I'm told that planes can actually stall when the airflow over the wing goes past Mach 1? Why does this happen and how do you design an aircraft to avoid it?
When the velocity of the airflow locally exceeds speed of sound above the wing a shock wave forms and the flow detaches beyond this shock wave.
Similarly to stall, the supersonic separation of flow removes the component of lift produced by decrease of pressure on the upper surface of the wing and so the effects are similar.
I causes reduction of lift and because centre of pressure is about quarter chord on the upper surface, but midchord on the lower, it causes a significant pitch-down moment, which might be impossible to recover even if the post-stall lift is otherwise sufficient to keep the aircraft flying straight. This effect is often called Mach tuck. Supersonic planes often have all-moving elevators to have sufficient control authority to compensate for it.
A difference from normal stall is that after supersonic flow separation the lift remains proportional to angle of attack and so the aircraft continues to behave more or less normally except for the change in trim.
Mach tuck may occur as low as Mach 0.7 depending on aircraft design, because the air moves faster over the wing. It can be delayed by using swept wings, because the shock waves only form when the air velocity component perpendicular to the wing exceeds speed of sound.
Supersonic aircraft eventually encounter supersonic flow separation, but the lower surface lift is sufficient to balance the aircraft weight at that speed and altitude and the aircraft can continue flying with the upper surface flow separated.
That's not the case with most subsonic aircraft, which for efficiency reasons tend to cruise at altitudes where they have very small margin to stall. As the stall speed increases with altitude while speed of sound slightly decreases with the lower temperature there, the stall speed will eventually equal critical mach number, which creates the coffin corner and absolute ceiling for the aircraft. For most civilian transport aircraft the range between stall speed and critical mach number (where drag increases and supersonic flow separation would begin) reduces considerably at cruise altitude, but they usually don't have enough engine power to reach the actual coffin corner.
Aircraft stall when the wing cannot produce enough lift to sustain flight. This can happen for two reasons:
- Flow separation due to high angle of attack. The lift curve slope, which is positive and linear at low angles of attack, becomes negative, such that an increase of the angle of attack results in lower lift. This is caused by viscosity effects.
- Flow separation and heady buffeting due to strong shocks on the wing close to Mach 1. This is caused by Mach effects. In many cases lift loss is not yet critical but buffeting will become uncomfortably severe and dangerous to structural integrity.
The first is a low speed stall, but it can happen at any speed. High speed stalls are the second variety. They can be provoked by
- flying faster at the same attitude, normally in very thin air and at speeds around Mach 0.8.
- Or they can be provoked by demanding more lift at the same speed, e.g. by initiating a turn.
- A third way is by climbing, such that the air gets thinner and colder, and the wing needs a higher angle of attack to produce the same lift as before.
In all cases, the initial flow over the wing was locally mildly supersonic and produced a weak shock. Either by increasing speed (more precisely: The flight Mach number) or angle of attack, the shock becomes stronger and can cause flow separation, such that the wing produces less lift than before. The aircraft stalls. In case of the U-2, the tail would still work, only the wing would produce less lift, so the plane pitches down and accelerates. Acceleration makes things worse, because now the shocks grow even stronger. Now the pilot is locked into a dive which he cannot end. A lot of fun when a few kilometers below the pilot sees MiG-17s, performing pull-ups from high speed to get close enough to open fire.
Thankfully, the drag increase due to the stronger shocks limits the speed increase, and after dropping for maybe 2km, the air becomes dense enough for the U-2 pilot to successfully stop the dive and start the climb back to the safer, higher altitude.
How to avoid this? Generally, this is impossible to avoid altogether, it can only be mitigated. Wing sweep is the most effective way to limit lift loss due to local shocks, but even a highly swept delta wing will have a reduction of its maximum lift coefficient around Mach 1. An optimized airfoil shape helps to push the limit up, but when this new limit is exceeded, the lift loss will be heavier. In the end, most modern aircraft resort to limiting the flight envelope and using an electronic FCS to ensure that the envelope is not exceeded.
This can happen at high altitude. At high altitudes, the air is less dense, (there are fewer molecules per cubic inch). Lift is generated by the force of air molecules hitting the surface of the airfoil. So at altitude, it takes a higher angle of attack on the airfoil to generate enough lift to hold the aircraft in level flight. Therefore, even if supersonic, there may not be enough molecules hitting the wing to generate sufficient lift at the stall angle of attack. ergo, the aircraft, in attempting to maintain level flight, increases AOA past the stall AOA. Even then, of course, all the pilot has to do to recover from the stall is reduce AOA below the stall AOA to resume controlled flight. The aircraft will of course begin to descend and he will have to deal with a dive recovery once he reaches an altitude where the air is dense enough to support 1 G level flight.