Deep Stalls

September 1, 1993

A turbojet-powered airplane with a T -tail is on a constant heading in a level-flight attitude. Forward speed, however, is nil, and the aircraft is sinking vertically at an alarming rate. The flight controls are ineffective, and the earth continues to rise. The fate of the aircraft is sealed.

Impossible? A dream? It is neither. It can happen, especially when flying a swept-wing airplane with aft-mounted engines. It is a deep stall, one of a jet pilot's worst nightmares. It is an excursion beyond the envelope and from which return may be impossible. (Although rare, deep stalls also have occurred in T-tail light airplanes and high-performance sailplanes.)

The T-tail became popular in general aviation design circles in the 1970s because it gives an airplane a jaunty, rakish appearance. It also is more effective than a conventional tail because it operates in the free airstream above the disturbed air left in the wake of propellers and wings. A T-tail allows airplane designers to use smaller horizontal surfaces to do the same job. The result is less drag and an increase in performance, which makes T -tails particularly advantageous in the design of jet aircraft. (It is not always so advantageous, however. In some small aircraft, the use of a T-tail reduces longitudinal stability and degrades handling quality about the pitch axis.)

Raising the horizontal tail surfaces high above the fuselage also makes it easier for the designer to accommodate aft-mounted turbofan engines.

During an approach to a stall in an airplane with a conventional tail, the pilot usually is warned of the impending stall when the horizontal tail surfaces are affected by the wake of the wing, which is what happens in a stall-producing nose-high angle of attack.

In most cases, the pilot feels and sometimes hears the buffeting that this creates. If he fails to take appropriate action, it is likely that the nose will fall and automatically begin the process of stall recovery. The pilot, of course, must continue the process and also avoid a secondary stall by releasing back pressure on the control wheel and reducing the wing's angle of attack.

Many pilots do not understand what causes the nose to drop. It has little to do with the wings losing lift — stalling — and mostly to do with the tail. During normal flight, the horizontal tail surfaces produce a down load (negative lift, if you will) that prevents the nose from pitching down on its own. But when these surfaces are lowered into the low-energy wake of the wing, the tail loses effectiveness and can no longer produce the down load. As a result, the nose pitches downward.

When that happens, the elevator also loses much of its authority, which prevents the pilot from forcing the aircraft into a deeper stall.

Things can happen differently with a T-tail. As the pilot applies back pressure to raise the nose, the horizontal surfaces might still remain above the wake of the wing even after the wing has begun to stall. The pilot might not feel any buffeting or natural stall warning. With the elevator still in the free airstream, it remains effective and allows the pilot to unwittingly drive the wing into a deeper stall at a much greater angle of attack.

Finally, the wing reaches an extreme angle of attack, yet the elevator remains effective. The pilot continues to pull back on the wheel and buries the horizontal tail surfaces in the wing's wake. The elevator rapidly loses effectiveness. Also, the disturbed, relatively slow air behind the wing might sweep across the tail at such a large angle that the tail itself stalls. The pilot loses all pitch control and is unable lower the nose.

The fuselage is another factor to consider. At very large angles of attack, the fuselage acts like a fat, stubby wing and produces a modicum of lift. The trouble is that this "wing" has an extremely small aspect ratio (the ratio of span to chord). When the fuselage is at a high angle of attack, it also creates induced drag in the form of "wing tip" vortices. These are shed from along both sides of the fuselage and behave like the tips of an absurdly proportioned wing.

These fuselage-generated vortices attack and interfere with the horizontal tail surfaces. They also can exert a downward pressure along the top of the stabilizer, forcing the tail down farther and deepening the stall.

It would seem that aft-mounted engines contribute to the deep stall because of the additional and seemingly excessive weight that they add to the rear of the aircraft. But this weight has no direct influence on deep stalls. The designer maintains the aircraft in perfect balance simply by moving the wing aft. The result is a strikingly long fuselage section ahead of the wing — a defining characteristic of some aircraft such as the McDonnell Douglas MD-80.

At large angles of attack, the lift produced by such a lengthy, forward-fuselage section has substantial leverage. It tends to hold the nose high during a deep stall, which exacerbates the problem. Also, the fuselage — unlike a real wing and more like a barn door — does not stall. Instead, it produces increased effective lift with increasingly large angles of attack (even after the wings become fully stalled).

(Aft-mounted engines can contribute to a deep stall for a different reason, however. At large angles of attack, their nacelles generate vortices that can interfere with the tail surfaces.)

Another factor that can contribute to a deep stall tendency is wing planform. It is characteristic of swept wings that — as angle of attack increases — the wing tips tend to stall first. And because the tips of a swept wing are on the aft part of the wing (behind the center of lift), it follows that a loss of lift near the tips causes the center of lift to move forward. This forces the nose farther up and the wing more deeply into the stall.

The designer, however, attempts to preclude the possibility of undesirable tip stalling. He does this by reaching into his bag of aerodynamic tricks for some goodies that help to ensure that the wing roots stall before the tips. Such tricks include washing out the wing tips (reducing their angles of incidence), adding stall fences, using a combination of airfoils, and so forth.

Even if the designer is successful in delaying the tip stall, however, it is possible for the tips to eventually stall more fully than other areas of the wing. The result is that — despite the designer's best efforts — the swept wing might still contribute to a deep stall condition.

(Another disadvantage of a tip stall is that it can involve the ailerons and erode roll control.)

Assume that upon entering a deep stall, the pilot pulls back so far on the control wheel that the tail is somehow brought below the wing's wake and below the vortices spawned by the fuselage and the nacelles. Because the free airstream below the aircraft has much more velocity than the low-energy wing wake, it can force the tail to rise back up and into the disturbed air that caused the problem in the first place.

After the aircraft enters a deep stall, increasing drag reduces forward speed to well below normal stall speed. Sink rate increases to possibly thousands of feet per minute. The aircraft eventually stabilizes in a vertical descent. The angle of attack approaches 90 degrees, and indicated airspeed is virtually nil. The emergency ends when the aircraft pancakes into the ground in near-level attitude. The crash signature of a deeply stalled aircraft impacting the earth is unmistakable.

Some pilots speculate that they might break the stall by rolling in one direction and then kicking bottom, rudder to get the nose down. It probably would not work. Flight controls are generally ineffective when the wing is at a 90-degree angle of attack, no matter how aggressively the pilot moves the control wheel and rudder pedals (although this might have some aerobic value) .If rudder alone is effective, the pilot might use it to yaw the aircraft and select the view that he will have while crashing.

This does not mean that a pilot should give up and accept the role of being a passenger in his own airplane. Anything is worth a try. Cycle the controls, the landing gear, the spoilers, and the flaps. Try every possible combination. It might also be worthwhile to have the passengers move forward because an aft center of gravity worsens the problem.

Can a pilot power his way out of a deep stall? Probably not. Air passing vertically in front of the engine inlets undoubtedly will result in severe compressor stalling and prevent the development of meaningful thrust.

Nor is the deep stall anything that a pilot can practice (not that he would want to). In this respect, it is like a flat spin. A pilot's first is likely to be his last.

Fortunately, deep stalls are easily avoided as long as published limitations are observed. Airframe manufacturers go to great lengths to provide sophisticated stall warning systems such as stickshakers and -pushers. A pilot does his part by heeding these warnings and not making a flight when a required stall warning system is inoperative.

In addition to a deep stall, the pilot of an aircraft equipped with a variable-incidence (or trimmable) stabilizer might discover that it is possible for the aircraft to become "locked" at the other end of the speed spectrum In other words, it will appear to him that he cannot recover from a high speed dive.

A trimmable stabilizer is used on many jet aircraft because it creates less trim drag than a fixed horizontal stabilizer that requires deflecting the elevator to provide pitch trim. But there is one problem with the trimmable stabilizer. The stabilizer usually is accompanied by a relatively small elevator because all of the real work is done by moving the stabilizer .The elevator is used only to make small pitch changes to the trimmed attitude.

Assume that a pilot encounters turbulence that causes the nose to pitch up, and he (or the autopilot) reacts by applying significant nose-down trim. Immediately thereafter, however, the nature of the turbulence changes, and this — combined with the nose-down trim — forces the nose down sharply.

The pilot hauls back on the yoke. The trouble is that the feeble elevator is unable to overcome the more aerodynamically powerful stabilizer. The pilot then jabs at the thumb switch on the yoke in an effort to apply nose-up trim. But the trim motor may not be strong enough to override the opposing force created by nose-up elevator.

The best and perhaps only way to recover from the resultant high-speed dive is to release back pressure on the control wheel as if attempting to recover from a stall. This relieves the trim motor and allows the pilot to reposition the stabilizer.

It seems that no matter how sophisticated the airplane and no matter how esoteric the maneuver, the control wheel remains little more than a house-size controller. Pull it back, and the houses on the ground get smaller; pull it back farther, and the houses get bigger.