[American Institute of Aeronautics and Astronautics 8th Lighter-than-Air Technology Conference - Jacksonville,FL,U.S.A. (05 October 1989 - 07 October 1989)] 8th Lighter-than-Air Technology Conference - Estimation of the flight dynamic characteristics of the YEZ-2A

ESTIMATION OF FLIGHT DYNAMIC CHARACTERETICS OF THE YEZ-2A K.R.Nippress, Airship Industries (UK) S.B.V.Gomes, Cranfield Institute of Technology (UK) AIM 8th Lighter-Than-Air Systems ~echnology Conference Jacksonville, FL I October 5-7, 1989 D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 KR. Nippmss, Akrrhip lndutbkr (UK), and S.B.V. Garrm, C n n W lmtitute of technology (UK) This pper presents an outline of the research conducted to enable estimation of the flight dynar k characteristics d the Y U - 2 A Airship The m d t s obtained from conelation of theoretical tedmiques. wind tunnel work and avai'able flight data will be presented and their i m ~ rct cm the devdopmerrt programme d the airship MI be discussed lJ Free stream m a n vdocfi;l in wind tunnet test section (m/s) j . Infroduct ion Airship Industries are currently under contract to the US Navy to design a large m-rigd airship. designated the YE.-2A, which will democlstrate the dfectivecress 04 such a vessel in the marttime airborne early warning rde. In order to fulfill this rde, the airship will need to operate in a hostile weather environment for rdatrvdy long periods, without sacriftcing mission effectbmess or placing an undue wwkload on the flighf m If the important performance parameters of range and erdwarwre are to be kept at or near optimum dues, the airship must be designed tor adequate strength coupled with minrmum weight It must also have an efficient fl~ght cuntrd system incorporating stability augmentation and an auto-pilot function so that pilot wockloed b kept st a manageable lave( and the probsbllity d generating high maneuver W s is reduced The structural strength and contrd design aspects each require that the airship's fligh: dynamic characteristics can be estimating with a high level of confidence for flight in both calm air and turbulent, gusty atmospheric conditions Previous design experience indicates that gust response dominates the structural design and maneuvering flight in moderate turbulence at high angles of attack and/or stdeslip generat& by hign contrd deflections dominate control system design It is evident therefore, that a necessary design tod is a mathematical model of the airship flight dynamics incorporating high 'fidelity aerodynamic data whict? should extend up tc high angles of attack, sideslip, and control deflection that is a non linear, wide angle model This paper will describe the strategy used b) Airship Industries to gair; the necessary aerodynamic data to establish confulence in the 'mdd A brief description illustrating the use of the modd in establishing the air loads encountered in manewering, gusty and turbufent night will be given When discussing the strategy for modelling airship response tt is instructive to compare any proposed amhip sirnuiation with its heavier than air counterpart. The m s o n for this is that there has been an immense amount of work in conventional aircraft simulation in recent years and any applicable expenerencc should be utilized. The model for both types of cratt will comprise the fdlowing basic elements: a) Kinematic Relationships, In which the forces and moments acting on the aircratt are related to Its motion with respect to a flat. non-rotating Earth D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 The kinemgtic feh&mh@ for an airship are theorettcaltyidenticalto-fordherm and so wll not be dtsarssed in decal hem. In pcactb the only d#rscence betwaecl airships and heavier than ah craft is that vbtual lnertla effeds. which m y be negteded for conventional aimaf!. are signkant for airships. These virtual inertia t e r n m y simply be induded h the mass matrbc for simulation work. tt shouid however, be noted that virtual berth effects are really aerodymmlc forces exerted on the cxah due to its mation through the air and their indudon as added mass terms is stridly a mathematical convenience. This k an imponant potnt to be h mind when considering load distribution data for stressing MSeS 2 2 Airshi9 A e r o d m r n ! ~ The aerodynamics d airships are significantly difierent from fixed wing aircraft. For the former, the forces hduced on the hdl due to angle of attack, for example, re comparable to those produced on the fins, whereas for the 4t:er the wing dominates Also, the treatment of gusts and turbulence differ in that a fixed wing aircraft i r ~ genera! has a wing chord significantly smaller than ttw scale length for tumulence and therefore moves relatively rapidly through it. bmequently the turbulence velocfty components may be assumed to be cms!ant over the wing. Such a dmpifficatkm Is not possible for airships. Because d their great hull length and relatively d w speed, the spatial distribution of twbulerrce dong the hull and fins must be considered. h is evident that atthough the charactertzation 01 the aerodynamic forces and moments acting on airshlps may be markedly different from their heavier than air counterparts, them & merlt in considering a similar development test strategy. 3. Test A tYpfcal test pqjramme for a header-than-air c m w o u ~ ,M ol the fdlowlng m m n t s a) A wind tuvrel test pro~ramrrpe to pfajuce force and moment data e) Match d t!w model tesportse to the night response and update the aerodynamic data accordingfy. The strategy adopted for the YEZ-2.9 was to:- i) mqoutwlndtWtestsonamodelof the existing Skyship 60(3, match the data for the SKS 600 to flight dab obtained from the 'fly by light' test programme on SKS 60044 ii) devdop the YET-2A airship in the wind tunnel li) use the experiencs gained in correlating the SKS 600 data with full scale when developing the YEZ-2A mathematical model.A block diagram d this strategy is shown in Fig 2.1. This paper will describe ttre wind tunnel vmrk In support of the YEZ-2A and the process d conelattng the theoretical, wmd tunnel, and flight data for the SKS 600. --. 3.1 Wind Tunnel Proummm Wtth the establkhmt of cooperation batween Airship Industries UK and the Wkge d Aeronautics. CrenReld. UY a programme of wind tur#lel work otlerrtated towards establishing the dynamic cbracreclstics of the airship was devised. The objedkes of this programme were as fdl-:- a) lnvestlgate techniques for testing airship models in the w(rd tunnel, in particubr scale effects cd the now over hull and fins, XI that correlation wtth existing fufl scale data could be attempted D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 c) Derive the YEZ-2A amfigurations and establish the pressure distributbn rowd the hull and fins and the f m and moment d) Having established the steady state character- bks, evailrate the frequency and damping cbaractertstlcs ot the airship model Ltjng free and torcred oscillation tests e) Investigate the response of the airship to turbulence in typical flight conditions, that is the lower layers d the atmosphere. The inv8s!cgation described above was performed at the College of Aero~ut ics as five distinct tasks With the exception of the turbulence testing. all d the test work was c a d ot? in the 8fi by 6?! wind tunnel at Cranfield Task 1 . Scale Effect Investigation The p u m d this task was to investigate the effect of surface roughness on the wind tunnel data obtained on a 1/75 scale model of the SKS 600 in order to sdect a value of surface rough- which gave resutts representative of the full size craft. Several techniques for obtaining dtffering degrees o: roughness were tned, induding covering the model with women's stockings, as can be seen from Fig 3.1. The resldts were compared with existing data (References 1 and 2). The minimum degree of roughness. found to be the most appropriate, ensured that: - (i) a turbulent boundary layer remained amched over the model length (ii) cross-flow vortices formed as expected on the MI size airship (iii) (pve drag data marginally above the values expected from existing information. The minimum degree of surface roughness was therefore used for all subsequent YEZ-24 wind tunnd w& The force and moment data W i n e d trwn this task formed the basis for the aerodynamic data in the SKS 600 mathematicst model. Task 1A: YEZ-2A Optimization Study The purpose of this task was to investigate the effect of rear hull shape and fin combinatims. A 1 /75 scale model was wed in the Cranfieid Bft by 6ft wind tunnel. Three types of hull rear end were tested in conjunction with two fin positions and two fin types. The results indicated that a fatter rear end and slightly smaller fins than originally proposed could be used. This would give extra gas lift with no discemaMe change in drag, whilst still maintaining good stabilrty and control characteristics , Task 2: YEZ-2A Force and Moment Data This task was intended to provide the basic force and moment data for the YEZ-2A design so that air loads and stabilrty and control characteristics for the cratl c o w be estimated The test programme was designed to provide overa!! forces and moments, pressure distribution, a m contrd hinge moment data for the airship (Fig 3.2). The tests were carried out at a Reynolds Number of 6 million, which is consrdered sufficiently large to minimize scale effects. This was borne out by the reasonable drag estimates obtained from these tests (Fig.S.3A). For those interested in wind tunnel statistics. the steady stat9 test programme carried out in Tasks 1 and 2 yidded approximately 6000 data points from the force and moment tests and 16000 data point; for the pressure distributions. -% Task 3. YEZ-ZA Dampicg Derivat ~ e s In order to estimate the damping derivatives for the YEZ-2A a special Dightweight 1/75 scale model was constructed and mounted in the 8fl by 6fl tunnel in such a way that it was free to oscillate in both pitch and y ~ w (Fig. 3.4). Free and forced oscillation tests were carried out in the manner described in Reference 3 and the damping derivatives were estimated The results gave an indication of good stability characteristics of the design tested. Task 4: Turbulence Testing In order to evaluate the airship's response to turbulence, a 1/200 scale model was tested in the 8tI by 4f i Bounda b y e r Tunnel at the Cdlw,e of Aeronautics. 7 his tunnel Is designed for atmospheric turbulence simulations D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 In previous putlrshed work in this M, both of a t h e o m w s n d e>cpertmentJ- (Refererrces 4.5 and 6). the authors have used the cltmospherlc turbulence nwdel more approprfste to aMtudes above 5.000 R In this work the technlquas for stmuiating Icm attitude turbulence outlined &I Wind Engbwhg Science ( R e f ~ c ~ n c ~ s . 7 and 8) were used, as this ecn/konment was cmsjdered more relevant to the mission profile for the YEZ-2A. A ca&d process of t u r b t ' m scaling in the wind tunnel test sedion w.is carried out before the airship model wJd be installed. Thk irndvad the use of Counihan teeth and case bted walls for vortex generation, and locf b m l s mrryfng toy buUiing bricks to simldate the lower atmosphere's bwnjary layer (Fig. 3.4). The W e arrangement was adjusted untii the tunnel test section yielded a scale factor as dose as possible to that of the airship mode!. The 1 / 2 0 scale model was built d uhra-iigh: material so as to minimize inertk, containirs an intemal strain-gauge balance located a: tts Center of Volume. capable of measuring the dynamic response in terms of lift and pr?chiw momenl. By cross condating the turbulence measurements. mad= simdanecusly with hot-wire anectornetry, to the airship model response. the tracsfer functior, beween turbulence and airship response was determined as a fundion of the wave number (Fig. 3.5). In addition, the pl-ma shift and coherence functions between ltte turbulence input and the airship model response were also determined as a function of fi .The peaks in the transfer function curm were then correlated with possible phase shifts and high values of the coherence function. In this way, the turbulence originated peaks could be discriminated against those generated by o?b: ursteady aerodynamics phenomena and the wave numbers at which maximum response was expected could be noted once it k applied at the appropriate scale. The design of the auto pilot and autostabilization systems can incorporate these data. Airship response under gusty and turbulent corrdttions implies encounters with high cross- flow angles (angle ot attack or sideslip) and its variation along the hull and fins Wind tunnel data d d only be economically obtained at It is intended to use the twbuience data obtained f m the wind tunnel as hput to the mathematical mode4 so that the validity of applying essentially steady state cmcepts to non-steady m j h i o n s can be assessed. 4.1 Hull Aerodvn~mics The hull awodynamics were obtained by cornpartson of the wind tunnel data with the work of Munk (Reference 9) and Upson (Reference 10).BcQh Munk's and Upm's papers consider invoiced flow deus only. which whlst ghring a reasonable approximation to the hull pitching or yawing moment. produccs zero lift or s i d e f m Wind tunnel testing indicated the early formation of vort~ces along the hull To generate this lir, term the crossflow drag concept due to von & m a n was employed (Reference 11) A typical match between the n o m l force distribution generated by this 'Modified Upson' theory and that generated in the wnd tunnel at moderate angle of attack is presented in Fig 4 1. It can be seeii that good agreement Is obtained Figs 4.2 and 4.3 compare the theoretical and measured overall normal force ard pttchinb moment coefkients as functiors of angle of attack. The agreemem throughout the angle oi attack-. range is excellent 4.2 Fin Aerodvnamic~ Airship confiuratms are diflererrt from t h of convent~onal aircraft in that the ratb of body radius to fin span is sufftciently large that fin- hull interference effects are s@nifiip,n! Although the 'fin plus h&!' !ads are sufficierri for rz~;sl-~se calcutations, the ratb d the icad induced on the fins to that Induced on the hutt is important when assessing structural integrdy Available techniques such as Weber and Hawk (Reference 12) depend on conform transformarion techniques end are therdore only applicable at low angles of attack or sideslip A wind tunnel investrgatlon using fins mounted on strain gauged rods has shown that the theoretrcal techniques indeed provide good D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 Wimates at law crr>sstlow angles but over- esttmste the Rn panel contrtbutiorr at higher craMw sngles, as can be seen from Fig 4.4. The fin hult interference problem at high angles d attadc or ddeslip me&! further work. The mathematical model constructed using the aerodynamic data obtained from correlation between the wind t u n d and theory was compared wi?h data obtained from the 'fly by light' test programme on SKS 600-04 in WeeksvUle, USA. These fl+$~t data were obtained from both longitudinal and latzral maneuvers exched using both small amplitude and largs emplitwfe contrd inputs. A typical match to the flrght data is shown in Figs 5.1 to 5.4. These present time histories of measured and predicted data of airspeed, normal accei- eratiorr and longitudinal acceleration for a high amplirde pull push maneuver. It can be seen that the model response adequately matches the measured response, giving cor?iidence in the wind tunnel data and the theoretical basis of the m d d No refinement of the correlated aerodynamic dam was necessary to achieve this match. indicat~w tha! the care taken in obtaining representatwe flow over the wind tunnd models pad dhrdends in reducing one element d uncertaimy in correlation. The success of this firs1 attempt at coneiat~ng the SKS 603 model with flight data has rndically changed the philosophy for further cierodynamic investigative work for !he Y E - Z A p;Ge,-1. R=!t.&i than perform further wind tunnel testing on more detailed modeis of the airship. the detai! design of an Airship Industries new airship. designated the Sentinel 1000. was adjusted to be geometrically congruent with the chosen YEZ-2A shape At a linear haH scale. this ship - now in build - will be flight tosted dud??$ : m. This airship ' d i be equipped with a stated-the-art, micro computer-based recording system whict~ will otfer facilities for simultaneous recording and data processing whilst airborne. A comprehensive set of high qualtty sensors will be fitted to the airship to enabie 'maximum Ilkdihood' statistical techniques such as those used at NASA Dryden, USA (Reference 13). and Delft Universtty, the Netherlands. (Reference 14). to produce minjmum error esthnatos of the mathematical model para meters This should provide in a HI@[ dynamics database for tho YEZ-2A of the qualrty hitherto only found in high techndogy fied wing aircraft projects and shadd result in a very efficient development programme for this airship. 6. Analysis of Air Loads In addition to work on the Y U - 2 A and Sentinel 100 . Airship Industries has been engaged in certification d the 5WHL airship to both CAA and FAA requirements This process invdves the estimation of loads induced by maneuvering flight and gust encounters and the airship mathematical model has proved to be a valuabie tool for this work. A noticeable feature of this analysis is that the gust encounter case always generates the limit load cases for the design Also. the variation between the respective design loads is considerable, depending on whether the CAA. FAA or MIL Spec requirements are used in !Miz analysis It must be inferred from the above that there is a strong possibility that the present requirements possibly result in over strength. and therefore overweight airships Since weight is a critical factor in airsh~p performance, it is possible that design and development of these craf? is being impeded by over conservative requiremews Ratification of such design criteria is theretore felt to be necessary One possible strategy for &?stking this prnberr, wouid be to - (i) consder the nature of atmospheric drsturtances using the b o d y of data collected in the USA and Europe ... (11) Def~ne an event as a flight case cs~si~iirlg of manewc: combrned wrth gust combined w~th random turbulence (iii) for events of equal probability level. estimate the loads induced on the airship and plot the maximum loads against probability of occurrence Such an BM~YSIS wotild reveal those parameters whch have the greatest effect on the induced loads The des~gn requirements - co J d then be reassessed in the I~ght of thrs ex- penence The work on the YEZ-24 programme, correlated with the Sentlnel 1000. Skyship 600 and Skyship 500HL analyses provides a sound toundat~on tor reassosmenr of the various agenoes I equrrements D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 Use of the rnathgmatical model h estimating a:fioads for certification d existing airships has indicated that the gust and maneuver kwd requirements of the CAA, FAA and Mil Spec complbnce documents shouM be reas& in the light of current knowledge about atrrospheric disturbances and airship response ~h~ractertstics D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 NCDE-NCE EGREESI X-SKS-600 WIND TLNNEL DATA CT&, 1) .- YEZ-2A WND W E 1 DATA (TASK 2] 0 - SKS-500 F!JGHT TEST DATA ' I P A E I FtGLlRE 3.1A - M A S CAMPARISON D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 Fig. 3.3 - The Task 3 1/75 scale model of the YEZ-2A.airship In the test section of the 83 x 6fl wind tunnel The electric driving motor k on the far left side, and there k on; spring on each sde of the connecting rod attached to the model stem. D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 Fig 3.5 - Transfer function, phase s h i f t and coherence values of L i f t as a function o f the wave nunber f o r the 0 degree of yaw and p i t c h a t t i tude . These a r e p a r t of the Task 4 resu l ts . D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 COMPARISON OF THEORETICAL & W / T RESULfS For Hull L o a d i n g Distribution of SICSdOrJ 10 d e g r c ~ Angle of Attack x/L Figure 4.1 COMPARISON O f M E O ~ W & 1-fl RESULTS SKSBOO. Hull Only CZ vs Angle of Attack Figure 4.2 COMPARISON OF TIIEORR'ICAL k W / T RESULTS SKS600. Hull Only CM vs Angle of Attack 0.4 ti r 0.35 - C - 0 0.3 - Wind Turmel Result. Iheorsticd M m o t e I l l 1 I I 0 6 1 0 1 5 9 0 2 5 3 0 3 6 4 0 4 Angle of Attack (degrees\ Figure 4.3 . D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 S K W RESPONSE TO PULL-PUSH MANEUVER SKS600-04 RESPONSE TO PULL-PUSH UANEUVEP. SKW0-04 REljPONSE TO PULL-PUSH MANEUVER U - *nunun) -- Klw y u l A r n ~ i , ! I D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 FIGURE 4.4 D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73 References 1. Bailey, D.B. - 'Patrol Airship Concept Evaluation', Naval Air Development Center, Warminister, Pennsyfavania, Report No. NADC 85019-60, Vol. 1, 15 March 1985. 2. Hurst, 0. - Wind Tunnel Tests for Airships Industries (UK) Limited', University of Southampton, Dept. of Aeronautics and Astronautics. December 1986. 3. Simmons, L.F.G. - 'Note Relating Two Oscillation Methods in Use for Determining Rotary Derivatives of Models', Aeronautical Research Committee, Reports and Memoranda No. 71 1, London, January 1921. 4. Calligeros, J.M. and Mc Davitt, P.W. - 'Response and Loads on Airships due to Discrete and Random Gusts', Bureau of Aeronautics, Department of the Navy, Contract No. NOas 56-8254, M.I.T., Technical Report 72-1, February 1958. 5. Lagrang?, M.J.B. -'Aerodynamic Forces on an Airship Hull in Atmospheric Turbulence", University of Toronto Institute for Aerospace Studies, Report No. 277, Toronto. Canada, April 1984. 6. De Laurier, J.D. and Huj, K.C.K. - 'Airship Survivability in Atmospheric Turbulence', AlAA Lighter-than-Air systems Technology Conference, Annapolis, Maryland, July 8-1 0, 1981, AlAA Paper No. 1323. 7. Cook, N.J. - 'Determination of the Model Scale Factor in Wind Tunnel Simulations of the Adiabatic Atmospheric Boundary Layer", Journal of Industrial Aerodynamics. 2 (1 977/78) pp 3 1 1 -321. 8. Cook, N.J. - 'Wind Tunnel Simulation of the Adiabatic Atmospheric Boundary Layer by Roughness, Barrier and Mixing-Device methods", Journal of Industrial Aerodynamics,3 (1 978). pp 157-1 76. 11. Von Karrnann, T. - 'Calculation of Pressure Distribution on Airship Hulls", TM No 574. NACA, 1930 12. Veber, JA. and Hawk, A.C. - Theoretics! Load Distributions on Fin Body Tailplane arrangements in a Side Wind/", R and M No 2947, ARC, 1953 13. Maine, R.E. and Illif, K.W. - Identification of Dynamic Systems" AG-300-VOL 2, AGARD 14. Mulder, J.A. - 'Estimation of Aircraft State in Non Steady Flight" CP 172 AGARD, 1975 3. De Laurier. J.A. and Jones, S.P. - 'Aerodynamics Estimation Techniques fo! Aerostats and Airships", A I M Lighter-than-Air Systems Technology Conference, Paper No. 81 -1 339, Annapolis, Maryland, July 8-1 0, 1981. 1. De Laurier, J. and Schenck, D - 'Airshrp Dynamic Stability", AlAA Lighter-then-Air Systems Technology Conference. July 11-13, 1979/Palo Alto, Ca. A I M paper No. 79 - 1591 2. De Laurier, J.A.. and Podleski, S. - "A Semi-Empirical Method for Estimating the Subsonic Aerodynamic Characteristics of Finned Axisymmetric Bodies". Proceeding. Tenth AFGL Scientific Balloon Symposium, At: Force Geophysics laboratory, Aug. 1978. 9. Munk, M.M. - 'Aerodynamics of Air:.hIpsn, Aerodvnami~ Theory. V d . 6, W.F. Durand, ed., Julius Springer, Beriin, 1936. 10. Upson, R.H. and Klikoff, W.A. - 'Applications of Practical Hydrodynamics to Airship Design'. Repop No. 405, NACA, June 1931. 4. Dorinelly, S.L. - ;An Analogue Simulation to Investigate Airship Dynamic Stability and' Response", MSc. Thesis, College of Aeronautics, Cranfield lnstitute of Technology, Cranfield, September, September 1988. D ow nl oa de d by O L D D O M IN IO N U N IV E R SI T Y o n N ov em be r 18 , 2 01 4 | h ttp :// ar c. ai aa .o rg | D O I: 1 0. 25 14 /6 .1 98 9- 31 73