LOADS
LOADS
LOADS TESTING
Landing and taxi tests are necessary to produce and measure gear loads, providing data to verify that the gear has been adequately designed to meet service requirements without exceeding design limit loads. This testing takes the form of landings at different sink rates and attitudes, and taxi tests under various conditions. Experience has shown that most pilots have difficulty landing at specific sink rates (usually because of ground effect) within the tolerance required to avoid gear overloads. Thus, it may be unwise to attempt to landing at the 100 percent vertical load sink rate. Two-point and three-point attitude braking roll, unsymmetrical braking, reverse braking (backing condition with reverse thrust), turns, and towing are specifically required to be demonstrated. As with any loads test, an appropriate build-up is mandatory. The taxi tests can become quite complex and will be covered in greater detail.
Strain gages are the preferred form of instrumentation for gear loads. MIL-A-8871 (Military Specification) stated that
“... the landing gear and backup structure shall be sufficiently instrumented with strain gages to measure vertical, drag, side, and torsion loads. Gear structure can vary greatly between aircraft types, but the resolution of vertical, side, and drag (longitudinal) loads are still the common results sought regardless of the design.”
For the taxi tests, the aircraft may be taxied over surfaces of various roughnesses. This roughness can be simulated on a runway by securing elevated articles (2x4 wood, for example) or by creating uniform “ruts” or spalls in the surface. Tests over various configurations of 1-cosine function bumps and dips (Figure 1) as well as 1 or 2 inch steps were also mandated. In the past these functions have been simulated by AM-2 runway repair mats or stacked plywood. A specific series of these concrete 1-cosine and step functions were in place in the Edwards AFB lake bed for such tests. The results of such tests are compared with the analytical model as well as ensuring that measured loads are within the design limits. The height and wavelength of the 1-cosine function may be selected to create the worst case excitation of rigid and flexible modes at selected taxi velocities, or a number of more benign configurations tested to verify the analysis. If the match is adequate then the model may be accepted to clear the worst case condition(s) if it predicts loads within limits.
Miscellaneous loading conditions produced during towing, turning, and pivoting may also need to be tested. Gear dynamics tests (without regard to airframe response) consist generally of taxiing the aircraft over a discrete bump, such as a single 2x4 piece of wood. This will check for any susceptibility to shimmy (discussed later) or other instabilities. In addition, loads produced by hard landings (high rate of descent), high touchdown ground speed, gear side loads (landing in a crab), gear spring back, wheel spin up, and operations on rough fields may also require testing to verify analyses.
Soft surfaces will create higher gear drag loads than a hard, concrete or asphalt surface. For aircraft intended to be operated on soft fields, gear testing at various surface densities is typically necessary. California Bearing Ratio (CBR) is the accepted measure of surface density. A CBR of about 32 is equivalent to concrete, with softer surfaces having progressively lower values.
Landing gear tests may be done at gross weights, taxi speeds and braking forces up to the maximum. Various center of gravity positions may be run to place the aircraft at different ground attitudes and gear static load states. Tire heating generated by brake rotor radiant heat and simple friction with the ground surface can create the hazard of tire bursting. A hydraulic leak can result in this inflammable fluid contacting the hot surfaces and igniting. These dangers require that brake and tire temperatures be monitored during these tests as well as taxi distance. These measurements are cross-referenced against flight manual brake energy and taxi distance charts. Most wheels incorporate “fuse plugs”. These are metal plugs that melt down to relieve tire pressure when they reach a specified temperature. A flat tire is much easier and safer to deal with than a burst tire. Most modern aircraft brakes also incorporate an anti-skid system that automatically relieves brake pressure if a sudden drop in wheel speed, indicative of a potential skid, is sensed. A skidding tire abrades rapidly and a rupture becomes likely.
The anti-skid also creates the potential for an instability known as gear walk. This is created by an elastic fore-aft deformation if the gear structure creates a false wheel speed change indication. In applying or relieving brake pressure in response to the speed indication the anti-skid may actually promote further elastic response and false indications. Thus, a sustained fore-aft gear oscillation can be created that can produce dangerously high bending loads in the strut. The solution is generally a change in brake hydraulic valve orifice or changes to the anti-skid digital control system. The normal brake testing should reveal any such instability.
SHIMMY
Shimmy is the undamped oscillation of landing gear during ground operations. Shimmy oscillations can produce severe vibration in the aircraft and result in overload failures in the gear assembly or other parts of the aircraft subjected to high inertia loads. The oscillations are typically dominated by torsional deformation, coupled with one or more other dynamic modes of the landing gear system or airframe structure backing-up the gear. Shimmy is normally experienced in an asymmetrical gear (such as a single-wheel fighter aircraft nose gear). It is excited by lateral force on the wheel such as induced by a sharp steering command, an abrupt brake input (main gear), or sudden unevenness (rut or obstruction) of the taxi surface. Factors which play into the problem are the tire and back-up structure elastic characteristics, gear structural stiffness, the rotational inertia of the wheel, the distance of the wheel rotational center from the pivot point (the “trail”), the steering angle (if relevant), lateral strut deflection, unfavorable gear system mass distribution, gear system freeplay, and the airframe motion. All of these factors imply a complex system of springs, dampers and freeplay in multiple axes.
The initial taxi trials of prototype aircraft are carefully performed to ensure that no free-rolling landing gear dynamic instabilities occur which could cause a structural failure and hazard to the aircrew. The free-rolling (no external excitation or braking) structural damping of the landing gear was required by Mil-Specs to be greater or equal to 0.12 G total system damping. Vibrations from tire imbalance (explained later), steering feedback and gear walk should not be confused with shimmy.
SHIMMY TESTING
Where possible, shimmy tests should proceed first flight of a prototype aircraft. For aircraft designed to operate on a variety of surfaces, the surface yielding the lowest additional torsional resistance should be used for the shimmy testing. The antiskid system should be active for the tests to evaluate any potential for the system to play in an instability and because this is the normal operating state for the aircraft. Tires should be new and all servicing (tire pressure, shock strut pressure, lubrication servicing, bolt torques, etc.) should be nominal. Shimmy tests are typically performed by making abrupt brake and steering inputs while taxiing or by rolling the aircraft over discrete obstructions, such as a 2x4 piece of wood oriented at 45 degrees to the path of the machine. The intention is to make a sharp torsional input to the gear. Tests are performed at ever increasing speeds. Testing on an unobstructed surface may be acceptable for aircraft with no off-field operation (landing on dirt or sod). Rudder pedal and brake pulses should be used in these case. Data sample rates should be selected for the dynamics anticipated. If shimmy is excited, an immediate deceleration is necessary. Applying brakes may aggravate the shimmy but is normally advisable in order to slow the aircraft rapidly. The shimmy may not cease immediately, even after decelerating slower than the shimmy onset speed. Shimmy testing and data analysis are much like that for flutter testing.
Because of the many variables introduced by a landing impact ? such as shock strut stroking, truck rotation, spring back, and tire deflection ? landings are not a suitable excitation source. A true free-decay of the gear dynamics cannot be obtained from landings. However, landings may introduce more energy into torsional modes than taxi test methods. Gear dynamics should be monitored during landings until confident that these inputs will not excite shimmy.
The common solution to shimmy is the rebalancing of the gear to alter the modal frequencies. This is usually done by adding a ballast weight to the appropriate place on the gear. A redesign of the gear to redistribute mass and revise component stiffnesses can also be done, but is seldom practical. An alternative method, used when redesign is impractical or a quick solution is required, involves the addition of a damper to the system which has been tailored to the shimmy frequency. Shimmy dampers are usually viscous oil dashpots, porting fluid between sides of a piston through a metered orifice (determining the damped frequency) as the piston strokes with the torsional displacement of the gear truck or wheel assembly.
Wheel oscillations may be observed during ground operations, especially during take-off and landing roll, when the wheels are out of balance or the tires out-of-round (flat spots or other uneven wear). These dynamics may produce moderate airframe vibration when the wheel oscillations excite a natural mode of response in the gear or aircraft structure. The wheel-driven dynamics can be differentiated from shimmy by examining the frequencies associated with the oscillations. If the predominant mode is the same as the frequency of the tire rotation (a function of tire diameter) then wheel-driven dynamics should be suspected.
REFERENCES
Military Specification Airplane Strength and Rigidity - Flight and Ground Operations Tests, MIL-A-8871A.
Norton, William J., Captain, USAF, C-17A Landing Gear Dynamic Stability Testing, AFFTC-TR-93-03, AFFTC, Edwards AFB, CA, May 1993 and Norton, William J., Captain, USAF, C-17A Landing Gear Shimmy Testing, Proceedings of the 24th Annual Symposium of the Society of Flight Test Engineers, July 1993, page 6-39 to 6-53.
Moreland, William J., “The Story of Shimmy”, in Journal of the Aeronautical Sciences, Vol 21, No. 21, December 1954.