10� was best
b. a lower pitch rate was safer, but difficult to define
c. increased Vr speeds reduced AoA (but chewed up much runway real-estate)
An analysis of the results turned up the most effective and practical solution, which was a combination of a low pitch-rate to an initial target pitch angle of 7.5� to 10�. This was chosen because it matched well with a built-in Fokker 70/100 inclination and a natural pilot tendency to hesitate during rotation at 8� to 10� of pitch. This is a tailplane moving into ground-effect proximity characteristic.
The mission then became to bring the Improved Flight Director’s (IFD’s) commanded pitch-rates into line with these targets. The requirement was to avoid higher pitch-rates and maximize the margin, throughout the takeoff rotation, between actual AoA and the unknown stalling angle of attack for a contaminated wing.
The “holdover time” (at the end of which a return to ramp for further de-icing is required) is a gray area. It’s a function of the type of precipitation being encountered and the type of fluid used. No flight crew can be expected to determine this period with great certainty, and the new wild card is infra-red de-icing which is being used now at a number of airports.
The Fokker 70 and Fokker 100 have very sophisticated Flight Directors, so it was decided to incorporate the new operating techniques into their two control laws (airspeed and angle of attack), and thus introduce a less aggressive rotation and built-in guaranteed safety margins against a stall. Fokker ensured that the unseen bias in the Improved Flight Director would inject a takeoff safety margin into any such scenario, including the engine-out case.
During a take-off with a slightly contaminated wing, the pilot can still rotate to 10� pitch without restrictions. Depending on the degree of contamination, the speed at 10� pitch will be somewhat lower than normal due to the greater drag resulting from the contamination. Consequently, the IFD will stay at 10� longer. The pilot will be limited in pitching up the aircraft (until speed and angle of attack are “within limits”), and then through the guidance of the IFD, the speed and angle of attack are maintained within safe limits during further rotation towards 18� of pitch.
Assuming a typical degree of contamination, the Fokker 100 and Fokker 70 will stall at an angle of attack around 10�. That is what happened in the Fokker 100 accident at Skopje.
That stall occurred at an angle of attack between 10� and 11� and was preceded by heavy buffet just before the stall. The stall itself is characterized by a sharp roll-off, followed by severe wing rock. The crew of the (not de-iced) accident aircraft concentrated on controlling the wing rock, but at the same time kept pitching up towards the FD cue above the current pitch angle of 10� -12�. Each time they pitched up towards the FD, the aircraft stalled again.
With respect to the FD, two lessons were learned. First, the pilot should be able to recognize the heavy buffet. In itself this buffet is not hard to recognize. However, as with the Challenger, the problem is that the buffet is followed immediately by the severe roll-off. Therefore, the pitch-rate when approaching the “critical” angle (near 10� pitch) should be low to create more time to allow the pilot to correct that pitch angle. This effect is now there with the modified IFD.
Second, the FD should not be displaying above the controlled pitch angle, tempting the pilot to increase the pitch angle, when the aircraft is in fact close to a stall and the only escape is to lower the pitch angle. Ideally, the FD should then command a lower pitch angle. This effect is also gained with the modified IFD.
Although the stall margin is significantly increased by using the new IFD, it should be realized that it is still possible to stall the aircraft if the wings are significantly contaminated (see “Pau crash” later). However, the crew is allowed more time to recognize the pre-stall buffet and to react to it. Simulator training is required to reinforce this. To this end all Fokker 70/100 simulators have been modified to provide heavy buffet when approaching a stall with a contaminated wing.
The IFD’s introduction automatically covered the dangerous scenario where a pilot boards his plane under a clear blue sky, not noticing that the overnight temperatures and a super-cold fuel load have coated his wings with a dangerously uniform and transparent coating of invisible frost. Time and again it’s been proved that a light coating of contamination can be missed or disregarded as inconsequential.
This scenario is apparently what happened to a Challenger 604 crew at Birmingham UK and a CRJ200LR crew at Baotou China. The Challenger 850 crew in Moscow (Vnukovo Airport) on Feb. 13 and the CL604 at Montrose appear to have been victims of an excessive hold-over time and snow-showers.
Fokker’s early 90’s workaround solution of changing their IFD’s control laws’ speed bias seemed to have worked. For years, Fokkers weren’t figuring in the takeoff accident statistics. However, a recent Jan. 25 fatal accident to an Air France Fokker 100 at Pau (AF7775, F-GMPG) seems to have all the hallmarks of a contaminated wing and an excessive rotation. Temperature at the time was 0�C and dew point was -1�C and the aircraft was not de-iced prior to departure.
But what about the Canadair (now Bombardier) Regional Jets? Out of eight corporate and nine regional jet (CRJ100/200) write-offs, we have nine (i.e., over half) that are associated with the basic unslatted wing’s characteristics. The Challenger 850 that crashed Feb. 13 in Moscow is based on the CRJ200LR (the type that crashed at Baotou). It doesn’t have leading-edge slats.
With the success of the CRJ200, airlines wanted a larger capacity CRJ that would allow them to offer lower seat-mile costs. Bombardier developed several: the CRJ700, CRJ705 and CRJ900. The Bombardier CRJ700 was developed as an evolution of the CRJ200, and incorporated an all-new wing with full span leading edge slats. Those full span leading-edge slats are to be seen in the CRJ700 photo at tinyurl.com/2lk3xq.
These, according to Bombardier, “allow excellent airfield performance with only a small increase in span from the CRJ200”. A cynic might say that the real reason those lift- enhancing LE slats are now there, and perhaps should have been there in the first place on the CRJ2, was to nullify the lethal leading edge and upper wing surface contamination problem on takeoff.
If that Moscow crew had been flying the Challenger 870 (based on the CRJ700LR), they would have had that significant extra buffer against the deadly contamination stall on takeoff. If the crew mistakenly taking off on the short runway 26 at Lexington had slats at the time, they too would’ve dodged the grim reaper.
There’s not a lot to be said against LE slats – and much to be said for them. They do much, much more than just “go with the flow”.