REPORT 1154 CONTENTS : SUMMARY... 1079 1080 1080 ost Dynamies of Syater. 1081 orecs in Shook Str 1082 1081 1085, 1085 1086, EQUATIONS OF MOTIO! 1087 Motion Prior to Shock-Strat Deeotion.. 1088 Motion Subsequent to Beginning of Shock-Strut Deflection. 1000 SOLUTION OF EQUATIONS OF MOTION. 1000 ‘Numerical Integration Procedures. 1000 ‘Use of Tire Foree-Deflotion Characteristice_ 1001 Exeet of Drag Loads. 1001 EVALUATION OF ANALYSIS 1001 1002 1002 1002 1005 1008 1008 Bilt of Oriice Diseharge Coolant. 1008 Effect of Air-Compression Process. 1100 SIMPLIFICATION OF EQUATIONS OF MOTION. aoe Bvalustion of Simplifications. 1103 Generalized Treatment... < 1uos ‘Wauations and solutions. 1103 Applicability of solutions. 1107 SUMMARY OF RESULTS AND CONCLUSIONS. 109 APPENDIX A—NUMERICAL INTEGRATION PROCEDURES. > Linear Procedure. = > ins Quadratic Procedure > ine Runge-Kutta Procedire.. 2 ane APPENDIX B—SOURCE OF EXPERIMENTAL DATA. 2 120 Equipment. ‘Test Specimen. Instrumentation 2 130 REFERENCES. 13 BIBLIOGRAPHY. naz5 REPORT 1154 ANALYSIS OF LANDING-GEAR BEHAVIOR * By Bewanty Miuwiraxy and Francis B. Coor SUMMARY ‘This report presents a theoretical study of the behavior of the conventional type of oleo-pneumatic landing gear during the process of landing impact. The basic analysis is presented in «general form and treats the motions of the landing gear prior to.and eubsequent to the beginning of shock-strut detection. In the analysis of the first phase of the impact the landing gear i treated as a single-degree-of-freedom system in order to deter- imine the eonditions of motion at the instant of initial shock-strut deflection, after which instant the landing gear is considered as ‘a system with two degrees of freedom. The equations for the two-degree-f-freedom system’ consider euch factors as the Iydrautie (pelocity square) resistance of the orifice, the forces due to air compression and internal friction in the shock strut, the nonlinear force-deflection characteristics of the tire, the wing Lift, the inclination of the landing gear, and the effects of wheel spin-up drag loads. The applicability of the analysis to actual landing gears has been investigated for the particular ease of a vertical landing gear in the absence of drag loads by comparing caleulated reculte with experimental drop-test data for impacts with and ‘without tire bottoming. The calculated Behavior of the landing gear was found to be in. good agreement with the drop-test data. ‘Studies have also been made to determine the effects of varia- tions in such parameters as the dynamic. force-deflection characteristics of the tie, the orice discharge coeficient, and the ppolytropic exponent for the air-compression process, which might not be known accurately in practical design problems. ‘The study of the effects of variations in the tire characteristics indicates that in the case of « normal impact without tire botloming reasonable eariations in the force-deflection character- istice have only a relatively small effec on the ealculated behavior of the landing gear. Approximating the rather complicated ‘force-defection characteristics of the actual tire by siniplified ‘exponential or linear-segment variations appears to be adequate ‘for practical purposes. Tire hysteresis was found to be relatively unimportant. In the case of a severs impact involving fire bottoming, the use of simplified exponential and linear segment apprazimations to the actual tire force-deflection characteristics, which neglect the effects of tire bottoming, although adequate up to the instant of bottoming, Jails to indicate the pronounced. increase in. landing-gear load that results from bottoming of the tire. The use of exponential and linear- segment approsimations to the tire characteristics which take into account the inereased stiffness of the tire which results from bottoming, however, yields good results. The study of the orifice indicates that the magnitude of the discharge coeficient hae a marked effect on the ealoulated behavior of the landing ‘gear; a, decrease in the discharge coeficient (or the produet of the discharge coeficient and the net orifice area) results in an approzimately proportional énerease in. the mazimum upper mass acceleration. ‘The study of the importance of the air-compression process in the shock strut indicates that the air springing 4s of only minor significance throughout most of the impact and. that variations in the effective polytropic exponent n between the ‘isothermal value of 1.0 and the near-advabatic value of 1.3 have only a secondary effect on the ealoulated behavior of the landing gear. Even the assumption of constant atr pressure én the strut ‘equal to the initial preseure, What is, n=0, yields fairly good resulls which may be adequate for many practical purposes. In addition to the more exact treatment, an investigation, has been made to determine the extent torwhich the basic equations of motion can be simplified and still yield acceptable reeulte, Thie study indicates that, for many practical purposes, the air-preseure force in the shock strut can be completely neglected, the tire force-deflection relationship can be assumed to be linear, and the lower or unsprung mass can be taken equal io zero. Generalization of the equations of motion for this simplified system shows that the behavior of the system ie completely determined by the magnitude of one parameter, namely the dimensionless initial-velocity parameter. Solutions of these ‘generalized. equations are presented in terms of dimensionless variables for aide range of landing-gear and impact parameters which may be uaefl for rapidly estimating landing-gear performance in preliniinary design. ‘Sopeiea NACA TW 6 “Anlyof Landing Gear Heri” by Denia MWtky and Fra E Cok, 10%1080 INTRODUCTION ‘The shock-absorbing characteristics of airplano landing gears are normally developed largely by means of extensive tril-and-error drop testing. The desire to reduce the ex- pense and time required by such methods, es well as to pro- vide & more rational basis for the prediction of wheel-inertia drag loads and dynamic stresses in flexible airframes during landing, emphasizes the noed for suitable theoretical mothods for the analysis of landing-géar behavior. Such theoretical methods should find application in the design of landing gears and complete airplane structures by permitting (a) the determination of the behavior of a given landing- gear configuration under varying impact conditions (velocity ‘at contact, weight, wing lift, etc.) (b) the development of a Ianding-gear configuration to obiain 2 specified behavior under given impact conditions (©) smorerational approach to the determination of wheel spin-up and spring-back loads which takes into account the shock-absorbing characteristics of the particular landing gear ‘under consideration (@) improved determination of dynamic loads in flexible airplane structures during landing. This problem may be treated either by calculating the response of the elastic sys- tem to landing-gear forcing functions determined under the assumption thet the airplane is a rigid body or by the simul- taneous solution of the equations of motion for the landing gear coupled with the equations representing the additional dogross of freedom of thostructure, Tn many eases the former approach should bo ‘sufficiently accurate, but in some instances, particularly when the landing-gear attachment points experienco large displacements relativo to the nodal points of the flexible system, tho latter approach, which takes into account the interaction between the deformations of the structure and the landing gear, may be required in order to represent the system adequately. ‘Since many aspects of the Ianding-impact problem are so intimately connected.with the mechanies of the landing gear, the subject of landing-gear bebavior has received analytical treatment, at various times (see bibliography). Many of tho earlier investigations, in order to reduce the mathematical ‘complexity of the analysis, were limited to consideration of highly simplified linear systems which have little relation to practical landing gears. Some of the more recent papers ‘consider, with different degrees of simplifieation, more real- istic nonlinear systems, ‘Tho present report represents an attempt at a more complete analysis of the mechanics of practical landing gears and, in addition, investigates the im- portance of the various elements which mako up the landing gear, as well as the extent to which the systam can be reason- ably simplified for the purpose of rapid snalysis, ‘Tho basic anslysis is presented in a general form and takes, into account such factors as the hydraulic (velocity square) resistance of the orifice, the forces due to air compression and internal friction in the shock strut, the nonlinear force- deflection characteristics of the tire, the wing lift, the inelina- tion of the landing gear and the effects of wheel spin-up drag Toads. An evaluation of thé applicability of the analysis to actual landing gears is presented for the ease of a vertical landing gear in the absence of drag loads by comparing cal- culated results with drop-test date, REPORT 115{—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS Since some parameters, such as the dynamic foree- deflection charactaristics of tho tire, the orfico discharge coefficient, and the polytropic exponent for the air-compression process, may not be accurately known in practical design problems, a study is made to assess the effects of variations in these parameters on the calculated landing-gear behavior. Studies are also presented to evaluate tho extent to which the dynamical system can be simplified without greatly im- pairing the validity of the calculated results. In addition to the investigations for specific cases, generalized solutions for tho behavior of a simplified system are presented for a wide range of londing-gear and impact parameters which may be ‘useful in preliminary design. SYMBOLS ‘ Ay pneumatic area A hydraulic area A, tree of opening in orifice plate A internal eross-soctional area of shock-strut inner cylinder Ae external cross-sectional area of shock-strut inner 7 cylinder A, cross-sectional area of metering pin or rod in plane of orifice A,“ neborifice area Ge orifice discharge coefficient a overall diameter of tire y Pheumati¢ foree in shock strut y hydraulic force in shock strut x friction foree in shock strut ih total axial shock-strut force A normal force on upper bearing (attached to inner eylinder) A normal force on lower bearing (attached to outer cylinder) Fg foreo normal to axis of shock strut, appliod at . axle Fy, vertical force, applied at axle Fu, horizontal foree, applied at axle Fo, resultant foree, applied at axle Fy force parallel to axis of shock strut, applied to tire at ground Fy force normal to axis of shock strut, applied to tire at ground Fr, ‘vertical force, applied to tire at ground Fu, horizontal force, applied to tire at ground Fog resultant foree, applied to tiro at ground t gravitational constant Ky lift factor, L/W L lift force h axial distanco between upper and lower bearings, for fully extended shock strut & axial distance between axle and lower bearing (attached to outer cylinder), for fully ox- tended shock strut, abmr constants corresponding to the various regimes of the tire-defloction process o combined constant, ad mw combined constant, md” 2 polytropio exponent for sir-compression process in shock strutANALYSIS OF LANDING-GEAR BEHAVIOR R Roynolds number De air pressure in upper chamber of shook strut Pa Ihydraulic pressure in lower chamber of shock strut “ Q volumetric rate of discharge through orifice t radius of deflected tire 3 shock-strut axial stroke r wheel inertia torque reaction t timo after contact 1 time after beginning of shock-strut deflection ° air volume of shock strut te Polar moment. of inertia for wheel assembly about axle rr vertical velocity Ve horizontal velocity wv total dropping weight w weight of upper mass ebove strut Ws weight of lower mass below strut Ea horizontal displacement of lower mass from position at initial contact a vertical displacement of upper mass from posi- tion at initial contact a vertical displacement of lower mass from posi- tion at initial contact uy dimensionless upper-tnass displacement from position at initial contact cA dimensionless lower-mass displacement from position at initial contact « dimensionless shock-strut stroke, 1M a dimensionless time after contact) ° angle between shock-strut axis and vertical ™ shock-strut effectiveness, S#;-——— te lnnding-gear « timo interval in numerical integration procedures » cooflicient of friction between tire and runway 4 coofficient of friction for upper bearing (attached to inner cylinder) 1 coefficient of friction for lower bearing (attached to outer cylinder) ? mass density of hydraulic fluid @ angular acceleration of wheel Axes: 2 vertical axis, positive downward 2 horizontal axis, positive rearward Subscripts: ° at instant of initial contact , at instant of initial shock-strut deflection au at instant of wheel spin-up max maximum value Notation: 1Ol absolute value of ( ) Or estimated value of (-) ‘The use of dots over symbols indicates differentiation with respect to time ¢ or 7. Prime marks indicate differentiation with respect to dimensionless time 8. 1081 MECHANICS OF LANDING GEAR In view of the fact that landing-gear performance appears to be relatively unaffected by the elastic deformations of the airplane structure (see, for example, refs. 1 and 2) par- ticularly since in many cases the main gears are located fairly close to the nodal points of the fundamental bending mode of the wing, that part of the airplane which acts on a given gear can generally be considered as a rigid mass. As a result, Ianding-gear drop tests aro often conducted in a jig where the mass of the airplane is represented by a concentrated weight. In particular instances, however, such as in th caso of airplanes having large concentrated masses disposed in an outboard. position in the wings, especially airplanes equipped with bieycle landing gear, consideration of the interaction between the deformation of the airplane structure and the landing gear may be necessary to repre- sent the system adequately. Since the present report is concemed primarily with the mechanics of the landing gear, it is assumed in the analysis that the landing gear is attached to a rigid mass which has freedom only in vertical translation. The gear is assumed infinitely rigid in bending. The combination of airplane and landing gear considered therefore constitutes a system having two degrees of freedom (see fig. 1(e)) as defined by tho vertical displacement of the upper mass and the vertical displacement of the lower or unsprung mass, which is also the tire deflection. The strut stroke ¢ is determined by the difference botween the displacements 2 and 2 and, in the caso of inclined gears, by the angle y between the axis of thestrut and tho vertieal. For inclined gears, compression, ‘of the shock strut produces horizontal displacement of the axle 7, From consideration of the kinematics of the system In the analysis, external lift forces, corresponding to the aerodynamic lift, are assumed to act. on the system through- out the impeet. Tn addition to the vertical forces, axbitrary diag loads are considered to act between the tire and the ground. ‘The system treated in the anslysis may therefore be con~ sidered to represent either a Ianding-gear drop test in a jig where wing lift and drag loads are simulated, or the landing impact of a rigid airplane if rotational motions are neglected. Rotational freedom of the airplane, where significant, may be taken into account approximately by use of an appro- priate effective mass in the analysis, Figure 1 (b) shows a schematio representation of a typical oleo-paoumatic shock strut used in American practice. ‘The lower chamber of the strut contains hydraulic fluid and tho upper chamber contains air under pressure. ‘The outer eyl- inder of the strut, which is attached. to the upper mass, contains a perforated tube which supports a plate with a small orifice, through which the hydraulic fluid is forced to flow at high velocity os a result of the telescoping of the strut. ‘The hydraulic pressure drop across the orifice thus produced resists the closure of the strut, and the turbulence created provides ® powerful means of absorbing and dis- sipating « large part of the impact energy. In some struts the orifice area is constant; whereas, in other eases a metering it cam be seen that a= and =e sin p= (%—a)tan ¢.1082 cK “1 REPORT 1154—NATIONAL ADVISORY COMMOTTEE FOR AERONAUTICS *y lie @ (@) System with two degrees of freedom. a * (4-29) for = am A —, () Schomatio representation of shook strut. Fravax Dynamical eystem considered in analysis ppin or rod is used to control the size of the orifice and govern the performance of the strut. ‘The compression of the strut produces an increase in the ‘ir pressure which also resiats the closure of the strut. In figure 1 (b) pa represents the oil pressure in the lower chamber and p, represents the air pressure in the upper chamber. In addition to the hydraulic resistance and sir-pressure forces, internal beating friction also contributes forees which ‘can appreciably affect the behavior of the strut. ‘The forees created within the strut impart an acceleration to the upper mass and also produce an acceleration of the lower mass and a deflection of the tire. Figure 1 (c) shows the balance of forces and reactions for the wheel, the inner inder, and the outer eylinder. It is clear that the strut and the tire mutually influence the behavior of one another andmust be considered simultaneously in analyzing thosystem. From consideration of the pressures acting in the shock strut it can be readily seen from figure 1 (b) that the total axial force due to hydraulic resistance, air compression, and bearing friction can be expressed by Fe=py(Ar—A,)+ Pade Ar) Pep Ey where A, internal cross-sectional area of inner cylinder ‘A; external cross-sectional area of inner eylinder ‘A, cross-sectional aren of metering pin or rod in plane of orificeANALYSIS OF LANDING-GBAR DEMAVIOR 1083 \ le Y \ NO CB \N va Y N Noma 1 ag} y ™ Z yy Y Y y Fores on ole der / oe ae Forees on ner eyinder © () Balance of forees and reactions for landing-gear components. Fravan 1.—Coneluded.1084 ‘Thia expression can also be written fs Fs=(pr—pa)(Ar—Ay) + Pade Fy (Pr—P.)Art Peat Fy =P R+F, a where Pr—Be pressure drop across the orifice Ay hydraulic area (,—A, for the strut shown in fig. 1) ‘A, pneumatic area (4s for the strut shown in fig. 1) In this report the terms (p4—2,)4y aud ped, are referred to as hydraulic foree Fy and pneumatic force F., respeo- tively. For the strut shown-ia figure 1, the hydreulic and pneumatic areas are related to the strut dimensions as previously noted. In the case of struts having difforent internal configurations, the hydraulic and pneumatic areas may bear somewhat differont relations to tho dimensions of the strut. In such cases, however, consideration of the pressures acting on the various components of the strut should permit these areas to be readily defined. Hydraulic force—The hydraulic resistance in the shock strut results from the pressure difference associated with the flow through the orifice. Ia a landing gear the orifice area is usually small enough in relation to the diameter of the strut so that the jet velocities and Reynolds uumbers are sufficiently lange that the flow is fully turbulent. As result the damping force varies as the square of the tele- scoping velocity rather than linearly with the velocity. Since tho hydraulic resistence is the major-component of the total shock-strut force, viscous damping cannot be reasonably assumed, even though such an assumption would greatly simplify the analysis, ‘The hydraulic resistance can be resdily derived by making use of the well-known equation for the discharge through ‘an orifice, namely, anced, where Q — volumetricrate of discharge Ge coefficient of discharge Ay netorifce area Bx hydraulic pressure in lower chamber Pa _aipressure in upper chamber mass density of hydraulic uid From considerations of continuity, the volumetric rate of discharge can also be expressed as the product of the tele- scoping velocity $ and the hydraulic area Ay Qn Ae Squating the preceding expressions for the discharge per- mits writing the following simple equation for the pressure drop across the orifice ‘Phe hydraulic resistance F, due to the telescoping of the steut is given by the product of the differential pressure | metering pin is used, can vary with strut stroko; hat REPORT 1154—NATIONAL ADVISORY COMMUTER FOR APRONAUTICS Pa—Ps and the area A, which is subjected to tho hydraulic pressure, as previously noted. ‘Thus ase HCA * ‘Equation (2) can be made applicable to both the compres- th ® sion and elongation strokes by introducing the factor to indicate the sign of the hydraulie resistence; thus = 1 eA FT AGAy ea) ‘The net orifice area A, may be either a constant or, when =A,(s), where A, is the area of tho opening in the orifice plate and A, is the area of the metering pin in the plane of the orifice. At the present time there appears to bo some tendeney to eliminate the metering pin and uso a con- stant orifice area, particularly for large airplanes, in which case Aw=A,. In the goneral case, the orifice dischargo coefficient might be expected to vary somewhat during an impact because of changes in the size and configuration of the net orifice aree, changes in the exit conditions on the downstream face of the orifice due to variations in tho amount of hydraulic fluid above the orifice plate, changes in the entry conditions due to variations in the length of the flow chamber upstream of the orifice, and because of variations in tho Reynolds number of the flow, so that, in general, j= Q4(, Fi) Although the individual offects of these factors on the dis- charge cocfiicionts for orifices in shock struts have not been evaluated, there is some experimental evidence to indicate appreciable variations of the discharge coofficent during impact, particularly in the eage of struts with motering pin It might be expected that such variations would be con- siderably smaller for gears having a constant orifice area. In order to evaluate the precision with which the orifico lischarge coefficient has to be known, a briof study is presented in a subsequont section which shows tho offect of the discharge coofficent on the calculated behavior of a Janding gear with a constant orifice area, under the assump- tion thet the discharge coefficient is constant during tho impact. ‘Phe foregoing discussion las been concerned primarily with the compression stroke of the shock strut. Most struts incorporate somé form of pressure-operated rebound check valve, sometimes called a snubber valve, which comes into action after the maximum stroke has been attained and closes off the main orifice as soon as tho strut begins to elongato, 90 that the fluid is forced to return to the lower chamber through small passages. ‘The action of the snubber valve introduces greatly increased hydraulic resistance to dissipate the energy stored in the strut in the form of air pressure and to prevent excessive rebound. ‘The product C4. to bo used in equation (2a) during the elongation stroke is generally uncertain. ‘Tho exact area A, during elongation is usually somewhat difficult to define from the geometry of the strut sinco in many cases the number of connecting passages varies with stroke and theleakage area around the piston may be of the seme order of magnitude as the area of the return passages. Furthermore, the magnitude of the orifice discharge coefficient, and even. possibly the nature of the resistance, are questionable duo to the foaming state of the returning fluid. Formately, the primary interest is in the compression process rather than the elongation process sinco the maximum lond always occurs before the maximum strut stroke is reached. Pnoumatio force—Tho air-pressure forco in the upper chamber is determined by the initial strut inflation pres- sure, tho area subjected to the air pressure (pneumatic aree), and’ the instantaneous compression ratio in accordance th the polytropie law for compression of gases, namely te pmm() Pa air pressure in upper chamber of shock strut Pao air pressure in upper chamber for fully extended strut 2 air volume of shock strut ty air volume for fully extended strut Sinco tho instantaneous air volume isioqual to the differonco, between the initial air volume and the product of the stroke and pnoumatic area Ay 2—Pa(s 245) ~The foree due to the air pressure i the pneumatic area: Fem tede(s es)” @ In the preceding equations, the effective polytropic coxponent n depends on the rate of compression and the rate of heat transfer from the air to the surrounding environment. Low rates of compression would be expected to result in values of n appronehing the isothermal value of 1.0; whereas higher values of n, limited by the adiabatic value of 1.4, would be expected for higher rates of compression. The ‘actual thermodynamic process is complicated by the violent, of the highly turbulent efflus.of hydraulic fluid and the air in the upper chember during impact. On the one hand, the dissipation of energy in tho production of turbu- lence generates heat; on the other hand, heat is absorbed by the acration and vaporization of the fluid. ‘The effect of this mixing phenomenon on the polytropie exponent or on the equivalent air volume is not clear. A limited amount of experimental data obtained in drop tests (rofs. 3 and 4), however, indicates that the effective polytropic exponent may bo in tho neighborhood of 1.1 for practical cases. A Driof study of the importance of the air-compression process and tho effects which different values of n may have on the calculated behavior of the landing gear is presented in a subsequent section. Internal friction force —In the literature on machine sign the wide range of conditions under which frictional resistance can occur between sliding surfaces is generally classified in three major categories, namely, friction between dry surfaces, friction between imperfectly lubricated surfaces, and friction between perfectly lubricated surfaces. In the ply the product of the pressure and ANALYSIS OF LANDING-GEAR BEHAVIOR 1085 case of dry friction, the resistanco depends on the physical characteristics of tho sliding surfaces, is essentially propor- tional to tho normal foree, and is approximately indopondent of the surface area. ‘The coefficiont of friction x, defined as the ratio of the frietional resistance to the normal force, is generally somewhat greater under conditions of rest (static friction) than under conditions of sliding (kinetic friction). Although the coefficient of kinetic friction generally de- creases slightly with increasing velocity, it is usually con- sidered, in first approximation, to be independent of velocity. Hf, on the other hand, the surfaces are completely separated by a-fiuid film of lubricant, perfect lubrication is said to exist. Under these conditions the resistance to relative motion dopends primarily on the magnitude of the relative velocity, the physical characteristics of the lubricant, the arca, and the film thickness, end is essentially independent of the normal force and the characteristics of the sliding surfaces. Perfect lubrication is rarely found in practice but is most likely under conditions of high velocity end relatively small normal pressure, where the shape of the sliding surfaces is conducive to the generation of fluid pressure by hydro- dynamic action. In most practical applications involving lubrication, a state of imperfect lubrication exists and the resistance phenomenon is intermediate between that of dry friction and perfect lubrication. In the case of landing-gear shock struts, the conditions under which internal friction is of concern usually involve relatively high normal pressures and relatively small.sliding velocities. Moreover, the usual types of hydraulic ‘fluid used in shock struts have rather poor lubricating properties, and the shape of the bearing surfaces is generally not con- ducive to the generation of hydrodynamic pressures. Tt would therefore appear that the lubrication of shock strut bearings is, at best, imperfect; in fact, the conditions appear to approach closely those for dry friction. In the present ‘analysis, therefore, itis assumed, in first approximation, that Ue internal friction between the bearings and the cylinder ‘alls follows laws similar to those for dry friction; that is, the friction force is given by the product of the normal fores and a suitebly chosen cosfficient of friction. ‘With these assumptions the internal friction forces pro- duced in the strut depend on the magnitude of the forees on tho axle, the inelination of tho gear, the spacing of the bear- ings, and the coefficient of friction between tho bearings and tho cylinder walls. Figure 1(c) schematically ilustrates the balance of forces acting on the various components of the Innding gear. The total axial friction in the shock strut is the sum of the friction forces contributed by each of the bearings: 7 Raq lAl+aleD F, axial friction foree m coefficient of friction for upper bearing (attached to inner cylinder) F, normal force on upper bearing (attached to inner cylinder) ts cocfficiont of friction for lower beating (attached to outer cylinder)1086 ih normal force on lower bearing (attached to outer eylinder) fe 1 ‘During the interval prior to the beginning of shock-strut motion the frietion forees depend on the coefficients of static friction; after the strit begins to telescope the coefficients of Kinetic friction apply. ‘From considerations of the balance of moments it can be seen from figure 1(c) that, factor to indicate sign of frietion force AeFe, te and so that ; [Fed [ounotst +m] “ where: Py, sin ¢—Fy, 008 ¢ (4a) and . Fyq_ force norma) to strut applied at axle , Fy, vertial force applied at axle horizontal force applied at axle @ angle between strut axis and vertical 1b, axial distance between upper and lower bearings, for fully extended strut distance between axle end lower bearing (attached + to outer cylinder), for fully extended strut, The quantities Fig Fre and Fa, aro forees applied at the axle and differ from ‘the ground reactions by amounts equal to the inertia forces corresponding to the respective accelera- tion component of the lower mass. Since the inner cylinder generally represents only a relatively small fraction of the ower mass, the lower mass may reasonably be assumed to be concentrated at the axle, With this assumption, the rela- tionships between the forces at the axle and the forces at the ‘ground are given by : Pro(Frete B-™) ‘The normal force at the axle can therefore be expressed in terms of the ground reactions and the component accclera- tions of the lower mass by 7) Fao (Fag " m Byoo (Feet lt sin oe) oe» shee Fy, “vertical foree applied to tire at ground ‘x, horizontal force applied to tire at ground We effective mass below shock strut, assumed concentrated e . abaxle 4 horizontal acceleration of axle 2a__vertical acceleration of a In the case of an inclined landing gear having infinite stift- ness in bending, the horizontal displacement of the lower mass Z,isrelated to the vertical displacements of the upperandlower REPORT 1154—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS ‘massés by tho kinematic relationship 2=(2.—a)lan ‘g, 08 previously noted. Double differentiation of this relation- ship gives 2, (2,—z,) tan», Substitution of this express into equation (4b) give’ Fyg=Fr, sin o—Fe,c0s o+ 2? i, sin p—Wasin e (40) In equation (4c) the quantity 2, sin y represents the ac~ celeration of the lower mass normal to the strut axis when the gear is rigid in bending. In tho caso of a gear flexible in bending, the normal accsleration of the lower mass is not completely determined by the vertical acceleration of the upper mass and the angle of inclination of the gear. If it should be necessary to take into account, in particular cases, the effeots of gear Nexiblity on the relationship between the normal force on the axle and the ground reactions, the quan- tity 2, sin g in equation (4c) may be replaced by estimated. values of the actual normal acceleration of the lower mass as determined from consideration of the bending response of the gear to the applied forces normal to the gear axis. The effects of gear flexibility are not considered in moro detail in the present analysis. Figure 2 (a) shows dynamic foree-deflection charactoris- tics for.a 27-inch smooth-contour (type 1) tire inflated (o 32 pounds per square inch. ‘These characteristics were doter- mined from time-history measurements of vertical ground foree and tire deflection in Innding-gear drop tests with nonrotating wheel at several vertical velocities. As ean bo seen, the tire compresses along one curve and unloads along another, the hysteresis loop indicating appreciable energy dissipation in tho tire. ‘There is some question as to whother -the amount of hysteresis would be as great if the tire wore rotating, as in a landing with forward speed. ‘The forco- deflection curve for a velocity of 11.63 feet per second is for ‘© severe impact in which tire bottoming occurs and shows the sharp increaso in foreo with deflection subsequont to bottoming. In figure 2 (b) the’ same foree-deflection charactoristics are shown plotted on logarithmic coordinates. As can. be seon, the force exhibits an exponential variation with dofloc- tion. A systematized representation of the forco-dloflection relationship can therefore be obtained by means of simplo equations having the form Pramaran (3) ) where Fr, ~ vertical force, applied to tire at ground a vertical displacement of lower mass from position at initial contact. (radial detection of tie) ad overall diameter of tire tm, constants eorreoponding tothe various regimes of tho tire-deflection process combined constant, md”ANALYSIS OF LANDING-GEAR BEHAVIOR 14a? res| Lnwor-segment rexinaten oo Ke @ force on te, Fig ven 1087 (@) Uniform coordinates, - Fravar 2—Dynamie foreedflection charactorsties of tte, At may be noted from figure 2 that essentially the seme forco-defiection curve holds during compression for all impact ‘velocities, up to the occurrence of tire bottoming, and that in figure 2 (b) the slopes.of the curves in each of the severtl regimes of the tire-deflection process are also independent of velocity, except in the compression regime following tire bottoming. Figuro 2 elso shows simple approximations to the tire characteristics which were obtained by fitting straight-line segments (long-dashed lines) to the actual forco-deflection curves in figure 2 (a) for impacts at 8.86 and 11.63 fest per second, ‘These approximations, hereinafter referred to a8 Jinear-segment epproximations, are included in a study, presented in & subsequent section, to evaluate the degree of accuracy required for adequate representation of the tire characteristics. ‘The various representations of the tire characteristics considered and the pertinent constants for ‘each regime of tire deflection are shown in figure 3. EQUATIONS OF MOTION ‘Tho internal axial force F produced by the shock strut ‘was shown in a previous soction to be equal to the sum of the 3 Tee detection, 4,18 (©) Logarithmic coordinates hhydraulic, pneumatic, and frietion forces, as given by equation (1). Sines theso forces act along the axis of the strut, which may be inclined to the vertical by an angle 9, the verticel component of the axial shock-atrut foree is given by Fy cos ¢. Tho vortical component of the force normal to the shock strut is given by Fy, sin y. ‘These forces act. in conjunction with the lift force tnd weight to produeo an acceleration of the upper mass. The equation of motion for the uppermass is Fyos e+ Fy, sin ¢-+L—Wi ae a © ‘The vertical components of the axial and normal shock- strut forees also act, in conjunction with the weight of tho Jower mass, to produce « deformation of the tire and an scedleration of the lower mass. ‘The equation of motion for the lower mass is Ws Pecos e+ Fry sin 9+ We a Fr, (ed) where the vertical ground reaction Fy, is expressed as a.1088 REPORT 1154—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS: . oe . Oy 1 tt en Fg 12 gent veto: A+ (2A) Regime mr 6 OF Regime am © FO 76x10? 134 Wr @ esx 134 ~@ @ sox 29 + @ sis6xio 280 - © assoxic? aro S ‘max 90 @ tsaxe i @ ier us © es5xw 134 Ty ® exo 136 - oe : et erage . “stration Fy, o'( ZF) +0 Sppetinaon Fy 0'(ZF) +0 aaroxn} — bs-l0BxKF shri 8 Osseo! sxe © erat 0% -aioxto8 5 s i 2 ‘Ett exponent! Exponent, reyire D exe (to hystrss) Ur Sagan! opprsinaton ‘no tystaest) seo regime @ elena (a) Impact without tre bottoming, Vrg=8.80 feet per second. Vere oa on ie, i Tee deacon, 2, ft o (W) Impact with tire Bottoming, Vy,=11.63 foot per second ‘rowan 3,—Tire characteritios considered in solutions (logarithmic coordinates). function of the tire deflection 2, ‘The relationship between Fry, ond % hes been discussed in the previbus section on tire characteristics, By combining equations (6) and (7), the vertical ground foree can be waitten in terms of the inertia reactions of the upper and lower masses, the lift force, and the total weight. ‘The overall dynamic equilibrium is given by ® Conventional oleo-pneumatic shock struts are inflated to some finite pressure in the fully extended position. ‘Thus the strut does not begin to deflect in an impect until sufficient force is developed to overcome the initial preloading imposed by the air pressure end internal friction. Sines the strut is effectively rigid in compression, as well as in bending, prior to this instant, the system may be considered to have only one degreo of freedom during the initial stage of the impact. ‘The equations of motion for the one-degree-of- freedom system are derived in order to permit determination of the initial conditions required for the analysis of the landing-gear behavior subsequent to the beginning of shock- strut deflection. Since 2 % during this first, phase of the impact, ‘equation (8) may be written as Pe, (2)=—E 2 WK) @ where For the general caso of an exponential relationship between vertical ground force and tire deflection, equation (6) applies and the equation of motion becomes Bieter WD 0)ANALYSIS OF LANDING-GEAR BEBAVIOR 1089 ‘The shock strut begins to telescope when the sum of the inertia, weight, and lift forees becomes equal to the vertical components of the axial and normal shock-strut forces, At this instant, Fo=F,,-+F,, and equation (6) can be written as (FotF,) 008 0+ Fr,, sin e+ Ke W—M em en Wit ay Fag initial air-pressuro pretond f0re0, Pye F,, static friction at instant t, At the instant é, e=0 and equation (4) becomes Fy=[Foa|Be (ata) where Kem[oote bbe] ‘and py and 42 are coefficients of static friction. Since the strut is assumed essentially rigid in compression (and also rigid in bending), there is no kinematic displace- ment of the lower mass in the horizontal direction up to the beginning of shock-strut deflection, so that z4=0 and equation (Ab) becomes Fay (Foy tt tM) sin pF, 008 @ any) Incorporating equations (11), (11b), and (9) into equation (11) gives, In equation (12) wherever tho +: sign appears, the plus signs apply when Fy,,>0 and the minus signs epply when Fy, <0, ‘From equation (10) the vertical displacement of the system fat tho instant t, is given in terms of the corresponding acceleration by 1 wey sem{2[wa—m)—Z s,}} (as) Tntegrating equation (10) and noting that «)=0 provides the relationship between the vertical velocity and the vertical displacement of the system at the beginning of shock-strut deflection aa Sper ar Weve, | In view of the fact that tho tire force-deflection curve is essentially linear for small deflections, it may be reasonably gsumed that r==1 for the purposo of determining the time after contact at which the strut begins to telescope. With this assumption f, can be determined from tho relationship de ae i Ly $F 2+ wae | : (4K, sin ¢—c08 ¢) (Kz W—Wi)— Fa (+ K.cosy-+sing) Fak, tines 0) a2) where the general expression for the variable # is obtained from equation (14) without the subscripts r. Performing the indicated integration gives tony) {sins Ca—Ke)sin-* o[a—K—"e" |} where ye tl —Kast ‘The computation of t, can be greatly simplified by uso of ‘tho following approximation which assumes a linear relation- ship between velocity and time: Bey tng (150) Equation (158) should be a faitly good approximation in ‘View of the relatively short, time interval between initial eon- tact and the beginning of shock-strut motion Eaquations (12), (13), and (14) permit the determination of the vertical acceleration, displacement, and velocity, re- spectively, of the system (uppor and lower masses) at tho1090 beginning of shock-strut deflection. Equation (15) or (15a) permits caloulation of the time interval between initial contact ‘and this instont. ‘These equations provide the initial conditions required for tho analysis of tho behavior of the landing gear as a system with two degrees of freedom after the shock strut begins to deflect. If drag loads are considered, tlie solution of equation (12) requires knowledge of the horizontal gfound force Fn,, at the instant z. Since the present analysis does not explicitly treat the determination’of drag loads, values of Fy,, have to be-estimated, cither from other analytical considerations, experimental data, or on the basis of experioncs. REPORT 1154—NATIONAL ADVISORY COMMITTEE FOR ARRONAUTICS, [MOTION SUBSEQUENT TO BEGINNING OF SHOCK-STRUT DEFLECTION Once the sum of the inertia, weight, and lift forees becomes sufficiently large to overcome the preloading forco in the shock strut due to initial air pressure and internal friction, the shock strut can deflect and the system becomes one having, twwo degrees of freedom. Incorporatirig the expressions for the hydraulic, pneumatic, and friction forces (eqs. (2a), (8), and (4)) into equation (6) permits the equation of motion for the upper mass to be written as follows: w atfingtio® + age ( where ‘and, since Fy,=Fy,(2,), equation (4e) becomes 25) + TH Pol[oartud Peas] cos eH KL — Wit Fy, sin 90 (16) Bye Fr(edsin Fr, cos o+ 28, sin pW, sin 9 where Fy,(z) is determined from the force-deflection characteristics of the tire. Fr,(2)=ma/,, as previously noted. For the usual type of pneumatic tire, Similarly, the equation of motion for the lower mass follows from equatioi (7): Tes ft eat e Ua GAY a. + pada (Sas, A) + pylFinl ato Pepe toe cos ot Fe @d~Frgsin ¢— Weed a ‘The overall dynamic equilibrium equation is still, of Zourse, as given by equation (8) Wis Any two of the preceding ‘equations (eqs. (16), (17), and ()) are sufficient to describe tho behavior of the landing gear subsequent to the beginning of shock-strut motion. ‘These equations may be used to calculate the behavior of given landing-gear configuration or to develop orifice and metering-pin characteristics required to produce a specified behavior for given impact conditions. ‘They may also be ‘used as 2 basis for the calculation of dynamic loads in flexible airplane structures either by (a) determining the landing- gear forcing funetion-under the assumption that the upper ‘mass is a rigid body and then using this forcing function to calculate the response of the elastic system or (b) combining tho preceding equations with the equations representing the additional degrees of freedom of the structuro; the simul- taneous solution of the equations for such system would then take into account the interaction between the deforma- tion of the structure and the landing gear. aD Bet W (Ke—1)+ Fr, (29) SOLUTION OF EQUATIONS OF MOTION In tho genoral case the analysis of a landing gear involves, tho solution of the equations of motion given in the section entitled “Motion Subsequent to the Beginning of Shock- Strut Deflection,” with the initial conditions taken as the conditions of motion at the beginning of shock-strut doflee- tion, as determined in accordance with the initial impact conditions and the equations given in the section entitled “Motion Prior to Shock-Strut Deflection.” In view of the fact that the equations of motion for the landing gear subsequent. to the beginning of shock-strut doflection are highly nonlinear, anelytiea! solution of these equations does not appear feasible. In the present report, therefore, finite-difference methods are resorted to for the step-by-step integration of the equations of motion. Al-ANALYSIS OF LANDING-GEAR BEHAVIOR though such nusnerical mothods lack the generality of ana- lytical solutions and are especially time consuming if the calculations are carried out manually, the increasing availa~ bility of automatic calculating machines largely overcomes these objections. ‘Most of the solutions presented in this report were obtained with a procedure, hereinafter referred to as the “linear pro- cedure,” which assumes changes in tho motion variables to bo linear over finite time intervals. A few of the solutions presented were obtained witha procedure, hereinafter referred to as the “quadratic procedure,” which assumes a quadratic variation of displacement with time for successive intervals. Tho generalized solutions for the simplified equations dis- cussed in a subsequent section were obtained by means of the Runge-Kutta procedure. The application of these procedures is described in detail in appendix A. In order to obtain solutions for particular cases, itis, of course, necessary to have, in addition to information regard- ing the physical cheracteristies of the landing gear, some knowledgo of the forco-deflection characteristics of the tire. If extensive data regarding the dynamic tire character- fsties, such as shown in figures 2 and 3, are available, an accurate solution can be obtained which takes into account the various breaks in the force-deflection curves (logarithmic coordinates), as well as the effects of hysteresis. In view of the fact that the constants m’ and r have the same values, throughout practically the entire tire compression process regardless of the initial impact velocity or the maximum ond attained, these values of m’ and r, as determined from the force-deflection curves, can be used in the calculation of the motion subsequent to the beginning of shock-strut deflec~ tion until tho first break in the foree-deflection curve is reached prior to the attainment of tho maximum force. If tho conditions for the calculations aro the same as those for which force-deflection curves are available, the values of ‘m’ and r for each of the several regimes subsequent to the first breale can also be determined directly from the force- deflection curves. In general, however, the conditions will not be the same and interpolation will be necessary to estimate the values of m’ for the subsoquent regimes, Such interpolation is facilitated, particularly after the maxi- mum foree-deflection point has beon calculated, by the fact that each subsequent regime hes a fixed value of r, regardless of the initial impact conditions. ‘Tho use of the tire-deflection charaeteristics in the caleula- tions is greatly simplified if hysteresis is neglected since, the values of m’ and r which apply prior to the first break in the foree-deflection curves are then used throughout the entire caloulation, except in the case of severe impacts where tire bottoming occurs, in which case new values of mm’ and r are ‘employed in the tive-bottoming regime. A similar situation exists with respect to the constants a’ and 6 when the linear approximations which noglect hysteresis are used. ‘These simplifieations would normally be employed when only the 1091 tire manufacturer's statie or so-called impact load-defleetion data aro available, as is usually the case. Although the present analysis permits taking into aecount the offects of wheel spin-up drag loads on the behavior of the Tending gear, the determination of the drag-lond time history isnot treated explicitly. ‘Thus, if it is desiréd to consider the ffeots of tho drag load on the goar behavior, such as in tho case of a drop test in which drag loads are simulated by reverse wheel rotation or in a landing with forward speed, it is necessary to estiniate tho drag load, either by means of other analytical considerations or by recourse to experimental date, As a first approximation the instantaneous drag foree may be assumed to be equal to the vertical ground reaction multiplied by a suitable- coefficient of friction m; that is, Fu,= Fr, up to the instant when the wheel stops skidding, aftér which the drag force may be assumed equal to zero. (Phe current ground-loads requirements specify a skidding coofiicient of friction u==0.55; limited experimental evidence, on the othor hand, indicates that » may be as high as 0.7 or ‘as low as 0.4.) In some cases experimental date indicate that representation of the drag-load time history can be simplified even further by assuming a linear variation of the drag force with time during the period of wheel skidding. ‘Tho instant at which the wheel stops skidding can be estimated from the simple impulse-momentum relationship Tz. polar moment of inert of wheel assembly about axle 1, initial horizontal velocity a" radius of doficoted tire tee _ time of wheal spin-up ‘When the drag foree is expressed in terms of the vertical force, the value of the integral Fe, dt ean be determined as the step-by-step calculations proceed and the drag-foree term climinated from the equations of motion after the re quired value of the integral at the instant of spin-up is reached. EVALUATION OF ANALYSIS BY COMPARISON OF CALCULATED RESULTS WITH EXPERIMENTAL DATA In order to evaluate the applicability of the foregoing analytical treatment to actual landing gears, tests were conducted in the Langley impact basin with a conventional oleo-pneumatic landing gear originally designed for a small tary training airplane, A description of the test specimen * ‘and apparatus used is given in appendix B. In this section calculated results are compared with ex: perimental data for a normal impact and a severo impact with tire bottoming. ‘The vertical velocities at the instant of ground contact used in the calculations correspond to1092 the vertical velocities measured in the tests. Equations (22), (43), (14), and (15a) wero used to ealeulate the values of the variables at the instant of initial shock-strut deflection. ‘Numerical integration of equations (16) and (17) provided the calculated results for the two-degree-of-freedom system ‘subsequent to the beginning of shock-strut deflection. In these calculations the discharge coeicent for the orifice and the polytropic exponent for the air-compression process were assumed to have constant values throughout the impact. Consideration of the shape of the orifice and examination of data for rounded approach orifices in pipes suggested ‘value of G, equal to 0.9. Evaluation of data for other landing gears indicated that the air-compression process could be represented fairly well by use of an average value of the effective polytropic exponent n=1.12. In view of the fact that tho landing gear was mounted in a vertical position and drag loads were absent in the tests, friction forces in the shock strut were assumed to be negligiblo in the caleulstions, Since the weight was folly balanced by lift forces in the tests, the lift factor Ky, was taken equal to 1.0. ‘The appropriate exact tire characteristics (eee ig. 3) were used for each case. Figure 4 presents a comparison of caloulated results with: experimental data for an impact without tire bottoming at 1a Vertical velocity of 8.86 feet per second at the instant of ground contact. The exact dynamic force-deflection charac- teristics of the tire, including hysteresis, were used in the caloulations. ‘These tire characteristics are shown by the solid lines in figure 2 (a) and values for the tire constants sm and r are given in figure 3 (0). Calculated time histories of the total foree on, the upper mass and the acceleration of the lower mass are compared with experimental data in figure 4 (a). Similar comparisons for the upper-mass displacement, upper-mass velocity, lower- mass displacement, strut stroke, and strut telescoping velocity ate presented in figure 4 (b). As can be seen, the agreement botween the calculated and experimental results reasonably good throughout most of the time history. Some of the minor discrepancies during the later stages of the impact appear to be due to errors in measurement since the deviations between the calculated and experimental ‘upper-mass accelerations (as represented by the force on the upper mass) are incompatible with those for the upper- mass displacements, whereas the calculated upper-mass dis- placements are necessarily directly compatible with the calculated upper-mass accelerations. The maximum value ‘of the experimental acceleration of the lower mass may be somewhat high because of overshoot of the accelerometer. In addition to the total force on the upper mass, figure 4 (@) presents calenlated timo histories of the hydraulic and pneumatic components of the shock-strut force, as determined from equations (2) and (9), respectively. It can be seen that throughout most of the impact the force developed in the shock strut arises primarily from the hy- draulic resistance of the orifice. ‘Toward the end of the impact, however, because of the decreased telescoping velocities and feirly largo strokes which correspond to high REPORT 1154—NATIONAL ADVISORY COMAUTTEE FOR AERONAUTICS compression ratios, the airpressure foree becomes larger than the hydraulic foree. Figure 5 presents # comparison of ealoulated and experi mental results for a sovore impact (Vy,=11.63 ft por s0¢) in which tire bottoming occurred. ‘The tire foreo-deflection characteristics used in the calculations are shown by the solid lines in figure 3 (b). Region (1) of the tire forco- Aeflection curve has the same values of the tire constants ‘m! and r as for the case previously discussed. Following the occurrence of tire bottoming, however, different values of m’ and r apply. ‘Theso values aro given in figure 3 (b). It can be seen from figure 5 that the agreement between tho ealoulated end experimental results for this case is similar to that for the comparison previously presented. ‘Tho calculated instant of tire bottoming is indicated in figure 5. When tire bottoming occurs, the greatly increased stiffness of the tire causes a marked increaso in tho shook- strut telescoping velocity, as is shown in tho right-hand portion of figure 5 (b). Since the strut is suddenly foreed to absorb energy at a much higher rate, an abrupt increase in the hydreulic resistance takes place. ‘Tho further increase in shook-strut foreo immediately following the occurrence of tire bottoming is evident from the left-hand portion of figure 5 (a). The sudden inereaso in lowor-mass acceleration at the instant of tire bottoming ean also be seen. In this sovere impact tho hydraulic resistance of the orifice represents an even greater proportion of the total shock- strut force then was indicated by the calculated results for an initial vertical velocity of 8.86 feot per second proviously discussed. ‘Tho foregoing comparisons indicate that the analytical ‘tweatment presented, in conjunetion with reasonably straight= forward assumptions regerding the parameters involved in the equations, provides a fairly accurate representation of the behavior of a conventional oleo-pneumatic landing goar. PARAMETER STUDIES In the previous section éomparisons of calculated results swith experimental data showed that tho equations which have been developed provide a fairly good representation of the behavior of the landing gear for the impact conditions considered. In view of the fact that the equations aro somewhat complicated and require numerical values for several parameters such as the tiro force-deflection constants am and r, the orifice dischargo coefficient Q,, and the poly- tropic exponent n, which may not be readily or accurately mown in the case of practical engineering problems, it appears desirable (a) to dotermino the relative accuracy with which these various parameters havo to be known and (b) to investigate the extent to which tho equations can bo simplified and still yield useful results. In order to accom- plish these objectives, calculations have been made lo evaluate the effect of simplifying the foree-deflection charac- teristics of the tire, as well as to determine tho effects which different values of the orifice discharge coefficient and the cffective polytropic exponent have on the calculated behavior. ‘The results of these calculations are discussed in the present section. ‘The question of simplification of the equations of motion is considered in more deteil in a subsequent section,ANALYSIS OF LANDING-GHAR BEHAVIOR 1093 7108 Belo Lonar-mass acoseraton locker, 18 Tina er cone, 80 (@) ‘ime histories of forees on upper mess and Jower-moass acceleration, ‘Froune Comparisons between caloilated refulta and experimental data for normal impact; solution with exact exponential tire characteristics, Vyy=8.86 feet per second; C4m0.0; n= 1.12. tor 1, : br 6 Uppers contin she, ft 2 Bi r7 ri 75 Zo °° oF 05 Te 76 20 Tone ter conte), se () Time histories of landing gear velocities and dieplacamonts. Fiouns 4 Concluded,1094 REPORT 1154—NASIONAL ADVISORY COMMITTEE FOR AERONAUTICS 12x 10? “2 "0 ~9} © Eee 29 24 i : : stone 5 a4 id : 3 é i ° 3a 08 2 76 2 0 ro 05 ie 16 Zo Time after contact, see () Time histories of forces’on upper mass snd lower-mass acceleration Flavin 5.—Comparisons between calculated results and experimental data for impact with tire bottoming; solution with ‘esuct exponantial tire charactoristca. Vyjer11.63 fest per second; Cy=0.9; no 1.12, Sr 6 Skeut vei ots we 0808 Expeiantl a 0 oF Be 1 is 20 ° om a 12 i = Tine ler canto, sac () Time histories of landing-goar velocities and displacements. ‘rouse 5.—Coneluded.ANALYSIS OF LANDING-GEAR BEHAVIOR REPRESENTATION OF TIRE FORCE-DEFLECTION CHARACTERISTICS In onder to evaluate the degree of accuracy required for ‘adequate representation of the tire foreé-deflection charae- teristics, eompatisons are made of the calculated behavior of the landing gear for normal impacts and impacts with tire bottoming when the tire characteristics are represented in various ways. First, the force-deflection characteristics will be assumed to be exactly as shown by the solid-line curves in figure 2 (b), including the various breaks in the curve and the effects of hysteresis. ‘Theso characteristics fare referred to hereinafter as the exact exponential tire characteristics. Tho effects of simplifying the representa tion of the tire characteristics will then be investigated by considering (a) tho exponential characteristics without hysteresis; that is, tho tire will be assumed to deflect and unload along the same exponential curve, (b) the Tinear- segment approximations to the tire characteristics Cong- dashed lines), which also neglect hysteresis, and (c) errors introduced by neglecting the effects of tire bottoming in the caso of sovere impacts. ‘The calculated results presented in this study make use of the relationships between vertical foreo on tho tire and tire deflection, as shown in figures 3 (a) and 3 (b), Figure 6 presents « comparison of the calculated results for a normal impact at a vertical velocity of 8.86 feot per seeond, whereas fiuro 7 permits comparison of the solutions 3.07 ! -25| s & 8 i E g Tee choraterties considered: voc exponent! Exponent! (no hysteress) near segment (ro ystresis) Exponential gi sit) 2 oF 16 20 1095 for a sovere impact, involving tire bottoming, at a vertical ‘velocity of 11.63 feot per second.” In figures 6 and 7 the solid-line curves represent solutions of the lending-gear equations when the exect exponential relationships between foree and tire deflection are considered. Since these solu- tions were previously shown to be in fairly good agreement swith oxpetimental date (Gigs. 4 and 5), they are used as a _-basis for ovaluating the results obtained when tire hysteresis is neglected and the force-deflection characteristics are repre- sented by either simplified exponential or linear-segment relationships. As in the calculations previously described, the sotutions wore obtained in two parts. During tho first stage of the impact the shock strut was considered to be rigid until sufficient force was developed to dvercome the initial air- pressure force. The calculations for the landing-gear behav- ior subsequent to this instant were based on the equations which consider the gear to. have two degrees of freedom. ‘Time histories of the upper-inass acceleration calculated on tho basis of @ rigid shock strut are shown by the dotted curves in figures 6 and 7. ‘These solutions show tho greatest rate of ingronso of upper-mass acceleration possible with tho exponential tire force-deflection characteristics con- sidered. Comparison of these solutions with those for the two-degree-of-freedom system indientes the effect of the shook strut in attenuating the severity of the impact. % Be 08 20 Tine afte cone, see (j) Time historic of uppor-mass acceleration and lower-mars acceleration. Frovne 6—Effect of tre characterises on caloulated landing-gear behavior in normal impact. V+j8.86 feet per second; C4=0.0; n=1.12,Stat ste, 2 v0 8 2 ope ose s g4 3 2 ~ | Te doris coins —— Bact exponentiot ‘N Se Sea oe ess) TIT Ep fo mata SS ° 04 08 a2 Fr 20k 0) (24 08 32 6 ‘Tine after conor, see () Time histories of landing-gear displacements and velocitiee, ‘rove 6.—Continued, 4 = ‘Tee checcttsties conser: | oct exponential = Exponential ro nyse) Gwar segment (ro hysteresis) © 3a 88 ie is 20 oF 8 2 15 Zo Tine after contort, s4¢ (Time histories of shock-strut stroke and velocity ‘Froune 6,—Coneluded.ANALYSIS OF LANDING-GEAR BEHAVIOR 1097 “12 +0 i elt Sahih Sr g 1 3-6 7 g : i 3 | 2 + Tea choctstes ensieed: 0 ‘oct epee HS Boone 9) (ro boning, no hysteresis) 2 ine sagen (ch boning, nyse “5 — Uaesrsegnen! 7 (co btaring, no ysl) expres i a) a . eo (Oita oan oe meristereeveriielz01 © Ofer Od) Tina oer contact, sac gee eee e0) (4) Time historias of upper-mase acceleration and lowor-mass acceleration, ‘Froune 7—Lifect of tro characteristics on calculated landing-goar behavior for impact with tire bottoming. Vrp=11.68 feet per second; 0y=0.9; n=1.12. ‘Tie chovctrsics conidereds sot exponent = Eiponentt (00 toning, no hysteresis —-— Lineor-sognent {ath botoming, co hysteresis) ~4) LUreor-segest (fo botoming, no hysteresis) 6 Oe oe Tine tee conc, 6 (©) Time histories of landing-gear digplacements and velocities Fiovmn 7—Continued, 210065001098 REPORT 1154—NATIONAL ADVISORY COMAUTTEE FOR AERONAUTICS 5 . ae s| 5 4] 4 = £ a 2 $3 gs F : & 4 ‘Tire cherocteristics considered: 7 Exoét exponeetioh —— Someta ; (ro bettening, fo hytes) tre bron, no hte) a ° 24 38 2 75 Zo 0 Dae 08) 3 16 20 “Time efter canoe, tee (6) Time histories of shockatrut stroke and velocity. . ‘Froume 7. ‘Normal impact—In the caso of tho normal impact at a vertical velocity of 8.86 feot per second, figure 6 shows that, the solution obtained with the exponential force-defiection variation which negleots hysteresis and the solution with the Jinear-segment approximation to the tire characteristics are in fairly good agreement with the results of the caloulation based on the exponential representation of the exact tire characteristics. ‘The greatest differences between the solu- tions aro evident in the time histories of upper-mass and lower-inass accelerations; considerably smaller differences are obtained for the lower-order derivatives, as might be ox- pected, With regard to the upper-mass acceleration, the threo solutions are in very good agreement during the early stages of the impact, In the caso of the simplified exponen- tial characteristics, neglect of the decrensed slope of the- force-defiection curve between the first breal: and the maxi- ‘mum (regime © in fig. 3 (a)) resulted in the calculation of a somewhat higher value of the maximum upper-mass acceler- ation than was obtained with the exact tire characteristics. For the simplified exponential and linear-segment character~ istice, neglect: of hysteresis resulted in the calculation. of somewhat excessive values of upper-mass acceleration sub- sequent to the attainment of the maximum vertical load, It is of interest to note that the calculated results for the exponential and linear-segment characteristics without hys- teresis were generally in quite good agreement with each ‘other throughout the entire duration of the impact, although the assumption of linear-segment tire force-deflection char- acteristics did result in somewhat excessive values for the maximum lower-mass acceleration. On the whole, the simplified tire force-deflection characteristics considered per- mit calculated results to be obtained which represent the behavior of the landing gear in normal impacts fairly well. Concluded. Impact with tire bottoming —In the caso of tho severe impact at a vortical velocity of 11.68 feot per second, the effects of tire bottoming on the upper-mass acceleration, the ower-mass acccleration, and tho strut telescoping velocity are clearly indicated in figure 7 by the calculated results based on the exact tire characteristics. As can be soon, the limear-segment approximation to the tire deflection charactor isties which takes into account the effects of tire bottoming resulted in e reasonably good representation of the landing- gear behavior throughout most of the time history. On tho other hand, as might be expected, the calculations which noglected the effects of bottoming on the tire force-deflection characteristics did not reveal the marked increase in tho ‘upper-mass acceleration duo to the increased stiffness of tho tro subsequent to the occurrence of bottoming. It is also noted that the discrepancies in the calculated upper-mass acceleration duo to neglect of hysteresis in the later stages of the impact are more pronounced in this case than in the impact without tire bottoming previously considered, again as might be expected. REPECT OF ORICE DISCHARGE COEPFICHENT In view of the fact that there is very littlo information aveilablo rogarding the magnitude of dischargo cooffcionts {for orifices in landing gears, it appears desirable to evaluate the affect which differences in the magnitude of the orifico coefficient can have on the calculated results. Figure 8 presents comparisons of calculated results for a range of values of the orifice discharge coefficient Cy betiween 1.0 and 0.7. ‘The four solutions presented are for the’same set of initial conditions as the normal impact without tire bottoming previously considered and aro based on the exponential tire force-deflection characteristics which noglect hysteresis.ANALYSIS OF LANDING-GEAR BEHAVIOR 1099 C ° Ba 85 a a Ba 5 1520 : Tina ater conto, see (@) ‘Time histories of upper-mass acceleration and lower-mass acceleration. Frovnn 8—Bifect of orice discharge cooficint on landing-gear behavior; calculations with exponential tie characteristics without hysteresis. Vr,=8.86 feet per second; n=l 12. 10 ‘oper moss Upper mass 4 7 : iy ic . 3 or 2| = 1 3) = ° a Time ofler contot, eae (b) Time histories of landing-gear displacements and veloitis. Fravan 8—Continued,1100 Stat stroke, ° Baas ie is 20 REPORT 115{—NATIONAL ADVISORY COMMITTEE FOR ABRONAUTICS @ ° (oa eas ce ale 16 Time ater contoct, see (© Timo histories of shock-strut stroke and volosty. . Fiouma 8—Coneluded. ‘These calculations show that a decrease in the orifice dis- charge coefficient results in an approximately proportional increase in the upper-mass acceleration. ‘This vari- ation is to be expected since the smaller coefficients cor respond to reduced effective orifico areas which result in greater shock-strut forces due to increased hydraulic resistance. As a result of the increased shock-strut force acting downward on the lower mass, the maximum upward acceleration of the lower mass is reduced with decreasing values of the discharge coefficient. ‘The increase in shock strut force with decreasing discharge cocfficient also results {in a decrease in the strut stroke and telescoping velocity but an inerease in the Jower-mass velocity and displacement, fas might be expected. However, since the increases in lower-mass displacement and velocity are smaller than the decreases in strut stroke and telescoping velocity, the upper- mass displacement and velocity are reduced with decreasing orifice discharge coefficient. ‘These comparisons show that the magnitude of the orifice coefficient has an important offect on the behavior of the landing gear and indicates thet a fairly accurate determi nation of tho numerical value of this parameter is necessary to obtain good results. Since the nature of the air-compression process in a shock strut is not well-defined and different investigators have assumed values for the polytropic exponent ranging any~ where between the extremes of 14 (adiabatic) and 1.0 (sothermal), it appeared desirable to evaluate the im- portance of the air-compression process and to determine the extent to which different values of tho polytropic exponent can influence the ealculated results. Consequently, solutions hhave been obtained for three different values of ‘the poly- tropic exponent; namely, n=1.3, 1.12, and 0. ‘Tho value n=1.3 corresponds to a very rapid compression in which an adiabatic process is almost attained. ‘The ‘value n=1.12 corresponds to a relatively slow compression in which the process is virtually isothermal. ‘The value to=0 is completely fictitious since it implica constant air pressure within the strut throughout the impact. ‘Tho assumption n=0 has been considered since it makes one of the terms in the equations of motion a constant and permits simplification of the calculations. ‘The three solutions presented are for the samo set of initial conditions as the normal impact without tire bottoming previously con- sidered and are based on the exponential tire force-deflection sharacteristies which neglect hysteresis. ‘Figure 9 shows that the sir pressure contributes only a relatively small portion of the total shock-strut force through ‘out most of the impact since tho compression ratio is rela tively smell until the later stages of the impact. ‘Toward the end of the impact, however, the ait-pressure force Decomes a large part of the total force sinco the compression ratio becomes large, whereas the hydraulic resistance de~ creases rapidly as the strut telescoping velocity is reduced to zero.ANALYSIS OF LANDING-GBAR BEHAVIOR 1101 7xI0> 710, $ | * 6 = 2 i He 2 3 53 Bo i i 2 i o 4 ° D4 “08 ad 16 4¢ 04 08 2 dé Tine after cortct, soe (a) Timo histories of upper-massforee and lower-mase accolerntion. -Btlect of polytrople exponent; caloulations with exponential tire characteristics without hysteresis, ‘Vrp"8.86 foot per second; Cs=0.9. Upper mass eee eS ere ete er 2 ero ‘Tena ofter conta, se () Time histories of landing-gear displacements and velocities. ‘Fioumn 9.—Continued. 8 ate eas1102 oa as es ° ‘Tne after conte, see ees ea ata eget 20) (©) Time histories of shoskestrut stroke and velocity. roves 9.—Concluded. As o result, the calculations show that the magnitude of the polytropic exponent has only a very small effect on the behavior of the landing gear throughout most of the impact. For the practical range of polytropic exponents, variations in the air-eompression process result in only minor differ- ences in landing-gear behavior, oven during the very Intest stages of the impact. ‘The assumption of constant eir pressure in the strut throughout the impact (n=0), however, oes lead to the calculation of excessive values of stroke and of the time to reach the maximum stroke. ‘The time history ‘of the shockstrut force calculated on the basis of this assumption is, on the other hand, in quite good agreement ‘with the results for the practical range of air-compression processes. On the whole it appears that the behavior of the landing gear is relatively insensitive to variations in the air- compression process. ‘The foregoing results suggest thet, in many eases, fairly reasonable approximations for the landing- gear force-time variation might be obtained even if the air- Pressure term in the equations of motion were completely neglected. SIMPLIFICATION OF EQUATIONS OF MOTION ‘The preceding studies have indicated that variations in the tire forco-deflection characteristics and in the air-compression ‘process individually have only aelatively minor effect on the calculated behavior of the landing gear. ‘Theso results sug- gest that tho equations of motion for the landing gear might be simplified by completely neglecting the internal sir pressure forees in tho shock strut and by considering the tire force-deflection characteristies to be linear. With these as- sumptions, the equations of motion for the upper mass, lower mass, and complote system (eqs. (16), (17), and (8)) oan be written as follows for the case where the wing lift is equal to the weight and the internal friction is neglected: Bhat Ast Ti=0 Wes Par AG a) tastb—W as) 5M er tarrtino Bat BH stantb—0 where ant pat Ta] 2(CuAn)Fo08 and @ slope of linear approximation to tire foree-deflection characteristics 5 value of foree corresponding to zero tire deflection, as determined from the linear-segment approximation to tho tire foreo-doflection characteristics 2 ‘The motion variables at tho beginning of shock-strut de- Meection ean be readily determined in « manner similar to that employed in the more general treatment proviously dis- cussed. For the simplified equations the variables at the instant t, are given by ame oP ao) sniper In. most cases the term 4% 2, is small in comparison with és? so that é,=ANALYSIS OP LANDING-GEAR BEHAVIOR 1108 39 +9 : oat = @ > g y 2 i § ; 3 i i g7'5] = g t g B £ E i i | | r| ——— Sdiution of figure 4 -t | ER se Smeied goer, M20 j ° 04 “08 12 16 20. 5 OF “08 ae Ie 20 Tin ole conte, se (8) Time histories of uppor-mass acceleration and lower-toxss acceleration. ‘Froun 10.—Bvaluation of caloulated results for slmplifed systoms. Vr,~8.88 fect per second; Cy ‘Phe values determined from equations (19) are used as initial conditions in the solution of equations (18). ‘The fact that the lower mass is a relatively small fraction of the total mass suggests that the eystem might bo simplified ‘even further without greatly modifying the calculated results by assuming the lower mass to be equal to zero. With this assumption f4=0 and the initial values of the variables in ‘equations (18) correspond to the conditions at initial contact. In order to evaluate the applicability of these simplifica- tions, the behavior of the Innding gear has been calculated in accordanco with equations (18) for an impact with an initial vertical velocity of 8.86 feet per second. A similar calcula tion has been made with the essumption We~0. These results are compared in figure 10 with the more exact solu- tions previously presented in figure 4, which include con~ sideration of the air-compression springing and the exact exponential tire characteriaties. A time history of the lower- ‘mass acceleration is not presented for the case where Ws is assumed equal to zero sinco tho values of 24/g have no significanco in this ease. Figure 10 shows that the two simplified solutions are in quite good agreement with each other, as might be expected, ‘and aro also in fairly good agreement with the more exact, results. Neglecting the air-pressure forces and assuming a linear tire force-deflection variation resulted in the calcula tion of slightly lower values for tho maximum. upper-mass acceleration and somewhat higher values for the maximum stroke than were obtained with the more exact equations. ‘Tho offoct of neglecting the lower mass was primarily to reduce the lowor-mass displacement (tire deflection), as a zegult of the elimination of the lowar-mass inertia reaction. On tho whole, it eppoars that the assumptions considered permit appreciable simplification of the equations of motion ‘without greatly impairing the validity ofthe ealeulatedresults. CCHNERALIZED TREATASENT Equations and solutions —By writing the simplified equa~ tious of motion in terms of dimonsionless variables, gouoral- zed solutions can be obtained for a wide rango of landiog-gear and impact parameters which may be useful in pre- liminary design. If Wy is taken equal to zero and it is furthor assumed that the tire forcedeflection curve is represented by a single straight line through the origin (6=0 throughout the impact), eduatious (18) reduce to Wiss at—agee Ft AG aa0 Alér-#)—a2=0 @0) Ws, Pataz=0 gat whore, 4, and o are constants, as previously defined, and any to of the foregoing equations are sufficient to describe completely thebehaviorof thosystem. With this representa tion of the system, the shock strut begins to deflect atthe ia- stant of initial contact. (j=0). ‘Thus, the initial conditions for equations (20) are the initial impact conditions; namely, and Jymigmde1104 ——— sation of figure 4» Soi pte, 2131 = Sirti Stam, #6 =O oo Be “Time ele conto! 560 (®) Tle histories of landing-gear displacements and velocities. raven 10,—Continued. 6 soutien of tigre 4 Sinpied system, i= 13 1b == Seelifed Syston, Me +0 a) 2a 8 “Tin te ore, ee i 7% (©) Time histories of shock-strut stroke and velocity. Figuan 10,—Consluded. @ANALYSIS OF LANDING-GEAR BEHAVIOR As can be seen, with the equations ia this form, the solution depends on five parameters, namely, %, A, @ and the initial conditions iy and %. However, since 2=0 in all cnses, the number of variable parameters is reduced to four. Jn view of the fact that these parameters aro independent of one another and each may take on a large range of values, & freut many solutions and a large number of graphs would be required to cover the entire range of londiog-gear end impact parameters with the equations in the form of equations (20). ‘Tho oumber of independent parameters which heve to be considered may be greatly decreased by the introduction of generalized dimensiooless variables and the corresponding transformations of equations (20). In this ease, goneralized variables can be obtained which permit transformation of the ‘equations of motion to a form which does not involve any constants, With the equations iu this form, there is ouly one variable parameter, namely, tho initial velocity param eter, To determine the generalized variables which sati the aforementioned requirements, lot, and ‘Thus, and polly wa Bp Substituting these now variables permits equations (20) to be written as (u’—ws+(4 22) wy’=0 (a —w¥—(G 8) 0 w+) ‘The number of independent parameters will be reduced if all the combined constants in equations (20n) are set equal to one another, that is, let (20a) a a ofp le From this relationship, it can be seen that ae WE vo ‘Thus the generalized variables become A (Wi) wnt ni, (EE s21000—s2—10 and th 1105 and ¢ ‘With these new variables equations (20) ean be written as? where (4! 14!) 44!" =0 w 4+ H=0 ‘nt—t=0 en whore any two of these equations are sufficient to describe the behavior of the system. Inasmuch as equations (21) do not involve any constants, their solutions are complotely determined by the initial val- ‘ues of the variables. Sinco the displacements at initial contact 14, and tr, aro equal to zero and the initial velocities ‘ty,’ and 1 are equal, the only parameter is the initial dimensionless velocity wot where t=,’ Genoralized solutions of equations (21) are presented in figuro 11 for values of tu’ corresponding to a wide range of Janding-gear and impact parameters. Parts (a) to (e) of figure 11 show the variations of the dimensionless variables during the impect; parts (f) and (g) show tho maximum values of the more important variables as functions of tw’. Part (h) shows the shock-strut effectiveness 7, and the Janding-gear effectiveness ny ‘The shock-strut effective- ness, sometimes called “efficiency” and, in Europe, “plani- metric ratio,” is defined as [ihae Wo Foe where o=th—ts is the dimensionless shock-strut stroke, Since 1, represents the ratio of the energy actually absorbed by the shock strut to the maximum energy which the strut could possibly absorb for any combination of maximum acceleration (or load) and maximum stroke, it serves as a measure of tho extent to which a given combination of maximum load and stroke has been utilized to absorb tho energy of an impact. A similar measure of the energy absorption effectiveness of the landing gear as a whole is sgiven by ny which is defined by [ian Santis) may erased fo slap eguatien none rai by dratatng ‘ustoqantiennd eating or ate nceqston. ‘Thies (Hur. "By lntodng tee raia w=, sequen may be rca t ha mere rer qunln (oe ho mt eae fo ball ean vou aed y=,Dinensonlss unper-mats deplacement, 4, Dimenscnless loner-moss dplocemant, menses upper-moss aeceleron, REPORT 1154—NATIONAL ADVISORY COMMOTTER FOR AERONAUTICS z iz 24 28 is 20 Dimensiess tne, 8 (@) Relationship between upper-mass accaloration, lower-mass displacement, and time. Fiovna 11.—Generalized solutions for simplified system, = 12 15 25 2 co 32 Oimersnless tera, 8 : () Relationship between upper-mass displacement and time, * Froure 11.—Continued.ANALYSIS OF LANDING-GEAR BEHAVIOR 1107 4 S iz is 25) Dirensiness time, @ 2a 2 32 BE 40 (@) Relationship between upper-mass velocity, lower-mass velocity, and tine. Frouan 11.—Continued. ‘Phe generalized results presented in figure 11° can be used to estimate tho performance of a given Ianding gear of known configuration for particular impact conditions or to choose the dimensions for e landing gear when tho impact conditions and desired performance are specified. Applicability of solutions. —To illustrate the applicability of the generalized solutions, the curves of figure 11 heve been applied to the previously considered case of the normal impact at an initial vertical velocity of 8.86 feet per second for comparison with the more exact solution presented in figure 4. In order to make uso of the generalized sokutions 2 Alda time ry mit ae penta fir vai fw alan wt aa yam nts or raat. Teen bo hart of aslo noo att ogee toes th sans gry sta out tn tine ab peat ele pnp lo ‘Sart Soon diferent aaa ere adamaoraemothods weap ia ‘tempt to say hs pan fon pros arther et ixrumaip tt ange of ‘svar hs ean Pateoted eucetaleompitn a tho elon, it is first necessary to approximate the tire foree-delection characteristics by a simple linear variation. ‘Two such Vinear approximations which might be considered suitable for this purpose are shown in figure 12. Linear approxi- mation Tis « straight line through the origin having a slope a=18.5X10* pounds per foot (a’=ad=41.6X10" Ib). ‘This value of a and the other pertinent landing-gear and impact parameters result in a value of the initial dimensionless velocity parameter w’=2.57. Linear approximation IT is a straight line with slope a=21.3x10* pounds per foot (@'=47.9X10" Ib) which does not pass through the origin Dut intersects the displacement axis at a valuo of ree 0.0508 foot. With this value of a, ms'=2.39. ‘Since the solutions of figure 11 have been calculated only for integral values of wu’, curves for the foregoing values of ‘tw ware graphically interpolated by cross plotting. ‘These results were then converted to dimensional values by multi-2 f i i i 3 | ° T 2 s REPORT 1154—NATIONAL ADVISORY COMMOTTEE FOR AERONAUTICS 6 7 e 3 0 “Pines sirt sete, (@ Relationship between uppermas acslration and strut stroke ‘Ficune 11—Continued. plying tho dimensionless variables by the appropriate constants. ‘The results obteined aro compared in figuro 13 with the more exact solution presented in figure 4. ‘The ‘values based on linear approximation I have beon plotted exectly as determined from the generalized solutions. ‘The results for linear approximation II, however, have been displaced relative to the origin of coordinates as indicated in the following discussion. ‘The assumption of linear spproximation I implies that the system must move a distanco: equal to 2, 9) after inal contact (eb constant velosty snes the wag it i ‘taken equal to the weight) before any finite ground reection can develop. ‘The derivation of the equations of motion, ‘on the other hand, assumes that the ground reaction in- creases Tineatly with deflection from the instant of initi contact. As a result, the equations of motion do not apply ‘until after the system has attained « displacement equal to | Zeyeor Which occurs ab a time after initial contact treed Ve on to & coordinate system transformed so that the tire force-deflection relationship passes through the origin; that fs, a coordinate system displaced by 2,» Telative to In other words, the equations of motion the coordinate system originating at the point of initial contact. It therefore follows that the upper-mass and lower-mass displacements determined from the generalized solutions for the case of linear approximation II must be Increased by «constant amount equal t0 2, case 0.0508 foot, and all results must be time by ® constant increment Aim —C%, in this caso 0.0608, 8.86 contact. ‘Thése corrections have been incorporated in plot ting the curves for linear epproximation II shown in figure 13. ‘As can be seen, the results obtained by application of the generalized solutions, particularly by the method employing linear approximation II, are in fairly good ogreement with tho more exact solution. “The discrepancies which exist are attributable to the neglect of the shock-strut preloading and springing provided by the air-pressuro force, neglect of the lower mass, and to differences between tho very simple tire foree-defloction relationships assumed and the exact tire characteristics. On the whole, it appears that the general- ized results offer a means for rapidly estimating the behavior of the landing gear within reasonable limits of accuracy and may therefore be useful for preliminary design purposes. 0.0087 second, relative to the instant of initialANALYSIS OF LANDING-GRAR BEHAVIOR, ao Simensieniess wppar-mase cowleraian, of ° T z o + 1109 5 e T e 3 70 Dimensionless uppercmoss placement, uy (© Relationship between upper-mass acceleration and upper-mass displacement. ‘Fravar 11.—Contiaued. SUMMARY OF RESULTS AND CONCLUSIONS A theoretical study hes been made of the behavior of the conventional type of oleo-pneumatic landing gear during the process of landing impact. ‘The basic analysis is presented in a general form and treets the motions of the landing gear prior to and subsequent to tho begining of shock-strut dofiection, In the first phase of th impact the landing gear is treated as a single-degree-of-freedom system in order to determine the conditions of motion at the instant of initial shock-strut defection, after which instant the landing gear is considered as a system with two degrees of freedom. ‘The equations for the two-degree-of-freedom system consider such factors as the hydraulic (velocity square) resistance of the orifice, the forces due to air compression and internal friction in the shock strut, the nonlinear force-deflection characteristics of the tire, the wing lft, the inclination of the Tanding gear, and the effects of wheel spin-up drag loads. ‘Tho applicability of the analysis to actual landing gears hhas been investigated for the particular case of a vertical Innding gear in tho absence of drag loads by comparing calculated results with experimental drop-test date for corre- sponding impact conditions, for both s normal impact and a severe impact involving tire bottoming. Studies have also been made to determine the effects of variations in such parameters as the dynamic force-deflection characteristies of the tire, the orifice discharge coefficient, and the effective polytropic exponent for the air-compression process, which might not be known accurately in practical design problems. In addition to the moro exact treatment an investigation hhas also been made to dotermine the extent to which the basic equations of motion can be simplified and still yield useful results. Generalized solutions of the simplified equations obtained are presented for e wide range of landing- gear and impact parameters. On the basis of the foregoing studies the following con~ clusions are indicated: 1, The behavior of the landirig gear es calculated from tho basic equations of motion was found to be in good agreement with experimental drop-test data for the case of @ vertical landing gear in the absence of drag loads, for both a normal impact and a severe impact involving tire bottoming. 2. Astudy of the effects of variations in the force-deflection characteristics of the tire indicates that, ‘a. In the case of e normal impact without tire bottoming, reasonable variations in the forco-deflection characteristics of the tire have only a relatively small effect on the caleulated behavior of the Tanding gear. Approximating the rather complicated force-deflection characteristics of the actual tire by simplified exponential or linear-segment variations appears1110 Dimensions tine lo maximum cece ° r 2 3 REPORT 1154—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS o 4 5 6 7 8 Inet dimensiones velociy, ug) (9 Variation of maximum upper-mass acceleration and time to reach masimum upper-mass acceleration with intial velocity parameter. ‘Fiounn 11.—Continued. to bo adequate for practical purposes. ‘Tire hysteresis was found to be relatively unimportant. . In the case of a severe impact involving tire bottoming, the use of simplified exponential and linear-segment approsi- mations to the actual tire force-deflection characteristics ‘which neglect the effects of tire bottoming, although adequate up to the instant of bottoming, fails to indicate the pro- nounced inerease in Ianding-gear load which results from bottoming of the tire. ‘The use of exponential or linear- segment approximations to the tire characteristics which take into account the increased stiffness of the tire that re- sults from bottoming, however, yields good results. 3. A study of the importance of the discharge coefficient of tho orifice indicates that the magnitude of the discharge cooffcient has a marked effect on the calculated behavior of the landing gear; 2 decrease in the discharge coefficient (or the product of the discharge coefficient and the net orifice ares) results in an approximately proportional increase in the maximum upper-mass acecleration. 4. A.study of tho importance of the air-compression process in the shock strut indicates that the air springing is of only minor significance throughout most of the impact, and that ‘variations in the effective polytropic exponént n between the isothermal ‘value of 1.0 and the near-adiabatic value of 1.3 have only a secondary effect on the calculated behavior of tho landing gear. Even the assumption of constant air Pressure in the strut equal to tho initial pressure (n=0) yields fairly good results, which may be adequate for many practical purposes. 5. An investigation of the extont to which the equations of motion for the lending gear can be simplified and still yield ‘acceptable calculated results indicates that, for many prac- tical purposes, the air-pressure force in the shock strut can be completely neglected, the tire force-deflection relationship ean be assumed to bo linear, and the lower or unsprung mass can bo taken equal to ero, 6. Generalization of the equations of motion for tho simplified system described in the preceding paragraph shows thet the behavior of this system is completely deter- mined by the magnitude of one parameter, namely, the dimensionless initial-velocity parameter. Solution of these generalized equations in terms of dimensionless variables permits compact representation of the behavior of thes for a wide range of Janding-gear and impact phrameters, which may be useful for rapidly estimating Innding-gear performance in preliminary design. Lanousr Asnonavricat, Lanonatory, Namioxat Apvisonr Comnarree ron ABRoNavrics, Lanousy Freuo, Va., May 1, 1962.ANALYSIS OF LANDING-GEAR BEHAVIOR - un 6 7 Pert i‘ fa alee Pr 3 im | I i ew FT : a r Ht Daz i 5 id Ld [++] 5 @ + io imersioess-wey, oj (8) Veriation of masimum upper-mass displacoment, maximum lower-mase displacoment, and maximum strut stroke with intial veloolty parameter, Fravnn 11.—Continued. 100} Shocks ellecTnerats, 7 a g ‘Landing ator etlecineness mg Ellectveness, 9, percent & 8 ° r 2 5 7 8 3 a > Ilo! cimansonioss velaiy, ug! (h) Variation of shock-trut effectiveness and landing-gear effectiveness with initial velocity parameter. Ficune 1.—Concluded,112 7x Fa] neo opproxination ss / expen é See ny eens 24 A g Bs] 3 “ines operoimation 2: (2=21.3xI0° it Ok poe" = E = Tite deftoction, 2, ft Fravnz 12.—Linear approximations to tire force-dofletion character- {stice used in spplication of generalized solutions. REPORT 1154—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS 3 3 3 i -1sf : es H solution of figure 4 -s aprlizd stra — sgt Bor eal oK0) igs 239 (2130 L © Sear a 05 3 16 “Tina after conte, 20 (@) Time history of upper-inass acceleration. ‘Comparison of generalized regults and more oxact solu tion. Vyq=8,80 feet per second} Ca 0.0.‘Sint ste, 1 ANALYSIS OF LANDING-GEAR BEEAVIOR () Time histories of landing-gear displacements and velocities. Freuas 18—Continued, Sit ves, fos sation of tigre 4 Genertzed stations 57 (0 185 x10%) 39 (0 = 213% 10°) ae 28 7 16 Zo naa e| ie 16 20 ‘Time efter contact, #80 (©) Time histories of shock-strut stroke and velocity. : Frouns 18—Conoluded. 113APPENDIX A NUMERICAL INTEGRATION PROCEDURES As previously noted, most of the specific solutions presented in this report were obtained with a numerical integration procedure, termed the “linear procedure,” which assumes changes in the variables to be linear over finite time intervals. With this procedure a time interval =0.001 second was used in order to obtain the desired accuracy for the particular cases considered. A few of tho specific solutions presented were obtained by means of e procedure, termed the “quad- ratic procedure,” which assumes @ quadratic variation of displacement with time for suecessive intervals. ‘This pro- cedure, although requiring somewhat more computing time per interval, may permit an increase in the interval size for ‘given accuracy, in some cases allowing & reduction in the total computing ‘time required. In the case of the more ‘xact equations of motion the accuracy of the quadratic pro- cedure with a time interval of 0.002 second appears to be equal to that of the linear procedure with an interval of 0.001 second. Although the accuracy naturally decreases with increasing interval size, the loss in accuracy for pro- portionato increases in interval size appears to be smaller for the quadratic than for the linear procedure. In, the case of tho simplified equations of motion reasonably satisfactory results were obtained in test computations with the quadratic procedure for intervals as large as 0.01 second; whereas the linear procedure was considered questionable for intervals Innger than 0.002 second : ‘Phe generalized solutions presented, because éf the relatively simple form of the equations of motion, were obtained with the well-known Runge-Kutta procedure. A study of tho allowable interval size resulted in the use of ‘an interval A#=0.08, which corresponds to a time interval of about 0.005 second for the landing gear under consideration. In this step-by-step procedure the variations in dis- placement, velocity, and acceleration ‘ere assumed to be linear over each finite time interval «. ‘The method, 9s used, involves one stage of iteration. Linear extrapolation of the velocity at the end of any interval is used to obtain estimated values of velocity and displacement: for the next interval. ‘These values are then used to caloulate values of th acceleration in accordance with the equations of motion, Integration of the acceleration provides improved values of the velocity and, if desired, the displacement and sccelera~ In this procedure all integrations are performed by application of tho trapezoidal rule, ‘Tho following derivation illustrates the application of the linear procedure to the equations of motion for the Ianding gear, which apply subsequent to the beginning of shock-strat deflection at time t. In the examplo presented internal friction forces are neglected in order to simplify the deriva- nu tion, However, the same general procedure can be used if these, or other complicating affects, aro included in the equations. For the case under consideration the equations of motion (eqs. (16), (17), and (8)) can be written as followe: Bata aP+BU—Oe—el"+D=0 (AD) a AGy 7 a) Bll —Ole.—29))-*+ Fry (28) — Wa (a2) Ws s 3) we | Bae B tr ot Bao where a Ag Ta] 2(G.An)¥e08¢ B=pyAcose Ae Pecos D=Ki.W-W : E=WK.—1) Solving equation (A3) for &: gives (a4) G2, HF (ea) 4 where Integrating equation (A4) with respect to ¢ between. the limits 4 and ¢ and noting that £,=4,=2, gives Amit Fr—@is—2)-H if “My(eddr (AB) where 7=((—t). Integrating again and noting that 2, 24. gives +O leben Ede, H ff Fy ledde dr (so)ANALYSIS OF LANDING-GBAR BEHAVIOR Substituting for 4, and 2: in equation (A2) gives Wi (4 [a+o—artrr—aH fF (eddr] + 1115 ai-ofa+a (erbi edt FE a" Fr, (ends ar} "2, (eo+m) (an ‘Phe motion of the landing gear subsequent to the beginning of shock-strut deflection is determined by means of a step by-step solution of equation (A7). ‘This numerical procedure yields time histories of the lower-mass motion variables '%, 4, and 2, from which the motion variables for the upper mass 4, 4, and 2: can be calculated by means of equations (Ad), (AB), and (AB). ‘Phe initial conditions for the step-by-step procedure are (as) where 2, %, and % are the conditions of motion at the Deginning of shock-strut deflection as determined from the solution for the one-degree-of-freedom aystem. Estimated values of the Jower-mass velocity at the end of the first time increment ¢ following the beginning of shock- strut deflection ean be obtained from the expression yet, (as) or, as a first approximation, ‘The corresponding displacement is given by fyarmert§ até (a1) After the initial conditions and the conditions at the end of the first time increment are established, « step-by-step calculation of the motion can be obtained by routine opera- tions as indicated by the following general procedure which applies at any time r=ne after the beginning of the process. ‘Pho operations indicated aro based on integration by appli- ‘ation of the trapezoidal rule: Ht, tey ma Sng b5 Gay tes,,) (AID) ‘With tho estimated values 2, and 2f, the accéleration of the lower mass can be determined by"substitution in the appropriate integrodifferentiel equation for the system, equation (A7) in the present case. Thus 1 2Eq9 Te) (A13) ‘In equation (A7) the integral expressions can also be evalu- ated by application of the trapezoidal rule. For example, when Pr, (2 a aide mS (ef bone +... 22 142, 0 fo erat tte) A) and [of erar are fo fi OM fdr det B(foevars [erer) (ais) An improved value for the velocity is obtained from the oxpression ty . Gs, at Fs) aio) ‘This value is used inthe caloulation ofthe estimated velocity ‘and displacement 2f,.,, for the next interval. TF desired, improved values of the displacement and acceleration for the nth interval subsequent to the beginning of shock-strut deflection can be obtained as follows: Payment § reared AE By ch (a and 8 I (Egy 2359 70) (ais) whére flé4y) 21) Fx) i8 an appropriate equation for thesystem, such as equation (A7). With the values of 2, #,, and %, the motion vatiables for the upper mass 2, #,, and 21, can be calculated separately from equations (A4), (A5), and (A6), as previously noted. In setting up the numerical procedure used in obtaining the solutions presented in this report, an evaluation of the erzors introduced by the procedure indicated that it would not be necessary to calculate the improved values of the displace- ment 24, (eq. (A17)) or the acceleration z,, (09. (A18)). How- ever, improved values of the velocity 42, were calculated by1116 means of equation (A16) for the purpose of determining estimated values of the velocity z, and the displacement z (eqs. (A11) and (A12)) for the increment immediately following. In order to illustrate the application of the method, @ tabular computing procedure for the solution of the system represented by equations (A1), (A2), and (A3) is presented in table I. uapramio Procepuns In this step-by-step procedure a quadratic’ variation of displacement. is assumed over successive equal finite time intervals for the purpose of extrapolating values of the REPORT 1154—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS ‘motion variables from one interval to the next. With this assumption the displacement variation over two successive equal time intervals is completely determined by the threo ‘Values of displacement at tho beginning ond ond of each of the two intervals. By writing the quadratic variation in difference form, the velocity and acceleration at the midpoint of the double interval can be expressed in terms of tho threo displacement values proviously mentioned. Substituting for ‘the velocity and acceleration in the differential equations for the systom yields difference equations of motion in terms of successive jdisplacoment values which can be ovalunted interval by interval. ‘TABLE ‘LINEAR PROCEDURE Row Equation Proeedurot ° , ® . 4 fat Garth) @ fet § te tH) O+O> ® Fr(@) Peighection charactoratice rreteaee Equation (1) O.+$10+0. © | fratgea Bavaton (415 o-+ $0401 © = Bauation (47) Given by equation (47). @ ty best § Gerth) Ort FZ Ot@ne ® Fabien) AE Cert) OFFO+O.e ® Ban nt § Carthy) O+$1O+@.e ® Bquation (AS) Given by equation (44), ® a Bauation (A8) Given by equation (AB). ® ay Equation (A8) Given by equation (A6). +O, denotes value for provious time intervalANALYSIS OF LANDING-GEAR BEHAVIOR 7 ‘The following derivation shows how the procedure ean be applied to the determination of the behavior of the landing gear subsequent to the beginning of shock-strut deflection a om time t,. In order to simplify the derivation, internal friction forees are again neglected in setting-up the equations of motion, ‘Tho assumption of a quadratic varintion of displacement with time (constant acceleration) over tivo suecessive inter vals, each of duration «, permits expressing the velocity and Aacecleration at tho midpoints of the double interval (seo sketch) in terms of the displacement values at the beginning, midpoint, and end of the double interval by the equations (G00 ref. 5, p. 18): ayntety ect (19) and 5 afeHT 2a tee 7 (420) had where é, 34 and 2, are the velocity, acceleration, and displacement at the end of the nth interval (r=ne) after the beginning of shock-strut deformation and ,.1 and Zi: are the displacements at the end of intervals n—1 and +1, respectively. Substituting the differenco relations for 4, 2, 2, and % into equations (A1) and (A3) permits writing the equations of motion for the landing gear in difference form as follows: Wasaga Beg Fig a) FAW Cig gs Ag i Banga bP) PBLL— O(21, 24) ED =O (at) and G (22g 41222, 245.) Hel Free) +E] (A22) ‘where tho constants aro es defined in the previous section. Substituting for 2,,, in equation (A21) gives, agg BS as [20 TEA Wiat WW) GWE ana eac | (428) where ang 2Woe1,— Wits, 86[Fr(24,) +B] = 2W 24, +(Wi— Wider, , +2Wi( 214-2151) 8 [Fv (2%,) +E] and, to (wp —O(@1,—22,)]*+D) Equations (A22) and (A23) are essentially extrapolation formulas which permit the determination of values for the uppersmass and lower-mass displacements to come from the values of displacement already calculated, ‘Theso equations thus permit step-by-step calculation of the displacements as the impact progresses, starting with the initial conditions, from which tho upper-mass and lowermass velocities and accelerations ean be determined by means of equations (A19) and (820). Sinco tho ealculation of the digplacomente 2 and s, at any instant by means of equations (A22) and (A28) requires values for the displacements at two provious instants, tho routine application of theeo equations can bogin only ab the ‘ond of the second interval (r=2e) following the beginning of shock-strut deflection. Before the displacements at the end of the second interval can be calculated, however, it is nocossary to determine tho displacements at tho end of the first interval. ‘These values can be obtained from the conditions of motion at the instant of intial shoclestrut deflection by applying equations (A19) and (420) to the instant t=4,1118 At the instant of initial shock-strut deflection. (ara) Application of the difference equations (A19) and (A20) to the instant t=, (that is, n=0) gives tho following equations: (A25) Sinco the landing gear is considered as a one-degree-of- freedom system from initial contact up to the instant (=, REPORT 1154—NATIONAL ADVISORY COMMUTER FOR ABRONAUTICS the foregoing application of tho difference equations results in identical values for the upper-mess displacement and Jower-mass displacoment at the ond of tho first interval. Simultaneous solution of equations (A25) gives the following expression for the displacement at the end of the first interval: (426) Page Fiqes Arb ee tS Be With, the values for 2 and zea, equations (A22) and (A23) permit the step-by-step calculation of the upper-mass sand lower-mass displacements subsequent to the first interval following the beginning of shock-strut deflection. ‘The corresponding velocities and accelerations of the upper and lower masses can be determined from the calculated displace ‘ments by means of equations (A19) and (A20), as previously noted. A tabular computing procedure illustrating the application of the method is presented in table II. ‘TABLE IL QUADRATIC PROCEDURE Row | Quantity ® . © on ®, ® a, @ ® Paes ‘Baquation (423) Given by equation (423). ° Fans Equation (422) Given by equation (422). ® . ® a ® A, 5 Py 10 denotes value for previous time interval.ANALYSIS OF LANDING-GEAR BEHAVIOR RUNGE-KUTTA PROCEDURE In this step-by-step procedure the differences in the de- pendent variables over auy given interval of the independent variable are calculated from a definite set of formulas, the samo set of formulas being used for all increments, ‘Thus the values of tho variables at the end of any given interval are completely determined by the values at the end of the pre- ceding interval. Unfortunately, however, unless the equa- tions to be integrated are relatively simple, the method can become quite lengthy. ‘The following derivation illustrates the application of the ‘Runge-Kutta method to the generalized equations of motion (eqs. (21)) for the simplified system considered in the section: on generalized results, Since these equations can be readily reduced to the first order, they can be integrated by the step-by-step application of the general equatious given on ‘pages 301 and 302 of reference 6 for first-order simultaneous differential equations. ‘Pho generalized equations for the simplified system pre- viously discussed (eqs. (21)) are (uf —wyuto (uy) —w—14=0 uy +u=0 Inasmuch as any two of these equations aro sufficient to describe the behavior of the system, only the last two equa- tions are employed in this procedure. ‘These equations can be reduced to a first-order system by introducing the new variable =u! (a27) so that wan (428) and the equations of motion become ow! —ta=0; aa wpa Solving equations (A29) for 14" and w’, respectively, gives wf (30) te (asi) Applying the general procedure presented in the reference previously cited to the simultaneous equations (A27), (A30), 1119 and (431) gives Anau, =i, sf eet Bh bak FD fit Ibt eth) (a8) § (m+ 2m3+-2m, +m) where ae Fe=(warth) a0 (est) a0 = (onrtl)a9 hy (apart B08 (Hear) a0 m= [orrtl)— afi, Fm] 40 With this procedure, 1, 2, and a can be calculated in step-by-step fashion from the values for the preceding inter~ val, tho procedure begianing with the initial conditions. From these values, 21’, uw”, and t’ can be caloulated by ‘means of equations (A27), (A28), and (430), respectively.APPENDIX B SOURCE OF EXPERIMENTAL DATA ‘Following is @ brief description of the apparatus and test specimen used in obtaining the experimental data presented in this report. ‘Tho basic piece of equipment employed in the tests is the carriage of the Langley impact basin (ref. 7) which provides means for effecting the controlled descent of the test: speci- men, Tn these tests. the impact-basin carriage was used in much the same manner as # conventional stationary landii gear tet fg (se ref. 8). In order to simulate mechanically the wing lift forces which sustain en airpleine during lending the pneumatic eslinder erid cam eystem incorporated in the carriage was used to apply s constant liftorce to the dropping ‘mass and landing gear during impact. ‘The lift fores in these teats was equal to the total dropping weight of 2,642 pounds. ‘rest spRoneEN ‘The landing gear used in the tests was originally designed fore small military training sirplene having gross weight of approximately 5,000 pounds. ‘The gear is of conventional cantilever construction and incorporates a standard typo of oleo-pneumatic shock strut. ‘The wheel is fitted with a 27- inch type I (@mooth-contour) tire, inflated to 32 pounds per square inch. ‘The woight of the Imding gear is 150 pounds. ‘Tho weight of the lower mess (unsprung weight) is 131 pounds, In the present investigation the gear was somewhat modi- fied in that the metering pin was removed and the original orifice plate was repleced with one having a smaller orifice diameter. Figure 14 shows the internal arrangement of tho shock strut and presents details of the orifice. Other perti- nent dimensions are presented in table III. ‘The strut was filled with specification AN-VV-0-366B hydraulic fluid. ‘The inflation pressure with the strut fully extended was 43.5 pounds per square inch. In these tests the landing gear was mounted with the shockstrut axis vertical. igure 15 is a photograph of the landing gear installed for testing. ‘TABLE HT IMPORTANT CHARACTERISTICS OF LANDING GEAR, USED IN TESTS 05781 8,204 0.5821 Pea 2a 13 A variety of time-history instrumentation was used during tho testa. ‘The vertical acceleration of the upper mass was measured by means of an oil-damped clectrical strain-gage accelerometer having a range of 28g and a natural frequency of 85 oycles per second. A low-frequency (16.5 cycles per second) NACA. air-damped optical-recording accelerometer, having a range of —ig to 6, was used as a stand-by instru ment and as a check against the strain-gago accelerometer. Another oil-damped strain-gage accelerometer, having « range of -£12z and a natural frequency of 260 oycles per second, was used to dotermine the vertical acceleration of the lower mass. Tho vertical displacement of the lower mass (tive deflection) and the shock-strut stroke were measured separately by means of variable-resistance slide-wire poten- tiometers. ‘The vertical displacement of the upper mass was determined by addition of the strut-stroke end tire-defle measurements, ‘The vertical velocity of the landing goar at the instant of ground contact was determined from the output of an elemental clectromagnetic voltage generator. A time history of the vertical velocity of the upper mass was ob- tained by mechanically integrating the vertical acceleration of the upper mass subsequent to the instant of ground con- tact. Electrical differentiation of the current output of the strut-stroke circuit provided time-history mensuremonts of the shock-strut telescoping velocity. ‘The instant, of ground contact was determined by. means of a micro- switch, recessed into the ground platform, which closed a circuit as long as the tire was in contact with the platform. ‘The electrical output of the instruments was recorded on a channel oscillograph. ‘The gelvanometers were damped to approximately 0.7 critical damping and hed natural fre- quencies high enough to produce virtually uniform response up to frequencies commensurate with thoso of the measuring instrumentation. A typical oscillograph record is shown in figure 16. Tt is believed that the measurements obtained in th tests aro accurate within the following limits: ‘Measurement Accuraoy Upper-mass accoleration, £09 Force on upper mace, I 2 500 Lowersmass accsleration, # 2 £09 Vertical velocity at grouiad contact, fps 2 go Uppor-mass velocity during impact, fps arr ‘Upper-mass displacament, f £0.08; Lower-mass displacoment, f. £0.03 Shookeatrut eroke, ft £0.08 Shook-atrut telescoping velocity, fps. 206 ‘Timo after contest, 20. 2 0.008,ANALYSIS OF LANDING-GEAR BEHAVIOR 1121 @ ® 3 Fler pug ‘Frounn 14.—Shook strut of landing gear tested at Langloy impact basin.1122 REPORT 1154—NATIONAL, ADVISORY COMMOTTER FOR AERONAUTICS Strut-stroke telescoping velocity slide wire Strain~gage ‘accelerometer Tire=deflection] slide-wire ‘contact Frovar 15.—View of landing gear and instrumentation REFERENCES 1, MePherson, Albert K., Evans, J, Jr, and Levy, Samuel: Influence of Wing Floxbiity on Foreo-Time Relation in Shook: Steut Follow Ing Vertical Landing Impact. NACA ‘TN 1005, 1040, 2, Stowell, Elbridge Z, Houbolt, John C., and Batdorl, 8. B.: An Evaluation of Som Approximate Methods of Computing Landing Stresses in Airoraft. NACA TN 1584, 1048, 3, Walls, James H.: Investigation of tho Air-Compresslon Procoas ‘During Drop Testa of an Oleo-Pneumatie Landing Gear, NACA TN 2477, 1051. 4, Busty, Waller C.: A Study of the Response of an Airplane Landing Gear Using the Dilferential Analyzer. Jour. Aero, Sei, val. 17, no, 12, Des. 1960, pp. 755-704, 8. Southwell, R. V.: Relaxation Methods in Theoretical Phyates, ‘The Cinrondon Pros (Oxford), 1048. +6, Scarborough, Jamos B.: Numerleal Mathomatical Analysis, Second ‘ed, The Johna Hopkins Press (Baltimore), 1060, 17, Battarson, Sldnoy A.: The NACA Impact Basin and Water Landing “Testa of Float Model at Various Velocities and Weights, NACA Rep. 795, 1044. Supersedes NACA ACR LAT6.) 8, Milwitaky, Benjamin, ond Lindquist, Dean C.: Evaluation of the Reduced-Mass Method of Representing Wing-Lift Bifeots in Hree-Fall Drop Teste of Landing Gears. NACA TN 2400, 1061. BIBLIOGRAPHY Calleio, Pietro: The Shock-Absorbing Systom of tho Alrplano Land- {ng Gear. NACA TM 038, 1040. Kochanomiky, W.: Landing and Tasying Shocks With Oleo Log ‘Underoarriages, British Ministry of Supply, TPR 8/TIB 2 Trans lation No. GDC 10/5250 T, Nov. 1044. Stowell, Elbridge Z., Houbolt, John C., and Batdort, 8, Eval- lution of Some’ Approximate Methods of Computing Landing ‘Stresses in Aireraft. NACA TN 1684, 1048. Schlaefke, Kz Zar Kenntnis dor Kraftwogdiagramme von Flug- zeugfederbeinen. (On Foree-Daflection Diagrams of Airplano Spring Struts) 1. Toilberioht: Vergletdh von Diagrammen mit linoarar und qua- dratischer Dampfung. (Partial Rep. No. 1: Comparison of Diagrams With Linear and Quadratic Damping) Toot. Borichio, Ba. 11, Hoft 2, 1044, pp. 51-68. (Translation available from CADO, Wright-Pattarcon Air Foreo Bast, as ATT 27004) 2, Tellberiht: Naherungaverfahren sum Borechnen dor Kraft- ‘wogdiagramme mit niohtlinearer Federkenalinle und no- ror oder quadratischer Dimpfung. (Partial Rep. No. 2: Approsimation Mothod for the Calculation of Foree-Deflc- tion Diagrams With » Non-Linear Spring Chart and Linear or Quadratic Damping.) ‘Tech. Berichte) Bd. 11, Hott 4, 1944, pp. 105-100. (Tranaation available from CADO, Wright-Patterson Air Foros Base, ax ATI 27031.) 8, Teilbericht: Der Tandestoss von’ Olluftfederbelnon. (Tho Landing Impact of Air-Oil Shock Absorbers) Tech. Berichte, Bd. 11, Heft 5, May 16, 1044, pp. 187-141.ANALYSIS OF LANDING-GEAR BEHAVIOR 1123 MWe an velocity, Tire de HNN NL bei panade eo mt tH ee ANGI f kit baer Hh La Hs HR [ ea ‘i Froune 16.—Typieal oscillograph record obtained during test in Langley impact besin. Seblactke, K.: Zur Kenntois der Weehselwisiungen siviechen Federbein ‘und Reifen boim Landestoss von Flugzeugfabrwerken. (On Reelp- roeal Effects Botween Landing Gear and Tire in Landing Impact of Alrplane Undercacriages)) Tech, Berichte, Bd. 10, Heft 11, Nov. 15, 1983, pp. 308-967." ‘Marquatd, E., and Moyer aur Capellon, W.: Naherungsweise Berech- ‘nung der sivicchon Fabrgestoll und Rumpf beim Landen auftro- tendon Foderkrifte. (Approximate Caleulation of the Shock Strut Foress Oceurring Between Landing Gear and Fuselage in Landing.) FB Nr. 1787, Doutseho Lutfabrtforschung (Berlin-Adlorshof), 1943. Marquard, E., and Mayer aur Capelien, W.: Naherungsweise Berech- rung der zwischen Pahrgestell und Rumpf beim Landon auftre- tondon Federungskrifte. (Approximate Calculation of the Shoek- Strut Frees Occurring Between Landing Gear and Fuselage in Landing) FB. Nr, 1787/2, Deutsche Luftfaketforsehung. (Berlio- Adlershot), 1948. ‘Yorgiadis, Alozandor J.: Graphical Analysis of Performance of Hy- ‘raulio Shock Absorbers in Aireraft Landing Gears. Jour. Aero, Bel, vol. 12, no. 4, Oct. 1945, pp. 421-428, : ‘Tomple, G.: Prediction of Undorearriage Reactions, R. & M. No. 1027, British A.R.C., Sopt, 1044, Makovski, 8. A.: A Method of Shock Absorber Performance Prediction With a ‘Note on Its Application to ‘Trleyele Aeroplanes, TN No. SME. 185, Brith RAE, Sept. 1943. Frankdand, J.'M.: Ground Loads in Carrier Aireraft, Rep. No. 7002, ‘Chance ‘Vought Aircraft, Div. of United Aireraft Corp. (Dallas, Tox), Oct. 2, 1949, Readey, W. B., and LaPavor, 8. A An Analytical Mfethod for the ‘Design of Metering Pins and Predietion of Load Stroke History of Landing Gears. Rap. No. 1688, MeDonnell Airraft Corp, May 12, 1960, Bory, F. Ry and Prick, R. Ps: Thooretical Development of Load Factor Vs. Time Curves for tho DO-6 Main Gear. Rep. No. 21348, Douglas Aireratt Co,, Inc, Sept. 1, 1950. Hurty, Walter CA Study of the Response of an Airplane Landing Gear Using the Ditferontial Analyzer. Jour. Aero. Sei, vol. 17, no. 12, Dos. 1950, pp. 756-704. Masaki, Mamoru, Smilg, Ben, and Moore, C. K.: Tho Prediction of ‘Vertical Two-Wheel Landing Loads. MR No. TSEACS—1505-2-10, Air Materiel Command, Eng. Div, U. & Air Force, May 28, 1946. logge, W.: Landing-Gear Impact. NACA ‘TN 2748, 1052