https://ntrs.nasa.gov/search.jsp?R=19930087545 2020-06-17T09:26:49+00:00Z RESEARCH MEMORANDUM PRELIMINARY FLTGHT MEASUREMENTS OF THE DYNAMIC LONGITUDINAL STABILITY CHARACTERISTICS OF TEE CONVAIR XF-92A DELTA-.ZING AEPLELNE By Euclid C. HoIleman, John H. Evans, and William C. Triplett NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS WASHINGTON June 30, 1953 .- u NATIONAL ADVISORY COXMITTEE FOR AERONAUTICS - " RESEARCH MEMORANDUM PRELIMINARY FZIGHT MEASuRplENTS OF THE DYNAMIC LONGITUDLNAL STABILITY CHARACTEEISTICS OF THE CONVAIR XF-9 DELTA-WING By Euclid C . Rolleman, John H. Evans, and William C. Triplett SUMMARY Some longitudinal maneuvers obtainedduring the U. S. Air Forceper- formance tests of the ConvairXF-W airplane have been analyzedusfng by measured period and timeto damp tohalf amplitude and by Reeves Electronic halog Computer (REAC) study to givea preliminary measurementof the air- plane stabilityand damping at Mach numbersfrom 0.59 to 0.94. For the range of these tests, no loss in control effectiveness was shown, . thestatic stability Cmcr increasedwith Mach nmiber, the damping was light but positive, and the damping factorC&i 4- % was lm. INTRODUCTION The XF-PA airplane was constructed by the Consolidated-Vultee Aircraft Corp. to provide inf'ormation on the flight characteristicsof a 60° delta-wing configuration at subsonicspeeds. Increased interest in the delta-wing configurationfor supersonic flight prompted the replace- ment of the originalJ-33-A-23 engine witha J-33-A-29 engine with after- burner. Air Force demonstrationand performance tests have been conducted since this change with the National Advisory Committeefor Aeronautics providing instrumentationand engineer- assistance. During these testsrandm longftudinal disturbanceswere obtained which were considered suitablefor stability analysis although these maneuvers were not performed specifically to obtain thisof informa-type tion. Under certain flight conditions undesirable lateral and longitudinal oscillations have been observed and were believed to indicate the possi- bflity of cross coupling between the lateral and longitudinalof modes - motion. Presented in this paper are preliminary results obtained by 7 2 NACA RM L53E14 analyzing maneuvers at Mach numbers from0.59 to 0.94. It shouldbe emphasized that these results are prelFminaryand are to be followedby a detailed research program designed to investigate completely the sta- bility characteristicsof the airplane. transverse acceleration,g units normal accelerati.on, g units C,, C1, k, and b transfer-function coefficients CL lift coefficient pitching-moment coefficient about airplane center of gravity C dCdm La . C% I- dCddGe cN airplane normal-force coefficient - C mean aerodynamic chord, ft %/lo cycles for oscillation to damp to 1/10 amplitude acceleration due to gravity, ft/sec2 pressure altitude, ft 3 airplane moment of inertia in pitch, slug-ft2 t,: Mach nmber c m airplane mass, slugs P period of oscillation, sec S wing area, sq ft time for oscillation to damp to amplitude, T1/2 1/2 sec t time,sec v forwardvelocity, ft/sec a angle of attack, radian B sideslip angle, deg elevon control angle, deg, radian rudder control augle, deg 0 pitch angle, radian * P air density, slugs/cu ft roll angle, radian d/dt & &/at 9 d0/dt dddt Subscripts: L left R r i&t 4 NACA RM L53E14 The Consolidated-VulteeXF-w airplane is a single-place 60° delta- wing airplane powered bya turbojet engine with afterburner. TableI lists the physical characteristics and figure1 presents a three-view drawing of the airplane. The inertia values used were supplied by the manufacturer. Weights and cen-ker-of-gravity positions for the airplane were determined from the quantityof fuel remaining. The airplane is controlled longitudinallyfull-span by elevons asd laterally by the same surfaces operating differentiallyby a con-and ventional rudaer. Controls are operated an by irreversible hydraulic system. c Standard NACA recording instruments werewed and were synchronized by a common timer. Airspeed measurements were recordedfrm a total- pressure tube mounted aon boom approximately'3.4 feet aheadof the air- plane nose inlet. Center-of-gravity accelerationsand velocities were measured by direct recording accelerometersand rate gyros. Accuracies of the recorded quantities are: M ................................t 0.0 3 6, radians per sec ....................... *O.W Az, ad Ay, g units ....................... to. 05 Control position, deg ...................... b.1 Airplane weight, lb ....................... 2100 TESTS AM) METHODS OF ANALYSIS During Air Force performance testsof the XF-92A airplane several longitudinal maneuvers suitablefor dynamic stability analysis were obtained. Time histories of representative runs are.presented fn fig- ure 2. The maneuvers were obtained at Mach numbersof about 0.59, 0.80, 0.81, 0.91, and 0.g4 at 6,700 feet, 23,000 feet, 36,000 feet, 30,000 feet, and 35,000 feet, respectively. About 20 seconds of each record are shown to emphasize the natureof the airplane oscillation, the sensftivityof the control system, and the effectivenessof the control surfaces and to show the resultsof RFX studies. Since no flight tests have been made specifically to obtain stability data, maneuvers were selected which couldanalyzed be by either of two methods. The first was the simple method described in reference1 in which the period and time toof dampthe amlane motion are measured t directly from the control-fixed portionof the time histories. The second method malres useof the REAC (Reeves- Electronic AnalogComputer); actual 5 control deflections are used an as input to theREAC and a solution (time E, C1D + C, history in 6) for the transfer-function equation - = . 'e D2 + bD t. k (ref. 2) is obtained for a particular setof values of transfer-function coefficients Co, C1, b, and k. These coefficients are then varied as necessary until the output6 most nearly duplicates the actual flight record. Both methodsof analysis are basedon the usual assumptions that the aerodynamic forces and momentsvary linearly with certain variablesand that the forward velocity is constantduring the maneuver. In addition, the simple analysis is valid onlya freefor abplane oscillation with controls fixed. In each of theruns the pilot isattapting to damp the airplane oscillation; consequently, only the small-amplitude portionof the oscillation approacheda controls-fixed condition. To assure greater accuracy in theREAC amlysis it WRS necessary to use the large-amplitude portions of the flight records where control the motFons were of sfepifi- cant magnitudes. The effects of changes in the trim 6, due to Mach number and altitude change during the test were not in includedthe REAC computations. RESULTS AND DISCUSSION * Figure 2(a) presents a time historyof an amlane oscillation obtained in a climb at about36,000 feet and at a Mach number of about 0.81. The oscillationwas analyzed by the sfmple method beginning at time 12 seconds, whereas theREAC analysis was made from ttme1 to 14 seconds. Figure 2(b) showsa gradual dive recovery at about 30,000 feet with Mach numberwrying from 0.9 to 0.9 and with normal acceleration varying from lg to 2g. The inFtial disturbance appears to have been lateral with attempts to control this motion exciting the longitudinal oscillation. Analysis by the simple methodwas attempted beyond time 15 seconds. The results of the RFX! studies are shownfrom time 6 to 16 seconds. Figure 2(c) shows a time historyof a dive from 38,000 feet to34,000 feet ata Mach number of 0.94. Control deflection during the maneuver, likethe other example histories, are small. Fol- lowing time 18 seconds an analysis by the simple methodwas attempted. From tFme 2 to 16 seconds results of the REAC analysis are shown. Since the resultsof the REAC studies obtainedby ueing the longitudinal con- trol as theonly input to the eystem inare good agreement with actual flight records, it would appear nothat serious coupling between the lateral and longitudinal modes exist, althoughamlane the underwent lateral as wellas longitudinal motion. 6 L NACA RM L53E14 For the simple analysis the period and the to damp were measured directly from the controls-fixed portion of the time histories and for the REAC analysis the same information was calculated from the coeffi- cients k and b. These data were corrected to a standardaltitude of 35,000 feet and were cmbined to give the cycles required to damp to 1/10 amplitude. The results of these measurements are presented in fig- ure 3 and show the measured periods and ttme to damp for an altitude of 33,000 feet to decrease with Mach number. Although it is not possible to define clearly the variation of cycles to ARII1T) to 1/10 amplitude with Machnumber it is apparent that the airplane, at an altitude of 35,000 feet, doee not meet the Alr Force dynamic stability requirement that the short-perid longitudinal oscillation damp to 1/10 amplitude in one cycle(ref. 2). Figure 4 presents the results of figure 3 in the form of stability derivatives C and C C asfunctions Mach number. Also 92 %%f of shown is a plot of obtained from the REAC analysis. For the sim- ple analysis, and Cmi, + Cmb! were calculated frm Similar equations from reference 3 are used to convert the transfer-function coefficients C1, k, and b tostability derivatives, a8 follows: 2Cl Iv 2kIy c-- - pv2sc 7 - The lift-curveslope C& used inthe computation was obtained from flight data and is presented also in figure 4. The control effectiveness derivative Cqe has a value of about -0.55. No loss in control effec- tiveness is indicatedfor the range of these tests. The sfmple analysis affords no way of obtaining this parameter.Results of bothanalyses show C% to have a value of about -0.2 at the lower test Mach nder and to increase to about -0.5 at the highest test Mach number. Results of thesimple analysis show the damping factor C% + Cm& to be of the order of -0.3 with a positive value for the derivative at the highest test Mach number.