Pitch-up

In aerodynamics, pitch up is an uncommanded nose upwards rotation of an aircraft. It is an undesirable characteristic that has been observed mostly in experimental swept wing aircraft at high subsonic Mach numbers or high angle of attack.

Pitch up problems were first noticed on high speed test aircraft with swept wings. It was a common problem on the Douglas Skyrocket, which was used extensively to test the problem. Before the pitch up phenomenon was well understood, it plagued all early swept wing aircraft. In the F 100 Super Sabre it even got its own name, the Sabre dance. In aircraft with high mounted tailplanes, like the F 101 Voodoo, recovery was especially difficult because the tailplane was placed directly in the wing wake during the pitch up, causing deep stall (although the T tail was meant to prevent pitch up from starting in the first place). Deployment of the braking parachute and a considerable height above the ground were essential for a chance at recovery.

Wings generate pressure distributions on their upper and lower surfaces which produce a single force acting at a point known as the "center of pressure", or CoP, which is normally located between ⅓ and ½ of the way back from the leading edge. This upward and rearward leaning force is replaced by an equivalent pair of forces called lift and drag. The longitudinal position at which these forces act and the magnitude of the forces change with angle of attack. In addition a varying pitching moment exists for any force location other than the CoP. These changes lead to a requirement to trim aircraft as they change their speed or power settings. Another major consideration for aircraft design is a vector addition of all of the weight terms of the parts of the aircraft, including the wing. This too can be reduced to a single weight term acting at some point along the longitudinal axis of the aircraft, the "center of gravity", or CoG. If the wing is positioned so its CoP lies near CoG for the aircraft, in level flight the wing will lift the aircraft straight up. This reduces any net forces pitching the aircraft up or down, but for a number of reasons the two points are normally slightly separated and a small amount of force from the flight control surfaces is used to balance this out. In this case of a swept wing, changes to the CoP with angle of attack may be magnified. The introduction of swept wings took place during a move to more highly tapered designs as well. Although it had long been known that an elliptical planform is "perfect" from an induced drag standpoint, it was also noticed that a linear taper of the wing had much the same effect, while being lighter. Research during the war led to widespread use of taper, especially in the post war era. However, it had been noticed early on that such designs had unfavourable stall characteristics; as the tips were more highly loaded in high angles of attack, they operated closer to their stall point. Although this effect was unfavourable in a conventional straight wing aircraft, on a swept wing design it had unexpected and dangerous results. When the tips stall on a swept wing, the center of pressure, the average lift point for the wing as a whole, moves forward. This is because the section still generating considerable lift is further forward. This causes further nose up force, increasing the angle of attack and causing more of the tip area to stall. This may lead to a chain reaction that causes violent nose up pitching of the aircraft. This effect first noticed in the Douglas D 558 2 Skyrocket in August 1949, when a 0.6 G turn suddenly increased out of control to 6 G. This was not entirely surprising; the effect had been seen earlier in wind tunnel simulations. In addition, swept wings tend to generate span wise flow of the boundary layer, causing some of the airflow to move "sideways" along the wing. This occurs all along the wing, but as one moves towards the tip the sideways flow increases, as it includes both the contribution of the wing at that point, as well as span wise flow from points closer to the root. This effect takes time to build up, at higher speeds the span wise flow tends to be blown off the back of the wing before it has time to become serious. At lower speeds, however, this can lead to a considerable buildup of the boundary layer at the wing tip, adding to the problems noted above. Finally, while not directly related to the effects above, it was common during the early jet age to use T tail designs in order to keep the aerodynamic surfaces well clear of the jet engine area. In this case it is possible for a pitch up event to cause the turbulent air behind the wing to flow across the horizontal stabilizer, making it difficult or impossible to apply nose down pressure to counteract the pitch up. Aircraft with low mounted tail surfaces did not suffer from this effect, and in fact improved their control authority as the wing's wake cleared the controls surfaces, flowing above it. This was not always enough to correct for the problem, however; the F 86 continued to suffer from pitch up in spite of increasing nose down pressure from the tail surfaces.

As the primary causes of the pitch up problem are due to spanwise flow and more loading at the tips, measures to address these issues can eliminate the problem. In early designs these were typically "add ons" to an otherwise conventional wing planform, but in modern designs this is part of the overall wing design and normally controlled via the existing high lift devices. The first known attempt to address these problems took place on the platform where they were first noticed, the Douglas Skyrocket. This took the form of a series of vortex generators added to the outboard portions of the wing, breaking up the boundary layer. However, this was found to have almost no effect in practice. Nevertheless, a similar solution was attempted on the Boeing B 47 Stratojet where it proved considerably more effective. This may have been helped by the presence of the podded engines, whose vertical mountings acted as barriers to span wise flow. More common solutions to the problem of spanwise flow is the use of a wing fence or the related dogtooth notch on the leading edge of the wing. This disrupts the flow and re directs it rearward, while also causing the buildup of stagnant air inboard to lower the stall point. This does have an effect on overall airflow on the wing, and is generally not used where the sweep is mild. To address the problems with spanwise loading, a wider variety of techniques have been used, including dedicated slats or flaps, the use of washout or automated control of the ailerons. An unusual solution tried on the XF 91 Thunderceptor prototype fighter was to give the wingtips a wider chord than the wing roots. The idea was to increase wingtip efficiency and cause the wing roots to stall first. Angle of attack sensors on the aircraft can also detect when the angle of attack approaches the attitude known to result in pitch up and activate devices like the stick shaker to warn the pilot, and the stick pusher which overpowers the pilot and forces the nose of the aircraft down to a safer angle of attack. Twist or washout built into the wingtips can also alleviate pitch up. In effect, the angle of attack at the wingtip becomes smaller than elsewhere on the wing, meaning that the inboard portions of the wing will stall first. A commonly used solution to pitch up in modern combat aircraft is to use a control canard. Another modern solution to pitch up is the use of slats. When slats are extended they increase wing camber and increase maximum lift coefficient. Pitch up is also possible in aircraft with forward swept wings as used on the Grumman X 29. With forward swept wings the span wise flow is inboard, causing the wing root to stall before the wingtip. Although at first glance it would appear that this would cause pitch down problems, the extreme rear mounting of the wing means that when the root stalls the lift moves forward, towards the tips.