Ray Prouty: Accidents Waiting to Happen

By Ray Prouty | April 12, 2018

Looking back over the history of our industry, we should all be grateful that the early pioneers persisted in making progress, despite the number of unexpected problems they encountered. (Perhaps we should be even more grateful that they did not have high-quality crystal balls.)

However, even after the dedicated work of those pioneers, some of the original problems are still with us. Two types of mishaps that have destroyed many helicopters before the vehicle has even lifted off are ground resonance and dynamic rollover.

Ground Resonance


A destructive oscillation may be encountered if for some reason the blades move on their lead-lag hinges, thus placing their combined center of gravity (CG) toward one side of the rotor disc. In most flight conditions, such an unbalance will rapidly right itself as the individual blades sort themselves out. In this sorting-out process, each blade leads and lags in such a way as to spiral the combined CG in toward the shaft where it belongs.

The potential problem exists if the aircraft is not airborne. A gust of wind, a sudden control motion, or a hard landing can displace the blades. The resultant whirling motion due to the offset centrifugal force may be at just the right frequency to rock the airframe on its landing gear.

Ray Prouty conditions affecting ground resonanceFigure 42-1 illustrates the situation. Once that happens, the two motions get in step and instead of the CG spiraling gently inward, it spirals violently outward — producing a rotating force at the rotor hub that can shake the aircraft to pieces almost instantaneously.

Despite this dire possibility, ground resonance does not happen every time it has an opportunity — just often enough to scare everyone concerned. The first recorded ground resonance accident was in the 1930s, when a Kellett autogyro apparently hit a rock while taxiing. This accident attracted the attention of scientists, who eventually produced a mathematical and physical understanding of the phenomenon. They found ground resonance can be prevented with damping but that the damping must be both in the rotor around the lead-lag hinges and in the landing gear. Thus the most critical condition is just before the ship becomes airborne — since the landing gear is extended and can provide little damping — although there is still stiffness to the rocking motion.

As far as the pilot is concerned, prevention consists of making sure that all dampers are operational during the pre-flight inspection. If, despite this, the beginning of oscillation is detected, the safest action is to either shut down or (if up to flying rpm) immediately take off. When composure is regained, the pilot should make the gentlest landing possible to a high-friction surface. This will produce damping in case the gear begins a scuffing action.

Rotors with high in-plane stiffness and no lead-lag hinges are not susceptible to ground resonance and therefore do not need damping in their rotors or landing gear. Other hingeless rotors may or may not need dampers — depending on how high their in-plane stiffness is.

In some rare cases, the equivalent of ground resonance can occur in the air with a sling load. Operation in compliance with the operator's manual should allow the pilot to avoid this situation but jettisoning the load is a sure — if drastic — cure.

Dynamic Rollover

In flight, high bank angles are of no great concern because control around the roll axis is usually where the helicopter is at its best. On the ground, however, even  moderate bank angle can be disastrous if it is enough to tip the machine over.

Ray Prouty start of a possible rolloverThe primary helicopter upsetting moments are due to rotor flapping, with the resultant tilted rotor thrust and hub moments as shown in Figure 42-2. Sometimes tail-rotor thrust and wind on the fuselage also contribute. The moment that keeps the helicopter from tipping over comes from the weight acting between the two wheels or skids. If the helicopter rolls on its landing gear, this stabilizing moment diminishes — it goes to zero if the ship ever rises on one wheel far enough to put the CG right over that wheel. If the helicopter is sitting on a slope, it already has a reduced restoring moment and a lateral CH position (perhaps caused by fuel sloshing). A narrow landing-gear tread or a rolling deck compounds the problem.

A rollover can happen in calm air if the stick is being held off-center enough during takeoff, but a crosswind can make it even more likely. Even in a strong crosswind, there is little or no main-rotor flapping due to nonsymmetrical aerodynamics until the collective is raised for takeoff, then the nonsymmetrical aerodynamics produce flapping — sometimes referred to as "blowback." In addition, as the shaft is tilted against the springiness of the landing gear, the increased angle of attack generates even more flapping. Thus, if the pilot is not compensated for the disc tilt with cyclic pitch, he will find the upsetting effects increasing at the same time that the restoring effects are decreasing.

In a normal takeoff of most single-rotor helicopters, one landing gear comes off the ground first but, since this happens just as the aircraft becomes airborne, this action is not associated with a rollover. If, however, one landing gear comes off the ground with only partial thrust on the rotor, a rollover may be starting. In this situation, the pilot might try to hurry the takeoff by raising the collective. This is usually a mistake since the increased thrust in the same direction results in an increase of the upsetting moment.

Another choice is to apply lateral control to put the gear back on the ground — but this action may be too late, especially if the initial motion came as a surprise. If an appreciable rolling velocity has developed, it will take a second or two to stop the motion and by this time the helicopter may have tilted irrevocably beyond its critical tip over angle.

A reduction of collective pitch to get both gears firmly on the ground is the accepted cure for a dynamic rollover but this should be done gently. If the helicopter is dropped too fast it might bounce on the gear that was in the air and start rolling in the other direction.

Although pilot distraction or inattention is usually required to set up the conditions for a dynamic rollover, some accidents have occurred when the liftoff was attempted with one landing gear still stuck to the ground by mud, ice or a tiedown.

The ability of the pilot to roll a helicopter over on the ground is enhanced by very stiff hingeless rotors, since even at flat pitch a little out-of-trim cyclic pitch can produce a high, upsetting hub moment. In the Lockheed AH-56 Cheyenne, to discourage the pilot from holding the stick off-center, a device was installed that stiffened up the control centering springs whenever the aircraft had its full weight on the landing gear. The device was deactivated on takeoff as "squat switches" senses the partial extension of both landing gear oleos.

Mast Bumping/Droop-Stop Pounding

Juan de la Cierva's invention of the flapping hinge led to the successful development of rotary-winged aircraft — but it also opened the door to a problem. The mechanical design and the geometry of the total aircraft dictate that the downward flapping — or teetering — must have some physical limit.

The designer tries to allow for normal motions, plus some margin for abnormal motions, but, as per Murphy, "If anything can go wrong, it will." Thus we continue to have cases — depending on the type of rotor — where the teetering limits are exceeded, leading to mast bumping, or the flapping limits are exceeded, leading to droop-stop pounding.

Most of these cases are caused by trying to maneuver during low rotor load conditions. Since most teetering rotor obtain pitch and roll control by tilting the rotor-thrust vector (exceptions are helicopters with hub springs, such as the Bell 222), control power depends entirely on maintaining load on the rotor.

Ray Prouty low rotor thrust available for maneuvering during a pushoverWhen, as Figure 42-3, a pilot tries to maneuver with cyclic pitch during a pushover when the rotor load is low, he or she will find his or herself having to tilt the rotor disc much more to produce the desired control response than he or she would have had to do in level flight. Even at zero thrust, the rotor tip-path plane will happily respond to cyclic pitch inputs by tilting. As a matter of fact, it will override any mechanical teetering or flapping stop by simply bending the blades, since the aerodynamic effects are much stronger than structural stiffness — at least in conventional rotors. A similar dangerous low thrust condition can exist following the reduction of collective pitch during one entry into autorotation.

If the tilt being requested is more than designers had in mind, mast bumping will result. A similar argument applies to rotors with flapping blades, although with offset hinges, they maintain a somewhat higher level of power as rotor thrust is decreased. Thus, the pilot will not be tempted to use quite as much rotor tilt to perform the same maneuvers. At low thrust, a rotor with hinged blades does, however, have a  factor penalizing it, since the coning will be low and the blades will be closer to their droop stops than in normal straight-and-level flight. At any rate, both mast bumping and droop-stop pounding are dangerous because they generate high oscillatory stresses — sometimes high enough to break things — in the parts of the rotor that are banging together.

Ray Prouty hard landings inviting tail boom strikesThe steepness of a hillside suitable for landings and takeoffs may be limited by mast bumping or droop-stop pounding characteristics — or simply by the amount of cyclic pitch allowed by the design. This is especially true during the cautious planting of the downhill landing gear while trying to prevent the aircraft from sliding down the hill. The sequence is shown in Figure 42-4. Takeoffs from a slope, on the other hand, can usually be done with less flapping by making a "hop-takeoff" at a right angle to the slope. This will rapidly clear the landing gear.

Blade Strikes

Designers should always design rotors with the flapping freedom required for maneuvers (including those concerned with escaping from unexpected dangerous situations) while ensuring the blades will never hit any part of the aircraft structure. Designers should, but they have not always. Their most common error has been to underestimate the amount of blade flapping the pilot will ask for, which has led to predicting adequate clearance where subsequent flight test has shown otherwise.

The discrepancy has forced configuration changes in at least two helicopters. The Sikorsky S-55, or H-19, originally had a straight tail boom. Following a series of tail-boom strikes, the tail boom was angled down 30 deg to produce additional clearance. On a more recent project, it was found that snappy pushovers put the forward blades too close to the canopy, forcing a redesign that raised the main rotor.

One of the most common tail-boom strike conditions is a hard landing. Figure 42-5 illustrates two scenarios for this type of accident. In the first, the landing gear sets down firmly and stops the downward motion of the fuselage — but the rotor blades keep coming down. Low rotor speed and a quick reduction of collective pitch contribute to the downward flapping and bending of the blades.

Ray Prouty hard landings inviting tail boom strikesThe other type of tail-boom strike conditions is likely to occur during a run-on landing when the flare angle is high and the aft part of the landing gear hits first. This tends to bounce the rear part of the helicopter back up while the front part and the rotor continue to descend. Also, the sudden nose-down pitching motion makes the pilot instinctively pull his control stick back, causing the rotor to tilt even closer to the tail boom. Designing the landing gear to have a long energy-absorbing stroke alleviates these landing problems.

Other tail-boom strike incidents have occurred on some helicopters during the entry into autorotation. The rapid reduction of collective pitch makes the rotor flap down in front and produces a sudden pitch-down condition. If the pilot overreacts with aft cyclic stick, the combination of low coning and aft flapping may be enough to reduce the clearance to zero.

When the rotor is up to speed, the centrifugal forces stiffen the blades and limit their possible bending deflections — but at very low rotor speeds during startups and shutdowns, the centrifugal stiffening effect is low and gusty conditions can cause erratic, high-amplitude flapping and bending. For this reason, many rotors with flapping hinges have centrifugally operated droop stops. At low rotor speeds, these stops limit the down movement of the blades. However, once the rotor is up to speed, the droop stops move out of the way to allow more flapping for the now-stiffened blades.

A special case of droop-stop pounding once occurred on a parked Sikorsky S-56 when another S-56 landing beside it. The rotor wash from the landing aircraft lifted one blade of the parked helicopter high into the air, and then withdrew its support, letting the blade slam down on its droop stop. The result was a badly bent rotor blade, demonstrating why blades should be tied down.

Operation Near Obstructions

A white-water boater can see the currents and rough stuff that might upset him or her, but a helicopter pilot very seldom gets such clues. It is bad enough that the ambient wind gets distorted by nearby obstructions — producing unseen currents and rough air, but the helicopter, being a wind machine, can add to the confusion with the recirculation of its own rotor wake.

Ray Prouty wake recirculationFigure 42-6 shows one of the most noticeable effects that can occur when taking off near an obstacle. The recirculating wake produces an increasingly non-uniform inflow pattern at the rotor disc as thrust is increased to lift off. This will require a nonstandard lower cyclic stick position that might come as a surprise to the unsuspecting pilot.

For example, if, because of recirculation, the down flow is stronger on the left side, the helicopter will tend to move backward because of the 90-deg lag in flapping. Not only is the cyclic pitch affected, but the increased down flow looks like a climb condition that requires more power to the main rotor and therefore to the tail rotor also. This represents a decrease in ground effect and might keep a heavily loaded helicopter grounded.

Distortions of the inflow distribution also occur whenever the helicopter is maneuvered into a region of disturbed wind — such as around a rooftop landing pad, mountain ridge, drilling platform or ship's deck. In these cases, inflow changes might be either up or down but will generally come on suddenly during a takeoff or landing. To minimize the surprise when landing, some pilots recommend a slow crosswind approach, which gives the best chance of reversing the decision if things get too rough.

Finally, care should be taken around airports where large fixed-wing aircraft are laying down strong but invisible wingtip vortices that can persist for several minutes. Crossing through one of these can be very upsetting, even for large helicopters.

Receive the latest rotorcraft news right to your inbox