la, EEE AERODYNAMICS f JACK N. NIELSEN : i wu cen ncn crs Palo Alto, California ] Nae eee Lt teed eli a | DEUCE WAC MSea mer tuiuiicet aan eine tite | i CEU Ln Conical eg 5 ryMcGRAW-HILL SERIES IN MISSILE AND SPACE TECHNOLOGY H, Guyrorp Sreven, Consulting Editor SSS Bussanp ano DeLavaa «Nuclear Rocket Propulsion Nuwets, Sounding Rockets Ninn - Missile Aerodynamics MISSILE AERODYNAMICS JACK N. NIELSEN Formerly Aeronautical Research Sent Nolionel Advinry Commie for Aronantice McGRAW-HILL BOOK COMPANY, INC New York Toronto London 1960:RODYNAMICS Copyright © t960 by the Metiraw- Hil Book Company, Ine. Printed {athe United States of Arnica. All rights reserved. This book, oF prt thereof, may not be reprdioed in any fre without permission Gre publishers ibrar of Congress Catalog Cand Number 5014402 40550 10 GISELA AND DAGMARPREFACE [In recent years the great many persons who have become actively connected with missile seience and engineering have had to rely princi- pally on technical journals and papers for aerodynamic information ‘The literature in missile serodynamics is extensive and in many respects complete, but an overall view of the ficld is reserved to those few spe- cialiats familiar with the hundreds of excellent technical papers available. Howover, a large group of persons who would find such an over-all view useful in the performance oftheir duties eannot, for one reason or another, review the numerous technical papers. It is principally for this group that the present volume has heen written. The book attempts to present a rational and unified aecount of the prineipal results of missile aerodynamics, A missile is described by Webster as a weapon or object capable of being thrown, hurled, or projected so as to strike a distant object. One distinction betsreen a missile and an airplane is that, unlike an airplane, ‘4 missile is usually expendable in the accomplishment of its mission From a configurational point of view, the distinetion is frequently made ‘hata missile ie more slender than an airplane and tends to possess smaller ‘wings in proportion to its body. These distinetions are, however, sub- ject to many exceptions, In fact, the configurational distinetions hhetween missiles and airplanes seem ty narrow as the operational speeds increase, Therefore much of the missle serodynamies contained herein will be directly applicable to airplanes. ‘Since this book dravs on a lange number of technical papers for much, of its content, it is important that the policy with rogard to credit for technical material be elear. The author would like to quote original souroes in all eases, Such a course of action is, however, impractical because original sources are often impossible to ascertain, or not readily available, Thus the references to technical papers herein are those most, ‘convenient from the standpoints of availablity oF pedagogical usefulness, vor simply those most familiar to the author. ‘The book attempts to present a rational aecount of the principal sub- jects of missile aerodynamics, It further attempts to present adequate ‘mathematical treatment of the subjects for use in design. ‘The alterna- tive approach, of compiling a handbook of missile design data, was notattempted for severe] reasons in addition to the author's natural dis: inclinations. First, the wide range of missile configurations and their ‘continuous evolution render it dificult to specify design data of general utility. Second, design date are often classified. "The author has been influenced in his choice of subject matter by eon- sideration of his special competencies. However, in the interests of eom- pleteness, he has included many subjects in which he has no particular competence, Many subjectsare treated extensively from a mathematical ‘point of view, but many other subjeets of equal importance are either not Rmenable to mathematical treatment or are imperfectly understood. Nevertheless the author has chosen to treat such subjects qualitatively, ‘oven though such treatment may not enhance the elegance of the book. ‘The emphasis in the main is on supersonic speeds, although much mate- Fial applicable to subsonie speeds is ineluded. Such emphasis is eon- stent vith the faets that missiles fly mostly at supersonic speeds and that many excellent books on subsonie aerodynamics are already avail able. ‘Though certain subjects have been included in the interests of Completeness, no claim for completeness is made. ‘The sin of omission is considered preferable to inadequate treatment of more material Readers frequently wonder what motivates the author in his arrange- tment of material, The first chapter is purely introductory in character, ‘and the second chapter eoleets together for convenient use many of the results used repeatedly in subsequent chapters. The third chaptor teats the subject of slender-hody theory which the author considers the Dackbone of missile aerodynamies, Slender-body theory has the great advantage that it is mathematically tractable for a very wide range of missile configurations, In Chaps. 4 to 8, inclusive, an attempt is made to present missile aerodynamics in an orderly building up of a missile from its component parts, the body alone, the wing panels, the tail, and the control surfaces. Since the aerodynamies of a tail behind the wings fof'a misile depends on the flow field of the wing-body combination, such flow fields are diseussed iu Chap. 6 before the diseussion of wing-body-tail tcombinations in Chap, 7, The final two chapters of the book treat the important subjects of drag and stability derivatives. ‘The nature of ‘aerodynamie drag makes desirable a separate chapter devoted to drag. ‘The chapter on stability derivatives attempts to treat all forces and ‘moments on a missile (other than dag) from a general and unified point of view ‘The author would like to acknowledge the many contributions made by others to the book. Professors Holt Ashley, J.C. Hunsaker, and “Arthur Bryson reviewed parts of the manuseript and made a number of hhelpful suggestions. should like to thank those members of the staf of the Ames Laboratory ofthe National Advisory Committee for Aeronautis, Dean Chapman, Max Heaslet, Robert T. Jones, Morris Rubesin, Murray ‘Tobak, and Milton Van Dyke, who willingly reviewed those parts of the book of particular interest to them. Also, the author would like to pay tulbute to those members of the staf of the 1- by 3-foot supersonic wind tunnel branch with whom he has orked in the feldof missileaerodynarnies for many years, and particularly to Wallage Davis, branch chief, These co-workers of the author include Wallace Davis, Elliott Katzen, Richard Spabr, William Pitts, Leland Jorgensen, George Kaattari, Frederick Goodwin, and others. The exacting job of preparing the final manuseript was faithfully undertaken by Virginia Stalder. H. Guyford Stover has ‘been very kind in seoking out the book for his series and in lending general encouragement and adviee to the author. In conclusion, the author ‘would like to acknowledge his debt to the National Advisory Committee for Aeronauties, in whose laboratory much of the knossledge in this book ‘was originated, and without whose cooperation this book would not have bbeen possible, Jack N. NielsenCONTENTS Preface (Chapter 1. INTRODUCTION Il, Mile Aerodynamics vera Airplane Aerodynamice 12, Chssifention of Mises 13. Aves; Angle of Bank and Included Angle ToL Angles of Attack and Bidet 15. Glory of Special Terme (Chapter 2. SOME FORMULAS COMMONLY USED IN MISSILE ‘AERODYNAMICS. 24, Nonlieae Potetial Bquation 22. Lisearisationof Potential Pguation 2.4. Rerouli’s Pquation; Presure Coufisint es « Power Series in Velocity ‘Components 2-4. Classication of Various Theorie Used in Succoding Chapters 25, Line Pressure Source 258, Aerodynamic Characteristics of Rectangular snd Tvisngular Lifting ‘Surfaces of the Basis of Supersonic Wing Theory 21. Simple Sweep Theory 28. Conformal Mapning; Notation; Listings of Mappings and Flows 28. Bliptie Intgrals Chapter. SLENDER-BODY THEORY AT SUPERSONIC AND ‘SUBSONIC SPEEDS Slender Bodies of Revolution S:. Slender Bodies of Revolution at Zero Angle of Attack at Supersonic Spends; Sourees %2, Slendor Bods of Revelution at Angle of Attack st Supersonic Speed Doblete 83, Slondar-body Thoory for Anal of Attack ‘Slender Bodies of General Cross Section st Supersonic Sp ‘4. Solution of Potential Equation by the Method of Ward 85. Boundary Conditions; Accuracy of Velosity Components 55. Determination of ay(2) apd bye) 5.7. Prose Coeficiente. 548 Lift, Sidefore, Pitching Moment, and Yessing Moment 39, Drag Force S410, Drag Dus to Lit 3-11, Formula Beplicitly Bshibiting Dependence of Drag on Mach Number 2 18 aStender Bodies of General Cross Section at Subsonle Speeds 73, Lifton Tal Stoton and Tail Efsiney for Dierate Vories a Plan of Tal $12, olution of the Potentat Equation 6 z 180 ‘Tail Interirence Feetor $8.18, Determination ofa(2) and 62) 83 rs ie 7, Caleulation of Tail Lift Due to Wing Vor 14, Drag Forni for Subsonic yAlembert’s Parador 8 ig Vortices in 5-14, Drag Formaln for Subsonic Speeds; d'Alombert’s Patados 1. Use of Tevers-fow Method for Caleuating Aerodynamic Fore 00 “Til Stetion in Nowueiform Flo ‘Chapter 4. AERODYNAMICS OF BODIES; VORTICES oo ates ene 8 avid Flow Tnvineis chapter @, AERODYNAMIC CONTROLS 208 444 Liftand Moment of Snr Boies of Tevoton 6 crea : ‘TE Pressure Distribution snd Loding of Sader Bodies of Revolution; of Contmls; Conwetions i eee $2) Allanovable Controls for Planar Condgurstions 3 4, Slander Bodies of Pliptcal Crocs Section; Etisal Cones u 85. Alomraig Cntr Cony Conssos 2s (Quusheylindeieal Bodie 2 $4. Coupling Eiecte in Albyovable Control a hens 8. Tralingsedge Controle Bs ort $0. Some Nooliear Elects in Acrodynaunie Coteol 2 3 85. Notes on Estimating Hinge Me 2 465, Postns nd Strengths of Body Vortions 86 i Hinge Moments 2 Py nd Strengths of Boy = 38. Change ia Minile Attitude Due to bmpulsive Heh Control; Aliude {En Fors and Moments Dee to Body Vortices; Alle’ Croaow Theory EF. Muton ef Syrmetriesl Psi of Crosafiow Vortis in Preeance of Ci ‘lar Cylinder ote 250 48, Motion of Vertes in Presnce of a Noneirelar Slender Cotiguration. 94 Chapter 8. DRAG 261 £5, Lift and Siefore on Slender Confgurtion Due to Free Vortces 0 SL. General Nature of Drag Forces; Components of Drag 2 10, Boling Momeat of Slender Confgeration Due to Foe Vortces 101 ‘92. Anuyteal Popertica of Deng Curves 265 Pressure Foredeeg, er BODY INTER nm Cee hice dca 93, Proesure Foren of Sender Bodies of Given Shape; Dryg Due to Lif 200 St, Definitions; Notation a8 SL Pressure Foredrag of Nonlender Misule Noses st Zo Ange of Attack 275 54, Planar Wing and Body Interference a 85, Shaper of Bodies of Revolution for Lest Dressre Foredrag at Zero $5 Daten of Lit betneen Wing and Body; Penet Center of Pressure, 118 Bear nae a 5-4 Crvcifors Wing and Body Intesfrenoe m1 846, Pressure Drag of Wings Alone 2 EE Geet of Anele of Bante on Triangular Panel Chameteristis; Panel 857 Preuure Fosednag of Wing-Bedy Combinations of Given Shape at Zero Panel lterfcronee 135 ‘Angle of Attack a0 166, Summary of Results; Afterbody beets ry 98, Wings and. Wing-Hidy Combisations of Teast Cressie Foredrag at E>, apileaton to Nonlender Configurations; Calelaive Example 13 Zam Angle of huisel = [58 Simplifed Vortex Model of Wing-Body Combination Bs 9.9, Minimizing Pressure Drag of Winge and Wingdy Combinations Tesond Tht Due 1 Thiknese 0 chapter 6, DOWNWASH, SIDEWASH, AND THE WAKE ue ae 6-1. Vortex Moe! Representing Sknder Wing with Trailing Bige Normal 210, Physical Festus of Flow at «Blunt Base; Types of Flow au "Flow M5 S51, Bais foe Corelaton of Baso-presir Measremente a0 (62, Rolling Up of Vortex Shoot behind a Slender Wiow us ‘812, Corrlaton of Basepresure Measurements for Blanttraling-dge Ale C3; Calculation of Fndaced Velocities of Traling-vortx System 1s Tailed Bluot-bae Boies of Revoation sir (C4. Vortex Model of Planar Wing and Body Combination 16 Chor Variables Iafncing Base Prose Ey 5. Fstore Infocecing Vortex Pathe asd Wake Shape behind Panels of "Planar Wing and Body Combination 166, ‘Skin Friction ‘66, Factors Influencing Dewowash Field behind Panels of Planar Wing nd 8:14, General Considerations of Skia Fvaton at Saporsonie Speeds ws ‘Body Combination 169 ‘36, Laminar Skin Friton; Mean-enthalpy Mthot 0 6-7, Crueiform Areangements m ‘51, Turblent Skin Priton ae ST, Other Variables Tnfuencing Skin Friction ory Chapter 7. WING-TAIL INTERFERENCE ast Ccnapter 10, STABILITY DERIVATIVES 30 ‘Te, Wing-Toil Interference; Fat Vortex Sheet 182 FB. Presse auing on Tail Seetion Due to Disrete Vortiors in Plane of 1-1, Referenee Aras; Notation 350 is 102" General Nature of Aerodynamic Pores; Stability Derivatives 38 “Tal :10, Propet of Stability Desvatives Renting fom Mise Syme; ‘Maple Synge Analyse for Crasifrm Nines 10-4. Mapisynge Analysis for Tifora Mie aad Otber Mises 1008, General Hnpresius for Stabllty Deratves a Terma of Inert Colle et, ee oy 104, subity Derivatives of Sender Pat Triangle Wing 10-7. Geners Method of Evaluating Tnctia Cocficients and Apparent Mace 10, Tables Apparent Mase with Appleton tthe Stability Desivatvee i Cruaorm Triangular Wings 109, Further Example the Ue of Apparetsnus Table 1040, ets of Aspect Rai on Sablty Daivatve of Tvanglar Wigs TOat, Contrbution of the Expennage to Certain Stability Derivatives ‘Bmpennage Interference ets Nome Inder Sabet Inder 58 an 368 a7 a8 386 sor sa cuaprer 1 INTRODUCTION ‘One purpose of this chapter is to point out some of the differences between airplanes and missiles by virtue of which missile acrodynamies embraces subjects not formerly of great interest in airplane aerodynamics. Another purpose is to collet in one place for ready reference many of the symbols, definitions, and conventions used throughout the book. LL, Missile Aerodynamics versus Airplane Aerodynamics One of the principal differences between missiles and airplanes is that the former are usually expendable, and consequently are usually unin hhabited. For this reason increased ranges of speed, altitude, and naneuvering accelerations have beon opened up to missile designers, and these increased ranges have brought with them now aerodynamic prob- lems. For instance, the higher allowable altitudes and maneuvers accelerations permit operation in the nonlinear range of high angles of attack. A missile may be ground-launched or ai-launchesdl and in eonse= tquenes ean undergo large longitudinal accelerations, ean utilize very high ‘wing loadings, and ean dispense with landing gear.” In the absence of a pilot the missle ean sometimes be permitted to roll and thereby to intro- duce new dynamic stability phenomena. The problem of guiding the missile without a pilot introduces considerable complexity into the missile guidance system, The combination of an automatic guidance system and the air frame acting together introduces problems in stability and eontrol not previously encountered. Many missiles tend to be slender, and many utilize more than the usual two wing panels. These trends have brought about the importance of slender-body theory and cruciform aerodynamics for mistils, 1.2, Classification of Missiles Missiles can be classified on the basis of points of launching and impact, type of guidance system, trajectory, propulsive aystem, trim and control devieo, ete. An important elassifieation on the basis of points of launching and impact is given in Table 1 Another souree of distinction among missles is the guidance aystetn. |iv command system the missle and the target are continuously tracked2 MSSHLE anwovysaates from one or more vantage points, and the neeessury path for the missite to intercept the target is computed and relayed to the missile by some means such as tadio. A beam-ridding missile contains a guidance system to constrain it to @ beam. The beam is usually a radar illuminating the target so that, if the missle stays in the beam, it will move toward the target. lioming mivsile has a seeker, which sees the target and gives the necessary dinestions 40 the missile to intercept the target. The ‘homing missile can be subdivided into classes having active, semiactive, ‘and passive guidance systems. Tn the active class the missile illuminates the target and reecives the reflected signals. In the semtiactiee class the missle reecives reflected signals {rom a target illuminated by means extemal tothe missile. The passive type of guidance system depends on a reeciver in tho missile sensitive to the radiation of the target itseli, ‘Tonun Hel, Cuasncanios oF Massisss AM Aietoale mile asi Aieto-utface mile acne ‘etonnderwater missile Sorfaceto-air mine Serfacet-murace mile Underatertosunderwater misile Another method of classifying missiles is with regurd to the type of trajectory taken by the missile. A ballistic miseile follows the usual hullstie trajectory of a hurled object. A glide missile is launched at « steep angle to an altitude depending on the range, and then glides down fon the target. -A-akip misnile is launched to an altitude where the atmos- phere is very rage, and then skips along on the atmospheric shel, ‘On the basis of propulsive systema missiles fall into the categories of lurbajet, ramzjet, rocket, ee. Ifthe missile receives a short burst of power that rapidly aceeleretes it to top speed and then glides to its target it is a toostglide missile, Sometimes « missile is termed single-stage, doublee sage, ete. depending on the number of stages of its propulsive system, Farther differentiation among missiles ean be made on the busis of trim sac control devices. A canand missle has a small forward lifting surface ‘hat can be used for either trim or conttol similar to a tail-first airplane. A missile eonteolled hy deflecting the wing surfaces is termed a wing- ‘ontrol missile, aud one controlled by defecting the tal surfaces is termed 1 faiLconirol missile. It is to be noted that those definitions depend on ‘which set of lifting surfaces is taken as the wing and whieh is taken as the tail. For missles with twa sets of lifting surfaces, we will specify the ‘wing to be the main lifting surfaces and the tail to be the balancing sut~ faces, a distinction msintained throughout the book, In 2 cruciform inissile, sets of controls at right angles permit the missle to tura immedi- ately in any plane without the necessity of its banking. On the other sxtnopvertox 3 hhand s banh-to-urn missile, like an airplane, banks into the turn to being the normal acceleration vector as close to the vertical plane of symmetry 1s possible 1:8. Axes; Ange of Bank and Incted Angle Of de two goer ystems of anes edn the prtet book, the sound system dos not appear unt the nal chapters "The ist sem, sly tobe deveribd, one well adapted for ure with the they of comes ‘arable anc an euhy iota in lendetedy them. Elo ecole ‘petri the NACA Handard used in sch lds a sabiity dations and dynamic stability. It is described in detail in the final chapter. It would niopify matt if one et oases Were uo las of the ts ‘six Comidention wae ven to detning muct'n companies o 1Nes but theidoa was deared boa the Sette noid peoably be tyadd ane ‘Suet, hero too many systema srealy ‘exist. Also, a single eystem of axes repre- as sents to gata dapertue fom wage In the oe terture "The bse et of aser used inthe fet nine chapters iva st of bry axes, yy and =e ey in the missilo with minor notational dilfer- f ence for various tile piions The ¢ the Tongtdinal mil aie Thy x positive to the tight, icing frm and hes in the horizontal plane of semty when one exists. ‘The «usin sitive vertically upward and ithe vertical plane of symmetey i oe Csists. Thea aan xer shonin Pig Io ora ight haded stern, ‘The boy aes z, 2 take onal he posible rintations« mise can assume ina niform air team. The angles which eonverenty speaty the orientation ofa isle with respect to ight deetion depen on the use to which such angles ate to be put. Kor the purposes of tie took seta angles and pare convenient. Conidera mise mite in wind tunnel on a ing etacidene with the prolongatan of is ng tadinal axis. Let the mile be alguod paral to the wind elects with the wing panels in the zo bank atsitade Devote the hal axon in this intial poston by 2, f, and #." Now rotate (ite) the mise aout the 9 axis by an angie sas shown fn Fig. 1-2 0 thal nl veexpy the postions "and" The angle owl be termed the ince ‘ng ands the ange nee between the nisl’ ong anand the feestream velocity, Now let the mis be rotated in a locking Sireton facing forward about the x avs wo hat an? gointo yan The axe are ela hy the fellosing equations Fic, 1H, Body aes,4 MISSILE AERODYNAMICS 2 = Beosa,— Bsina veg aay = Esina, + 2008, os a; ~ Fain a 08 9 — Brin asin y — F605 a,sin y 02) sin g + 2 sin a, 608 @ + & 008 a. 008 ¢ From Eq, (1-2) the direction cosines between the #, 7, #and the 2, y, 2 axes can be readily found (Table 1-2). Tt is important to note that the 2 @ ® ‘Fo, 12, Axis conventions for itch and bank, (@) Pitch about g; (6) bank about 2 angle a. must be applied to the missile before ¢ is applied for the above direction cosines to be valid. ‘Thus, the pitch and bank operations are hot commutative. In particular, ifthe missile is frst banked about 2 and ‘Tanun 1-2. Dinucrion Costes oF Booy Axes vow ‘Goumiszo ae AND y DisruAceurs ? i : cote ° _— then pitched about 9, the @ axis will remain perpendicular to the ait stream, In other words, the missile will remain in a position of ero sSidestip. 1-4, Angles of Attack and Sideslip ‘The angles of attack and sidesip are defined here as purely kinematic quantities depending only on velocity ratios. As such, they measure velocity components along the body axes of the missile. Let the air stream velocity relative to the missile center of gravity be Vo with com- ponents u, , and.w along 2, y, and c, respectively. As defined, w, », and tw are flow velocities, and —u, 2, and —w are velocities of the center of gravity with respect to the air stream, ‘The angles of attack and sidoslip ha ‘boon defined in at least three ways. The small angle definitions are ‘The sine definitions are sing =F sin FE (4) ‘The tangent definitions are tan 8 = ae (5) Pio, 13, Anglos attack and sdelp The subscripts # and ¢ are used to differentiate between the sine and the tangent definitions. A graphic interpretation of the angles ay, 8, as, and 4, is shown in Fig. 1-3. Note that a positive sideslip angle occurs when the air stream epproaches from the right facing forward, For small angles, the angles of attack and sideslip do not depend on which definition is used. For large angles it is necessary to know which defini- tions have been adopted. Frequently, the sine definition is used for one ‘quantity and the tangent definition for another. It is a simple matter to relate the angles of attack and sidestip to the ineluded angle and angle of bank, With the aid of Table I-1, we have uw = Vocos (2,2) = Vo0os ae v= Voeos (fy) = ~Vosinassine 8) w = Vecos (22) = Vosin a, 08 ¢ For given values of a, and y, the values of a and 8, are exprosed by’ in a sin sn, Conversely, the values of a.and y necessary to yield a, and 8, are iven by sin’ a, = sin a, + sin® 3, tany = BD (8) For the tangent definitions, a set of rolationships exist similar to Eqs, (1-7) and (1-8):6 MISSILE ABRODYNAICS tan as = tan a, 608 ¢ tan 6, = tan ae sin ¢ tan? a. = tant ay + tan® Be a9) (1-10) (an) ‘This relationship has wide use in erueiform aerodynamies. It does not matter what the angle g i, 20 long as a, is small, Tt is noteworthy that Fas. (18) and (1-10) would be used Lo set a sting-mounted model in a wind tunnel to previously selected values of ay By, OF of ay Be twstrative Eeample value of a, 8, and 6 for an included angle of 30° anda bank. sin ac 008 ¢ = 0.500(0.906) 1 sin 8, = sin ay sin ¢ = (0.500)(0: B= 12.3" From Eq. (1-9) tan ay = tan a cos ¢ }) = 012 (0.5774)(0.428) = 0.244 1.5, Glossary of Special Terms Many special terms occur repeatedly in missile nerodynamics, Some of these terms are naw listed for ready reference. Body aces: set of cartesian axes fixed in the missile and parallel to the axes of symmetzy of the missile if such symmetry axes exist Crosgflow plane: a plane normal to the free-stream velocity CCrueiform scing: four siailar wing panels mounted together at come mon chord and displaced one from the next by /2 radians of are Fineness ratio: ratio of body length to body diameter (calibers) Horizontal plane of symmetry: the horizontal plane in whieh the lower half of the missile is the mirror image of the upper half Included angle: angle between freestream velocity and missile Jongi- ‘tudinal axis inrnonvert0s 7 Interdigitation angle: angle bebween the plane of a lifting surface and the plane of another tandem lifting surface ‘ormal plane: a plane normal to the missile longitudinal axis ‘Subsonie leading edge: a leading edge such that the component, of the free-stream Mach number normal to the edge is less than one ‘Supersonic leading ge: 6 leading edge such that the component of the free-streamn Mach number normal to the edge is greater than one ‘Symmetrical wing: a wing postessing a horizontal plane of symmetry ‘Tangent ogive: a missile nose having oonstant radius of eurvatuce in all plano through the longitudinal axis from theapex to the cireular eylinder to which it is tangent ‘reste plane: 9 SK tious crossflow plane infinitely far behind a missile or lifting suriuee to which the trailing vortex system extonds without viseons dissipation Vertical plane of symmetry: the vertical plane in which the left half of the missile isthe mirror image of the right half Wing panels: those parts of the main missile lifting surfuees exterior to the body svmois free-atream velocity missile body axes; a #0, ¢ #0 missile body axes; a. = 0, ¢ = 0 missile body aves; «, #0, ‘angle of attack ineinded angle angle of sidestip sine definitions of angles of attack and sidestip tangent definitions of angles of attack and sideslip angle of bank ‘REFERENCE. 1. Warren, C,H, Fs The Definitions ofthe Anges of Incidence and of Sep, Ae Tech Note Aco. 2178, Act, 1952CHAPTER 2 SOME FORMULAS COMMONLY USED IN MISSILE AERODYNAMICS "The primary purpose of this chapter is to collect together for ready reference certain formulae of theoretical uerodynamies and mathematics commonly used in missile aerodynamies. ‘These formulas are derived in {tail and discussed in other works, and their rederivations here will not be attempted, Since repeated use is made of the formulas throughout, the book, they are collected together in a single chapter for convenience, ‘and to obviate repeated explanation of the formulas and notation. The formulas include the potential equation and Bernoulli's equation in their nonlinear and linestized forms, A listing and classification of the principal theories used in the book is provided. Some common aero- flynamic forinulas are included for line pressure sources, rectangular and ‘riangular wings, and simple swoop theory. With regard to mathemat- jeal formulas, a list is given of conformal mappings used in the book, together with a list of the complex potentials of the lows to be used TThe terminology and notation of elliptic integrals is also included. 2-1, Nonlinear Potential Equation ‘The common partial differential equation underlying the of nearly all vss considered in this book isthe potential equation. potential equation is the partial differential equation for the velocity potential 4. ‘The velocity potential is a scalar function of position and time, from which the flow velocities ean be obtained by differentiation, For a discussion of the velocity potential, the reader is referred to Liep- mann and Puckett."* A number of eonditions determine the actual form of the potential equation used in any particular case. Some of these fonditions are (1) whether the fluid is compressible or incompressible, {@) the eoordinate system used, (3) the velocity of the coordinate system ‘with respect to the fluid far away, (1) whether the equation is linearized or retained in its noplinear form, and (3) the basie flow about which the ‘equation is Hinearized, For the first ease consider @ compressible fuid stationary at infinity. ft the cartesian axes &, 9, § (Fig. 2-1) bea set of axes fixed in the Aud he pressure and density for the compressible uid are related through + Superior numbers efor to item fa the bibliographies at the ends of chapters * FORMULAS COMMONLY USED IN MISSILE ARMODYNAMICS 9 E-@ a ‘y being the ratio of the specific heats Let @ he the potential function. The full nonlinear equation? for & is = D(6 ¢ EEE BY yy 9 Bo HE Meee A BMD ey + HAH) + WHE HH bes + Peby Hee t Syibby) + bby + bbe + bby) (22) The erm pet ine al he ef sand jn tho uni {bal ni 2 nin 2) cant covered te mass tiuon greg he pte of teow abot mise ing hoe “ifs oet would appear tom eb fand ontop nay des fast Ftd Fro, 2:1, Asea Bie in uid and azo fixed in mia for unifors tranlation ‘aces of interest in the theory of missle aerodynasnies, the luid velocity ‘at infinity ean be considered parallel and uniform, and the missile can be ‘considered stationary with respect to the observer. It i nov shown that the form of Bg. (2-2) is unchanged in this now frame of reference. With reference to Fig. 2-1, let 2, 9, and 2 be axes fixed in the missile at time ¢ with # parallel to the uniform velocity Vo of the Quid at infinity fs seen from the missle center of gravity. Also, choose the ¢ axis of the Enger system parallel to #, To obtain the potential equation for the 3, J, 2 axes with the Suid in uniform motion at infinity, we first convert the Bow as seen by a ground observer from the E{-r system to the 4.9.8 system with the fluid still stationary at infinity. Then we super impose a velocity Ve along the positive # axis to obtain the flow we seek ‘The transformation equations are acd eas es)0 MnssiLb AERODYNAaHICS rms £, 2, Superimposing A nev equation for & i then obtained in terms £, 9, r the velit Ve along the positive 2 axis changos the Bow pattern and thauges the potential # into the new potential ¢ in accordance with e-ve+e a» vena n(ae tee] Be Ges + 2Ub.Gyba5 + debates + dodrdes) + Adedu + boon + ordu) (2) eo) 5) now is completely similar to With this physical interpretation, Eq. (2-5) pl wo 1g, (2-2), In faet, the fist factor in each equation is nothing more t the square of the loeal speed of s0 2-2, Linearization of Potential Equation a. (28), we mst rece al terms “otinarite the potential equation, Ea (2-8), we mst Feat eee product of in polenta au ts drvaivs ean aaa darar ire Ab the ont it soak be sated tha hoe see ber of way ssn at th ert, Gece or et tivo the problem at hand. vay pst case he ilies Unearned problem should be exained t see if it ull the Hi ne eaton A parca wey of Hnestxing the ae ioe chs proved partial tail to the Aro ae nl Ganges th volo Held fom = un rm ht the ai of sped Po Sve vlosty components Veta get # oo 6. ‘where @, &, and «are small perturbation velosties. With the possible 00 i ns, our frst assump- exception of limited regions such as stagnation regions, ti thurs thatthe perturbation velocities are small compared to Vo 0 that FORMULAS COMMONLY USED IN MISSILE AEKODYNAMES 11 In this equation the symbol O( stands for “of the onder of magnitude of In contrast to its more precise mathematieal meaning, the symbol hhas the approximate physical meaning in the present cotection that the velocity ratios have numerical values of magnitude « much less that tunity. In the neighborhoods where the perturbation velocities are large, the solutions of the linear equation for stall perturbation velocities ean. hot be accurate, but, if these regious are limited in extent and number, it ‘ean be hoped that the solutions will be representative of the flow in the large. In connection with By. (28) we have also assumed that the perturba tion velovities are of comparable maguitudes, If, as is frequently the case, the lateral extent of the region of influence of the body on the potential field is approximately the same length as the longitudinal extent, then, on the average, the gradients of the potential will be the same in all directions, and @, 8, and a will be of compatable magnitude. ‘The validity of this assumption must be adjudged for the particular problem at hand. The velocity components are then of orders M+ OW) 6 o eV) = Ol) RY) (3 ew We now need some measure of the orders of magnitude of the time derivatives of the potential. Firs, since ban ay ey as am oe 0 ° rat? eae ean — : Vet = OV) Now consider the body to be undergoing some unsteady motion such as Periodie oscillations characterized by frequeney m per unit time. ‘The Perturbation potential will change 2n times per unit time so that = OV ibn) Let us introduce the number of eyeles per body length of travel xo ea) ‘Then, in terms of this frequency parameter, the time derivatives are of orders of magnitudeB utssines AERODYNAMICS = OleV w= 0(87*) oun aves or) ‘th forging oper sri to trina ae ee ters) and to dasa thse of seem so thd Lae ee dag entation vil be vad forall also xo srt han odor af agai iy of, specially el ow and he tncaized equation is (cote lve) ee = bet Peon+ Behn A) |u terms of the free-stream speed of sound tnd the freestream Much number M en uation (2-16) becomes ll = 1h + ont n= pet Moe CB “This equation is the essential equation of linear aerodynamies 2-8, Bernoulli's Equation; Pressure Coefficient as a Power Series in ‘Velocity Components ‘Bernoulli's equation for the compressible unsteady potential Sow of = sows dnaby i fant nly of the peste isn the Bie sys : [Brarh= co a9) where P= OP +o + ee Some interpretation of the above form of Bernoullis equation is interest- ing, Tn tat form, it holds for esch point in the unsteady flow for all timos, ‘The function € has the same value at all points in the flow at any particular time, but its value can change with time. However, if the flow at any’ point (such as at infinity) does not change with time, then € is ponstant with time also. Bernoulli's equation ean be thought of as. rela~ Tlonship between the pressure feld of the flow and the velocity field USED IN MISSILE AERODYNAMICS 13 ‘hus, i is known for a given for, the "hu bs aoa far ie ow, he rem ce tt on Ex Gx, amoltsaquaton branes nn Senay td Py _y me Fae B71 tT 220) The quanies with no subscript re for any gener pont, wile Mose with subserita refer to quantities a some rterene Sondition. ‘Cuticle ney ol esresion fo he presar sotigent i arma the aon eo 4, along the 2, 3, £ axon. For this purposo we define the prow coefficient P in terms of certain reference quantities: aiiate @21) wher pion, and i 1 pp, and Vx aro usually taken as the presure, density, a sls tthe esta ios th 88 nat em sas Vain the Gn system for eompleteaaogy betwee he “Sites. "To obtain the poner series for Pn vlocty compontst lets perform the expansion in the gtr system i the Emir stem and then transfor te ‘cgay system. With the ubserit refering to the cuton at oe ‘the Ef,7 system, integration of Eq. (2-19) yields 7 7 Pig De vate tate ak teed where @ and gare taken a8 seat ini to the Sah umber toy ne Po Ea ey w= er (228) [Expansion of Ka. (2-24) yields the power series Pa PaPe- 924% yy (Cts, VeaVe VF +uo(' 2 + 11c0(834)' where dante one of wt exe of magitade, This atone pra ae in powers of the derivatives of # in the En,t,r ‘ten. "ty sone 25 tothe yoy, we wl te reac cot bas ih tant Vale ed ie geal ehenne sa“ rotational one that the subscript. case On pane My iso change 0 nrc ie ptetia in he 4 8 ‘accordance with Eqs. (23) and (2-4) © emia) = oe — Vol 980 = Vek (2.26) Por the derivatives of # ave thus obtain aoe ae ware eo oe 1d be the perturbation velocities parallel to the 1, we further let 2, Gpand Z axes, we have noe tag bat = rs ee (228) We thus interpret Pq. & 44 MS Cay 2099, + tras of tind onder ne 22) ation is Hinarze, the sare terms i Se ee ehgisea at for instance, in slender-body theory. Beeb) ean egos ac. Chasiention of Various Theories Used in Suceeting Charters z Jie theories are utilized in succeed- ote rom umber of aera dl ie nn nt om en be ned fle nate of nests flows (1) Potential or nonpotential (2) Mach-uamber rang of applicability mete {@) Dimensionaity of fow: i, two-dimensional axial 27m (G) Shape of physieal boundaries consi « thore we wl ose re poten chen with ww Ne )| and the viscous erossflow theory Of the Newtonian theory (Ea. (9-50)] FORMULAS COMMONLY USKD IX MISSILE aunoDYNAMICS 15 (See, 4-6), With regard to the Mach-nnmber range of applicability: we will be eoneerned principally with theories valid in the supersonic speed range, slthough various of these theories are valid at subsonic speeds also, We will be interested in theories that apply to two-dimensional flows, axially symmetric flows, and three-dimensional flows. As for the shape of the physical boundaries, such shapes as planar surfaces, bodies of rovolution, airfoils, ete., are encountered in classifying the various theories. ‘Only steady flows are considered, Tans 21. Cuassiication oF Axnoorxamie Tutonies Usui 1s Text ] Foow Typiet | Seed ‘hry [Poets amenity | spre | ree | setet [Yu |Toodimemona [amom [art a Baseman Yer | Two-dimensional | Alfie IuSa. 4 Shockerpension [Yer | Twroslinensional | Aso wet ok Method of char=| Yes |Twostinensona, |Avolsand = 0S 1 teristics | astny amet | bein of rove wip VouayTwo-dimeesional Thieedien- Any AT onal Simple sweep [Usually Twoulisemsonsl [Swept wing and Any AFB sivepteplindens Supersonic wing |Yeo | ‘Three-dimensional | Wings wo1ie [Ye | Threeaimensional | Wings cones ASL ‘wally Supersonic |Yee | Threraineuonal | Wings built of S| Ting line | horeshoe Qusisptinder Yer Thedineaional Quinbeylindes ars |e ‘oly Sender body —|Yee | Thre-dimensioust “Setler"" bates Any Newtonian No | Threedimessional (Any shape Any MB espace ' A listing of the theories to be considered is given in Table 21. ‘The theories are classified in classes A, B, C, and D. ‘The first three elasses tare essentially potential theories but D is not. Class A is a class of ‘sentially two-dimensional theories; lass B is the elass of tyo-dimen- sional theories applied to throo-dimensional shapes, and class C isa class of essentially three-dimensional theories. ‘The theories of class A are arranged in order of increasing exactitude ‘The first three theories have been treated in a form suitable for engincer- ing caleulations.* The Ackeret theory embraces solutions of Eq. (2-18) specialized to two dimensions (let = Nou = n= 0 (eso) ‘Three-limensinal | Skeader bodies | Any 3F | D6 satssitat APRODYNAMICS respec gnsel teste {rl ns a er se Se rest ey a ete heroin ha, City tan die emp ie ae yrs a9 a it to carry out, {In many instanoes the graphical procedros ae ‘dapted to 4 omatie computing techniques. In such cases the methoc celeron perma Se era Ce ee Lan a on neh 4— ——~ Siow yetiveen strips, To cach strip is then ap- Fro, 22, Stip theory plied any two-dimensional theory or = ee gil fee tats Vineet fener re a on tions ical flow theory can be put into a form wit So yt we razaead yu ima aint ale (Mt = Vou — ey = on = 0 @ a. SeeUeTy Sknown for many different wing planforns for owesaies 17 aul nonlifting wings. Supersonic wing theory for lifting surfaces (no thickness) is termed supersonic liting-eurface theory, and some results for lwiangular snd rectangular supersonie lifting surfaces are included in See. 2-6, Conical low theory isa special form of linearized theory applica- thle to problems in which the flow quantities are constant along any Tine ‘emanating from an apex. ‘The supersonic flow over a cone or a triangular lifting surface are well-known examples of eonieal flow. ‘The Jones’ line pressure source described in See. 2-5 is another example, —Lagerstrom* hnas listed a large number of conical flows, The utility of conical flow ‘hoory lies in the large number of wing low Bilds that ean be eonsteet ot hay superimposing eonieal flow fields with diferent apex positions ‘The counterpart at supersonic speeds of the Prandtl lifting-line theory will be termed supersonic lifting-ine theory. The essential difference ix that supersonie horseshoe vortices are used (See. 6) instead uf subsonic hnorseshoe vortices. In this method, the lifting surfuce is replaced by o or more horseshoe vortices. In the process, the details of the flow in the vicinity of the wing are lost, but simplicity is gained in trying to caleulate the Row field at distances’ remote from the ‘The ealeulation of slownweash and sidewash velocities at distanees remote from the wing i= tractable only in a few cases with the full accuracy of supersonic wing hicory. Again in the ealeulation of the Mow field associated with wings boy combinations, the use of lifting-line theory is tractable where the full linearized theory is uot. Quasi-cylindrieal theory at. supersonic ~poeds is analogous to supersonic wing theory in that both utilize the same Partial differential equations, but in the former the boundary eoudtitions «re applied on a eylindrieal surface, rather than the 2 = O plane as in the latter. In this connection the eylinder is any closed surface generated by 1 Fine moving parallel to a given line, Many lifting surfaces ean he 20 xenerated. Herein we confine our applications of quasi-eslindieal theory to eylinders that are essentially eireula.* ‘The remaining theory of elass€ (sleuder-body theory, about which we will have much to say) is particularly adapted to slender bodies such as ‘many missiles, This theory, deseribed in detail in Chap. 3, is based on solutions to Laplace's equations in two ditnensions with the streamwise coordinate being manifest through the boundary conditions. ‘The ecurrence of Laplace’s equation renders slender-body theory particularly ‘amenable to mathematical treatment and males its application to three- ‘dimensional bodies tractable in many eases of interest. The theories of lass D are not potential theories and are discussed in Sees. 4-6 and 0-5, Line Pressure Source ‘As an example of a conical flow solution, we have the line pressure source of R. T. Jones, which is useful in problems of controls, drag, ete "Tho general features of the flow are readily shown, Consider the infinite1s MISSILE. AERODYNAMICS triangular cone shown in Fig, 28, Such a cone is the boundary formed Thy placing a line pressire source along the leading edge. The pressure ‘coefficient for a subsonie leading edge is 28 cosh Pe RP rants — BIR an AB = Ban *= laratany = T+ Gata = BAH OP and fora supersnie leading edge is p= RP eo ess) UB — tan 7 Here the designation RP denotes the real part of the inverse eosine or inverse hyperbolic cosine, The equations show that the pressure coefi- tients depend only on tan r, 4/2 and 2/2 quantities, which are constant ‘along rays from the origin. ‘The pressure field is therefore conical. ‘The wedge and pressure field are symmetrical above and below the 2 = 0 plane, ‘The pressure field shown in Fig. 24 is that for a wedge with a sub- soni leading edge. The pressure cooffcient is zero slong the left Mach line, inereasing as we move from left to vight. At the leading ‘edge, the pressure coefficient is the- 2, nite we inte ame Grey fe "To the i the leading edge, the pressure again falls from infinity to zero at the Mach line. The infinity ean be viewed as high positive pressure corresponding to stagnation pressure. A wedge vith a supersonie leading edge has a conieal flow field of the type shown in Fig. 2-5, The distinetive feature is the region of constant pressure between the leading edge and the Mach line. By superimposing line pros- sure sources and sinks, umber of eyrmmetrieal wings of widely varying planform eat be built up. 2.6, Aerodynamic Characteristics of Rectangular and Triangular Lifting Surfaces on the Basis of Supersonic Wing Theory {in contrast to the symmetrical pressure fields of symmetrical wings at zero angle of attack, the pressure Belds of lifting surfaces are asymmetri- fal: that is, the pressure changes sign between the upper and lower sur- faces. Since we will deal extensively with lifting pressure fields, it is desirable to st up notation and terminology for loading coefficient, span PORMULAS COMMONLY URED 19 49,28, Tk pone din on ine wr with wenn ating2 MISSILE ABRODYNAMICS loading, ete. By the loading caeficient of a wing or body, we mean the difference botween the pressure coelicient at corresponding points on the ‘upper and lower surfaces. AP = PY P- ss) ‘The superseript plus (++) refers to the impact pressures of the lower surface, while minus (—) refers to the suetion pressures of the upper surface, The distribution of AP over the surface is ealled the loading dis- tribution, The svetion lift coeicient isthe average over the local chord of the loading eoeficient at [fora 35) ‘The span-load distribution is the distribution across the wing span of the product of the loeal chord and the seotion lift coefficient cc. The center of pressure isthe position at whieh all the lift of a wing pane! ean be ¢on- centrated for the purpose of ealou= f lating moments. | Let us now summarize some of the results of supersonic wing the- ‘ory for tiangular wings. Por tri angular lifting surfaces with subsonic Trading edges (Fig. 2-6) the ift-curve slope has been determined by Stewart," to be aCe Qetane og Ga B= Piao O80) where B is the complete elliptie in- tegral of the second kind of mod lus (1 — BF tant 0) (see See. 2-9) bution is constant along rays from the apex Fie. 2-6, Notaton for triangles wins ‘The lifting pressure tatan w 1H tant ean ap @ ‘The lifting pressure field is conical with respect to the apex, and the pressures are infinite at the leading edges. The spun-load distribution is ‘liptical for triangular wings with subsonie leading edges. ccs = (odo( = B52)" ess) ‘The span loading at the woot chord (ede is (caja = Heerpene @39) FORMULAS COMMONLY USED IN MISSILE ARRoDYSaanios 21 ‘Because the flow is conical, each triangular element from the apes has its center of pressure at two-thirds the distanee from the apex to the base. All triangular elements have their eonter of pressure at the two-thirds root-chord axial distance and s0, therefore, does the wing. ‘The lateral position of the center of pressure for an elliptical span loading is at the 4/3r semispan position. The triangular lifting surface with supersonic leading edges also has imple aerodynamic propertios. First, its lift-curve slope is the same us ‘that of an infinite two-dimensional sirfoil ac, 4 Ta "B 240) ‘The loading distribution is conical and ean be ealoulated directly from the results for line pressure sourees in the preceding seetion since the upper anu lower surfaces are independent. ‘The slope ris simply replaced by « in Bq. (2-88). On this basis with a line souree slong each leading edge wwe have for the wing loading etn o/B— Btn» etn o/B + B tana stor ta ane) uation (2-41) yields a constant loading in tho region between the Much lines and the leading edges 4a oP groin es Forth rgionhetween the Mac lines, manipulation of Bg, (21) yields pee LB (ett m/ BY tnt tact» oP = ea tovan |! Fo Eee) es) ‘The span loading in this case is not elliptical as forthe lifting surface with subsonic leading edges but has a linear variation over the outboard see- tion and a different variation between tho Mack lines. For the linear part, we have with reference to Fig. 2-7 falta —y)eine te : c= Re gS ism OHH) ‘and, over the inboard section,"* {BY tan? ws + 60 Get VB tao2 wissiLa AERODYNAMICS ‘The center of pressure is still a the two-thirds root-chord axial location since the lifting prossure field is conical. "Turning now to the aerodynamic characteristics of rectangular lifting surfaces at supersonic speeds, we must differentiate several different eases, depending on the effective aspect ratio BA, For BA > 2 the tip Mach waves do not intersect, for 1 < BA <2 the tip Mach waves intersect “Xe io. 27. Loading distribution slong tailing edge, and span-oad dstibution for ‘Glangular wing with superanni lnding eds ‘each other but do not interseet the wing tips, and for 1g < BA <1 the tip Mach waves intersect the wing tips only once. ‘The lift-curve slope for cases 1 and 2 (Fig. 2-8) has the same analytical form. ar 4(,_ 1) ae - i ) BAZ 48) or ease $ the lit-eurve slope is We 81g 1) ae Sot ee (eee (14 gy) a- ane] M 2 so thut only the influence of one tip is felt over the length of any chord: O ya tho arg Cea) ‘eo er the sme colons Tho ow fh ny ree torte Th fw pst hay Ryn alse, te oe pasta bly Bsn 8 paral ean eblained by tse af Eq 00 through he tnnormaton of he type given by E(t, whi cm, vera By into By Inthe power ete fon rls a eid ‘sae tse sce tude and By (28) sues that th tangeey eonion i mantineddahing the etaormation Attar tase rn for bdr hone shapes ne foto of me Some ofthe Compe potenti we wl we ae sted in Table 2 Tamu 2 Conus Poverstais von Vaniocs Flows 8. Civedar einer in uniform tow 2B. Uniformly expanding cine Wile) = rev log 630 MIssHLA ABRODYNAMICS ‘Panis 23, Couruax Pornsriais vor Vantous Fuows ( . Bapanding lps of conta /® ratio: al +o wo =famt L Smt D. lipo i uniform fw I 18, Planar midcing ond boy combination GF voy nto 9 P. Blip bank with rorpect flea axis [@seom te (et yen FORMULAS COMMONLY USED IN MISSILE ABNODYNAMICGS SI 2-9, Elliptic Integrals ‘We shall have oveasion to use elliptic integrals a number of times, so it is desirable to dofine notation and usage. The elliptic integrals of the frst kind #(,6) and of the second kind (8) are dofined as definite Integrals rie = [caterer toa) = [00 tatareae [REED a ‘The angle g, which will usually le between 0 and #/2, is termed the ampli- Inde, and the parameter i is termed the modulus. The elliptie integrals are funetions of amplitude and modulus only. Ifthe amplitude isx/2, we II the elliptic integrals complete, and use the notation kay = F(4,3) no = (63) hus if the amplitude is not specified, it is assumed to be x/2, and the elliptic integral is complete. ‘Tables of the elliptic integrals ean be found ‘im Byrd and Friedman..* 08) SYMBOLS 4 aspeet ratio Bo Ge . Focal wing ehord fy velocity of sound in free stream in 2 seotion lift coefficient ey ——eomplex constr fe root chord velocity of sound at infinity in &mg system ry velocity of sound at stagnation point in 2,98 xystem eda span loading at root ehord funetion depending only on r Cx, life coffieiont based on wing planform area 1 elliptic integral of seoond kind PF elliptic integral of first kind © modulus of elliptic integral complex coustant K complete elliptic integral of first kind ve tending edge 1 characteristic lengthMach number normal to leading edge free-stream Mach number in £,J,2 system Voce cycles per second statio pressure free-stream static pressure in 7.92 system statie pressure at infinity in mg system pressure eooficient prescure coefficient on impact surface pressure coelfciont on suction surface loading coetfiient ‘magnitude of velocity free-stream dynamic prossure, 3poFot component of qe normal to leading edge polar coordinates; y = ros 8 = rsin 8 subscript, at reference condition real part of ‘maximum semispan of triangular wing time in 2,9,2 system trailing edge velocities along 2, g, and 2 axes ‘velocities along y and = axes velocities along & and 9 axes radial and tangential velocities in y.z plane free-stream velocity ‘complex potential in 3 plane ‘complex potential in ¢ plane body axes for triangular and rectangular wing streamwise distance to wing eenter of pressure Fig. 2-1 complex variable, y + t= angle of attack. angle of attack in plane normal to leading edge ratio of spocifie heats Ihalf angle of wedge polar angle in y,2 plane sweep angle of leading edge tan (y/2) mass density of Suid mass density of free-stream fluid in 3.9.8 system ‘mass density of fluid at infinity in &9,F system. Fig. 21 complex variable of physical plane, & + iy time in fy system FORMULAS COMMONLY USED IN MISSILE AERODYNAMICS 33 velocity potential in 49,2 system; also amplitude of elliptic integral velocity potential in gt system frequency parameter streain functi semiapes angle of triangular wing eens [REFERENCES 1. Liepmann, HL, and A. Puckett: “Acrodynamien of « Compreeible Fi” se. 7h ohn Wiley & Sone, Ine, New York, 147 12 Garrick, LE, and &. 1 Rubuow: Thoorstical Study of Air Fores on an ‘osating or Stendy Thin Wing in a Supersonic Main Stream, NACA Tech, Reps S72, 1047 ‘3. Staff of the Ames 1- by 3-fot Supersonic Wind-tunae Section: Notes and ‘bls fr Us inthe Ansys Supersonic low, NACA Tech Nolee 28, December, 1047 1 Ames Research Stal: Equations, Tables, and Charts for Camprenible Flom, NACA Tock Reps, 155, 1958 '& Sauer, Re "Theoretische Hinfohrang in die Gaxdynamik”” Eawasds Bros, ‘Ann Aibor, Mich 1047, 6. donee, RT: Efe of Sweopback on Boundary Layer and Seperation, VACA eck, Rept 884, 1017. 1 dons, Robert, and Doris Coben: Aerodynamics of Wings at High Speed sec. in "Atrodynamsi Components of Aircraft at High Speds," vol. VI "igh peed Aerodynamics and Jet Propulsion,” Frinecton University Pros, Princeton, 17, B. Lagersrom, P. A.: Linearized Suporsonio Theory of Conical Wings, NACA. Tech. Note 1685, 1050, '® Nieken, Jack Nu: Quasieylindical Thoory of Wing-Nady Interference at Superson Speeds and Comparison with Experiment, NACA Tech. Repl, 1282, 1055. 10. Jone, KT Thin Oblique Aisle st Supersnie Spd, NACA Tech Noter hor, 1916 11, Stowart,H, J.2 The Lift of s Delta Wing at Superson Speeds, Quart. Appl Math, ol 4, 20.3, pp. 246-254, 194, 12, "Roges, A. W Application of Two-dimensional Vortex Theory to the Pre ction of Flow Fields behind Wing-Body Combinations st Subwoni and Superson Sper, NACA Teoh, Notes 8227, 1, 18, Busemane, A Tnfisitesmal Conia! Superonie Flow, NACA Teck Mem. 100, voit 14. Milne-Thompeon, LB: "Theoretical Hydrsiynamis" 2 ed, pp: 138-140, ‘The Macmillan Company, New York, 1060, 436. Byrd, PF, and M, D, Friedttan "Handbook of Eps Integra for Eng seers and Physi," “Grundoren der Mathematichen Wissenschaften,” Baud UXVIIE, Springer-Verlag, ers, 198, InCHAPTER 3 SLENDER-BODY THEORY AT SUPERSONIC AND SUBSONIC SPEEDS ‘The prineipal purpose of this chapter is to derive a number of general formulas for slender bodies at subsonie and supersonic speeds having application to a wide range of slender missiles, ‘The formulas yield pressure coefficients, forces including drag, and moments for such con- figurations as slender bodies of revolution, bodies of noueireular cross section, wing-body combinations, and wing-body-tail combinations. ‘The basic results of this chapter are applied to nonslender missiles in sub- sequent chapters. Slender-body theory is greatly simplified if only bodies of revolution are taken into consideration, ‘Then the mathematical analysis ean pro- ceed along the intuitive lines of sourees and doublets. ‘The first part of this chapter including Sees. 3-1 and 8-2 considers the problems of deter- ‘mining the potentials for slender bodies of revolution. Tt also serves as fan introduction to the theory for bodies of noncireular section, the analysis of whieh isnot so direct, ‘The second part of the ehapter, Sees 344 to 3-11, is concerned with the more general analysis based principally ‘on the methods of G. N, Ward.t ‘The analysis for bodies of revolution suggests certain procedures used in the general analysis. ‘The third part of the chapter is concerned with slender configurations at subsonic speeds. No results for specfie configurations are considered here, but this subject is reserved for later chapters. ‘The emphasis is on the mathematical methods and general formulas, Therefore, the reader who would avail himself of specific results ean pass lightly over the mathematics herein, particularly the Laplace and Fourier transform theories. ‘The theory of this chapter is limited in application to that range of angle of attack of a slender missile over which its aerodynamic characteristics are essentially linear. Tt is further limited to steady flow in the missile reference system, SLENDER BODIES OF REVOLUTION 3-1, Slender Bodies of Revolution at Zero Angle of Attack at Supersonic Speeds} Sources In the study of bodies of revolution let us denote the potential at zero angle of attack, the thickness potential, by #, and that due to angle of 3 attack g.- To obtain a solution for the potential g of a slender body of revolution, its convenient fist to ret up the potential to the full aecursey of linear theory, and then to specialize the general results to slender bodies of revolution. ‘The basis for the linear theory potential is Eq, (2-18) for steady flow expressed in eylindvieal coordinates (Fig. 3-1) cur 8 (M4 MEL g an ar wherein BY = Me — 1, ‘The potential for a body of revolution at zero nglo of attack is constructed from axially symmetric solutions of Ea, ‘1), solutions not dependent on @, Some axially symmetric solutions of Bq, (1) are He 64 = RP cosh! . 1 on = RP Gr a 2) ss may be verified directly by differentiation. The second solution is the x derivative of the first solution. Tt is easy to see that g- and oy also satisfy Bq. (3-1) so that 2 and @ derivatives of solutions are also soltions ‘The solution @,, is sometimes termed the supersonie source with eenter at the origin beens of its abvious similarity to the potential for an ineom= pressible souree, 1/(e! + 1) It is intuitively obvious that a body of revolution in uniform flow ean be constructed by adding sources and sinks in just the right strengths along the axis of the body. Let the souree strength per unit length along the # axis be f(@). The continuous distribution of souroes (and sinks) represented hy J(g) ean be summed by integration to yield their combined potentials Se dé a= [eae PRR 3) “The sourees usod are of the ¢, type, and the limits of integration are pu posely not specified. ‘The limits ate established on the basis of certain arguments explainable with the help of Fig. 81. The Mach cone from point P will intersect the « axis at a distance x — Br downstream from ‘the origin. Downstream of this intersection no souree ean influence point P since the rogion of influeneo of a souree i8 confined to ite down ream Mach cone, ‘The upper lini is therefore 2 — Br. The sources start at z = Din the prosent ease, and f(é) = Oif €<0. ‘Therefore any lowor Himit equal to zero of less is possible. We therefore write a= ("ges on (=o F ‘Js to be noted that a potential Vy due to the uniform flow is additive to 6: to obtain the total potential,‘The source strength distribution #(@) must be determined from the shape of the body. To the accuraey required here the boundary eondi- tion yields with reference to Fig. 341 5) ‘The quantity S() is the erose-sectional area of the body of revolution. ‘To wiilize this boundary condition we must determine 36,/4r from Eq. Sgro cane 7 ‘Sytem F 4 : Zhi. i : a 5 [Reseed i, Ane and notation for badly of revatio a einen G4). Assume that J(0) is zero, and rewrite Ba, (B+) as ap 2 =F [10 come Tora “slender body," the body radius is small compared to 2, and the quantity (c ~ 8)/Br is lange except fora limited interval near the upper Timit, which we ean neglect, ‘The inverse hyperbolic cosine can then be expanded Sag co - 2a - 8) . cost et = og EEO + @ For a dender body, Ba, (8) therefore assumes the form * es wd [oie = pa - E[tome%e os ae 2. te en From Eq, (85) the source strength is directly related to the body shape seo = - "3 @10) and the potential from Bq, (8-8) is then 6 Se 12 [gy ogc Bat Piet 22 [i sowee oa orn SUBSONIC AND SUPERSONIC SLENDEL-BODY ‘THEOILY ar For purposes of physical interpretation, separate ¢, into a part depend- cent on r and a part independent of r. so log r + of) or ate) = + SO iogF - £2 ['s"@ tog e- Haz G1 With reference to Fig. 1-2 the thickness potential is the sum of a part which depends on the position in the erossflow plane A.A, and a part which has the same value for every point in the plane. ‘The part of depending on r is precisely the potential function for an incompressible souree flow in the erossflow plane. The flow velocities in the erossflow plane depend only on this term since g(z) has the same value all over the 2 Ay ® f Secon Aa Fro. 32 Source fw is cromaw plane of bay of revelation. plane, The g(e) term ean, however, influence the pressure coefiient which depends principally on 6,/a2, To obtain the function g(x), it ‘was necessary to specialize the full linear theory: potential to a slender body of revolution. We will consider next the eect of angle of attack, which is additive to that of thickness in a simple way, ‘The question of pressure coefficients and forees is left until later. 3-2, Slender Bodies of Revolution at Angle of Attack at Supersonic Speed; Doublets ‘The axis system and the body of revolution at angle of attack are oriented with respect to the uniform flows as shown in Big, 33. The component of velocity Vs eos a: along x causes ¢, a discussed in See oa ca selena, Fic, 33, Incompresble croailow around body of revolution at sngle of attack38 MISSILE. AEHODYNaMIES nd the component of the flow velocity oa, along « causes the potential component g now to be evalusted. Just as ¢ was constructed by dis- tributing sourves along the body axis, so 6. is constructed by superimpos- ing dipoles along the axis. First eonsider the dipoles formed from the axially symmetric solutions of Fa. (@-2). ‘The dipole is formed by placing ‘sink directly above a source of equal magnitude und letting the source ‘approach the sink, while keeping the product of the magnitude and disx tanies between source and sink a constant. This physical process is mathematically equivalent to taking the derivative of the source solutions with respect to 2. The constant multiplying the solution, the so-called ‘ipo strength, is of uo concern at this point; only the analytical form of” var the two source solutims of Ka. CE have the two corresponding dipole solutions: esine rin = RP fat P ba RPE pays Oe RP ep Consider now a superposition of dipoles along the body axis, If aCe) vere the dipole strength per unit length, we could form a dipole potential similarly as the souree potential was formed from ¢,, solutions ae ae) dg tities ee sing a f-™ _A(@r = | de “rt Jo (@- OF = BF Cau Fora slender aly of revlon, £3 Br, ond, (18 ako the se ome : a= hy aw) ‘he fnton Me) ow wo be determined Inter f the onda cee eet stack. Tie pte veel flow Pete ndinst nada ow in ect ne at he ds sce do ug tack ide ° Gan a dio + Vas) SUAOSIC AND SUPERSONIC SLENDER-ODY THEORY 39 with the result that Wz) = Vacant (1s) ‘The potential of « slender body of revolution due to angle of attack is thos simply sing Var 28 Guy) % ‘The physical interpretation of the potential gy is that of an incom- ppresible two-dimensional doublet in the erossflow plane. ‘There is no additive funetion such as (2) in Kg. (8-12) for the potential due to thickness. ‘The entire potential due to angle of attack could have been ‘constructed by eonsidering the flow in each erossflaw plane to be ineorn- presible, In fact, a simplified slendor-body theory based on this pto- cedure is deseribed in the next seetion, 8-8, Slender-body Theory for Angle of Attack ‘The distinguishing characteristies of flow about sender bodies was dis- cussed by Munk in his early work on the aerodynamics of airship hulls.* In this work he laid down the busis of Munk’s aitship theory which has subsequently been extended into what is now known as slender-body to, tt, As ut in Hendersbody theory. ‘theory. Consider a slender body, not novessarily a body of revolution, fying through still air at a speed Vo, at Mach umber Ms, and at an angle of attack a, and passing through a plane fixed in the fluid. The flow as viewed in the plane is nonsteady as the body passes through it. If, however, the plane is fixed in the misilo, the fow will appear steady. Let the 2,9. axis system be oriented as shown in Fig. 3-t with the # axis parallel to Vo, and let the crossflow plane correspond to 7 equsl a eon- tant, ‘The flow about the missile is governed by Eq. (2-81) for linear theory 2 9 woe oe Fe (320) If the body is sufliiently slender (or if the Mach number is close to unity), the fest term of the equation is negligible, so that we have tre + On @at40 snasite ApRopyNancs For an observer fixed with respect to the body, the flow in any plane normal to the 2 axis is thus the steady incompressible flow based on boundary conditions in that plane. Tt is independent of the erossflow in all other plaues, An exemple of the incompressible flow in s normal plane is shown ia Fig, 33. ‘The normal plane will be defined as the plane hhormal to the body axis. The planes # = constant are crossfow planes. For small angles of attack, the flow patterns in the normal plane and crossflow plane can be considered identical for slender bodies. ‘The foregoing simplified analysis of the flow about a slender body is generally applicable to the ealeulation of the potential due to angle of Attack as we have seen in the preceding section. However, itis not ade- ‘quate for obtaining the potential due to thickness existing at zero angle fof attack. The mathematical reason for this inadequacy is readily apparent, In descending from three variables in Eq. (8-20) to two vari- fables in Bq, (+21), we eliminated the possibility of determining explicitly the dependence of the potential on #. Por instance, any solution to Eq (G21) is still a solution if « function of # is added to it. Furthermore the tddition of a function £ will not change the velocities in the erosstiow plane, Such a funetion of 2 does, however, change the axial velocit fand therefore the pressute coefficient, which depends principally on this velocity. Tk turns out that the funetion of 2 is different for subsonic fand for supersonic speeds, This feature is the essential difference betwveen slender-bisy theory at subsonie and at supersonic speeds, as we shall subsequently see. SLENDER BODIES OF GENERAL CROSS SECTION ‘AT SUPERSONIC SPEEDS. 34, Solution of Potential Equation by the Method of Ward In the ensuing sections it is our purpose to derive the prineipal formulas of slender-body theory for supersonic speeds following the method of Ward.! Some attempt will be made to maintain mathemat- ical rigor and to carry order-of-magnitude estimations of the terms neglected in the analysia, ‘The escential method of the analysis is to find ‘4 general solution for the wave equation of linear theory, and to select those terms out of the general solution that remain under the assumption fof a slender body, In this way all terms that should appear in slender- body theory are found explicitly. “The body is assumed pointed st the front end, and is either pointed or blunt at the rear end. ‘Tho body length is taken to be unity, and the maximum radial dimension is . The angle between the free-stream direction and planos tangont to the body should be small, xs well as the rate of change of this angle with streamwise distance. The assumptions assure that there are no discontivuities in the streamwise slope, and hence SUBGONIC AND SUPERSONIC SLENDER-BODY TitBORY 4 vo sngulastes n gf one restos no singly in leat prsure ‘hich depends priaspallyon the aval derivative of, he must pose the aenal requirement of 0 Garon nserver TE nthe sain deta of any eros sion, the eurvetare nthe trom pan at ny pin on th bay whee convey outwand Sh be Of) Nosh martin noeary for points where the tedy in onoveinvardasina wing hehe, ‘Theloreysig eae toon simply ase thatthe prerbation velo du to the ye Soa conpard ohana ap ne py hee rn convexontvard and setinly of much greater ort tan ‘hesledarody theory gion infiteprtarbation vos sp peste coniintaat nach pin that ie ewnates of hs only mags ofthe tern eget in ended theory become inva Gries from trma nh vit and poly drag mn, poverty aceaty ihe oner of sgt det he fllorng oman il the lel prescott is pystally utenti it cool papers ree enrme epi ee enarree ape file The onder of ngnitde ofthe remainder tema nthe totais for the phil quantities dt the spproninatons ofthe anal al be given i ter ofthe masa ra dimmuon, which i cea sul etipared to nity (= 1y and which nds Let ¢ be the perturbation velocity potential for unit fre-stroam veloc- ity with the system of axes shown in Fig. +4. The perturbation velog- ities are then : a 4% % 3-8 322) [Exquation (2-18) specialized to steady low forms the basis of the present analysis Or + due — Bis = Bra Me —1 3) In th anata which ells ve wk a gnea solution of By. (2), a then pick ot the ters a expansion te goer luton spore eta sender boy” For supersne fw the mathematic Gal oe Yeniet for doing ths isthe Laplace tnfor theory. La Set rete Bay (2-23) torn of yaa edits Gen ‘The transformation we will use converts the potential #3, onverts the potential $(2,,0) into @ ‘transformed potential #(p,r,6) by means of the Laplace operator L. Helen ol = [Fo emeienn as (2252 sssIL aBRODYNANICS With reference to Churchill Lge] = 0 Mel oe (8-26) Ld) Lideal = 8 — pol0%.n@) ~ (047.9) ‘Since in supersonic flow there is po influence of the body for 2 less than zero and sincd # is continuous, we have that #(0*y,6) is aero, Also we y 2¢ (o+,7,8) is zoro on the basis of the following physical may assume that $8 (07,8) the basis of th phys snguent. 1628 (0%) jumps dont eng # = 0, we ean tke contin Dy an infiniti aking of te bad witout si 88 (or 20) as zero icantly intone mp. On hsb we take ‘ho phot argunent is not atuly resin, ad the mato tent wth 28 (7.9) not zero vil giv hese al reas 38 J waafermatin of Ba (20 fhe proved by Fraenkel.+ te eet At Bite oa) A solution of Fa, (3-27) ean easily be found by the method of the separa- tion of variables in the form + Se ip) sin nd + Delp) c08 n0l BPH) + Fe.) sinno + Pap cos mB) 28) ‘he funtion: Cp), af; ean be considered constant so far as Bi, (21) concede Asaly thay are arbitrary functions of p thes so as to ste the boundary conditions ‘The functions fy nd Ka ave modified Desel factions of th frst and recond kinds repedively For lange arguments they have the flloving asymp Sein Lon ~ espe (329) KG ~ (iq) @30) “The dominant term of the inverse transform of the [, function represents upstream waves increasing exponentially in strength along upstream Mach waves, 4+ Br = constant. The K,(Bpr) funetions, on the other SURSONIC AND SUPERSONIC SLENDEK-HODY 2HEORY 48 hand, represont downstream waves attenuating exponentially along the downstream Mach waves, — Br = constant. Tt is clear that we are thus interested only in the K,(Bpr) function’ except for possible rare cases, Por only the K,(Bpr) funetion Eq. (8-28) ean be written into the following compnet form by combining the sin n8 and cos nd by means of arbitrary phase angles 8,(p): # = Adpsaten + F AW@)KlBp om 0 + A] GBD ‘The next step in the analysis is to find the special form of Ka. (3-31) eppropriate to slender configurations. A slendor configuration is one characterized by the fact that its r dimensions are small compared to its F dimensions. We therefore seek a form of Eq, (331) valid for small values ofr. ‘The questions which then arise are: In what region will the neve form be valid, and how large an error oceurs in ® as eampared with from Eq. (831)? To obtain the form of Eq. (8-31) for small r, we note the following expansions of the Bessel functions for small values of r (ris Bale’ eonstant) Kelli) = = (4 ton EP) + 011 KuBer ~ 3 jel + 06+ ogni 2) KG = 2S GE 1 + oe ‘he dominant terms in therefore yield seo (re ee) +1), w-o1( We use the subseript zero to denote the value of © for small x, ‘The frac- ‘ional error iy is st most O(0 log r) and, if the Ay term is missing, then the error is OF). Inspection of Eq. (3-83) shows that the series con- verges ifr is greater than some value r;.. The series converges exteraal to a eylinder enclosing the body as shown in Fig, 3-5. The series will, Usually not converge inside this eylinder and does not represent: the soli tion at the body surface. Tt must be continued inside the eylinder by the ‘process of anslytical continuation. Sinee r: is some dimension of the same order of magnitude as f, the fractional error in by is O(¢ log (). ‘To establish the potential 4p in the j plane, we must take the inverse wform of Ey. (3-33) term by term. For this purpose denote the BY cere t ay 53) c“ ansive aboprstes ‘inverse transforms of the various terms in the equation as (a) = 11m Aafo wea) = 1] (y+ 622) Aun] G80 anaes 9-1 | GM (BY an ‘The inverse transform of Eq. (83) is then sige tng Seceratthtinet gay tis clear that dois the real part of the funetion 17() of a complex varie fee RPWG) sere 3) We) = alogs t+ DE oa) What we have shown is that go is solution of Laplace's equation in the erossflow planes, # equal to a constant, ‘The funetion by(2) is the Sy plane tot Fis, 85, Cylindrical contol aueace enclosing slender body. function left indeterminate in the simplified treatment of slender-body theory in Sec. 33, In the form of Eq, (3-35) the series converges outside a eylinder of radius r, enelosing the body. Although the series eonverges for lange values ofr, it does not follow that it represents the flow about a slender body for large values of r. This is apparent when we reeall that Eq. (3-88) was established by extracting from Eq. (3-81) for the full linearized theory those terms dominant for small values of x. Slendor- body theory’ is accurate, therefore, only in the field near the body. To ‘obtain solutions for slender bodies for distances far from the body it is SSUREONIC AND SUPERSONIC SLANDLA-HODY THEORY 6 necessary to retain the full linearized equation. ‘The slender-body poten- ‘ial gy has the fractional error in the form (g — 40)/6 of order # log t. 8-5. Boundary Conditions; Accuracy of Velocity Components First let us consider the matter of boundary conditions and then turn ‘oar attention to the accuraey of the Velocity components compared with those for the full linearized theory. Consider contours Cy aud C's in the crossflow planes corresponding to 2 and 2 + dz with the body as shown in Vig 3-6. Let the normal and the tangent to the contour in the erossfiow plane be v and, respectively. Consider a streamvise plane containing » Fic. 28. Boundary coition in streanive ple through sender bay. shown with section lining in Fig, $8, The plane is normal to and inter- scots an element of length dl of the body surface hetiveen Cy and C:. Lat vy be the outward normal tor and dl, Thus » is the normal to the surface and fies in the streamwise plane. In the streamvise plane the exact ion of flow tangeney is Voll + 09/88) _ Voas/av de ) a % (04 thie nthe cy HOU" ith esas eth on wa spel cate a a eo stn sn yong tty OG She mal tw enor unl aga sted cota ae bah Tintern 3) Ae 0). Wesco oa: silk nat ta (3-38) ay = Ol) a, = OLer8)46 MI8SILE ARRODYNAMICS Since differentintion with respeot to # does not change order of magnitude, ‘we have from Eq. (3-85) for the order of magnitude of the perturbation velocities (assuming by is not a dominant term in 36/02) 26 om =o eam $8 = 0 og ‘and for the potential $= Oe log) Let us now linearize the boundary condi the order of the error introduced thereby. Let us also consider the errors ‘due to the use of go rather than ¢ associated with the eomplete linearized ‘equation. We will then be able to tell which of the two simplifications tctually controls the accuracy of slender-body theory. “Linearising Bg. (3-38) yields simply 98 21 + O€F og 9) sf 340) 8 2s oy « 38 5 OW ogo For de we will se the inarized boundary condition doy Oy de Gan) Now the error due to use of for g was shown to be $= 404 OU loge) 2) “The error in ania velocity is the same since derivation by # does not change the anor of magnitude $8 = 20" + ote one ‘Phe exror in the erossflaw velocity components is 28 88 4 QC oF tO tow) a6 a (a3) 29 +. Ota 2" + Ol logs) sinoe the fractional error in the velocity components is the same as the fractional error in ¢o (as Ward has proved). It is seen that the error due to linearizing the boundary condition equation, Hg. (3-40), is the same as the error due to the use of gs for g, Eq, (8-13), 20 that the two simplifiea- tions are eompatible, Eq, (9-38), and estimate SUBSONI AND SUPERSOMIC StowonI-nODY THEORY a 8-6. Determination of a,(2) and by(2) It is possible to obtain the values of at) and bo(2) in Bry, (9-85) from the distribution of the body eross-seotional area along the body axis regardlessof the cross-sectional shape. The higher-order coefficients a(2) depend on the shape. The a log r term corresponds to souree flow in the crosstlow plane and is zero if the body’ eross section is not changing size To evaluate ay eonsider the contour K shown in Fig. 3-7. From the A Section da Fue. 37, Contour fr ovation of (2) integral of the ontilow neross KC berger [pear [C0 2a us) cas) rl Now invoke the linearized boundary condition (Hq, 3-41) aud reevaluate the contour integral about A Bongsu f Beste $asete = ferde=se) ey Here 3/2 nthe erneamtional aoa inthe eroiow plane, The na result for x se wn = 22 Da "he fueton 2) i wir fn any ernssetional plane ad yields nothing tothe eons veoity componente dors contribu tothe reste colicin bat not co tho loading. Now fom By. (2-4) 4a) = ttn) ~ tos [ Ay togE] ar The use of the convolution theorem' yields aay8 issiuss szRopyxanies where we have made se of the eondlition for a pointed body (0) This seen that by(2) depends on Mach number. 3-7. Pressure Coefficient With the magnitudes of the velocity components known, itis a simple matter to formulate the pressure coefficient from Eq. (2-20), For unit Vy, the pressure eooficient, including quadratic terms inthe perturbation velocity: components is Pa 20 Em + wet = 1) 9) If we ignore the last term, the error is O(0! log? ). Let us express the pressure coefficient with respect to velocity components u, x, and w along the hody aves t, y displaced from the axes 2, j,£ by pitch angle ae and Dank angle ¢ (Fig. 1-2), From Table I-l the velocity components 4, f, and ® are related to u,v, by Bw tasin g + wa, cose b= vcore + wsing 50) Direct substitution into Bq, (49) yields P= ~2(u— vaesin g + was cos) — (08 + a) + OU log?) 51) where we have disearded the terms aa, wus sin gy and nea, 008 ¢ 8 terms of higher order than ¢* log? if a, is O(). In terms of the angle of attack a and the angle of sidesip 8, we have P = -2(u ~ 28 + wa) — (+ 0) (se) Itis probably important to note that the superposition principle does not necessarily hold for pressure eoeficient in slender-body theory. The principle of superposition has been retained for the potential, however 8-8, Lift, Sideforce, Pitching Moment, and Yawing Moment ‘The lift Z and sideforce P ean be evaluated by taking the rate of change of momentum: in the and g directions through a control surface of the type shown in Fig. 3-5, The not transfer of momentum in the vertical direction through the eylindor r = ry and the base plane area Sy is tom [i oredpap er nsad)ase- fore( snd in the lateral direction Pa f (ve8888 + pons)as SUBSONIC AND SUPERSONIC SLENDEK-HODY THEORY 49 1s convenient to form the complex foree Fata 8818) (P)]as n) + ae af Bi 42\(8 4 28 2fE(1+ 38) 68+ 1F8) as, es) "The pressure coeffcent sequal in magnitude to th percentage change in sivolate presure, namely, O(log. By the hero rel : 3c isentropie relations the percentage change in density is the same. ‘magnitade 20 ‘that i so that Mare aff Fat + 0@ logo (8-56) Also we have the relationship ar, ay ~ ap aE ah 08, 08 oop ar se relationships reduee Hq, (3-55) to aout aos aa 2oedTT as, * [Brass [0g geet 2 [Cg +8) as. ore 238 Eq. (8-58) since the variation represent ition represents in part the error term, ‘The first term is simply handled by Stokes's theorem, The contours C and K ave shown in [Co +B) as, ~ if, as +69 +i, onan + ian =i feeots— [retin [ae en Hf we substitute Eq, (8-59) into Eq. (8-58), we have BoM fomas~ ff [° (ote - star aS: vanes + O(log?) (8-60)0 MISSILE asaoDy Naat ‘The terms 22%, UF, A re all velocities in the eroseflow plane, whi The terms 288, UE, A aro all volities in the erosflow plane which decrease as I/r or faster as r approaches @. As a result the double integral is zero in the limit. ‘Thus a2 dh + Oe) “The contour € is the outline of the base intended to pass around any Singular points that may oceur on the body surface. It is noted that the force depends only on the line integral of the potential around the base. Since $s depends (except for a constant) only on the base configuration and angle of attack, we have the simple result that the foree depends only fon the base characteristes and is independent of the forward shape of the body. ‘The formation of vortiees behind the position of maximum span ean modify this result for wing-body combinations. If jy is the center of area of the base, the complex foree F ean be ex pressed as in the following form (derived in Appendix A at the end of the chapter) f- say + 28'(aa(L) + (3-62) r op a tenPa, + 28009, +2500 + 009 % = (3-64) te Pay + 28e, + 280) + OC log? gingls ‘The quantity P is the sideforee, and Z is the lift "To obtain the moment we ean write Bq. (9-62) for the force at any axial distance and integrate the local loading times # to obtain the moments. ‘Thus, if AL, is the moment about the J axis (positive when 2 moves ‘toward 2) and 3s is the moment about the £ axis (positive when # moves toward J), we have wma [are s-i['rou 269 MO gin) — os [st exisesin etn Sat el 48) ‘ts oF the cut endo boy My = = te P (oi) + de® faa = 2IP {S"(1)ao(1) + S30) — SC3(D) My 4 RP (a) — Ae RP foe + 2RP [SA 30(1) + SC3/"() — S(T "The quantity My is the pitching moment, and Ma is minus the yawing (ot 3-9. Drag Force ‘The drag formula of slender-body thoory is a widely used result which {or special types of slender bodies exhibits elegant mathematical proper- ties. In the derivation of the drag formula, use is made of a eylindtical control surface as shown in Fig. 3-3. It is ensy to set up the drag foree in torms of pressure and momentum transfer. 0 fo ovtoneras~ fore (s +32) seas, = [fps ore (i482) ]oss— mse eon ‘The symbol py stands for the static pressure acting on the base. To simplify Ki, (8-68) we introduce the emniservation of mass By multiplying Eq, (3-69) by W's and subtracting it from E simplitivation is achieved within the framework of exactness. ne mas. ~ [prod teas, [foros (04%) ]os.— nse = be 86 _ (aed 4 1 (aan og = [Gey +8CBY] + oerme on With these approximations, Eq. (3-70) now become: none feaees + fe) + ey Jos Ge = PoS() $ Ot log? 9 (8-72) (468) some52 MISSILE: AERODYNAMICS “whore Pr is the hase-pressure eoolficient. From the results of Appendix B at the end of the chapter, Eq, (3-72) takes the form 2 ade! dt + Draatball) ~ foo 3dr = PrS{i) + OC lo) te fasted — arodinbaty ~ $65 de — PaS() + O(log?) (8-73) ‘The drag ean now be evaluated since the values of ag and fy are given Ped! [meeigoos outa [veh = fo 82de — PoS(t) + OUClog) (3-74) (ode ‘This is the Ward drag formula for a slondor body. It is interesting to note that the drag represented by the first two terms depends only on the asial distribution of the body eross-eetional area and is independent of ‘cross-sectional shape. The seeond two terms depend on the slope of the body eross section at the baso only. We will investigate the various terms of this drag formula at eonsiderable length in See. 9-3. ‘Two important classes of slonder bodies result in considerable simpli- fication of Eq. (474). ‘These elasses oreur when the base is pointed or when the body is tangent to the eylindrieal extension ofits base. Tn both instances the drag formula reduces to the symmetrical form Dip fy, 1 i Bad [fog Ly Ss @ deds + O01) — PaSL) 8-75) 2 zh gS"OS'@ de de + O4¢og?O ~ PuS(L) (48) Minimum drag bodies ate derived on the basis of this result in See. 95 3-10. Drag Due to Lift ‘The following treatment is good not only for supersonic speeds but also for subsonic spoeds, a8 we will subsequently show. When a body develops lift, it develops.a wake of one kind or another. A lifting surface usually develops a well-defined vortex wake. In this ease the contour C must be enlarged to enelose the vortices as shown in Fig. 386, Another kind of wake arises when flow separates from a surface under angle of attack as shown in Fig. $82, We imagine a dead water region to form in the separation region which is then enclosed by vortex sheets. ‘The wake ean then be considered a solid body extension and the contour deformed to enclose the dead water region, The foree acting on the body cnlarged to inelude the dead water will be th same as that on the solid SUBSONIE AND SUPERSONIC SLENDEI-RODY BHEORY 38 boundaries of the body since the resultant foree on the dead water region ‘An inspection of Eq. (9-74) reveals that the drag represented by the first two terms is independent of the lift, depending as it dees only on the axial distribution of body eross-seetional area. ‘Thus the drag due to lift is tobe found in the integral about C, negleeting any changes in the base- pressure coefficient due to changes in angle of attack. To evaluate the «drag due to lift we must inspect this integral under the conditions of no @ Fig. 3:8 Separator vortex Hows routing distortion of contour of integration [a) Body; (0) Winesbody combination Tit and of lit. Let the potential g be composed of a part gus at zero lift and a part gu; duo to lift = dob bon @6) ‘The integral about the contour ( of the base becomes 2, gu, $05 ae fou 2B dr +h, 628M dr tT) ‘Tho first integral is not part of the drag due to lift. ‘The second and third integrals are coupling terme betsceen the potential at lit. and zero lit, while the fourth integral is « “pure” lifting effect. The derivative 6m/4» in aecordance with the boundary condition, Eq, (8-41), represents ‘the change in the streamwise slope of the body surface due to angle of attack aif the angle of attack is arbitrarily taken to be zero at zero lifta issita: AERODYNAMICS With reference to Fig, 3-6 the potential gn must produce a velocity normal to the body sulieient to offset the component of the free-stream velocity normal to the body. Thus 280 = a 008 (rt : 280 — a cos (n8) (78) where cos (x2) isthe direction gosine of » with respect to the Zaxis, ‘The sevond integral then becomes §, 6088 “tu sen nes eo bene of (61 ne the ir for Fe Slat neg ot sls of Spend Ca th ond inc May the ede to it sree ald othe oth ier hh by Be af. tmom (adr = of, ond = 0 (70) 680) Sgnin Bq. 61) shows that the hit is Za 26, odd @sn “The drag duo to i for constant hase pressure is now « 0 : D=Di=$2~—f, sn8ttar as where Das the drag at zero lt, We ean put lq. (3-74) into the follow- ing form for lft D = De+ $24 APaS() + Ol log" 83) where AP» is the change in base-pressure coelicent due toangle of attack, ‘The physical significance of Eq. (882) is that the lift ereates a drag. 92/2 rather than aZ, which would be expected for a flat plate. ‘Thus the resultant force on a slender configuration due to angle of attack is inclined backward at an angle a/2 from the normal to the free-stream direction, 3-41, Formula Explicitly Exhibiting Dependence of Drag on Mach Number “Let us divide the drag given by Hq, (8-74) into parts dependent on and independent of Mach number, Examination of Eq, (3-74) shows directly that any part of the drag dependent on My must oecut as a result of the contour integral about C. To obtain this part, let us write ta bow ot sy ‘rhe discussion following Eq, (8-37) and Bas. (8-16) and (8-18) shows that all the dependence of yon Jf, enters through bo, 0 that ¢” isindependent fof Ms, Also by is uniform in the plane of C. "Thus dbe/@y is zero, and the contour integral can be written fowiternn$,% ‘The first integral is seadily evaluated by means of Ha. (8-15) ae fein ase hf Bftte = aad Introducing these relationships into Eq. (474) yields the desired drag equation ». =wur o Jog # — PS) tah [me © Ce epeoa §,e8e Foene® (89 "(SD dg de “The first two terms depend on Mach nuraber. If the base is pointed, S'(1) is zero, and the first term is zero as well asthe base pressure. The drag on the basis of slender-body theory is then independent of Mach ‘number (neglecting separation over the base). Ifthe hase is tangent to its own eylindrical extension, $'(1) is zero, and the only effect of Much amber on drag is through ita inftence on base pressure. The potential ‘6° in the equation is just that potential which would be obtained by plying Laplace’s equation to the flow: in the erossflow plane as described sn Bee, Se SLENDER BODIES OF GENERAL CROSS SECTION [AT SUBSONIC SPEEDS, 8-12, Solution of the Potential Equation ‘The treatment by Ward of supersonic slender-body theory’ has ite ‘counterpart for subsonic flow. Mathematically, the difference is one of ‘sing Fourier transforms instead of Laplace transforms. Actually, the ‘entire difference betsveen the subsonie and supersonic eases enters through the fy term. ‘Thus all results derived for the supersonic case not depend. ing explicitly on bs are unchanged for subsonie speeds. Let us now find the operational solution to the potential equation, Eq. (3-24), on ‘the basis of Fourier transforms. Consider the Fourier operator F and ‘the inverse Fourier operator F defined by the following pair of reciprocaloo ssn, axmoprNaatos relationships i Fetenol = [7° e oar de = wer) Peril 3 [2 meer de = een — 87) Note the use of the complex Fourier transform and the placement of the 2r factor. Integration by parts establishes the following transforms for 2B derivatives of @ r[2e]=eoll — it 1 [28] wom 28 | — sae a a It wo can invoke the boundary conditions abe) = a=) 0 erry 80) Bip) = Ba, (62) beeomes ow ate (3-90) Again, as in the ease of supersonic flow, a suitable general solution of Bq, (3-00) for the prosent purpose ean be obtained by separation of vari- ables, Tn fuet, the solution is of the following form in complete analogy to Bq. (3-28): oF nna) sn nt + Fle sn) 4 In{Boar)[Cafo) sin n + Dale) 008 n8) (3-91) ‘The value of w ranges from —- to +2, snd the arbitrary functions, Cala), Dales), ete, ae to be suitably chosen so that (1) the behavior of ¢ is not divergent as r—> -, and (2) ¢ is real. ‘Tho requirement that ¢ is not divergent as r—» = transforms to the requirement that not. be divergent as r+ =, sineo the transformation doos not involve r. We must discuss separately the cases for positive w and negative «For positive w, we have already seen that J,(Busr) vasies as e', and is not admissible on aceount of the first eondition, ‘Thus Cx(e) = Dye) =0 20 (92) For negative values of o, we rust make use of the relationship between Bessel funetions of negative and positive arguments. K(—2) = (“1)Ka(@) ~ xiT@) E(=2) = (= Dn) oe SUBSOSIC AND SUPERSONIC SLENDER-DODY TOMY 8 ‘The sin nd part of the solution then becomes Ble) K—Bust) + Cs(—0) Tal — Ber) = Es(—a}l(=1"K (Baer) ~ iT (Bar) + Cal—a)(—1)*Ta(Bosr) (8-04) “The [.(Busr) terms must have zero coefficient ifthe behavior asr-— = is not to be divergent. This behavior is assured if =riB,(—w) + (-1)*C.(~w) = 0 (3-95) We are then left with the solution Be) Ky Boar) + Cyl —0) (Boer) E{w)(—1)K (Boor) (3-06) If the coefficients of K,(Byer) are chosen to be new functions as follows, Bales) wo “D(-) a <0 Ada) = Fale) ao AUF(-w) a <0 Gon the general solution of Eq, ($01) with the correet behavior ean be expressed as 2 SK bilolr Ble) sin ns + Ado) cosn6] 08) Ie should be noted that Bg (0-97) des nt place any condition on Aa(s) rnd Bu) see (a) and Pu) are quite arbitrary. ‘The second condi tion that @ bo real ean be simply satin by choosing Ag(—0) = Ato) BAR) = Bx) oe Equation (98) is the solution in tho transformed plane of the ful Tieareed equation which is appropriate fr subsonie speeds. ‘The value fig it gives wil become sual as r= und will extend upstream and lownecteam, ‘The problem now is to extract from the full linearized solution the special solution suitable for slender couigurations, The problem i salve in exactly the sme manners for the supersoie ese by expanding as given by Eq, @-08) in a series valid for small , and rvtaning the dominant terns. In fact, the expansions for pareidential in form «u(2) log men + Dy sees gonne $ bar(e) sin m8 (3.100) 6 but the coeficients are now determined as inverse Fourier transforms38 user APRODYNADTES ‘rather than inverse Laplace transforms as(e) = =F Ate] boa) = -F by + tog 2 Ja) en 2 ‘The only term that ean differ from that for the supersonic ease is by). ‘The rest of the terms in ¢ are solutions to Laplace's equation in the eross- flow plane aud are uniquely determined by the boundary eondition in the ‘rossflow plane regardless of the Much number. 3-48, Determination of ay(#) and bu(2) ‘The value of ay(2) in this ease is presisely the same as for supersonic speeds since the part of ¢ involving as is independent of Mach number. Thus se (6-102) ate) Phe funtion bo(2) is obtained from Bq. (+101) on « purely operational basis bute) = (7 + toe!) ane) — Ft og ll 408) ‘The inverse transform of a product of transforms can be obtained by ‘means of the convolution integral PHGLaHCal = fF aeme — 8 te G09) ‘To insure the existence of the separate transforms let Glu) = wale) se) = 2 @-105) so that sa) = ig) and aca) = [77 lel onde os fm sai (2-108) With the help of Brddyi otal na) = dy + toga) a>0 107) nay = g(r tlogit) 2<0 SUBGONIC AND SUPERSONIC SLENDER-BODY ‘TitFORY 39 Finally with the help of Eq. (3-94) 4a) = ona) om 38 — 5 [ov ow 8 — 0 88 1 fa : 49 [Poste ice = 228 — reas tog t = Has") log (— 2) (6-108) 43-14, Drag Formula for Subsonic Speeds; d'Alembert’ Paradox ‘The drug fore for subsonic speeds willbe developed from Eq. (73). ‘Though wa have developed the formulas for subsonieslender-body theory othe basi of ponsle hunt bac, S(1) 0, auch a body ail nt ful tho requirements of slenderess A blunt baee in aubeonie flow ean send strong upstream signals, which it ennnot doin supersiniefow bootuse of the rule of forbidden signals. As'a consequenco we muct now assume that $1) and $C) are both zero; that is the base is pointed. With S'd) =, 62) booomes [fr $2) coutinuous and $0) = 0] Bs bu(2) = a0(2) log b [ae wee — 948 +} foc wee - 2.48 (100 By Bq, (2-78) the drag is then pene Be Lig BisP - OF sa ['s@ ete - pacae -ef @ ['s"@ toe eva +x [iso [som ce deas— f, 28a PaS(l) + O(log") 410) = fo, 28h + 0 og) coun) D [Noting the next to last equation of Appendix B which follows, we have with a4(1) equal to zero Puno o Slender-body theory thus yields d’Alembert's paradox in subsonie flow as it should, SYMBOLS (2) goeffcient of log term in expansion for do s(n) arbitrary function of p (2) eoeffcients in expansion for (3) x(a) arbitrary function ofRP si aie Ve wo we) ane BaF os SMHsSiLE. AERODYNAMICS cooficient in expansion for de Gre ~ a= arbitrary funtion of ding free drag force at zero Lit Source strength per tnt € distance Pi Foner transform operator, and inverse Fourier transform ‘operator dipole strength per unit & distance modified Bessa functions of fist and second kinds Laplace transform operator, and inverse transform operator My + iM free-stream Mach number moment about axis, pitebing moment tnoment about 2 axis, negative yvsing moment atic prewure, variable of Laplace transform free-stream static pressure thane static pressure pressure eoufcint, (p ~ 93)/a6 bse pressre eoficient fee-stream dynamic prewure radivn vector in J, 2 plane (also in 2 plane in Sees, 3-1 and 32) local body radius radius of eslindical control surface, Fig, 95 Feal part ofa complex funetion trea of slender configuration in erosslow plane ‘naximtem madal dimension of slender configuration perturbation velocity components along 2, 1,2 Perturbation veleity components along 2, 9,2 Fadial velocity component froe-stream velocity complex potential ¢ + conjugate complex potential, ¢ — #7 Drineipal body axes, Fig. 3-1 body axes for ac =O antl ¢ = 0, Fig. St Feoordinate of eeutzoid of S(2) Sideforee along 9 axis coordinate of centroid of S(2) forae along 2 axis, lit pee nei SUBSONIC AND SUPERSONIC SLENDER-DODY THEORY 6 ® angle of attack, a, 08 ¢ ‘te included angle between body axis and free-stream velocity, 8 angle of sideslip, a, sin g 7 Euler's constant, 0.5772 3,(p) phase angle ° polar angle in z, 9, # coordinates i ‘normal to body eontour in erossflow plane E variable of integration ’ local mase density’ » free-stroam mass density : tangent to body contour in erossflow plane ‘ ‘general potential solution of Faq. (3-24) ° angle of bank approximation to ¢ valid for sender configurations potential for a doublet potential for a source, an axially symmetric potential Laplace transform of Fourier exponential transtorm of @ transform of 4 vvariablo of the Fourier transform [REFERENCES 2, Ward, G. Ns Supersonic Flow Past Slender Pointed Bodies, Quart. J. Meck cand Appl. Moth, vl. 2, part I, p 94,1010. ‘2 Heaslt, Max aed Hatvaed Louas: Superonie and Transonie Smal Pestur- tation Theory, see. Din “General Theory of High-peed Aerdynamics" vol, VIF "“High-pond Aerodynaiies and Jet Propsion,” Princeton University Peay, Pinos to, 1998. ‘3 Munk, Max ML The Aerodynamic Forees on Airship Hull, NACA Teoh. ep. 184, 1024. ‘© Chureil, Ruck V-: “Operational Mathematics" 24 ed, MeCraw-Hill Book Company, Ine, New York, 1958. 1. Frat, L.E.: Oa the Operational Form ofthe Lincarised Equation of Super soni Flow, J deronout. Sey v9. 20, ne 9, pp. 047-048, Tenders’ Forum, 1053 1. Bry, A. (ed): “Tables of Integral Transforms,” vol. 1, MeCraw-1ill Book Company, Ia, New York, 1954. APPENDIX 5A In See. 3-8, the complex foree, F = P + 42, was put into the following oem g I log? f) Fa if, outs + Of eto ean) ‘vis possible to find 4 somewhat more appropriate form for ealeulativee seria: axmopyNanies purposes by replacing ge by IF ~ ie Ea -21f,was—2 fous (a2) ‘Some care must be taken in connection with Eq. (BA-2) because the ‘expansion for 1V(3), Eq, (8-87), contains a logarithmie term whieh is not single-valued WE) = ostogs + bo+ Fos eas) ‘To make the IV oto © as shown function single-valued, we put a eut in the j plane from ig, 34, and the argument of the logarithm inereases io. 3.0, Distortion of eoatour in et plane. by 2nf every time 3 crosses the cut. Now the contour @ encloses the cross section of the body base but indents auy singular points of 1V(3) a8 shown, ‘The nature of tho series in the expansion for W() is such that it converges if ls] is greater than the largest value associated with any singular point. The series for 17(@) will not converge on all of C, and s0 ‘we expand the contour to K’ on which Eq. (84-8) is convergent. ‘Then $owds=f,.W a= fp (arlogs+ urea = ao(2rijs) + 2riay (BA-4) Note that the value of the integral depends on where the cut starts, We will get « compensating term from the other integral SURSONIC AND SUPERSONIC SLEXDEI-DODY THKORY 8 For this socond integral, integration by parts yields $e oa = Wale ~ fae » wldale — 3 Fede GAs) since Gao) We havo the geometric integrals giving area and moment of area 80) fete» toe wan a a Beason = 6, i92 dr a8) ‘he nora hus tees $oteis=S0)n= SHAD = VAIS) Finally from Eq. (A-2) there is obtained F Aray(l) + 25's) + 2480) (Baio) APPENDIX 38 We now evaluate the two integrals of Eq. (3-72) = [228 a oi n £[@yeaegyje OP ag! 08 nd + Now @ = aalogr + bo + (3B) 0 that (noglecting the sin né terms) ae = fi (aay log rs + abe) dt @B4) +64 MISSILE AERODYNAMICS ‘The same result is obtained with the sin n# terms. Let us operate on Is by Stokes’s theorem to convert it from an integral over Ss to integrals about C, its imer boundary, and , its outer boundary. v= [AC <2(2) = = fo adttae se [a SBrte B5) a = fn (aslogrs +b Beat fn lee “Serge eye = alo og nF ab) Tous Ih Belatlogrs + ad) ~ fon 28er an) APPENDIX 3¢ ‘The integral to evaluate is that of Eq, (377) I do $. 02a acy where the contour Cis shown in Fig. 5. We ean subtract the integral of Bq, (3-70) from 7 since it is zero = fotuthtir = feouien 02 ‘Consider the contours Cand K enclosing Sy, and apply Green’s theorem to area Ss, 1a be (oot 8 ~ out) ar + = guP!on) dSs (CB) Here ¥* is the Laplace operator, and ie and gu; are solutions of Laplace's ‘equation. Hence, (acs) ‘Since the integral has been transposed to the contour K, we can use the ‘expansion for gu and gn. which converge on K but not on C, On K the SURSONIC AND SUPERSONIC ALENDEL-NODY siHLORY 6 expansions for gue and dry have the form (neglecting sn nd terms) bu = clog + ev + yen = ese 208 no where &, ey and dy are constants. The form of the integrand then Decomes (acs) Ms ose-+0(2) +0(PEE) aco = en ®t) ao = 0(8") on Since J does not depend on r; because the drag cannot depend on the radius r; of the control surface, we ean let ry approach = o(") acs) =0 aero 1cuaprer 4 AERODYNAMICS OF BODIES; VORTICES In the present and ensuing chapters we will be concerned with applica tion of the general results of the preceding chapter to various types of ‘eonfigurations such as bodies, wing-body eombinations, and wingchody- tail combinations, Coneurrently, it will be our purpose to investigate how departures from slenderness modify the slender-body results, as well as how viscosity introdices additional effects, some of whieh’ ean be treated by extensions of slender-body theory. In the first half of the chapter inviscid slender-body theory is applied to bodies of eircular and clliptical cross section. Also, the theory of quasi-eylindrieal bodies of nearly circular eross section is treated, No diseussion is included of non- Tinea theory or of nonslender bodies, since for zero angle of attack these ‘subjects are considered in See. O-t in connection with drag. ‘The appearance at high angles of attuck of vortiees on the leeward ‘of slender bodies constitutes one of the most important single causes of the breakdown of inviseid slender-body theory. However, in one sense the slender-body theory has not failed at all, But rather the slender-body model must be generalized. In fnct, if discrete vortices are introduced into the slender-body model to account for the effects of viscosity, itis not difficult to extend slender-body theory to include the vortex effects. ‘This is the principal purpose of the second half of the chapter. Results will be obtained for slender configurations with pancls present. INVISCID FLOW 41, Lift and Moment of Slender Bodies of Revolution, In Soes. 341 and 32, the potentials were derived for slender bodies of revolution at zero angle of attack, and at angle of attack by introducing the assumption of slendemess into the solutions based on the full linear~ ied theory of supersonic flow. The potential for a slender body of revolution due to angle of attack, Eq. (8-19), is independent of Mach ‘number, so that the distributions of ift and sideforee along the body’ are also independent of Mach number. Let us use the general formula, Eq. (8-62), to ealeulate the forces and moments on a slender body of revolution. ‘The body is taken oriented with respect to the , 9, 2 axes as ABRODENAMICE OP BODIES; VORTICES or in Fig. 4 at angle of attack ae Ite centroid then lies slong the tine p= iad. Sines S(2) and 4) are known, it remains only to doter- thine a: in Eg. (262) to obtain the forees. Although only the potential due to angle of attack creates lift or sideforce for a body of revolution, the cocticient oy arises as a result of both angle of attack and thickness because of compensating terms in Eq, (3-62). We can, however, ignore bet) in Eq. (3-47) since it has no contribution to a;. ‘The complex potential for a slender body of revolution consists of the part 1) existing at zero angle of attack plus apart I.(j) due to angle fof attack, ‘The part at aero angle of attack isthe sun of by() and a logarithmic term proportional to the rate of body expansion, With reference to Table 2-3, the equation for W,(3) with due regard for shift in crgin is WiG) = bolt) + ry log (3 — a) rot) + - 2 log Gs — ix) (4-1) where ri the loeal body radius, From Table 2-3 the complex potential for angie of attack suitably modified for shift in origin is Wa) = im [6= 30 42) ‘Tho entire complex potential with Ve = 1 is [eset —2)] ~# [6-0-7 Expansion of this equation yields the coefficient a;(2) of the g~* term: WO 4 am 1 FO 50 ws All the necessary quantities are now at hand for evaluating P and 2 from Eq. (6-02): +1 = ~28(2'@) = rate) Sage ws Fao Zarasie ‘The lift per unit axial distance along the span of a cone-ylinder has been caleulated by Eq, (4-4) and is shown in Fig. 4-10, A similar ealeula~ tion has boon made for a parabolic-are body and is shown in Fig. 4-16, Since (8) i linear in #for a cone, the lift distribution is Linear as shown,os sussiLe AzRopYNAMrcs Behind the shoulder of the body where 5"(2) falls discontinuously to zero, the lift distribution also falls to zero on the basis of theory. Unless the body is very slender, some measurable lift. would intuitively be expested to be carried over past the shoulder, and in practice such is the ease ‘The scoond example exhibits equal areas of positive and negative lif ‘The net lift on the basis of Bq, (4-4) is zero for this case in inviscid flow, since the base area is zero, The body boundary layer will usually not — ‘ @ : ‘ ;

ce (be) Feo, 4 Lilt disteibations for slender bodies of revolution. (2) Cone- er We are interested in obtaining the pressure cocficient as follows, = (09/80) Ve “whore L-1 denotes taking the inverse Laplace transform. Before taking the inverse transform, et us write Faq. (461) as exsatrocrr [(GzHe0 4 3) — 3] aan Cw) so that * Po Fe) 400) where we have let a ~ 1 without any loss in generality. ‘The technique how employed isto split the expression into two parts, one dependent of the boundary conditions as represented by Fa(p), aud the other independ- tent of the houndary conditions, as follows: Lfyle ~ r+ D] = Palpveree-2 469) Kun, 1 rate) Pr) 6 UW] oth (4-65) ‘The part independent of the boundary conditions represented by Pa. (4-65) has been made the basis for the definition of a st of characteristic functions Wr). Assuming that these functions are known, we ean Write the inversion of Eq, (4-08) by the eonvolution theorem that givet the inverso transform of a product of transforms, We thereby obtain the pressure coefficient from Eq. (4-62) 03 a ay, (ee [Ho (&-3 ai-g Ze cos nd (466) sukovyNastes OP BopmEs; vouTioEs 83 rs result has been ysritten for eylinders of any average radius for any Jlipersonie Mach ntimber, The equation can be used to ealoulate the pressure coeficient of any quasi-eylinder of neatly circular cross section Pgpocifed by Eq. (4-52), ‘The W(z,7) functions required for the ealeu- fation have been tabulated elsewhere. The caleulation is mado by numerical or graphical integration. In the reference the physical sg hifeanco of the TV,(r,7) funetions is discussed. ‘They represent down dream pressure waves associated with a sudden ramp on the body surface ustrative Example {ro shoxe how Eq, (1-66) might be used, let us ealeutate the pressure distution on an axially symmetric bump on a cireular eylinder as =] show in Fig. 4-U1. The equation of the bump is taken to be heosut(i-2) an eee 07) exe roa Kg, (152) the fae) funotions are 5(, 22 ma) ~13(1- 2) ose =o eo Only one term remains in the summation of Bq. (4466) for the pressure cveliciency sussiL AsnopYNaines fuli—te=e*a9] P @ @ war Lome iy tas tt HBL 0-2) bt aa} OS2se 00) fe B ‘The pressure coefficient on the body has been ealeulated for several values (of B, and the results are expressed in the form of BP/r in Fig. 412. Tho eymbol r indicates the initial ramp angle as shown in Fig. 411, Of Pie, 412, Prenure ditibuton on cies eyinder with axially aymmeteie bump. Interest is the fact that pressure eoeficent alway’ starts off with a value of 2r/B. Such a value corresponds procisely to the Ackoret value, the value to be expected for the full linearized theory and two-dimensional flow. Since the flow is essentially two-dimensional to start, the result it tobe expected, However, as the flow eoutinues downstream, it sees part of the bump in its forward Mach cone ns eurved rather than on a flat sur face, Ifthe bump had remained at, we would eoutinue to have only the first term of Eq. (4-09). ‘The seoond term thus represents the influence oh the curvature of the surface on whieh the bump is fitted. In this senst the second term represents throo-dimensional influences. If Mf is large ‘the second term is small, Such a result is in accordance with the fact that the upstream Mach cone has a narrow field of view and eennot “ste” ‘much curvature of the body. As B approaches infinity, the upstream Mach cone “sees” only a planar strip of body so that the ealeulatee pressure eoeficiant has tho local two-dimensional value everywhere, ARNODYNAMUICS OF MODIS; VORTICES 8 voRTICES 4-6, Positions and Strengths of Body Vortices ‘The subject of this second half of the chapter is body vortices. ‘The appearance of vortices in the flow can cause significant departures between experiment and inviscid slender-body theory. One of the most direct ways of illustrating the effects of vortices is to examine the pressure distribution around a body of revolution at high angles of attack. Such eaee Fi 4.13. Pressure distributions sround body of revolution; comparison of theory and 4 pressure distribution taken from Perkins and Jorgensent is shown in Fig 4-13. In this figure the experimental pressure distribution is com- pared with the theoretical distribution predicted by inviseid slendor-body theory, Eq. (4-25). According to slender-body theory, the pressure dis- tribution on a nonexpanding body section is symmetrie above and below the horizontal plane of symmetry; that is, the positive pressure existing on ‘the windward face ofthe body is aso resovered on the leeward face of the body. An examination of the data points reveals that no such prossure recovery appears. In fact, somewhere near the side edge of the body the pressure change ceases, and a fairly uniform pressure level exists over the {op of the body. The lack of pressure recovery is ascribed to the body86 AUSSILE AERODYNAMICS boundary layer, which soparates from the body with the resultant forma- tion of a “dead water region" of more or less uniform pressure on the leoward side of the body." The boundary layer itself rolls up into vortices, Let us now examine the vortex formation in greater detail. ‘The goueral features of flow separation on bodies of revolution at supersonie speed have beon studied by Jorgensen and Perkins,* Raney,? % 4 "SSS, (th i io. G14, Crossflow vortices of boy of revolution. }— « —j ro, 4.16, Location of vortox seperation for body of revelation. and others. These features are illustrated in Fig. +14, As the boundary layer flows from the underside of the body around to the leeward side, it ‘separates along a lino of separation shown on the body. Aftor separating, the boundary layer continues as vortex filaments, whieh rise above the ‘body and curl up into strong body vortices on each side of the body. As the body vortices proceed downstream, more vortex filarsents originating at the separation lines feed into the cores and increase their strengths AuROUYSAMICS OF RoDIES; VORTICES 7 Cone of the pertinent questions is: At what distance zs behind the body ups do the body vortices Srst form? The distance will depend strongly tn the angle of attack; but, sinee the controling phenomenon isoundary- layer separation under presure gradients, the Reyuolds number and Mach number are also involved, a indood is the shape of the body ite. Some data exist® forthe dependence of zs on as. These data are repro- ‘ced in Fig. 415 for an ogive-ylinder combination ata Mach suruber of 2 AC the higher angles of attack, the vortices tend to originate at the ‘ody shoulder. This is reasonable, ince the expansion of the ody in front of the shoulder tends to thin out the boundary layer aud inhibit Fie. 4.16, Nondimensiona vortex strengths for bodies of revolution separation, ‘The precise location of vortex formation eouldl not he ascer- ‘aincd, but rather a region of vortex formation was obtained. tis possible to obtain a nondimensional correlation ofthe strength and position of the body vortex cores as a funetion of z and a, on the basis of ortain plausible arguments. Consider the body vortioes as seen in planes normal to the body axis. Assume that the change in the pattern of the flow with changes in is analogous to the change in the low pattern about a two-dimensional eyliuder with time if itis impulsively moved normal to itself at velocity Vy. If zero time corresponds to the distance “e, then time and distauee ate related by Vat aos Vue (470) ‘The nondimensional parameter which characterizes the impulsive flow88 MISSILE AERODYNAMICS ApNODYNAMICS OF BODIES; YorTICES 89 05; any eT = By analogy the corresponding dimensionless number for our ense is oe of yesenso (a2) 2 Bsr cinger 2 Gin cinder (nodes) Ifthe anslogy is correct, then the vortex strengths and positions in non dimensional form should eorrelate on the basis of NV alone for different values of 2 and a. L ‘The analogy has been tested,* using data from Jorgensen and Perkins,* oa and Raney.’ The measure of the nondimensional vortex strength is sete T/2rVviow. ‘This parameter is shown asa funetion of N'in Fig. 416. A - rough correlation exists. It must be remembered that correlation is hampered by experimental difilties of measuring T. The vortex posi- tion® are simply specified by tho nondimensional quantities ys/e and v/a, These quantities are correlated as funetions of N in Fig. 4-17, and the correlation is considered fairly good 4-6, Forces and Moments Due to Body Vortices; Allen's Crossflow Theory Since the body vortiges ean significantly influence the pressure dis tribution, they will have lange effects on the body forces and moments in certain eases, Tt is our purpose now to present the theory of Allen for such effects, The theory is based on the concept of the erossflow drag cooficient (ee If dN/dz is the normal foree per unit length (viseous croesforee per unit length) developed normal to an infinite eylinder of radius @ at anglo of attack «,, then the erossflow drag coefficient is 80 defined that = (eel2a)quat (473) ‘he estoy dag cans of» mami of ie sins ve teed a ered Ln By ante vr estrs dtthe derby seth mete ey noe hae : Sons made sent ised ae then en ® citer Soe deere temeee tee 2 apa + Gon2ant ar whore S is the body cross-sectional area, Integration then gives the0 AUSSIE AERODYNAMICS total body normal force N= PgmSa + CiqiaSe 79) where Suis the body base area, and Se the body planform area subjeet to viseous erossflow. The area Se is behind the body eross section corre sponding to 2s (as given by Fig. 4-15, for instance). The tacit assump- tion in the integration of Eq, (74) is that (ey) is uniform along the body length. ‘There is some evidence that (c,). is not uniform," but an average value of (¢,), hos been assumed, Tt is clear that the pitching moment an easily be caleulated since Bq, (1-74) gives the body normal foroe distribution, aire 475g | feeeeiee| Mn? daa + E | Liecoticent.¢ xgerinent 1 senate sory Crreton thay Fro, 418, Comparison of measured and price body ceo ‘The lift coefficient and center of pressure of » body of revolution have been calculated on the basis of slendor-body theory and of Allen's eross- flow theory. ‘The ealeulated values are compared to experimental values in Fig. 418, ‘The actual body is of very high fineness ratio, and the viscous erossforce for such a body is much greater than the it predicted by slender-body theory. The large rearward shift of the eenter of pres sure with inerease in angle of attack is noteworthy. Generally speaking, the lift predicted by slender-body thoory acts on the expanding sections of the body in front of the vortex separation region, and the viseous eruss- foree acts behind the region of vortex separation. "As the angle of attack Increases, the viseous erossforee increases approximately as a, while the slender-body lift increases as aq ‘The rearward shift of the center of pressure is the result. is eharneerein, Aunopesamtes oF nopies; vornicus 1 4-7, Motion of Symmetrical Pair of Crossflow Vortices in Presence of Circular Cylinder Many problems of interest in missile acrodynamies require a detailed nossledge of the vortex flow due to bodies or lifting surfaces. In this fection we will explore the behavior of a symmetrical vortex pattorn of thro vortices in the presence of a circular eylinder. As pietured in Fig, 4-14, the vortiity is moving along the ae feeding sheets into the cores at all Lest times. IF we negleet any influence a resents of the feeding sheets in comparison | i wvith that of the cores, then we ean Dan iss idealize the low model as shown in Vig. 419. Two extomal vortices coeur with equal vortex strength but ‘opposite rotation, and with the vortex steengths changing with time, Inside the body are Toeated two image vortices to insure that the body sur- face isa streamline. The right image vortex has the opposite sense of rata- ee Sion of the righ external vortex but ! the same magnitude; with a similar 120. 419. Symmetrical vortex pair in result for the left vortices. If the ex- Presence of erela elindr, temal right vortex has position jy, then the image vortex must be located by the method of reciprocal radi, namely, so that (4-76) where jis the coordinate of the image vortex. ‘The complex potential for 2 vortex of strength P counterclockwise at position jo is Eos 6 - 30 ‘Tho complex potential for the model of Fig. 4-19, including potential erosslow and four vorties, is WG) = o+ ie : eon va(s— Sos — Bog [$= te - in Fo see PERE \ number of intresting special cases ofthe ena eas will now be ‘spl ‘One question which might be asked is whether thoro exist eombinations of vortex positions and strengths for whieh the resultant veloctios at the | am2 MassnLs AERODYNAMICS external vortices are zero, Such @ question was studiod by Foppl® “Tho velocity ve ~ iwe of the right vortex is given by no tao = tin ZL 176) + Z toe - 30] ) ‘The resultant velocities at the vortex are = lim oe m)| 2atVeaaes 4 TP [ zo rot Be ree — at Gt — a + day ofr + SUES =) (79) r 1 us uolre + a) Fl an ee TG a Fae where jem guch ita rah («30 ‘The condition that 2 — is be zero leads to the condition, after eliminat- ing the vortex strength Gade = 0*)* = saoGn + (si) After reduction to polar form, this equality yields ry oF = Bry 008 (4-82) e equilibrium or Foppl positions and strengths. The vortex strongth T' corresponding to ry is ane 7 (83) ‘See Milne-Thompson’ for details of the derivation. ‘The locus of the equilibrium positions given by Eq. (4-82) is shown in Fig. 4-19, For ‘equilibrium positions far feom the body the vortex strength is large, the strength ineroasing in accordance with Bq. (4-83). One thing to remem= ber is that, though the equilibrium positions are points of zero flow veloe- ity, they are not stagnation points of the erossflow in the usual sense, since the flow velocity changes discontinuously from infinity to zero as ‘tho points are approached from any direction. Another relationship of interest is that between the vortex strength ‘and vortex position when there is to be a stagnation point inthe erossflow ‘on the body at the point specified by 5, = ae. ‘The total velocity in the Auropyxanics oF popnes; vouTICES 98 rossflow plane is may (eye (es ff -wn(is$)+E atin 1 ~ incre) or zero veosty on the body at Rese eee ie Ve.” neh Behe Manipulation of Bq, (185) and the requirement that 1 is real yields r ame [glee + otro) — a 00s (06 — 03)134(r0 + a*/re) + 260s (0 + 0) Bly — a2/ra) 008 @ (6-80) Fora given vortex strength P and stagnation point, Ba. (4-86) will yield a curve on which the vortex must be located. “The actual streamlines in the erossflow plane of the vortiees depend on how the vortex strength varies with time. Actually to consider variable sirength of the vortices without including the feeding sheet leads to a physically inconsistent model, One important ease for which the vortex streamlines ean be found analytically is that for eonstant vortex strength, If the function gy isthe stream function of the vortex streamline, then ah = Mayet eee at cs Poo tn) tie ‘hn den fo inode i ‘The integration with the aid of Eq. (4-79) yields = (1 — 2) — gf wr | += 4(-8)- aa (ata ame] Oe constant ‘The constant is to be evaluated from the knowledge of one point om a particular vortex path. A different set of streamlines occurs for each value of the nondimensional vortex strength I'/trV.aa,. For a value of this parameter of unity, the vortex streamlines have the general pattern shown in Fig. 4-20, Vortices near the body move downward against theo rassiLn Astorwamtes flow, and thowe far from the body move with the ow. The F&ppl posi tion forthe given vale ofthe nondimensional vortex strength is nlided inthe figure, A ropion of eeulatory flow exist about the Foppl point. The asyinptaic lateral positions of tho vortices at infty ys shown in Fig 4-30 can be obtained inplity fram Bq, (88) a follows B12) — sth a) ya) ~ Seen 98 ye For the general ease in which the vortox strong is cbanging with time, ee a anyteal sltion fr the vortex th Met a oY 10g Ye (4-80) Asymotoes f peeeieaet path seems not to he generally possible. Tn faet, a stream fane- tion for the vortes path in the usual sense doos not exist for this ease. ‘To obtain the paths we must inte- grate Eq, (4-79) numerically, using small time increments. Another problem which is also analytically intractible except in special eases is ‘the determination of the positions jo and 2) as functions of the time, To obtain such relationships the following equations must be solved. Fis. 420, Paths of symmetrical vortex ) pai in presen of eirculareylinde. (4-90) ‘The functions ve and wy are to be taken from Bq, (4-79). For the special ease of Vo = 0 and two symmetrical vortices as shown im Fig. 419, Sacks," has determined the time explicitly from Eq. (1-90) 4-8, Motion of Vortices in Presence of a Noncircular Slender Configuration [Let us consider a pair of vortices not necessarily of equal strength in the presence of a noneircular slender configuration as shown in Fig. 4-2 ‘The number of vortives considered is of no importance since the method is valid for any number of vortices. The external vortices induce veloc ties normal to the body and panels, Single image vortices of the type ‘considered in connection with cizeular eross sections will not be adequate in this case, Tn Euet, a complicated image system is required. For this reason it is easier to transform the body eross section into a cireular one for which the image system is known, and then to relate the vortex velocity in the 3 plane Lo that in the ¢ plane (Fig. 4-24), Let I(e) be the eomplex potential for the complete flow in the ¢ plane. AERODYNAMICS OF BODIES; VORTICES 9% With referonco to Table 2-3 we have wea) = Hae («— Ee) (eo) Lot the transformation equations between the j and ¢ planes leaving the flow at infinity unaltered be «= att) “The complex potential for the flow in the physical plane is now 1F(o(3)) Ho) (492) revels esi Fio, 421. Transformation of missile cross section into cick: ‘The vortices are transformed as vortices. Look uow at the velocity 6) ~ fie of the vortex at jin the physical plane. a y= fy = Tim ed [Wea + Row — a0] ee The veloity ofthe vortex in th « plane is denoted by ps ~ én montagne +Bee=a] wen iv) modish wu96 aUSSHLE AERODYNAMICS (4-98) as tif [eon + og ~ oi] + tim $ [toe 6 — 59 swe 4-05) + im § (Fie 2= 3) 4-96) ‘The logarithmic term ean be evaluated by differentiating and using the ‘Taylor expansion, ee ertava it or ee = t= 00 (9 Fu-w+Ou- wt (m7) ‘The Taylor expansion required is fie), 1a E-@)tG)08 “The limit is thea simply Hn) +0G- 5) 4-08) a Ldto/dy Inghesin- - Fat 9) “The vortex velocity inthe physieal plane is now as _ Tae n=O edie, OO ‘The term involving the second Aeretv arena an ailinn tot ik term which would be anticipated if the vortex velocities transformed in the same manner as ordinary flow velocities. ‘The ealculation of vs — iw; for the vortices in the presence of a goneral rose section will ususlly proceed streamovise step by step in a numerical solution. The initial vortex positions and strengths Ti, in, Ts, and te are given, The positions j and js are transformed into ¢: and 2s. ‘Then the velocity of the vortex in the transformed plane, px — is, is computed by the method of Sec. 47. ‘The vortex velocity in the physical plane is ealoulated from Eq. (1-100). The change in vortex position is then obtained by assuming tho vortex velocities uniform over the time or dis- tance interval chosen for the calculation. The eyele is repeated in a step-by-step caleulation to establish the vortex paths. The vortex path in theg plane is not the transformation of that in the 3 plane. Variations in body cross section and in vortex strength are easily accounted for in a step-by-step ealeulation, 4-9, Lift and Sideforce on Slender Configuration Due to Free Vortices free vortices follow their natural streamlines in flowing past a slender configuration, the lift and sideforee due to the vortices can be established mies oF noviEs; voRTICES a simply in terms of the vortex strengths and positions. Sinoe a method Tor caleulating vortex paths was deseribed in the previous section for a slender configuration of general erosssoetion, the possibility is at hand of Astermining the lift and sideforee distributions along such a missile. It is the purpose of this section to derive the necessary formulas in terms of vortex strengths and positions. Consider a single free vortex of strength 1 developed by a vortex generator (Fig. 422), or any other means such as hoily vortex separation, The vortex is fre to follow the goneral flow past the winged part of the configuration. Before starting the derivation Fo, 422, Control areas for eseulating foros and momenta du to free vortices. of the formula, itis desirable to determine the magnitudes of the lateral ‘velocities due to the vortex, and then to compare them with the magni- Indo of the velocities without vortices. ‘The complex potential due to a vortex of strength T's at 3: is, WG) = Eos (= 39 (10 ‘and the lateral velocity components are given by aWy =i a f= (4-102) ‘The bars on 6 and 1; indicate that the velocity components are along the and 2 axes. Equation (1-102) will yield the magnitude of the lateral Velocities if the magnitude of Tis known, Tn this matter we must dis- tinguish between wing-induced vortices and body-induced vortices. If the vortex is body-indueed, then with reforonce to Fig. 4-16 avi ~ (4) 4-103)as sassiLe auovynaauies whore {i the body length. For unit body length Q Ve e (4-108) Sinee the angle of attack is 0(), and the lateral dimensions such as G = 3) are also O(0, ste find that 0, — 10s is OW) for Ve = 1. Here Cis the maximum radial dimension of the slender configuration of unit length, For a vortex induced by a wing of semispan s, at angle of attack a., Eq. (@21) gives Pio Pate (4-105) Since the body is slender, a4 is 0(0 just as a. Equations (4-104) and (4-105) show that the vortex strength is of the same magnitude for a slender configuration whether body-induced or wing-indueed. ‘Thus, for Vo = 1 both types of vortices produce lateral velocities O() just as the lateral velocities without vortices. What this means is that we ean use the order-of-magnitude estimates of Chap. 3 in developing formulas for Lift and sideforee due to vortices. With reference to Eq, (3-38) and Fig. £22, the generalized foree PHidis Pato nese fle Pees OW as, + O(log?) (4-106) [a ‘To evaluate the forces requires a knowledge of the complex potential We ‘without vortices and 17, due to the vortices, ‘The eumplex potential has the general form renee t6.+ ces and the complex potential due tothe vortex plus its image ary Wa = SP og 6 — a) + WG) (4-108) Actually, the procise form of W,() is hard to write down in the plane unless the cross section is some simple shape like a eile. Tt is easier to ‘transform the missile cross section into eizelo of radius r in the ¢ plane, while leaving the field at infinity undisturbed, ‘The transformations SenopysaMies OF woptes; vouttens 9 between j and o under these eireumstances have the forms (4-109) ‘We ean now write the complex potential 17; in the # plane explicitly W,G@) = SE h (4110) See Fig. 428. ‘To make the complex potential singlo-valued, we must put cuts into the planes. First, in Wo(3) there isthe log 3 torm which is indeterminate apne = plne Fro, 42%, Cate and contour of intgrstion. to multiples of 2 The logarithm term arses because of sources within the body eros setion, ‘Thus a source cul must extend from some point within the body to infinity as shown in Fig. 4.25. So long as no path erosses over the cut, the Wj) funetion will be single-valued. If any path eroses the eut, then W(Q) must be increased or decreased by 2riz, depending on which direction the cut ie erosed. If S'2) is 20m, m0 logarthmie term occurs in Wj). Two logavithms appear in the teem W5G). Actually, a eorex cu from o, to a wll ender 1() single-valued 1t-can easly be shown that Ys continuous crossing the ent but that 4s tas the value —T/2 0m tho right side of che eut and 1/2 on the at side. ‘Examine now the integral over 8; given in Eq. (4-106). ‘The area Ss is enclosed by the contour QQ’MNP'PO, which has been chosen to eros over no euts, Let the eontour K be the outer ereleof radius 7, let Co be that part ofthe contour next to the body, let Cy be the eontou consisting10 MISSILE AxKODYNaMICS of segments QQ’ and P’P, and let Cz be the contour segment MV. The ‘whole internal contour from Q to P is denoted by C, and Cate enn Applying Gten'sthoormn to the are, yields aw [r= if. 00- if, 08 an) The contour integral about A ean be simply rewritten ax Greta sf Se Bretaoas is) anv then ntoducd ito Eq, (4106) to yield lagu +f (ga6aR _ awa), i =f Ce oi) aeae oan Now ¥ +12 eannot depend on r;,and,sinee the frst integral is independ- cent of ri, s0 must the double integral be. ‘The integrand of the double §ntegral is O(1/rs) so that the integral approaches zero as.ry—+ «© with or ‘without vortices present. We now have PED we What has heen achieved is that the quadratie integrand of the double integral has disappeared, and the contributions to P -+ i ave linear in W. Thus, if Y, + i2; is the contribution due to the vortex, an expression, for this quantity ean be written down immediately Hai fo) = 20h, Was 26,60 (ets) Hee RAB og a 2f nh GMO ‘The integral around C of Hy ean be distorted to K since W,G) is an analytic function in Ss, and K can he transformed into K, in the ¢ plat femora $, wie $,, wiley Bde conn where K. isin a large contour into which Kis transformed in the ¢ plane. ‘The expansions as Barto @) rae) «Be a()vo(%) ano AERODYNAMICS OF BoDIES; YoRTICES 101 permit the contour integral easily to be evaluated by the residue theorem (8). a0 ‘The intogral around ( of yx is zero because ysis constant on Co and is eon tinuous neross the euts bracketed by the contours Cy and Cy. ‘Hauations (4-116) and (4-119) thus yield the final result Pe Wa) as =Ti—o) 4119) Ifo, ~ 6, is the point 1.6% in the « plane, the sideforee and lift are then aan ‘These simple formulas provide a means of ealeulating the forees due to the vortices up to any axial position in terms of the vortex positions and strengths, However, their use presupposes a knowledge of the vortex positions, Such knowledge is obtained by a step-by-step ealeulation of the type deseribed in the previous section. The offocts of many vortices ray be found from Eq, (4121) by superposition, Any coupling between the vortex elfects enters through mutual interference between vortex paths. Te isinteresting to note that, if the contribution to a of W(3) had been introduced into Kg. (3462) derived on the basis of no vortices, exactly the vortex contributions found here would have arisen. Sacks! ‘makes an equivalent statement, Also, Bq. (J-121) is obviously applicable to the determination of the force between any two erossflow planes due to ‘one or more vortices, whether they originate on the missile or not 4-10. Rolling Moment of Slender Configuration Due to Free Vortices Tis possible to derive a formula for the rolling moment developed by free vortices passing a slender configuration in terms of quantities in the plane of the base analogous to the lift and sideforce formulas of the pro- reding seetion. For eonvenienee consider the same cireumstances as ‘hose prevailing in Figs. 4-22 and 4.23, exeopt that in Fig. 423 transform ‘he body eross section so that the center of the eirele falls on the origin in the e plane. ‘The pressure forees ou control surfaces Se and S; do not contribute to the rolling moment. Only the transport of tangential ‘momentum geross areas Sand Ss ean eause rolling moment, and, of these, it tarns out that only Sy has # contribution. ‘The rolling moment L! is = Ve, ood + Ve f, Lb aDeodss (129102 MISSILE AERODYNAMICS with positive L’ taken in the negative # sense, 2—+ 9. Division by goand the use of the density relationship, Bq. (3-56), yields Ba 42 oobdse+2 f ordse+ 060g cei29 ‘To show that the integral over S: contributes nothing to the rolling moment, rewrite the integral as fiootaie= [a0 [2 ntetao cata) ‘he general form ofthe potential function ineuding vortex est canbe tren in the following form convergoal oa ntour K enlong the vortices ‘The source eut in this ease is of no importance since ¢ is continuous aerase the source ext. The vortex cut is important for that part of @ dus to vortices. On Ss the values of ¢y and 6, ean be calculated by diferentia- tion of Kg. (1-125). If the values of gs and gy are substituted into Eq. (124) and the integrations carried out, itis found that the integral is Consider now the contribution of the area S;. At this point lot us confine our attention only to that part of the rolling moment due to the vortex, This is now possible because the remaining integral in Eq (4-123) is linear in 4. While the rolling moment due to the vortex ean be evaluated in terms of the vortex position in the base plane, all cam- ponents of the flow will influence this position. The surfues integral over Sis taken over the area within the dashed contour in Tig, 423. ‘The area integral is converted to contour integrals by means of Green's tron [oosso$ fete $f eter ‘he onto €ncmped ofthe at Cn proximity wth the ay, he post Cy spend of gens and @'@ sat tsar the far Cs compog sont SIV stat the verter eae og Suund siesta condant, These futon fave tote ‘oro and age ha cot val on Grands annus ser any ta $ovate ‘Thus, the integral over S; for the part of e due to the vortex ean be 0 12) ARNODYNAMLCS OP BODIES} VORTICES 108 written [Sas -Afetot von =L¥,anioue 9 ‘The evaluation of the contour intogral eannot be made diveetly by the residue ealeulus beeanse the integrand is not analytie. Let us transform. the contour in the 3 plane into a circle of radius», in theo plane with 3, st the origin, (4129) ‘The field at infinity suffers finite translation only. ‘The coefficients dy are usually eomples, and the function fle) can usually be written in finite form for most eases of interest. ‘Tho integral about C can be broken up into two convenient parts with tho aid of the following identity = GWG —L) + G3 +49 — a (4-130) With the following notation n= bf, wrodss +10) 1e= 3 f, Waals ~ WG - Corn) veo see that J+ is the contribution when jy = 0, and sis the additional contribution when jy is not equal to zero. f%es- hth (4-482) Confine the analysis to the evaluation of fs for the present, ‘The integral Tyan be written hen gemod tage Also, since Ys is constant on Co and continuous across the cuts fern fbit ae (4133) ° (41s) and 1, ean be written Na WRG WG) + aR FWaedds 485) nntour C ean be transformed into the plano and then enlarged into 8 large circular contour D, centered on the origin and enclosing the body ‘and vortex eut, We ean then expand the integrand in a series in # and104 MISSILE ABRODYNAMES integrate term by term. In the ¢ plane the complex potential is WiGGo) = FE loge ~ 00) lowe ~ 201 agg) oe ‘The expansion valid on Dis Wwe (137) “Thus f, Walon ds ~ 138) Sineo only the ~! term contributes to the intogeal $, Wise) & = — Ke (139) ‘The value Js thus [= NU RP 6) teIPo— ol 4) From Eq, (4-120) nwo Pe igt oy so that u--Ge- i) 1a) ‘The evaluation of I, requires different treatment from that of fy, Ibis ‘irst decomposed into integrals over Cy and C; since the soutoe at C1 i of xno concer here, af, Meas - 96-5 +4 J, mous — 0 — I (143) ‘The integral along the vortex cut is easily evaluated since ¢ = —1/2.0n the right side and 1/2 on the let. 1 5 [woes — ines - Ea le I) = Fn = uP = tae a Dna = ~Sot— a) as) ‘The integral around Cy is transformed into an integral about Cy, in the ¢ plane. In the « plane genoraly @-mo-w and on Cy in puticular 6-6-9 =s0() (146) [By using the sories of Eq. (4-129) we can expand the produet in « Laurent Sleisie) (4145) on wnd—w = (e+ 84) (E4 S42) bine au) ‘The coefficients ky turn out to be k=) Seen positive (Ges) nenogative with dan 4-00 mt G9) We shall confine our attention to thos cases wherein thn series eonverges on Gy, although its derivatives are of no concer. The integral around Cy! now becomes on integration by parts Sf. MGM — 3G - 8 Te ele From the series expansion for (3 — am (iit 1y we -EA+t +s ( diteot integration yields fey = NG =H) die C130) (6 = $4) aud that for AV de HG) em 3 [, Wald — 396 = 59 = - Bre He(eenet Stee San) cm106 MISSILE AERODYNAMICS . fee Zotar + S[ae—an + A+ DA] tsa wherein we have made use of the relationship Bare (154) ‘The final result for the rolling moment is u we = HU oe eS (eB) wasp maT w where Vi is no longer unity. Tt should be remembered that this result, contains any moment de to the vortex generator (Fig. 422). The roll ing moment betwcen 6 erossflow planes ean be found by differencing as shown in the following example Miustrative Example Caleulate the rolling moment de to a free vortex of strength Pas it passes a triangular wing as shown in Fig. 424 ‘This example is a essn wherein the serios are finite. The rolling [pr a Fic, 424, Foe vortex passing teangular win ABRODYNAMICS OF BODIES; VORTICES 107 moment is given from Big, (4-155) a8 a @-@.. The transformation taking the wing cross seetion into a circle with center at the origin is dian From Bq. (148) the values of keel beth! bent ‘G: ‘he qty §; isnot independent of jo. In faet, axis determined from the initial position so by a step-by-step ealeulation of the vortex path, As result the rolling moment is eer eeee eat = Ferd = inh + Be svnocs « sean radios of quas-eylinder @. major and minor axes of elise {tls} coefficient of log term in @ expansion .——oefciont of term in expan (je _-partaf a du to anate of attack a0 Goelsient ef r= term in @ expansion (ie park of due to thicknoss ho" itive finetion of 2 in ¢ expansion B (re = ¢ length of bump in ercalar body (3. eromfow drag eooiient fe coeicints fm expansion for © a (be (Co). drag eoticiont of cieular cone (Cr). rag eosin of litical eone Cet evetient ©.) functions of p a body diameter 4 coelcient in expansion for 5 D, rag of litical cone108 a) Fo) 7% Junctions of = specifying Laplace transform of fo(2) Imaginary part caliber of tangent ogive modified Bessel funetion of second kind length of slender configuration length of tangent ogive Laplace transform operator, and inverse transform operator rolling moment about # axis froo-stroam Mach nuraber moment about 9 axis, pitehing moment moment about 2 axis, yawing moment hnumber of Fourier eoraponent of quasi-eylindrieal body summation indices dimensionless number for normal force due to viscous crosstiow ‘order of magnitude of fin physieal sense variable of plane of Laplace transform; local statie pressure free-stream static pressure ‘complex eonjugate velocity of vortex Py in ¢ plane pressure coofliient, (p ~ pe)/qe pressure cooficient due to angle of attack ‘pressure coefficient due to thickness (a, = 0) pressure coefficients on impact and leeward surfuces loading coefficient, PY — P~ free-stream dynamie pressure polar coordinates : radius position of right external vortex of 2 symmetrical pair radius of control surface radius of base of body of revolution radius of circle in ¢ plane vortex polar coordinates in Foppl equilibrium condition loeal radius of body of revolution 1/rg for tangent ogive radius of curvature of tangent ogive; loeal radius of quasi- cylindrical body raximum semispan of wing panel body’ cross-sectional are control surfaces, Fig. 4-22 ase area of slender body body planform area subject to viseous erossflow ‘maximum lateral dimension of slender configuration for unit length; time perturbation velocity components along 2, y, and © pe of quasi-eylindrical body cous erossflow; also normal foree ABRODYNAAIICS OF NObIES; vouTICES 109 perturbation velocity components due to thickness perturbation velocity components due to angle of attack perturbation velocity components slong 2’, y/, and 3” ‘complex conjugate velocity of right external vortex of a sym= metrical pair ‘complex conjugate velocity of vortex Pin 3 plane freo-stream velocity velocity of flow normal to eylinder volume complex potential, 6 + i ‘complex potential due to vortex Fy ‘complex potential due to image system of vortex Ts ‘complex potential at zero angle of attack ‘complex potential due to angle of attack axis systems deseribed in Seo. 1-3 axial distance to vortex separation points of body axis systems described in See. 1-3 2/L. axis systems deseribed in See. 13 axial distance to center of pressure position of right external vortex of a symmetrical pair value of ys when 2» = 0 value of yo when 25 = = forees slong y and = forces due to vortex forees along g and 2 tie ‘external position of right vortex of a symmetrical pair position of vortex Fy position of controid of body eross section internal position of right image vortex of a symmetrical pair position of separation point on body surface included angle between free-stream direction and body axis, radial distance to vortex Py in plane vortex strength strength of wing eireulation at root chord strength of vortices vortex strength of Féppl equilibrium position strength of body vortices polar angle in construetion of ellipse, Fig. 46 also height of ‘bump on eylinder, Fig. 4-11 polar angle of vortex T' in ¢ plane ‘semiapex angle of elliptical cone in plane of major axis variable of integration; also 2 — 2,10 MISSILE ABRODYNaMteS » froe-stream density ° polar angle in 3 plane % polar angle of right external vortex of symmetrical pair as polar angle of stagnation point on body » aan elliptical distance, Fig. 48 A lis = a Aw ii ~ io, Pig. 428, ° variable of transformed plane fe2 positions of vortices P und I in ¢ plane = position of image vortex for Py + distance along tangent direetion to body eros: seation; also ramp angle 6 velocity potential ° Laplace transform of g ¥ stream funetion for complete flow ve stream funetion for vortex path ° semiapex angle of eiroular cone REFERENCES 2. Taslor, G.I and J. W. Maccoll The Air Prewure on » Cone Moving at High Speeds, Pre oy. Soe. Laman A, vel. 138, pp. 278-31, 185. 1 Kahane, &, and A Solarkis Supersonic Flow about Sled ross Seton, J Avro Sk, vol. By no. % pp 513-02, 1868. 3, Fraeukal, .E: Superomi Flow post Slender Beis of Elite Cross Seton, Brit, AHC B & 31 25, 1085. 4 Nieke, Jack N-i Tables of Characteristic Functions for Solving Boundary ‘value Problems ofthe Wave Equation with Application to Superonie Taterferenee, NACA Tech, Note 873, February, 157 '. Perkins, Edward W., and Leland H. Jorgensen; Comparison of Expevieatal ‘and Theoretical Novmalforee Dstsbutionsnlading Reynolds Number Elects) on ‘an Ogivevylinder Body at Mach Number 18, NACA Tach, Nols $716, May, 1066 6 Jorgensen, Leland HL, and Edward W. Perkins: Investigation of Some’ Wake Vortex Charsctrstire of an Tnsined Oive~linier Body ut Black Number 18, NACA Keach Mem. ASSES, gist 135. 1 Rane, D. Jz Measurement of the Crows Mow around an Inclined Body at Mach Number of 191, RAL Teck, Nae ster, 2357, Jaman 1986 '& Niche, Jack N,, and Geonge E. Katine The Lees of Vortex and Shook expansion Fide on Bitch and Yaw Instabilities of Supersonic Airplanes, Tat Aeron. Se. Preprint TA, 157 2 Lindsey, WF: Dog of Cplinders of Simple Shapes, NACA Tech, Ret. 618, 10, Allon, HE, J, and BW. Purkios: A Study of Bests of Vsoity on Flow over Slender Inlined Bois of Revolution, NACA Tech, Lops. 104, 1951, 1M, Goldstein, &." Morn Developioatsin Fhid Dynamics?" vol. If, pp-A18-2, (taseaden Pre, Oxon, 1938 U2 Foppl, L Witbelbewepung hinter einen Kisisaylinder, Sitter. ayer Abad Wise, 18 Posies of Elliptic AmoDYNaMIOS OF BoDINS; YorTH mn 18, Socks, Alvin H.: Theoretical Lit Dus to Wing Incidence of Sender Wing ady-Tail Combinations at Zero Angle of Attack, NACA Tech, Notey 006, 1950, 4. Milne-Thompson, I, M. "Theoretical Hydrndymamiy, 2d ey pp 881-882 ‘The Macmillan Company, New York, 1950. 18. Beyson, Arthur Ty, Jez Evaluation of the Inertia Gocficients of the Cross section of a lender Dod, J. Aeron, Sei, vol. 21, no. 6, Meadors Forum, pp Ber 1954, 16. Sicks, Alvin H: Vortex Interference on Stender Airplanes, VACA Terk Note 525, November, 1856. 1H Lin, C. Gs On the Motion of Vortices in Two Dimensions, University of ‘Toronto Press, Toroato, 1943.cuaprer 5 WING-BODY INTERFERENCE, ‘The purpose of this chapter is to present methods for predicting the jwerodynamie characteristics of configurations formed by the addition of; lifting surfaces to a body. The lifting surfaces ean be wing panels, ‘empenniage panels, ete.,and will be termed panels forshort. The primary: focus here is on planar and cruciform wing-body combinations. By a planar wing-body combination we mean one with two wing pancls, usually, fof the same shape and size, symmotrieally disposed to the left and right. sides of the missile. By a cruciform combination, we mean one with four panels of equal size and shape, disposed sround tho missle 90° apart. Configurations built up by the addition of panels of unequal size as in fan empennage are treated in Chap. 10. ‘Traditionally in airplane design the aerodynamic characteristies of the wing-body combination have bean! viewed as dominated by the wing as though the body were not there, For subsonie air frames where wing spans are usually large compared to the body diameter, the traditional assumption ean be defended. How: fever, the use of very small wings in comparison to the body diameter, which characterizes many missile designs, requires a different approach. ‘The point of view is taken that nether the panels nor the body necessarily havea dominant influence on the aerodynamic characteristies of the wing- body combination. Rather, the over-all characteristies result from the body and wing acting together with mutual interference between exch thor. ‘The chapter starts in See. 5-1 with an enumeration of the various] dofinitions and notations, and then in Sees, 5-2 and 53 takes up the sub-/ ject of planar ving-body’ combinations for zero bank angle. ‘The load- ings, lila, and centers of preasure are determined for the pressure fields) acting on the panel and body. In Sec. 5-4 the characteristics of banked cruciform combinations are investigated. The influence of the angle of} bank of the interference between panels is treated in See. 5-5 for both planar and eruciform configurations. In See, 5-6 the results are summee| ized for a complete wing-body configuration. The question of thel application of these results to nonslender configurations and a ealeulative) ‘exumple illustrating tho thooretieal methods are the subjeots of See, 5 Finally, the chapter concludes with a discussion of a simplified vot 12 WING-DODY INTERFERENCE 113 soe of wing-hody combination wtul fr such purpost as calling the Bow fel about te wingbody combination 54. Defistons; Notation or purpoeas of wing-body intrornse, eh wing ane wl be taken Elpida epee me Moses blanketed by theta, Ther when th exposed ing panel di tppent, so dows te ing alone. "The body alone th wing body eon thea hese the wing ancl Actually the prea doions ould i \. roteboey \ Weget serie sisty ia. 54. Section of wing-body combination requiro a specification of how the panels are parted from the body, but we will forego this refinement. The interference ean be specified once the wving-alone and body-alone definitions are specified. ‘The interference for any quantity is the difference between the quantity for the complete Wing-body eombination less the sum of the quantities for the wing alone sand the body alone. For instance-—the interference potential would be 8. bo = (Ow + 0) Ga) Where the subseripts i, C, W, and B refer, respectively, to interference, combination, wing alone, and body alone. ’If the wing-alone definition is chauged, it is clear that the interference will change since the character- isties of the eomplete combination are independent of how the wing alone is defined. ‘The interference potential ean influenee part or all of the body or wing, The values of ¢: at the body surface account for the effect of the wing on the body, and the values of g, at the wing surface account for the effect of the body on the wing. ‘The various sections of a wing-body combination are illustrated in Fig. 51. For convenience, the various sections of the body are subdivided into the forebudy infront of the wing panels, the winged section of the body With the wing panels, and the afterbody behind the trailing edge of the wing panel,i ussiia: AEWODYNaMCs ‘Tyo sets of axs re of importance in so far as forees and moments are! eoneorned. ‘The axes 2’, y', 2" correspond to the principal body axes of, symmetry for ¢ = 0 but a, not equal to zero, ‘The axes 2, y, = aro the principal body axes under all combinations of ¢ and aq. The forees on {the body due to the wing or on the complete configuration will generally be ruferred to the ',y/e" systema, The foree along isthe lift the foes. C along u" is the eross-vind force; and the moment about y'is the pitehing ‘moment. We will also be interested in the panol forces which, for y not equal © zero, are not conveniently ; L pecifiod with respect to 2, y'2" axes, With reference to Fig, 5-2, the panel normal foreo eoofficient is denoted Py (eu ten moment coofiiont in ‘snd the Ringe line i — % __ taken normal tothe body axis at the ~€7-¥ samelocationasthepitehing-moment Je” referee axis {Before eousideration of the appi- cation of lender-body theory to Mie. wing-body interference, itisprobably ‘well to montion that wing-hody inter= ference problems can in certain in stanees be solved by full linear Fin. 62, Fore and moment coefficients theory Fy aa fe panels and compte cenfgumnton, ‘be0ry. For rectangular wings and, cireular bodies, for instance, the {formal boundary-value problem presente by thefulllinear theory has been solved. Alco, another solution for part of the interference field is given by Morikawa? However, these methods are gencrally too complex for actual engineering use, but they do serve as uscful yardsticks for evalu- ating more approximate but simplor enginovring methods, One such ‘method is the essential subject matter of this chapter. A goncral survey of the subject of wing-body interference has heen presented by Lawrence and Flax.* 5-2. Planar Wing and Body Interference ‘The utility of slender-body theory is never better exemplified than in its application to wing-body interference. From it we can derive the loading coefficients, span-load distribution, lift, and moment of a wing-body combination, ¢ well as the components of these quantities acting on the panel and the body. Consider a planar wing and body combination at zero anglo of bank as shown in Fig. 5-8, for which the perturbation complex potential will now be constructed, Let the body radius a and the semispan s be functions of x. The complex potential wixe-nopy iNtenrenesce us, can be separated into two parts: 17,3) due to thickness, which exists at f attack, and W(7) due to angle of attack. The part of the sees coee dnc ena aly ao wig = tra ff(r2) (rf) ‘he complex potential duc to thickness is precisely that due to the body of revolution taken to be the budy alone. Thus the entire perturbation fomplex potential for unit Ve is we) «er +o ta ((04 2) (442) Sinco the wing panels have no thickness, they have no contribution to Fic. 88, Planae wing snd boty combination st zero angle of bank ‘The velueity components entering the loading coefficients differ for the wing and body. The velocity components ty, #4, tte are those due WG) with Vs of unity and a of unity, Correspondingly we have %, tajdie so 1743), ‘The superscript ++ indicates the lower impact surface. an — the upper suction surface, The loading on the body is not infl- enced by thickness effects as discussed in connection with Eq. (4-15) Thus from Big. (12) (Pom (Pr = Phe Balter + ate — tan = atte) oy ‘The symmotry properties of the missile yield so dat (Pay = tent 65) For the wing pane in the presen of the body, we have fron Ba, (34 (Pivw = (Pie + Prodein HA Btaae 4 ek (awe ba] “Blouse er Cann + wa) = loeb 00s) + (ove + awa) “Flor + ane) -k (wi + a v1 Go)us Mussina axnouywamies whieh for the following symmetries in the velocity components for the wing woe le able OD yields (AP )won = —Aatat — Aavetort (6-8) We note that the wing loading has « quadratie form while the body load- ing does not. ‘The velocity components needed to obtain the loading eoelicients ean be obtained simply from Eq. (5-8). For the body we obtain anita pony + (v4) [B(0-2) +228] TSF ae? = 4a et . 2asin20sin@ OT TGF t/a)? = dat cont 69) ie aed 2a es oon se eT ae ed For the wing the perturbation velocity components ste ee ner: tt Tet ara + ere at yl = at/y') (5-10) ° RF OPT oe ea eres aie where we have assumed that the wing has no thicknoss in calculating the thickness velocity components, ‘The loadings as obtained for the velocity components are et OP non = Ea aR [(-)es (Pw = EAT WINE ATVI ~ Se talG-1)+(-9)]} em {vis noted that the loadings on both wing and body depend on the expan- sion rate of both wing semispan and body radius. It is interesting 10 compare the loading for body eross sections of identical shape but for wivenopy ivrenemiasce 1 ddujde = Oand da/de x 0. Such a comparison is made in Fig. -4, whieh shove loadings on a combination of o triangular wing and a circular eylinder, and « combination of a trinngular wing and a cone. ‘The influ fence of body expansion on the shape of the loadings is not important in this ease. "These loadings with da/de = 0 are the same a8 those obtained by Lenuerts as a solution to 9 problem of minimum indueed drag. ‘The problom, one of subsonie flow, is based on Trofftz plane methods. The vortex wake is assumed to retain the general shape of the body in end ° a io, 54, Pressure diferences at Unilin eges of slender wing boy combinations view in moving backward to the so-called Trefftz plane. Here the terion of minimum drag is that the vortex wake move downward m torted. Mathematically the problem isto solve the Laplace equation for the cross section of the wake moving downward with uniform speed. Wis mathematically equivalent to the present problem with no body expansion, ‘The details of the solution are given by Durand. Consider now the total lift of the wing-body eombination as given by Bq. (3-61). Lot ¢m be the maximum span of the combination, and let ve the accompanying body radius, ‘Then the lift up to this axial station ‘comes out to be Le a ‘The lift includes that developed by the missile forebody. Actually, the ‘otal lift of the combination is given by Eq, (5-13), independent of the shape of the combination in front of the axial position for 8) oF of the 13)us AUSSILE AEHODYNastes shape of the wing pancls behind this axial position. The loadings given bby Eas. (6-11) and (6-12) do depend on the planform through ds/dz, but the total intograted lift doos not. If the trailing edge of the panel is normal to the flow at the axial postion for ge, then no question of lift due to additional wing area behind this position arises. However, eve wing area with s < 2m does occur behind this position, no inereasod lift ‘ocours on the basis of slender-body theory. The reason for this behavior ’ is diseussed in See. 7-1. Actually, the trailing-vortex system from the panel trailing edge induces down- ‘wash om the area, which just offsets the angle of attack. ‘The precise role of the body expansion isnot 30 clear. Ifthe body is expanding up to the axial position for &q, then Ea, (518) is corteet, Body’ contrac tion aft of this position may influ Fic, 5.5, Triangular wing and boy ence the total combination lift, bt Cepemeas 1 consideration of this problem is beyond the seupe of the present srork. In fact, we shall assume that the afterbody isu cireular eylinder in our succeeding discussion of wing-body interference 5-8. Division of Lift between Wing and Body; Panel Center of Pressure It is of interest to vee how the total combination lift is distributed between the panels and the body. For this purpose, assume that the body is a circular eylinder si that we have no body expansion term, Also, for purposes of dofiniteness, assume that the wing is triangular, ulthough this assumption will shorty he rlaxed, With reference to Fig 5-5 the lift on the panel is brn sata f on I. ‘One integration yields (one panel) tee f[(ers)-Go9] oem ‘The integrand gives the shape of the span loading. ‘The span loading is very elusely elliptical, as discussed in connection with Table 6-1, ‘Though the integration has been carried out fora panel of triangular planform, the span loading is independent of the exaet shape of the pane! for a slender configuration, What follows is therefore vali for panels of goneral plan- form. The total lift on the wing panels is eonveniently expressed as & = aye ae 0 oan aT ous) WINe-noDY INTENPERBNCE ne fraction of the lift developed by the wing alone Lw: Bm bralse — a) (18) ‘The lift rato is denoted by Kir, and the value as found from Eqs. (5-15) and (16) is vat 7) ‘The lift ratio is a function solely of 4/6 ‘Tanur S, Suasoennoor Panaueteas ron Loxpive Due 10 Parew* | oon) , 0.8/0 on! oan? | aan o2! ose | aaa 03 | oss | cals oe cat | vas a8 cam | oat oe ses | ote of ves | nats os | oe Hd Oe 8 | 0 tetas “The aeowacy of te tabulated roan eatated to be 20.02, ‘Triangle pol [An analogous lift ratio to Ky also serves to specify the lift on the body ue to the wing: ke = las ow Pe ton he ty do he nn neat sie Laer) = Le = Lwin) — Le (19) or the tit on the iste yen bY 1 = at =) cs120 MISSILE AERODYNAMICS ‘Phe value of Ky turns out to be Ke- (142) ke 21 s0 that Kp and Ky are both funetions solely of a/sm. ‘They are given in ‘Table 6-1, and plotted versus a/sy in Fig. 5-6. ‘The values of Ka and Ky shown in Fig. 5-6 roveal some of the salient ‘rs facts about wing-body interference, Ata value of a/&q of zero, the value of Kir is unity because of the way in which Ky has been defined, and Kp is zero because there is no body. Hosaver, at the upper limit of 44/sa of unity, the panels are very small and are effectively mounted om an gape infinite reflection plane. From the Ty i potential ¢. given by Eq. (8-19) itis easy to see that the body produces a Tocal angle of attack slong its side edge of 2a, since the velocity here is twice the ~velovity of the main flow normal to the body. The wing panels therefore de- velop twive as much lift as they would at angle of attack a go that Kw is 2, ‘Thus, interference of the body on the wing through upwash has inereased the panel lit to twice its usual value. As rough rule of thumb, Fig, 56 shows Fro. 56, Interferance lit retie for thatthe fractional inerease in wing Iitasweiated with pitch panel lift due to body upwash is a/ ‘The parameter a/s is thus the primary measure of the importance of interference on lift. ‘The nature of the lit on the body due to the wing panels represented by Knisof interest, Actually, the lift is entirely transferred or “carried over” onto the body from the wing. ‘The wing is the primary generator of the lft, and eertain of the lift is carried over onto the body because of its proximity to the wing panel. For a very small panel and a very large body that prevails as a/¢, approaches unity, there is a large expanse of body to “eatch”’ the lift generated hy the wing. This area accounts for the fact that the body “catches” as much lift as aets on the wing pauels ‘themselves, a8 a/sq becomes unity. ‘The application of the ratios Kw and Kz to nonslender eonfiguration is shown in Seo, 5-7. In addition to the division of lift between body and panels, the center of pressure of the panel is of some interest. The center of pressure of the lift on the body due to the wing is significantly influenced by afterbody length and is discussed in See. 5-8 where alterbody effects are considered. The lateral center of prossure is denoted by (Js) ia, and the longitudinal position by (#,)wa measured behind the leading edge of the wing-body WING-DODY INTENTERINCE ma jjuueture, Singe the shape of the span loading is given by tho integrand of Ba. (5-15) itis easy to write down the expression for (Ja) rin * (me + at/tn)* = a/ yy y dy i Hele wre Ga lon + 0'/80)* — y+ aby) PE dy au” RRR TRIAO + O=¥ = DEW FOTO DEW +0 =D G2) whore K(8) and E(B) are complete elliptic integrals of modulus b nai tA ean ‘The values of (Go — 0):r./(% — a) depend solely on o/s and are given in Table 51, The lateral enter of pressure does not depart significantly from the value of 4/3 that is obtained for an elliptical span loading This result, independent of wing planform, really shows that wing-body interference does not influence the lateral center of pressure appreciably Tt can easly be shown that the streamwise eenter-of-pressure positon js definitely not independent of panel planform, as isthe lateral position For instance, to the extent that slender-body theory ean be applied to a rectangular wing panel, slender-body theory would place its center of pressure on the leading edge. Tt i worthwhile ealeulating the eentor-of- ‘pressure location for a triangular wing to see what effect interference has ‘on the location as far as its axial position is concerned. ‘The values of (Za/e-}wcn have been ealeulated from the loading of the panel as given by ‘The ealeulation is not reproduced here, but the values are ‘ble 5-1. Actually, the variation in the value of (@a/e) wm from the value of wo-thieds forthe wing alone is very small for triangular panels, In fact, the olfect of interferenee on both the lateral and longi- ‘udinal eenter-of-pressure positions ean be negleeted for most purposes on the basis of slender-body theory 5-4, Cruciform Wing and Body Interference ‘The load distribution and the lit and eross-wind forees will be ealeu- lated for a erueiform wing-body combination formed of a fat- ‘and a cieular body. Actually, the vertieal panels ean possess a semispan 12) different from the horizontal panels, which have semispan s(x). As shown in Fig, 57, the configuration is pitched through a. and banked by angle ¢, s0 that the eombination is at angle of attack a = ax cos ¢ and at angle of sideslip $ = asin e. ‘Tho fact we shall use to establish the flow is that the flow field due to « will be wnaltored by the presence of the Eq. G sven ini MISSILE AZRODYNANCS vertical panels, and that due to é will be unaltered by the horizontal panels. As. result, we nood only compound two flow fields for a planar ‘wing-body combination at right angles to obtain that for a cruciform con- figuration. This addition follows from the fact that potential fanctions and flows can be added linearly in slender-body theory. We must, how- ever, perform an analysis ta see what happens to the pressure eoeficient under those circumstances. ‘To study the pressure coefficient, let us eonsider the total potential ‘Sunetion for the perturbation velocities / to be composed as follows, b= bt ad, + 09 ‘where 6, and gy are the perturbation potentials for unit velocities in the a and 8 direction of Fig. 5-7. Then the perturbation velocities are of the form fate + Bis (6-26) ‘The form of the pressure cootsient equation appropriate to the prevent problem i P= =2(u + aww — Be) — (0? + 4) 25) Fg. 57. Ass msm for ersorm where the velocity perturbation come mise "under combioct itch and Ponentsarvalongthes,,e4xes. Let us now apply Eq, (5-26) to caleulating the loading on the riht horizontal patel given by P* — Por (SP)n ‘The perturbation velocity components possess the following symmetry propertios and boundary values forthe wing panel (with panel of 20 thicknoss) wt = ay wt wit = wr = 0 tt yt w= 1 62) ae et = og wy! = -wr = 0 values it is easly established that CaP)» = dane? — desteyé + dare = ey") 6-28) ‘The first two terms correspond precisely to those for a planar eonfigura- tion as given by Bq, (548) for g = 0. The loading of the erueiform eon figuration is thus the same as the planar configuration for ¢ = 0. For not equal to zero an additional term arises: a term proportional to a8. ‘This third term represents the effect of bank angle on the panel loading. Tits nature is discussed in the next seetion for both plane and erueiform \WING-RODY INTERFERENCE 128 configurations, and it is termed tho af coupling term. Note that it is symmetrical from left to right, "The velocity componenta involved in the panel loading, tet, x and 1, havo already been given in Eqs. (5-0) and (5-10). The remaining velocity component vs* ean be obtained from Rg. (5-8) by appropriately interchanging » and w. a (w/D(t = at/y4)_ = Fa) + WA + ‘The complete loading for the right horizontal panel is now wesel(-s) osteo (2) ee WU as = GSN FTP Aye) (W/N(A = at/y)iae sin g 008g [Because ofthe seeond term the normal force on the right panel is increased as it moves downward with positive y, and the upgoing left, pavel loses the same amount of normal force. It is elear that the loading ean be ‘obtained on any’ panel from Bq. (5-80) by changing bank angle or intor- changing # and ! |A similar analysis of the loading ean be carried out for the body. Howover, the boundary conditions for the body result in different rela- tionships for the velocity components than for the wing panels. With the symmetry properties of Eq, (5-27) (but not the boundary values), the loading on the body becomes (AP Yaw) = haut — dalwwe + ratect + wate) deat” — vet + nctost + westuy®) (6-31) ‘Tho seoond term is z0ro since the velocity in the erossflow plane due to thickness «1 + ivy is perpendicular to the velocity due to a, lavct + ia(1 + we")]. The first term exists at all bank angles and is the sume term given by Eq. (5-5) for a planar wing. The third term is t coupling term fr the bodly loading analogous to that for the panels. The velocity components ta and we* oecutring in the coupling term are given by Bq, (5) as for a planar configuration. The velocity components ty" and wy are easily obtained from symmetry considerations from the results for vat and w.*. 2a sin @ sin 2 rer — tater +1 2a sin 20 cos ai) = aa 632)14 MISSILE, \ERODYNaMteS ‘The body loading is =e ‘The body loading contains in the first term a part proportional to rate of body expansion and a part proportional to rate of change of wing sexni- span. However, neither of these quantitis influenoes the loading associ- ated with combined pitch and sideslip. ‘With regard to the total forees on a erueiform configuration, it has been noted that the coupling term proportional to 43 eaused a mich increase in loading on the right panel for positive g as it eaused decrease on the left pancl. Likewise, the coupling term in Bq. (5-38) eauses a siailar situation with respeet to the right and left halves of the body. In eonse- quence the coupling terms prrduee no net lift but only eause an asym- metrical loading. The total foree on the eonfiguration ean be ealeulated by adding the forees due to two planar configurations at right angles just 1s the flow can be similarly constructed, This is rue sinee the gross forces are independent of the coupling terms. ‘Thus, the foree Z along (aP)20n = tos ito Be 6 2 ton(1- #42) Zz 3 (5-34) vo ane (1 “in the ton ig 3.) b= Zens Yeine hie orate engin, 4 ~ a some b= ma(1 #4) 638) and the cross-wind force along the’ axisis zero. Its seen that changing the bank orientation for a eonstant value of ae does not change the lift force, nor does it develop any crose-wind foree. Thus, as the missile rolls, it will continue to devetop lift in the plane defined by the relative wind and the missile axis. This characteristic, an important property of the cruciform configuration, is also true of the triform configuration. ‘The serodynamies of slender eruciform configurations have been studied by Sproiter and Sacks. WING-nODY INTERPERENCH 125 5-5. Bifect of Angle of Bank on Triangular Pane! Characteristics; Panel-Panel Interference For a cruciform or planar sing-body combination, banking the mis- sileina positivedireetionintroduces an additional laadingon the right pane! proportional to a8, and subtracts alike loading from the loft panel. The ‘amount of asymmetrie panel loading so produced is given for planar and cruciform configurations by the third term of Hq. (5-28). For erueiform ‘configurations, the loading is given explicitly by the second torm of Eo. (5-30). It is the purpose of this seotion to evaluate the asymmetrical panel normal forees due to bank angle. The difference in the results for the planar and cruciform configurations is an illustration of pane!panel interference, Consider now the second term in Hq. (5-80) for the loading and desi. nate it by Py daa! sin g cos o(6? = 1% Pe BG — BYE = YF ‘ ty aeaee wan) With the notation of Fig. 55, the total normal foree on the right panel due to Py is Me A fm yy fp. ae ane [ey [Poa act sin g cos g ft/s" (54 — 1) ote [ea ay de [Or weet 6 "The integration with respeet to + yields elliptic integrals na" WM ae © Sar etek) + Feeds 30) (/0=) ys OED ee 2TH a0) coupe CHD yy 8 (aa + 188 A further integration to obtain the panel norma foree appears formidable, and the evaluation has actually been performed numerically. ‘The normal foree Z» is conveniently made nondimensional in such a way that it ie apeciied by a it ratio K, depending only on a4/@. Bintan K Boe. (6-41)126 MISSILE ARKODYNAMIES Here Zp is the normal foree on the pavel as a part of the wing alone at angle of attack a. Ze w 8 (GH8), (At), and (5-42) give = rl, ale (a2 eee aga Ke aap | Sage" iva + Pees G48) for a cruciform wing-body combination, 09 Ak oe +— i tN , 04 © 1 | oa} {|_| a | Fo, 58, Interference Wt ration for lang sociated with bank angle Fora planar wing-body combination the expression for K, ean be deter- tmined in the same fashion as fora cruciform combination.” The loading coefficient due to a3 is obtained from the second tern of Eu. (5-80) with ‘The equation for Kis Se aegaaie f sata — 5 \6 Falla? = 178, ‘The values of K, are tabulated in Table 5-2 for ready use and are plotted in Fig 5-8." The difference in K, between the planar and eruei- form cases is astociated with a form of panel-panel interference, In Eq, (6-28) itis seen that the force associated with Ky depends on a coupling between the sidewash velocities r, and a duo to angles of attack and side- slip. ‘The presence of the vertieal panels betvreen the horizontal panels in the cruciform apparently has the gross effect of diminishing the coupling and reducing the value of Ky. Tn the illustrative example whieh follows, the nature of Ky for a triangular wing will be explained, 1d (Got) wrxe-nopy INTERFERENCE 2 ‘but first let us consider the center of prossure of the loading associated with Ky "The conters of pressure associated with the Ky loadings of planar and cruciform combinations with triangular wings have been calculated rnumerieally and are listed in Table 5-2, The centers of presture so ‘Tan 62, Suawonnaooy Panawrias rox Losoiso Dow 70 Bax; “Taisxovtan Paweta* ] Planar Cruciform ' (,. | 5) 0 oman ene oar |. ot |r| 8 cr |e Sa jos Sr so |e. 88 [oem fan fe [og 3h jose 8 oe rape ete) oe ta [ote [oe oa of [ose eae fe fa 0am Be oa o9 | 0.128 Oras 0.625 to [°")[oteo [etare Rog [eran “The accuracy ofthe taulated rola etimated to br 0002 caleulated aro useful for predicting the variation with hank angle of the rolling moments and root bending moments as well as the panel hinge moments, Comparison of Tubles 5-1 and 5-2 shows that tho migration of panel center-of-pressure position with a/éy is much greater for the Ky ppavel force than for the Ky panel foree. No integrated results are pre- sented for the loading on the body which is asymmetrical with respoct to the vertical plane of symmetry. It is elear that the body loading has no net effect on body normal forve, rolling moment, or pitching monet Mtustrative Example Lot us examine the variation of the foree on the panel of a triangular Wing as it sideslips at constant angle of attack. Calculate the fractional change in the panel foree, and compare it with the change computed on the basis of slender-body theory using Ky. With regard to Fig. 5-9 consider a triangular wing with no thickness of semiapex angle «, at augle of attack « and angle of sideslip 8. The pressure of distribution on the wing is eonical with respect to the apos, ‘and the loading of the right panel is greater on the average than that of the left panel for positive sideslip. The ehange in the panel foree with sideslip can be ealeulated on the basis of linear thoory from the results of8 AUSEILE AERODYNAMIC Fro. 610, Changs in ending of panel of tiangulr wing det sideslip, ‘A. L, Jones and A. Alksne® The results for the pressure distribution hhave been integrated to obtain the panel normal force coefficient (C')p. Tet (ACz)» be the change in force eoellcient due to changing the angle of sideslip for © to g. Then (ACz/Cz), is the fractional change in panel force due to sideslip. Normalize the sdeslip angle by forming the pararn- WING-HoDY INTERPEAENCE 19 ler tan 8/tan& ‘Then, for a valuo of the parameter of unity, the left ide edge is streamwise.” ‘The values for the triangular wings are show, iu Fig. 5-10 for two different conditions, For tane = 0.5 and My = 20, the right edge becomes supersonic for afew degrees of sideslip. Actually the foree gained by the left panel is not precisely counteracted by the foree lost by the left panel, but the balance is nearly precise. Lot us now apply the Ky factor to ealeulate the panel force on the basis cof slender-body theory. From Eq, (S-f1) the force coefficient ratio is ‘AC2\ _ BKy Cz Jr ~ tame os) Lot us substitute tan # for so that /aCs\ _ yg tang ¢ -), = Kine 40) ‘The meaning of Kis now elear sine itis the slope of the eurve shown in Fig 5-10. For a planar wing Table 5-2 yields a value of K, of 2/r. The sight line shovsn in this figure has this slope and therefore represents slender-body theory. It is surprising that slender-body theory fits the results of linear theory so well when the large semiapex angles and angles of sidestip are considered 5-6. Summary of Results; Afterbody Effects ‘The previous results apply as dorived to slender planar and erueiform vving-body combinations, It is the purpose of this section to gather together the results into a compact form for application to gotta missiles, ‘The formulas are illustrated by application to a erueiform missle under a Fhunked condition iu the next section. Before suramarising the results, let us note that the panel forees and moments ate not all referred to the some axes as the forces and moments of the other components, The two axis systems and the corresponding notations are given in Fig. 52. For simplicity, the hinge axis of the panel is assumed to have the same longi- {tudinal position as the eenter of moments. Transfer of hinge moments to any other axis ean easily be made. ‘The results for the right panel apply 1 all panels sinee the bank angle is arbitrary. Planar Contguaton Pet nd momen of i pe sina Re (S22), ec BE (LE), atin gence 32), Gam (Cie = Ke wean‘wo ey (Con = =e (S42), 2 a Ke (4962) oem, Fin 9 008 9 2 da )w Forces and moments on body due to wing: Wann = Bo (!2), sings (Cod (5-48) (Conny = ~Ku( 2), 2212 a oot g (Conn = 0 Forces and moments of complete configuration (oe = Coe + Comm + Ke (2), sore (ax + Cann + Ke (EE), sn goons (e+ Coan ~ Rei) oe = =v (M2), on ay (5-49) Cue Crutorm Confgwation Fores nomen tp (coe te (Cie Same aati for a Ea (7 Fores and mometon bay de wig ‘cn ane = Bo( M2), (Coin = 0 (Cane =~ (22), 28 (Chaar = 0 ee a pe ae (5-50) (eae =o + s+ Ro (#2), oe (Coe a Gene (51) dem + Reiger) (2), 5 WING-RODY INTERFERENCE 131 ‘The quantities due to the missile nose can bo calculated by any method applicable to bodies alone. "The results for tho over-all forces and moments on the eruciform missile show several interesting properties. First, the resultant force is inde pendent of bank angle in magnitude and direstion, being always in the plane of a Second, the rolling moments ofthe individual panels add up to zero. These two factors produce an air frame, the characteristics of which are independent of bank attitude in contrast to. planar wing-body combination. The technological importance of the erueiform eonfigura- tion is assoriated in part with this result. ‘Before discussing the application of the foregoing formulas to an actual nonslender ease, let us be concerned with the values of the it ratias and centers of pressure to be used in the thoory as given in Tables 5-1 and 5-2 Actually three lift ratios are cos cemed:Ky,Ky,and Kp, The values of Kir and Kp'as derived do not de- pend on the wing planform although K, does. Nevertheless, asa first ap- proximation to the af coupling term, it is believed that K, ean be used for panels other than triangular. With regard to the panol center of prossure, the values of # and Je are very close to the values for the wing alone for the triangular panels considered in the derivation. Actually, the value of Je does not depend on the planform and should apply to panels other than triangular. For panels ‘other than triangular, itis recommended that the eenter of pressure of the wing alone be used for ., since wing-body interference has little effect fon panel center of pressure for a triangular panel. For rectangular panels some linear theory calculations aro given by Pitts et al* to show the effect of wing-body interference on z» for @ rectangular panel. At worst, interference causes s few per cent forward shift. The values of tho center of lift on the body due to the wing are open to some ertieism ‘when ealeulated by slender-body theory in eertain instances. Let us now consider afterbody effects from the point of view of Ky and (t.)a>- For slender configurations, the length of the body behind the wing should not have an important effect on the body lift or center of pressure. However, for nonslender configurations, the existence of an afterbody can have a large influence on the values of Ky or (Ea)n- In slender- body theory itis assumed that the loading on the body due to the wing carries straight across the body diameter along AA os shoven in Fig. 5-1. Actually the pressure waves travel around the body and follow the hulices intersecting the parallel generators of the body at the Mach angle. Fic. S11. ‘Transfers of LL funy sing to body.132 MISSILE AxRODYNAMICS he pressures on the body are thus transforred a distanee downstream snywhere from zero at the juncture to rBa/2 on the top of the body. ‘The importance of this effect depends among other things on afterbody length and Mach number, Behind the Mach helices from the wing traile ing edge, the wave system from the trailing edge causes a decrease in afterbody loading. A swept trailing edge further complicates the prob- Jem. An approximate model for ealeulating the loading and center of pressure on the body is shown in Fig. 5-12. The body is assumed to be planar and to act at zero angle of attack to “eatoh” the lift developed by ED utcatenng ees: ano iS pe © B io, 52. Planar models for ealelating sfterbany elects. (a) No alterbody; ho. 5 del sterbody (a) No atterbody; the wing. If no afterbody is prosent, the loading on the body is inte- grated only over the region in front of the trailing edge. However, if the ttailing-edge Mach waves interseet on the afterbody, the region in front of the waves is considered to be effective in litt ‘The pressure field duo to either panel is considered to be the pressure field of the isolated panel, With reference ta the coordinate system of Fig. 5:12, the presue field fora supersnie edge is (Ea, 2-35) Baym og 1 §/B + By ‘ 2 FoR TR OO me Cae ‘and fora subsonie edge is p = Sa(Bm)* (1/B =i ‘4 Posten (aera) ow In the application of these fields to the wing-body combination, it has been assumed that the Mach wave from the leading edge of the wing tip falls behind the trailing edge of the wing-hody juneture, This assump- ton, which insures that no tip effects fall on the lifteateking areas, leads to the condition saan (i+ dyes ‘The values of Ke and (24/e;)xcm caleulated on the basis of the planar ‘models are shown in Figs. 5-18 and 5-14. It is apparent that the effect wINo-nODY INFEMFENENEH 133 se Bahari) 4 -1) Hare , - — Bacsn(qy+t)=4 _R ot ont A Lr z is g 2 jie: a = Se bs I ) o Fic, 548, Valu of Kx based on planar mode. (a) No afterbady: @) sfterbods134 MISSILE AEMoDYNanes of the afterbody depends prineipally on the value of the perameter 2Ba/e,. For large values of the parameter the large lifting-eatehing area behind the wing trailing edge causes largor values of Ks and more rear ‘wand positions of the center of pressure. ‘The importance of afterbody effects inervases with Mach number if the afterbody is sufficiently long so that more afterhody area falls in front of the trailing-edge Mach helices, ® Fic. 644, Valu of a/e)as beat on planar mode. (a) No afterbody: () tesbody. 5-1 Application to Nonslender Configurations; Calculative Example ‘The results for the forces and moments summarized in the previous soetion depended on the quantities read from either Tables 5-1 and 5-2 of Figs, 5-13 and 3-14, However, it is noteworthy that the forees and ‘moments, with the exception of those due to the missile nose, are all pro- ortional to the lift-curve slope of the wing alone, ‘The theory was deliberately set up in this fashion; that is, all interference lifts were normalized by the lift of the wing alone. As Tong as the wing-body com binations are slender, the formulas apply without much question. But, if the wing-body combinations are not slender, can the theory be applied with any confidence? It turns out that the anéwer is yes for the following reasons. It is reasonable that the ratio of the interference lift to the ‘wing-alone lift will be better predicted for nonslender configurations than \WING-RODY INTERFERENCE 135 the abssnta magnitsdo ofthe intctornce Ht tall, In fast it tasomol Tha iftrato ed cna of proure ae acartelyprdisted by Sinderboly they, then te Toregnng formas apply diety to an. Snr cowguratots, provid an saute vale he ieee dope — : 1 37S ff ‘Experiment ~202 | L rece Eee acaece Wing aspects Fe, $18. Comparison of theory and experinent for teangular wing and Indy fembinatins. 420 | —§ 400 2s Fo, 5-10, Dimensions of model used in ealoulative example. ‘of the wing alone for the nonslender wing is used. (For this lift-curve slope, either the value from the linear theory or an experimental value will do.) The proof of the assumption lics in being able to predict accurately the measured lift and moments of wing-body combinations by ‘he method, Actually, the method has been tested successfully for large136 MISSILE AERODYNAMICS ‘numbers of wing-body combinations at subsonie and supersonic speeds.* In Fig. 5-15 a favorablo comparison is made between the predictions of the formulas and the measured charueteristies of a systematic series of triangular wing-body combinations varying from slender to nonslender ‘These data are those of Nielsen, Katzen, and Tang.* Calewlative Example: Caleulate the forces and moments of the right panel, the body in the presence of the wing, and the complete eanfiguration for a wing-body come bination with the dimensions shown in Fig, 5-16. Take «, = 0.3 radian, © = 22.57, My = 20. This is the configuration studied by Spahe."™ Asa first stop, let us evaluate the quantities oeeurring in the formulas as given by Eqs. (5-17), (6-50), and (5-51). From the dimensions, we have «on 2985 oan Table 1 then avs ke ois an 08 een For the loading due to a8 coupling, Table 5-2 gives 050 =~ osa7 Qa ome tte tia an = he figure then gives WwINe-HoDY INTERPERUNeH 17 1 value considerably greater than the slender-body value of 0.556. ‘The final quantity required to evaluate the forces and moments is the lift- curve slope of the wing alone. From Bq, (2-36) (2), - Rpg (1 Bint 00 acts = 3" Ac.) _ mt [et us now evaluate the foree and moment coefficients for the right wing panel as given by Ha, (17). ar = 1.25 2 avon 8 : +050 (211) (9 )x0.924) (0888) = 009 (de = = 0.28) 234 (0.618)(0.3)00.92H iiNet se = 228 (25) son aaonen0as = -435 (ins ~129284 (3) oman = 029 218) Eos 0.920 0389 33 0.100 ‘The coefficients for any other panel can be culeulated as ifthe right-hand panel had heen rotated by angle ¢ into its position. ‘The force and moments for the body in the presence of the wing are siven by Bq. (5-50) (Coda = 0.39(2.14)(0.8) (Cuda, = =0.80(2.14)(0.85)(0.3) 0212 (Colm ~ (Cine ~ 0 ‘The forees on the complete wing-body combination are given by Bq (on, (do = Cady + 0.80 + 1.2)(2.19(0.8) = Cady +108 (Coe (Cade = (Ca) = [0.99(0.85) + 1.28(0.688))(2.14)(0.) = (Cy) — 0.725, (Coe 0138 MISSILR AKRODYNAMICS 6-8. Simplified Vortex Model of Wing-Body Combination A simplifod vortex model of a wing-body eombination is useful for many purposes, and such a model is illustrated by Fig. 5-17. Consider the eizeulation distribution aeross the wing panels shown in the figure ‘The actual shape of the distribution is given by the integrand of the integral in Ba, fT is the ciroulation at the wing-body juneture, c= Sstcton A ‘Sectin BB ‘ig. 547. Simpliod vortex model of slender wing body combination then = a fent — yt) tribution curve and is distributed continuously across the wing span, being concentrated toward the wing tips. According to the discussion of See. 0-2, the tralling vorticity soon rolls up into @ conoentrated vortex near the center of gravity of the vortox shect. ‘The enter of vorticity for the present circulation distribution, which is nearly elliptical, lies WING-RODY INTERFERENCE 130 very close to #/4 of the panel semispan from the wing-body juneture Seo Table 6-1, Assume therefore that the external wing panel® are replacod by a bound vortex in the panel plus a trailing vortex on each side as shown in Fig. 5-17, The presence of the circular afterbody requires san image Vortex system to eancel the velocity normal to the body induced bby the external vortices. Or, from another point of view, the bound ‘vortex in the wing as to be terminated inside the body’ in some fashion, Jn so faras the flow in each erossflow plane ean be considered independent ‘of that in other erossfiow planes, asin slender-body theory, we ean satisfy the body boundary condition Dy the inteoduetion of the image trailing vortieos as shown. ‘The image vortices must be 50 located that 655) It is possible to complete the vortices by extending ther forward to form horseshoe vortices as shown in the figure. It is to be pointed out that the foregoing model is not aceurate in the immesdiate neighborhood of the wing because many vortex lines lie on the wing surface, Nevertheless, the model accurately predicted the division of lift between wing and body. Sinco we have replaced the wing-body combination by a pair of horseshoe vortices, we have a uniform loading long the part of the vortex normal to the stream, the so-called lift- ing line. The load per unit spanwise distance of a lifting line in pV, ‘and the lift on the body is represented by the length of the line inside the body, and similarly for the lift on the wing. Thus tex antin(v.—2) = astra 1 will be recognized that this equation is a special ease of Hq, (4-121). "The vortex strength is (5-56) } aat(l = aP/aut + otf Hey r 57) The ratio of the lift on the body to that on the wing panel is, (5-58) ‘The values of K's/Kw obtained from the simplified model are compared with the corresponding values from slender-body theory for several values of the radiusemispan ratio in Table 53. These values are based40 MISSILE APRODYNaMCs fon the value of (Ye ~ a)/(Om ~ a) of a/4. It is interesting that the ‘approximate model prediets a division of lift between body and wing very ‘lose to that of slender-body theory. Behind the trailing edge, vortices roll up and follow the streamline sven implicitly by Bq, (4-88). Actually there is developed a load on the ffterbody because of the motions of the vortices. The actual lond can bie ealenlated by Bq, (4-121). As the vortices pass along the body in the ‘Tanin 6:, Vattns oF Ka/Be 0] of | ot | 06 oa | ost | 0.610 0.22 | 0.350 | 0.650, oitm Senr-body theory Vortex modet downstream dinection, their lateral spacing decreases. Tt can readily be eon from Eq. (+121) that the afterbody loading is then downward, that is, negative. The problem of afterbody loading for a symmetri- cal vortex pair in the presence of a cireular eylinder was studied by Lagerstrom and Graharn.2! symnors, * body radius a body radius occurring with ss 1 aspect ratio of wing alone tu(2) —_unetion of # oeeurring in complex potential B Gare = & chord at wing-body juncture ¢ cerosewind force, Fig. 5-2 Ce tross-wind foree coefficient, Fig. 5.2 G binge-moment coefficient of wing panel, Fig. 5-2 C rolling-moment cooffiient Ci/da —lift-eurve slope per radian Ce lift coefficient, Fig. 52 Cn pitebing-moment coefficient, Fig. 52 G Z foree eoellicient, Fig. 52 B ‘complete elliptic integral of second kind & ‘modulus of eliptic integral ¥ complementary modulus, (1 — #)* K complete elliptic integral of first kind Ke ratio of lift on body in presence of wing to lit of wing slone, 0 Ke ratio of Hit of wing panels in presence of body to lift of wing ‘alone, ¢ = 0 o WING-nODY INTERFERENCE Mi lift ratio specifying additional wing load due to sideslip at eon- stant angle of attack reference length Tift force in plane of Va and tangent of wing semiapex angle free-strearm Mach number local static pressure pressure eoellicient, (p — pe)/ae ‘additional pressure coefficient due to sideslip angle of attack pressure on impact surface (positive a) pressure on suction suefacn (positive a) PP free-stream dynamie pressure radial distanee to external vortex rial distance to image vortex local semispan of right wing panel ‘maximum semispan of right wing panel local semispan of vertical panel ‘maximum semispan of vertical pane! perturbation velocity components along 2, y, 2 tively, for unit Vo perturbation velocity component at a = 8 = 0 for unit Ve perturbation velocity components due to angle of attack for unit Vand unit « perturbation velocity components due to angle of sideslip for unit Vo and unit 8 free-stream velocity complex potential at a= 0 ‘complex potential duc to angle of attack missile axes of symmetry rissile axes of symmetry for angle of attack with y= 0, Fig. 52 ‘coordinates of center of pressure for lauding due to angle of attack coordinates of center of pressure uf additional loading due to sideslip at constant angle of attack lateral position af concentrated vortex, forees along y and 2 axes vis angle of attack, a, 608 & ineluded angle between Vs and mai wing angle of attack angle of sideslip, a in © Iongitudinal axis constant respee- le longitudinal axisM2 MISSILE AERODYNAMICS ry value of P at wing-body juncture Ty) circulation distribution ‘ vie “ semiapex angle of wing slone e polar angle » ‘9/a; also panel taper ratio + sat ba Fig. 5-12 on interference potential & potential due to thicknoss, @ = 0 ee potential due to angle of attack os potential due to angle of sideslip 4 angle of bank Yude Ea, (40) Subscripts: B body alone BOW) ody in presence of wing panels e ‘complete configuration N missle nose or forebody P swing panel w ‘ving alone formed by joining exposed wing panels together W(B) wing panels in presence of body REFERENCES 1. Nishen, Jack N.: Quasbalindscal Theory of Wing-Body_ ote Superionie Spode and Comparnon with Experiment, NACA Tech, ep 2 Morikona, G. Ky The Wing-Body Problem for Linaried Supersonic Plow Aoctoal these, California Tnatitte of Technology, Pasadena, 109. Lawrence, Hy Tey and Ay H. Flax: Wing-Body Interference at Subsonic and Supersnie Sede: Survey and new developments J. Aeronaut Si, vol. 21, n0. 5, 1954 “€ Darand, Willam Frederick: “Aerodynamic Theory” val. IV, pp. 152-157, rsrand Repriting Conant, Cliornia Tnatitite of Technology, Pasadena, 1083. TE Spreter, Joba Rey and Alvin H. Sacks: A Theoretical Stady ef the Aceo- aynamies of Sender Ceultertewing Arrangements and Thcir Wakes, NACA Tech ‘Rept. 1296, 1067 Pitts, Willan C, Juck N. Nicken, and Goorge B. Kaattri: aft and Center of Prenure of Wing-Body-Tail Combinations st Subsonic, Transonic, and Supe tonic Speeds, NACA Tech, epi, 1307, 1957 "t Lageatoin, B.A: Linearized Theory of Coninal Wings, NACA Tech, Note 1085, 1918, 8. Nicken, desk N., Eliott D, Katecn, and Kenneth K, Tang: Lift and Pitching- ‘moment Interference betwen a Bolnted Cylinrial Bady and Trinngular Wings of YVerious Aspect Rios at Mach Nutra of 1.0 und 209, NAGA Terk, Note S796, 18 wino-nopy INrERWERESCE 3 8. Jones, AL and Alterta ¥. Allane: The Load Distribution Dus to Sidetip oa Tiangular, Teapezoidal, end Related Planforma in Supersonic Flow, NACA Tech Nate 207, January, 1960. 40, Spat, J. Richard: Coptribution ofthe Wing Panel tothe Fuses and Moments of Superson Wing-Body Combinations st Combined Angles, NACA Zech Netee 410 danasey, 1958 11, Lierstrom, PA. and M, H, Graham: Aerodynamic Interference in Supersonic Minis, Dower Avera Co, Repl SM1874%, 1050,