Return To Title List
Wheel Shimmy:Time for a Better Approach?
How could it be that in 2007, the aeronautics industry is still applying a 65 year old "solution" to an incorrectly defined problem that continues to dog us? Could a much simpler and more effective answer lie in the forces that stabilize bicycles?In 1939, Germany's Luftwaffe requested a twin-engine "Destroyer" airplane. Messerschmitt responded with what became the Me 110. The aircraft was a disappointment when used in combat with single engine opponents. It was retired to more mundane functions until, when fitted with radar, found success as a night-fighter against Britain's RAF Bomber Command.The Me 110 was an aircraft that didn't possess the panache' of the Bf 109, or the FW 190, but in a way, its very existence had an unanticipated but significant impact that haunts us to this day: Its tail-wheel shimmied.We have all watched the track of a golf ball as it travels toward the hole, only to see it lip the cup, accelerate, and roll away.Basketball shots can seem to go down into the hoop-roll sideways and then climb up and out. "Hooked", or "sliced" baseball and golf shots are a common sight. A stubborn grocery cart with a shimmying front wheel is an all too frequent annoyance.These are all examples of our daily encounters with gyroscopic forces.As aviators we depend on the gyroscopes in auto-pilots, artificial horizons, turn and bank indicators, Dg's and HSI's to make flying safer, and in some situations, possible.The wheels on our vehicles, including airplanes, once in motion, become gyrostats.Our bicycles, and motorcycles could not be made to stand erect on in-line wheels, were it not for the gyrostatic stabilities that are developed from their spinning wheels.So, what do these seeming unrelated topics have in common? What does the prior existence of a largely forgotten airplane have to do with gyroscopes?As mentioned above, the Me 110's tail-wheel shimmied. It shimmied to such an extent that there was detailed study of the phenomenon by German scientists.Shimmying wheels are certainly not unusual. When airplanes abandoned the tail-skid, and utilized wheels instead, many but not all applications were subject to shimmy. Shimmy is the un-commanded oscillation of a tire/wheel assembly about the steering axis of that assembly. It may start spontaneously, and develop un-wanted vibrations throughout its vehicle. It seems to be related to speed, load or outside-forces.It is indeed unfortunate, even now, that the Me110 shimmied.When the European war ended, the victorious Allies were greatly impressed by the advances seen in German aeronautical engineering. Operational jets, rocket airplanes, swept wing concepts, and rocket propelled bombs demanded the scrutiny of all things "German".One area that attracted interest was the attention that had been paid to the analytical studies of pneumatic tires. One of those works devoted itself to the tail-wheel shimmy problem of the Me 110.Since shimmy was (and is) mysterious and the airplane was in mass production, the study was largely to determine if any damping could be provided by the tire used on its tail-wheel. To that end the German engineers undertook specific, and somewhat unique, measurements of that size tire. Although shimmy is a dynamic process, the measurements were of necessity made on static tires.These reports were studied by the United States facility which was most concerned with things aeronautical-the National Advisory Committee for Aeronautics. (pre-cursor of today's N.A.S.A.).In 1956, N.A.C.A. 1299 was published. (Its bibliography acknowledged the prior German analytical efforts). This has become the "Bible" for those studying shimmy. Its thrust is not to discover what causes shimmy, but how to study pneumatic tires and especially the possible role their hysteretic characteristics might play in damping shimmy tendencies. In 1958, NACA published " Tech. Report 4110", which suggests a plethora of static testing to determine the visco-elastic properties of tires. This report does not dwell on the word "shimmy", although its footnotes again cite those prior works on the subject. It mainly seeks tire test data so as to verify equations which may roughly describe the multi axis interactions taking place within and without the distorted tire. Over the years, it has been used as a "cook book" and caused vast quantities of data to be collected and collated.Certainly the telescoping struts of a landing gear afford little mechanical advantage to prevent the inner pistons from twisting in their outer cylinders. The scissors joint at their apices, and the "torque" links which attach them, does provide some leverage, and a site for the variety of shimmy dampers that are installed. In some respects the basic arrangement is not all that different from the front fork set-up of the ordinary bicycle. The bicycle, and its cousin the motorcycle, cannot tolerate shimmy.Somewhere in the distant past, and probably by trial and error, a workable combination of friction and forward caster was discovered. If the caster angle were too great, the bicycle could not be steered. If it were too small, the bicycle became difficult to ride.The nascent automobile industry passed through this same terrain. Early cars were prone to front-end shimmy, and not until the three-dimensional aspects of camber, toe-in and caster were tied together did a workable "truce" evolve.Prior to "1299", aircraft designers sometimes used the bicycle approach-witness the forward cant of the nose struts on the C-54 (and derivatives)*, the Lockheed C-69** and other World War 2 tricycle geared designs.Documents written subsequent to "4110" have increased the amount of tire "shimmy data" required for airframe certification.Each airframe company has someone who devotes time to shimmy. He or she is tasked with developing computer programs that will predict the shimmy tendencies of proposed landing gear systems.These programs look at discrete properties of the pneumatic tire under consideration, as well as the mass and stiffness of the metal supporting structures.No programmer has yet shown sufficient confidence in these predictions so as to recommend that the new airplane be built without shimmy dampers.Some airplanes that have been built, with a passing grade using these procedures, have had such severe initial shimmy problems that the prototype has had to have mass damping added to the landing gear in order for flight testing to begin.The landing gear of an airplane is a part-time necessity. For most of the airplane's duty cycle it is excess baggage. It is stored somewhere that does not interfere with flight systems. This reality can and does make the caster angle presented to the tire assemblies something less than an "exact science". The angle is not specified, and is whatever the gear attach point, strut length, and height subtend. This can result in unstable geometry, where any excitation from anomalies in the rolling surface, or even a worn tire can trigger shimmy. This applies to both auxiliary and main wheel applications. Some airliners have had a history of almost 100% pre-wear out removals of nose tires for "shimmy-vibration" complaints. This is a waste of un-used rubber, labor for tire and wheel changes, and serious reductions in dispatch reliability. In some airplanes that have experienced torque-link failure, the event cascaded into catastrophic shimmy excursions of their main wheels. The wheels and tires doing 360's about their strut.The computer is a tremendously valuable tool in the design of product. However it is limited to the in-puts of its programmers. It is not a practitioner of logic.Unless engineers elect to take courses in logic, they may not have been exposed to its virtues either-- only the extrapolation of what they have been taught.Were the Me 110 to be placed in production today, and its tail-wheel tire made to "benefit" from the battery of testing needed for current application approval, its shimmy potential would be no less than that of its 1940's clone.In the early days of aviation, there was puzzling phenomenon known as "the mystery dive". Airplanes could somehow drop their noses, and in spite of up-elevator pressures, continue a steep descent into the ground. The airplane might also revolve as it dived.One day, someone trapped in this fatal maneuver, and feeling that he had nothing to lose, applied down elevator pressure and was relieved and astounded to find that he was regaining control of his airplane. This was truly an intelligent experiment. Only much later, were the physics of the aeronautical "stall" understood, and the preventative and recovery measures enunciated.The brothers Wright were bicycle men who learned something of aeronautics by conducting intelligent experiments. It is to be hoped that the aeronautical community may begin performing well designed experiments and learn the secrets of the bicycle's front wheel. This will allow the writing of gyroscopic equations descriptive of zones of inherent stability applicable to sleeved joints with spinning wheels. We could then not only fix the Me110's problem, but vastly improve the aircraft that we are designing for the future!