ebook img

The Impact of Structural Vibration on Flying Qualities of a Supersonic Transport PDF

12 Pages·2001·1.1 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview The Impact of Structural Vibration on Flying Qualities of a Supersonic Transport

AIAA 2001-4006 The Impact of Structural Vibration on Flying Qualities of a Supersonic Transport David L. Raney, E. Bruce Jackson, Carey S. Buttrill, and William M. Adams NASA Langley Research Center Hampton, VA 23681-2199 AIAA Atmospheric Flight Mechanics Conference 6-9 August 2001 Montreal, Canada For permission to copy or to republish, contact the copyright owner named on the first page. For AIAA-held copyright, write to AIAA Permissions Department, 1801 Alexander Bell Drive, Suite 500, Reston, VA, 20191-4344. The Impact of Structural Vibration on Flying Qualities of a Supersonic Transport David L. Raney*, E. Bruce Jackson _,Carey S. Buttrill*, and William M. Adams* NASA Langley Research Center Hampton, VA 23681-2199 ABSTRACT in the design of the Boeing High Speed Civil Transport (HSCT). Constraints imposed by flight A piloted simulation experiment has been at supersonic speeds resulted in a very large but conducted in the NASA Langley Visual/Motion relatively light and slender vehicle design that Simulator facility to address the impact of exhibited unusually low-frequency aeroelastic dynamic aeroelastic effects on flying qualities of modes. The resulting low frequency cockpit a supersonic transport. The intent of this vibrations had significant potential to negatively experiment was to determine the effectiveness of influence the pilot's ability to maneuver the several measures that may be taken to reduce the aircraft, not only due to the degradation of ride impact of aircraft flexibility on piloting tasks. quality but also due to adverse coupling between Potential solutions that were examined included the human pilot's control dynamics and the structural stiffening, active vibration configuration's aeroelastic dynamics. suppression, and elimination of visual cues A piloted simulation experiment was associated with the elastic modes. A series of therefore conducted in the Langley parametric configurations was evaluated by six Visual/Motion Simulator (VMS) facility to test pilots for several types of maneuver tasks. address the impact of dynamic aeroelastic effects During the investigation, several incidents on flying qualities of the High-Speed Civil were encountered in which cockpit vibrations Transport. The primary objective of this due to elastic modes fed back into the control investigation was to determine the effectiveness stick through involuntary motions of the pilot's of measures that may be taken to reduce the upper body and arm. The phenomenon, referred impact of aircraft flexibility on piloting tasks for to as biodynamic coupling, is evidenced by a the HSCT. The secondary objective was to resonant peak in the power spectrum of the establish preliminary guidelines for designing a pilot's stick inputs at a structural mode structural mode control system for an HSCT frequency. concept. The results of the investigation indicate that An earlier motion-based simulation study structural stiffening and compensation of the using a dynamic aeroelastic HSCT model in the visual display were of little benefit in alleviating Langley VMS facility revealed a significant the impact of elastic dynamics on the piloting reduction in the ease with which piloted tasks, while increased damping and elimination approach and landing tasks were performed of control-effector excitation of the lowest when dynamic elastic modes were included in frequency modes offered great improvements the model. 1 An even earlier investigation by when applied in sufficient degree. Schmidt and Waszak had also been conducted in INTRODUCTION the Langley VMS, illustrating the potentially detrimental effects of dynamic elasticity on the As commercial transport aircraft designs flying qualities of a flexible B1 aircraft become larger and more flexible, the impact of simulation in 1985) The potential for such aeroelastic vibration on the vehicle's flight effects had been noted in a 1983 report by dynamics, flight control, and flying qualities Ashkenas, Magdaleno and McRuer that increases in prominence. The consideration of cautioned of implications of structural flexibility such effects assumed unprecedented significance for the flying qualities of large aircraft. 3 *Research scientist, Dynamics and Control Branch, MS 406, Senior Member AIAA. Research scientist, Dynamics and Control Branch, MS 149, Senior Member AIAA. t Research scientist, Systems Development Branch, MS 125B, Senior Member AIAA. $Distinguished Research Associate, Dynamics and Control Branch, MS 406, Senior Member AIAA. Copyright © 2001 by the American Institute of Aeronautics, Inc. No copyright is asserted in the United States under Title 17, U.S. Code. The Government has royalty free license to exercise all rights under the copyright claimed herein for government puqooses. All other rights reserved bythe copyright owner. American Institute of Aeronautics and Astronautics Theapproacohfthepresenintvestigation trailing-edge flaperons (four per wing), and wastovarycertaipnarameteinrstheaeroelastic leading-edge flaps. modetloprovideasimplifierdepresentatoiofn severaplotentimaleanosfalleviatintgheimpact The vehicle is designed to cma2¢ 300 ofdynamiaceroelasticoitnypilotingtasksT.he passengers a distance of approximately 5,000 potentiaslolutionthsatwereexaminecdonsisted nautical miles. The aircraft has an operating of 1)increasintghefrequencoyftheelastic empty weight of 280,000 lb and a maximum modes2,)increasintghedampinogfvarious takeoff gross weight of 650,000 lb. Final cruise combinatioonfselastimc odes3,)eliminatioonf weight is expected to be 385,000 lb. controleffectorexcitationof the lowest The length of the HSCT configuration is frequenceylasticmodesa,nd4)eliminatioonf approximately 310 ft, with a wingspan of 130 ft. visuaclueasssociatweidththeelastimcodes. The HSCT's length and slenderness result in very Approximarteepresentatiofonrseachof low-frequency structural modes, with the first thesepotentiaslolutionwseregenerateindthe symmetric bending mode occurring at approxi- simulatiomnodeblydirectlyparameterizthineg mately 1.25 Hz. As a point of reference, the modalfrequencieasnddampinga,swellas length of the Concorde is approximately 204 ft, controilnputandvisualoutpuet ffects.By the length of the Tu-144 is 215 ft, and the B1 is explorinpgarametvriacriationinseachofthese approximately 135 ft. The HSCT planform is factorsin,formatiownasgainedregardintghe shown in Figure 1, super-imposed with a effectiveneosfesachapproacahn,dthedegreteo planform of the B1 configuration for scale whichitmustbeexerciseindordetroachieve comparison. thedesirefdlyingqualitieims provemeAntt.otal of20parametrciconfiguratiownesreevaluated B1RideControl bysixtestpilotsrepresentinthgeFAA(1), Va_e(RCV) Boein(g1),NASA(2),andCalspa(2n). HSCT SIMULATION MODEL This experiment used a mathematical RkteContro_Vanes simulation of the so-called "Cycle 3" version of (RCV)andChi_ Fin (CF)addedat nose the "Reference-H" supersonic transport design. 1 The model was published by Boeing Figure 1. Comparison of HSCT and B-1 Commercial Aircraft Group in the summer of planforms. 1996 as the fourth major release in a series of increasingly detailed math models of the The B 1uses an active vibration suppression Reference-H configuration. The simulation control system to improve ride quality at the model was based upon a combination of wind cockpit station. Small active control surfaces tunnel and computational fluid dynamics studies called ride control vanes are located near the of the Reference-H design, ranging from low pilot station of the B1to damp vertical vibrations subsonic to Mach 2.4 supersonic wind tunnel at the cockpit. The Boeing HSCT design is studies. equipped with similar devices for active Finite-element structural models were used suppression of vertical vibrations as noted in to predict the effect of steady flight loads upon Figure 1. Additionally, a vertical "chin fin" is aerodynamic stability derivatives, referred to as included in the HSCT design for active quasi-static aeroelastic effects (QSAE). A key suppression of lateral cockpit vibrations. feature of the math model was the inclusion of dynamic aeroservoelastic effects (DASE), which Dynamic Aeroelastic Effects required additional states to represent the flexible The dynamic aeroelastic portion of the modes of the aircraft structure. model used in this simulation experiment contained six flexible aircraft modes, 3 General Configuration Description symmetric and 3 antisymmetric. The mode The Reference-H vehicle design has a shapes and their associated in-vacuo frequencies cranked-arrow planform with a conventional aft are shown in Figure 2. The model was generated tail and four under-wing engines. The control for a flight condition of Mach 0.24 at aweight of devices include an independently actuated 384,862 lb and a cg location at 53.2% of the horizontal stabilizer and elevator, a three- mean aerodynamic chord, which constitutes the segment rudder on a fixed vertical fin, eight landing condition. 2 American Institute of Aeronautics and Astronautics Symmetric Modes (SideView) Antisymmetric Modes (TopView) variation probably extends beyond the conditions Mode AN1: 1.39Hz that would be physically practical for this design Mode SY 1:1.25 Hz because of the weight penalties associated with producing the stiffer structure. Aircraft weight L Mode SY 2:2.01 Hz Mode AN 2:2.13 Hz was not increased in the simulation as the modal frequencies were increased. IL. Mode SY3:2.70 Hz Mode AN 3:2.82 Hz The "stifl" condition of 1.45 Hz for the first symmetric mode was considered to be most representative of the actual configuration, and so was used as the modified baseline configuration Figure 2. Aeroelastic mode shapes and in-vacuo for all other parametric variations. Acceleration frequencies used in simulation. time histories from the motion-based simulation were used to verify that this parameterization Dynamic aeroelastic modes contained in the method produced the desired effect. model could be excited by turbulence and by control effector movements. No inputs from Configuration Frequency Stiffness 1stSYMode Ratio Increase Frequency landing gear reaction forces or engine pylon baseline 1.00 -- 1.25Hz reaction forces were included in the model. still 1,1_} -.35% 1,45H_t stir2 t,30 ,-85% t,80Hz Visual effects of the structural flexibility stir3 t,_0 ,-150% 2,00}-tz were provided in the simulation. The out-the- 0.10 0.08 0.06 0.04 002 -- damping ratio window scene that was presented on the cockpit monitors moved in relation to the Head-Up Display (HUD) to represent local perturbations in pitch and yaw at the pilot station. The overall effect was that the out-the window scene appeared to bounce slightly both vertically and laterally in response to elastic excitation. These Mode 4 (AN) visual perturbations were typically on the order of + 0.1 degrees during maneuvers performed ........ t with dynamic aeroelastic effects. Mode 2 (AN) Parameterized Aeroelastic Characteristics The baseline aeroelastic model described above was parameterized to provide the ability to systematically vary several characteristics of the 0 .......... -......... -......... •....... X:..X........... X..... piloted simulation. These modifications allowed ::::::::_::::::: ............................... ::!i::i::::......... the impact of structural stiffening, modal ...":::::'.'.'== damping, modal cancellation and visual cues to -5 "'i ......... i......... i.... ..... i.... i" ........ i 0.5 -3 -z_ Real Axis -o_ o be evaluated from a piloted control standpoint. Variation of Structural Stiffness Figure 3. Migration of elastic mode poles with The variation of structural stiffness was structural stiffness variation. represented by multiplying the frequencies of all Variation of Modal Damping six elastic modes by a given frequency ratio. The representation of structural stiffening by directly The portion of the test matrix that varied the manipulating the model in this fashion is clearly modal damping level actually targeted several approximate, but appeared sufficient to capture issues. The first issue was the effect of the the basic effect. Frequency ratios of 1.0 amount of damping that was applied to the (baseline), 1.16 ("stifl" configuration), 1.36 dynamic elastic modes. Damping ratios of 0.07, ("stif2") and 1.60 ("stif3") were chosen. This 0.15, and 0.30 were selected based on feedback selection produced first symmetric bending obtained during an HSR Dynamic Aeroelastic mode frequencies of 1.25 Hz, 1.45 Hz, 1.80 Hz Model Working Group meeting that was held in and 2.0 Hz. Aeroelastic simulation models were August of 1997. produced for each of these conditions. The second issue dealt with the frequency Figure 3 illustrates the migration of the range of the modes to which damping was dynamic elastic poles that occurred as the applied. In the first variation, damping was stiffness level was varied. The total range of applied only to elastic modes with frequencies 3 American Institute of Aeronautics and Astronautics lessthan2Hz. Thisrangeincludetdhefirst Nz ps / Elevator (g/deg) symmetrimcodeat 1.45Hz andthefirst : : ': / antisymmemtriocdeat1.62Hz.Inthesecond variationd,ampinwgasapplietdoelastimcodes o_..................._ base!ine withfrequencileesssthan3 Hz. Thisrange /- oo7 : / includetdhefirsttwosymmetraicndthefirsttwo .....,.............. ............ antisymmemtriocdeTs.heintenwt astoexamine theeffectof dampinognlythefirstfuselage bendinmg odesascomparetodthefirstand seconfduselaghearmonischsowinnFigure2. Figure4 illustratetshemigrationof the 0.US dynameiclastipcolesthatoccurreadsincreased dampinwgasappliedin thesetwofrequency 0 rangesT.hefrequencreysponspelotsshowinn 0 Frequency, Hz 4 Figure5illustratteheattenuatioonftheelastic Ny ps / Rudder (g/deg) responstoecontroilnputsthatresultfsromthe increaseddampinlgevelsapplietdomodewsith frequencileesssthan3Hz.Verticaalcceleration 0.Z5 atthepilotstatio(nNzps)inrespontsoeelevator inputasndlateraalcceleratiaotnthepilotstation 02 (Nyps)inrespontsoeruddeinrputasreshown. 0.1 Anadditionaislsuethatwasaddressbeyd thisportionofthetestmatrixwastherelative 0. importancoef suppressinsygmmetrimc odes versuasntisymmemtriocdesI.nbothfrequency 0.05 rangesa,dampinrgatioof0.3wasapplietdothe symmetrmicodeaslone(configuratio"ndsamp4" 0 " " _ .... ; and"damp9a")ndthentheantisymmemtriocdes 0 Frequency, Hz 4 alone(configuratio"dnasmp5a"nd"dam1p0"). Figure 5. Variation of frequency response to Conflquration Dampinq Ratio Modes control inputs for various damping levels. ........s..t.ir!...........................n..o..m...i.n.a!..............._............................. A total of 10 parametric conditions were included in the damping portion of the test matrix. Again, since the model was directly manipulated to produce the desired damping levels, the representation of an active mode suppression system is approximate and lacks nonlinearities and additional filter dynamics that might be present in the actual system. Z5 ........ 0.i5 - 0.0'7 _- ...... 0.30 °Mode AN3 Impact of Modal Cancellation zo ...,. .:_:"Y_"i..-ModeSY3 Another portion of the test matrix examined ...................."::=........ i...... -Mode AN2 the impact of modal cancellation. This term " _ : Xio. refers to the elimination of the control effector ,CO 15 ...... __- ............_:-_.x =_-:.,"Mode SY2 excitation of a particular elastic mode or modes. 10 i .--ModeAN1 Such a modification allows the control system to b .....ii...... _ _ _-. "Mode SY1 pitch, roll, or yaw the aircraft without exciting ....... .;.i.i.... the specifically targeted modes. It is intended to represent the effect that would be produced by .............:.............................. using multiple control effectors in appropriate ,_ 0 proportion (canard and elevator for instance) to '... . ... pitch the vehicle without exciting the first -5 •.., i i " i ...... "w. .....,.,.,........ii fuselage bending mode. In the lateral case, it -I Real Axis -Z 0 represents the use of rudder and chin fin in Figure 4. Migration of the dynamic elastic poles combination to avoid excitation of the first with damping level variation. antisymmetric mode. 4 American Institute of Aeronautics and Astronautics Configuration ModesCanceled ModesDamped require more control effectors. This representa- stifl none none tion is inherently approximate since it cannot depict the effect of nonlinearities that would be iiiiiiiiiiiiiii iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiSpr e!seiinitii i"ini!iiainiiiiaicitiuiaiiliiiiimiioidieiiiiiisiuipipiirie1ssiioin iiii iisiy site!m iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii such as control saturation or rate limiting. Nz ps / Elevator (g/deg) __ THE LANGLEY VISUAL/MOTION SIMULATOR o._ .................. _.stJf.!.. (ba.se!j.n.e The Langley Visual Motion Simulator cancl (VMS), shown in Figure 7, uses a synergistic hexapod motion system. The motion platform 0.15 ............ i.i..... L.:__./\i[-canc2 ,i, ........... provides up to +0.6 g acceleration cues vertically o.1 ................ _ I_..... _../.._....O_i_!!C_'. ...... within a 5.75 foot travel envelope; lateral and longitudinal acceleration limits are similar. 4The o.o .........i..../........_.-./.i........ :i.:.c.a..n.e..4....... angular limits of the platform are +30/-20 ° pitch, i i/: :: +32 ° yaw, and +22 ° roll. (Positive pitch is in the nose up direction.) 0 Frequency, Hz 4 Linear Accelerations, g Ny ps / Rudder (g/deg) Surge: + 0.6 Sway: + 0.6 Heave: + 0.8 Angular Accelerations, deg/s 2 0.25..............:.:.............:.:.............::.............. Roll: + 50 Pitch: + 50 Yaw: + 50 • I 0.1 ............. i........... i........ i.............. 005 .................. 0 Frequency, Hz 4 Figure 6. Variation in frequency response to control inputs for mode-canceled configurations. The effect of such a mode-canceling control system was approximated by eliminating the Figure 7. Langley Visual/Motion Simulator elements of the B matrix in the dynamic (VMS) and its acceleration capabilities. aeroelastic model that represent the control effector excitation inputs to the first symmetric The cockpit configuration at the time of this and first antisymmetric fuselage bending modes. experiment included a left seat pilot's station and Mode-canceling configurations were generated a right seat observer's station. A four-lever for each of the three damping levels so that the throttle quadrant was located between the two test matrix would include direct comparisons of stations. A McFadden left-handed side stick cancellation on and off for each of these controller was used for all maneuvers performed damping conditions. during this experiment. The pilot station Figure 6 illustrates the change in frequency included an armrest that was adjustable to provide appropriate forearm support for the left responses that resulted from this modification. arm of the evaluation pilot. The plots show that the first symmetric and first antisymmetric modes can no longer be excited Refinements were made to the motion drive by control inputs for the canceled configurations. algorithms to improve the suitability of the The modal dynamics remain, however, and are simulator for representing the aeroelastic motion subject to excitation by turbulence or by cues. The motion commands produced by the coupling from other elastic modes. Cancellation dynamic elastic portion of the aircraft model was applied only to the first symmetric and first generally bypassed the motion washout filters to antisymmetric modes, since cancellation of avoid any attenuation or delay of the elastic higher fuselage harmonics would probably vibration cues. American Institute of Aeronautics and Astronautics Figure8 wastakenfroma 1973report completing a sufficient number of runs to rate a whichdocumentethde frequencyresponse particular configuration, the test pilot's capabilitieosf the LangleyVMS Motion evaluations and comments were recorded. The facility4.Theinput/outpaumt plitudreatiosfor pilot's task performance, in terms of touchdown vertica(lz)andlatera(ly)sinusoidinapl utosf1.8 parameters or flight-director tracking accuracy, incheasreshowna,longwiththeresultinpghase was provided on a cockpit display immediately lag,forinpuftrequenciferosm0.1to12rad/sec. following each run. This information provided Alsoshowonnthespelotsisthefrequencraynge an indication of whether desired or adequate ofthedynamieclastimcodetshatwereincluded performance tolerances had been achieved, thus intheearlie(rLaRC1.)assessm1ent. helping the pilot to navigate through the Cooper- O Harper decision tree. 1.2_ NASA TN D-7349, 1973 In addition to Cooper-Harper Ratings, the pilots were asked to provide an assessment of the extent to which dynamic elastic effects adversely impacted their control inputs and ride quality. - 0 - 1.8 in peaks The rating scales for these assessments are shown in Figure 9 and Figure 10. .5 I. o DASE INFLUENCE ON RQR RIDE QUALITY Cquoaclkitpyi.t vibrations do not impact ride [1] Cockpit vibrations areperceptible but not IF ,__1 objectionable -no improvement necessary. IL2JI Cockpit vibrations are mildly objectionable - IF Q_I improvement desired. ILr.JI OX Cobojcekcptiiotnvaibblreatio-nsimparroevemmoednetratwealyrranted. [4 l 30 - Oy Cimopcrkopvietmviebnrattiorenqsuiraerde.highlyobjectionable- [5] Cockpit vibrations cause abandonment oftask - improvement required. IkOj Figure 9. Evaluation scale for dynamic Figure 8. Frequency response of Langley aeroelastic influence on ride quality. Visual/Motion Simulator as documented in NASA TN D-7349, 1973. DASE INFLUENCE ON PILOT'S CIR CONTROL INPUTS The dynamic elastic portions of the model caused the motion platform to operate at the Preilsoutltdooefsairncoratfatltfelerxciboinlittyro.l inputs asa [1] threshold of its capabilities. The simulator Pilot intentionally modifies control inputs r_ provides reasonable performance (0.8 amplitude toavoid excitation offlexible modes. Ik'-JI ratio vertically and 1.0 amplitude ratio laterally, Cockpit vibrations impact precision of r,l_ with about 15 deg of phase loss) at the lowest voluntary controlinputs. structural mode frequencies (1.25 and 1.39 Hz). Cockpit vibrations cause occasional r/l_ The next two dynamic elastic modes (at 2.0 and involuntary control inputs. Ik"JI 2.1 Hz) lie marginally within the motion Cockpit vibrations causefrequent r_ platform's capabilities with about 25 deg of involuntary control inputs. Ik"JI phase loss. Based on cockpit accelerometer time histories, the facility appeared to provide a Cockpit vibrations cause sustained involun- rc:'_ tary controlinputs orloss ofcontrol. Ik"JI reasonable representation of the modes that were included in the dynamic aeroelastic model. Figure 10. Evaluation scale for dynamic EVALUATION SCALES aeroelastic influence on pilot's control inputs. The familiar Cooper-Harper rating (CHR) These supplemental rating scales were scale was used during pilot evaluations of the designed for this experiment and are intended to parametric aeroelastic configurations. After target the acceptability of a particular aeroelastic 6 American Institute of Aeronautics and Astronautics configuratioanpartfromdeficienciethsatthe signal containing segments from various pilotmayperceivienthenominafllightcontrol maneuvers that had been examined in previous system.Forinstanceif,a pilotawardsthe HSCT simulations. These segments included a landintgaskaCHRof4,butprovideasDASE localizer capture, a glideslope capture, a ControInlfluencReating(CIRo)f1andaDASE descending turn, and a rapid pull-up as RideQualityRating(RQRo)f1,thenwemight performed during an aborted landing task. Flight concludtheatthedeficienCtHRisduetopilot path and track angle command segments from dissatisfactiwoinththebaselinfelightcontrol these tasks were combined with varying order systema,ndnotwiththedynamiaceroelastic and sign to produce a flight director behavior characteristics of that particular configuration. that was not easily anticipated, but was still representative of actual flight maneuver tasks. The CIR Scale shown in Figure 10 bears further discussion. The scale was developed Each of the parametric aeroelastic based on pilot comments from the earlier configurations were evaluated using each of the (LaRC. 1) piloted assessment of dynamic 3 maneuver tasks described above. All tasks aeroelastic effects. Pilots sometimes indicated were performed in the presence of mild that they were "reducing the gain" or "backing turbulence that was produced using a Dryden off'' on their control inputs to avoid excitation of spectra turbulence model. the elastic modes. Several time histories from the LaRC.1 test suggested that the cockpit RESULTS AND DISCUSSION vibration environment had sometimes corrupted the precision of pilot control inputs, or even Configuration Preference Ranking caused occasional involuntary stick inputs that have been referred to as biodynamic feedthlT_. 3 Figure 11 presents a ranking of the 20 The Control Influence Rating Scale was parametric aeroelastic configurations based on developed to specifically address this issue apart the average of the DASE Control Influence from the pilot comfort or ride quality issue. Pilot Ratings and Ride Quality Ratings that were feedback was incorporated during the design of assigned by all the pilots for all three maneuver the CIR and RQR scales prior to this experiment. tasks performed with a given configuration. Cooper-Harper Ratings did not seem to EVALUATION MANEUVERS discriminate among configurations as clearly as the CIR and RQR ratings which specifically The tasks that were evaluated during this targeted DASE effects, and their impact has not investigation included an up-and-away flight been included in this ranking. director tracking task, a nominal approach and landing, and a lateral-offset landing task. The .......w...L offset landing was the most challenging of the three evaluation maneuvers. This task was initiated at an altitude of 750 ft with a 300-ft lateral offset and 580 ft longitudinal offset of the g 4.5 cancellation& instrument landing system (ILS) approach glide- c slope from the nominal approach path. The pilot rr" LU 3.s was directed to fly down the offset glideslope to C9 < an altitude of 250 ft. At this point, the test 46 conductor called "Correct," and the pilot _o_ 2.5 executed a lateral correction and vertical descent © to re-acquire the runway centerline. The pilot then executed a flare and attempted to achieve 1.5 touchdown within the tolerances required for qsaeO __ desired performance. The task required an o.s aggressive lateral maneuver due to the low 1 2 3 4 5 6 8 9 10 11 12 13 14 15 16 17 18 1_ 2_ altitude at which the correction was initiated. PilotPreferenceRanking The lateral offset landing is a standard evaluation maneuver used in Calspan's Total In-Flight Figure 11. Pilot preference ranking of Simulator (TIFS) aircraft flight tests. configurations based on DASE ratings. The intent of the flight director tracking task The ranking exhibits a number of trends that was to cause the pilot to exercise awide range of would appear to make intuitive sense. First, the control input frequencies. The flight director baseline aeroelastic configuration (base0), with presented on the pilot's HUD was driven with a no structural stiffening or active mode 7 American Institute of Aeronautics and Astronautics suppressionw,as rankedas the worst characteristics. On the basis of the average configurationA.lso reassuringlyth,e rigid ratings, the first four configurations (qsae0, (quasi-staetifcfectsonly)configuratiwonithout canc4, canc3, and damp8) appear to be in or on anydynamicaeroelasteicffects(qsae0w)as the border of the acceptable ride quality region. ranketdhebest. 0.5 Differentiatioanmongconfigurationiss greateasttthestartoftherankingin, themost 1.5 desirabrleegiona,ndtaperosfftonear-tieastthe undesirabelendofthespectrumTh.estiffened casewsithouatctivemodesuppresswioenreall 2.5 rankepdoorlys,uggestitnhgatthisapproacwhas noteffectivaetreducintgheimpacotfcockpit 3.5 vibrationonpilotingtasks.Theconfiguration withthegreatensutmbeorfmodesdampe(d4) 4.5 atthe greatesletvel(0.3)andwithmodal cancellatio(cnanc4w)asratedthebestofthe 5.5 flexibleconfigurationbsu,t wasstill very noticeablydifferentfromtherigid aircraft 6.5 (qsae0T).henext-becsatsewasidenticatolthis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 onewiththeexceptioonfdampinlgevelw, hich _*--_: Pilot Preference Ranking worst wasreducetdo0.15(canc3T).hefourthranked configuratihoand0.3dampinogn4modesb,ut Figure 12. Average ride quality rating vs. hadnocancellati(odnamp8). preference ranking of parametric configurations. The orderof rankingprovidessome Also shown on the plot are the maximum interestinignsightsregardinpgotentiatlrades and minimum ride quality ratings assigned to the betweemnode-canceclionngtroalndadditional configurations. On the basis of the maximum dampingA.notheirnsighits gainedwhenwe rating, even the most highly mode-suppressed compatrheerankingosftheconfiguratiwonhich configuration (canc4) provides only marginally hadonlysymmetrimcodesdamped(damp9, acceptable ride quality at the pilot station. It ranked18th)withtheconfiguratiownhichhad should be noted that these ratings were provided onlyantisymmetmricodesdamped(dampl0, during tasks that were performed with mild ranke9dth).It iscleatrhatthepilotsfoundthe turbulence, and that the ride quality acceptability undampeadntisymmetmricotionstobemore will probably vary with turbulence level. This problematthicanundampseydmmetrmicotions. plot provides a subjective basis for the judgment A comparisoonfthedisp0configuration of an acceptable level of mode suppression from withthedampcl onfiguratioinndicatethsatthe the pilot's ride-quality perspective. eliminatioonfvibration-indupceerdturbatioinns thevisualscenperovideldittleornobenefit Control Influence Ratings vs. Pilot Preference accordintogthisrankingT.heseconfigurations This chart provides an analogous ranking to wereidenticainl allrespecotsthetrhanthelack the previous chart in terms of the Control ofvibration-indupceerdturbatioinntsheout-the- Influence Rating instead of Ride Quality Rating. windowscenfeorthedisp0configuration. The CIR assigned by the pilots is shown adjacent to the control influence axis, along with shading Ride Quality Ratings vs. Pilot Preference to indicate the transition from acceptable to marginal to unacceptable configuration charac- The ranking in the previous figure provides teristics. Based on discussions with the test insight regarding the order of pilot preference for pilots prior to the experiment, it was decided that the 20 parametric configurations, but does not the unacceptable threshold should be placed at indicate the point in the ranking at which the point at which cockpit vibrations impact the dynamic elastic characteristics make the precision of voluntary control inputs. configuration unacceptable. Figure 12 shows the average Ride Quality Rating assigned by the On the basis of the average control influence pilots for each configuration plotted against the ratings, the first four configurations (qsae0, Pilot Preference Ranking from the previous canc4, canc3, damp8) again lie within the figure. The RQR assigned by the pilots is shown acceptable region. Each of the acceptable elastic adjacent to the ride quality axis, along with configurations applies a damping level of 0.3 to shading to indicate the transition from accept- the first two symmetric and first two able to marginal to unacceptable configuration antisymmetric modes. On the basis of the American Institute of Aeronautics and Astronautics maximum rating, the most highly mode- accelerations in g's (dashed line) and lateral stick suppressed configuration (canc4) again lies in deflections (solid line). Although the units on the marginal control influence region. the two quantities differ, the scaling of +1 is 0.5 convenient since it represents the maximum throw for lateral stick deflection and since lateral g's commanded by the simulation remained in 1.5 the range of + 1 g. The plot in the lower left of the figure shows the power spectral density of 2.5 lateral accelerations and lateral stick deflections applied to a 6-second segment of the time history 3.5 (from 37 to 43 seconds). The frequency spectrum of the pilot's 4.5 voluntary control inputs during this period lies primarily below 1 Hz. The frequency spectrum 5.5 of the lateral accelerations at the pilot station shows some content at the first antisymmetric 6.5 mode frequency of 1.6 Hz and the second 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 antisymmetric mode frequency of 2.5 Hz due to m:_ Pilot Preference Ranking [Wo_ minor turbulence excitation of these structural modes. (Recall that the frequencies shown in Figure 13. Average control influence rating vs. Figure 2 were multiplied by a factor of 1.16 to preference ranking for parametric configurations produce the "stifl" modified baseline.) The border between acceptable and marginal control influence on this plot is somewhat Time History: Offset Landing Maneuver Task, stir I Configuration arbitrary, since it might be perfectly acceptable for the pilots to intentionally modify their control inputs to avoid excitation of the dynamic elastic modes as long as their ability to precisely control the aircraft is in no way hindered by this ,l practice. However, recorded time histories of 35 40 45 50 time, sec pilot stick deflections indicate that pilots were Power Spect=ra sometimes unaware that cockpit vibrations were L ] ........ in fact impacting their control inputs. Sustained I Feedback The occurrence of biodynamic feedthru of __L_ of Cockpit [_T_ Vibrations cockpit vibrations through the pilot's arm and back into the control stick is involuntary and zI: ] [",--"LateralStictk1 ,=="LateralAccel therefore may indeed be unnoticed by the pilot in Frequency, Hz I,[ _,'/\: I minor instances. Use of the Control Influence Voluntarylnputs 0 Frequency,2 Hz 4 IO_o!y''y/ Rating scale shown in the figure requires the Input Contamination [ Frequency, Hz pilot to be aware of the occurrence, and therefore the CIR ratings may sometimes be optimistic. Figure 14. Power-spectrum analysis of There were, however, a number of profound biodynamic coupling incident for pilot B. instances of frequent or sustained biodynamic The power spectrum of a later 6-second feedthru of cockpit vibrations into the control segment of the time history (from 41 to 47 stick as indicated by the maximum CIR ratings seconds) indicates the bulk of the pilot's input shown on the figure. Frequent or sustained spectrum remains below 1 Hz, but it also shows feedthru of cockpit vibrations through the pilot's arm and back into the stick will be referred to as some frequency content of the pilot's inputs in the range of the lateral elastic modes. Once the Biodynamic Coupling (BDC) in this report. pilot begins to move the stick at the resonant Example of Biodvnamic Coupling Incident frequency of the first antisymmetric structural mode there is tremendous potential for the lateral Figure 14 presents a power-spectrum mode to be excited by the control inputs, analysis of a lateral offset landing run in which producing larger lateral accelerations at the pilot the pilot experienced biodynamic coupling while station. These lateral accelerations can move the flying the stifl configuration with no additional pilot's frame in a fashion which produces damping or cancellation. The time history at the involuntary control inputs that further excite the top of the figure shows lateral cockpit structural mode. 9 American Institute of Aeronautics and Astronautics

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.