1、I NASA Technical Paper 1350 the F-111 Tact Research Aircraft Alex G. Sim and Robert E. Curry OCTOBER 1978 NASA Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-TECH LIBRARY KAFB, NM NASA Technical Paper 1350 Flight-Determined Stability and Control Der
2、ivatives for the F-111 Tact Research Aircraft Alex G. Sirn and Robert E. Curry Dryden Flight. Reseurch Center Edwurds, Culiforniu National Aeronautics and Space Administration Scientific and Technical Information Office 1978 I Provided by IHSNot for ResaleNo reproduction or networking permitted with
3、out license from IHS-,-,-FLIGHT-DETERMINED STABILITY AND CONTROL DERIVATIVES FOR THE F-111 TACT RESEARCH AIRCRAFT Alex G . Sim and Robert E. Curry Dryden Flight Research Center INTRODUCTION The F-111 transonic aircraft technology (TACT) aircraft is the latest of a series of research aircraft to inco
4、rporate supercritical wing technology. Unlike previous supercritical wing designs, the TACT wing was designed to provide improvements in transonic maneuver capability without degrading the F-111A aircrafts cruise and supersonic performance. A research program was conducted jointly by the National Ae
5、ronautics and Space Administration (NASA) and the U . S . Air Force. During the envelope-expansion phase of the flight program, the flight-determined derivatives were used to update the analysis of the vehicle dynamics to insure safety of flight. One goal of the TACT program was to provide stability
6、 and control derivatives to establish a data base with which experimental and analytical prediction techniques could be improved for this class of aircraft. To lend credence to the flight-determined derivatives and to indicate the deviation from potential theory, some of the major derivatives were c
7、alculated based on computer models of the aircrafts geometry. These are referred to in this report as analytical model derivatives. This report presents the flight derivative data base for the F-111 TACT research aircraft. The flight derivatives are correlated with the analytical model derivatives f
8、or specific flight conditions and aircraft configurations. SYMBOLS The stability and control derivatives, as presented, are partial derivatives representing standard NASA coefficients of forces and moments. A right-hand sign convention is used to determine the direction of forces, moments, angular d
9、isplacements, and velocities. Except for angle of attack, the data are referenced to the vehicle body axis. Angle of attack is referenced to the wing reference plane for consistency with wind tunnel data and other TACT flight data, and thus, it is lo higher than it would be if it were referenced to
10、the vehicle body axis. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Physical quantities are given in the International System of Units and parenthet- ically in U . S , Customary Units. CLB CLDA rolling moment coefficient with respect to angle of s
11、ideslip, per degree rolling moment coefficient with respect to aileron deflection, per degree CLDR rolling moment coefficient with respect to rudder deflection, per degree CLDS rolling moment coefficient with respect to spoiler deflection, per degree CLP rolling moment coefficient with respect to ro
12、lling rate, per radian CLR rolling moment coefficient with respect to yawing rate, per radian CMA CMDE CNA CNB CNDA pitching moment coefficient with respect to angle of attack , per degree pitching moment coefficient with respect to elevator deflection, per degree pitching moment coefficient with re
13、spect to pitching rate, per radian untrimmed normal-force coefficient with respect to angle of attack, per degree yawing moment coefficient with respect to angle of sideslip , per degree yawing moment coefficient with respect to aileron deflection, per degree CNDE untrimmed normal-force coefficient
14、with respect to elevator deflection , per degree CNDR yawing moment coefficient with respect to rudder deflection, per degree CNDS yawing moment coefficient with respect to spoiler deflection, per degree CNP yawing moment coefficient with respect to rolling rate, per radian CNR yawing moment coeffic
15、ient with respect to yawing rate, per radian CYB side-force coefficient with respect to angle of sideslip, per degree 2 I II Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-C YDA CYDR CYDS C n a M a h side-force coefficient with respect to aileron de
16、flection, per degree side-force coefficient with respect to rudder deflection, per degree side-force coefficient with respect to spoiler deflection, per degree section normal-force slope, per degree Mach number angle of attack with respect to wing reference plane, degrees angle of wing leading-edge
17、sweep, degrees AIRCRAFT DESCRIPTION The TACT modifications to the F-111A baseline vehicle included a new wing planform with a supercritical airfoil and a new high-lift system, a modified wing seal, a modified overwing fairing, and a fixed-structure glove. The general arrangement of the F-111 TACT ai
18、rcraft is shown in figure 1, and the aircrafts physical characteristics are given in table 1. Additional description of the TACT aircraft, as well as a comparison with the F-111A baseline vehicle, is given in references 1 and 2. The TACT aircrafts control surfaces were controlled by an irreversible
19、hydraulic system. The pilot controlled the aircraft through a conventional center stick and rudder pedals. For pitch control, the horizontal stabilizer was deflected symmetrically by either the pilot or the rate command augmentation system. A similar arrangement with the rudder was used for yaw cont
20、rol. However, for roll control, the pilots inputs activated both the differential horizontal stabilizer and the spoilers, while the rate command augmentation system activated only the differential stabilizer. INSTRUMENTATION Data were obtained at 20 samples per second through a 10-bit pulse code mod
21、ulation system. All the data were calibrated and analyzed after the flight using a ground-based computer. Angle of attack and angle of sideslip were measured using a vane flow angu- larity sensor system. This system and its calibration are described in reference 3. 3 Provided by IHSNot for ResaleNo
22、reproduction or networking permitted without license from IHS-,-,-I, . , In addition, angle of attack was estimated along with the flight derivatives for correlation with the measured angle of attack. Angular positions were measured with a stable platform, angular rates were measured using rate gyro
23、s , and linear accelerations were determined from linear accelerometers. Control positions were measured using control position transducers. Corrections were applied to the airspeed data to obtain true velocity, Mach number , and dynamic pressure. Linear accelerations , angle of attack, and angle of
24、 sideslip were corrected for displacement from the center of gravity. FLIGHT CONDUCT Before each flight, a detailed flight plan (checklist) specifying particular maneuvers was prepared. The flight was then monitored by chase aircraft pilots and control room personnel. This procedure not only insured
25、 safety of flight but also allowed the research engineer to monitor the flight data in real time, thus giving him insight into the adequacy of the maneuvers and the instrumentation data quality. The flight plan flexibility was sufficient to allow a maneuver to be repeated if necessary. Longitudinal
26、and lateral-directional maneuvers from which aircraft data were obtained were performed throughout the flight envelope. Data were obtained for angles of attack from approximately 3O to 14O for a Mach number range from approx- imately 0.25 to 1.70. The data for the higher angles of attack were obtain
27、ed at elevated load factors. The longitudinal maneuver consisted of a horizontal stabilizer doublet , followed by 2 to 3 seconds of no pilot input , followed by a second horizontal stabilizer doublet. Because the vehicle normally exhibited an over- damped longitudinal response , the second doublet w
28、as necessary to increase the amount of statistically significant transient response information. One of the objectives in the lateral-directional mode analysis was to obtain independent derivatives with respect to spoiler and aileron (rolling tail) . However, it was difficult to separate the aileron
29、 and spoiler derivatives with the roll augmentation off because the two control surfaces operated nearly in phase. Because the roll augmentation acted only through the aileron, out-of-phase aileron and spoiler motion could be obtained with the roll augmentation on. With both roll and yaw augmentatio
30、n on, the resulting airplane motion was heavily damped. The maneuver finally selected (fig. 2) was performed with the roll augmentation on and the yaw augmentation off. It consisted of two pilot-initiated roll doublets followed by a rudder doublet. This maneuver proved to be adequate for the derivat
31、ive extraction process. One flight was made without the use of the spoilers during maneuvers to evaluate the possibility of obtaining more consistent sets of derivatives with lower uncertainties. Better results were obtained from these maneuvers. 4 Provided by IHSNot for ResaleNo reproduction or net
32、working permitted without license from IHS-,-,-Some of the maneuvers analyzed were not performed to obtain derivatives. Examples of these include structural excitation (using stick raps) and handling quality evaluations. Although specific derivatives could be obtained from these maneuvers, a consist
33、ent, complete set of high-quality derivatives could not. DERIVATIVE ANALYSIS Flight Data A digital computer program employing a maximum likelihood estimator method was used to determine sets of derivatives and uncertainty levels for the longitudinal and lateral-directional modes from flight data. Th
34、is computer program and its theory, mathematical model, and practical application are documented in references 4 to 6. The derivative analysis was usually performed within a week of the flight using a “best estimate? set of moments of inertia. After the completion of the derivative-extraction flight
35、s, the moments of inertia were experimentally determined using the Air Force Flight Test Centers Moment of Inertia Facility. The derivatives for presentation in this report were adjusted to reflect the experimentally determined moments of inertia. In addition, the derivatives were adjusted to a refe
36、rence longitudinal center of gravity. The reference center of gravity varied as a function of wing sweep, and the variation is documented in table 2. A variable reference was used to provide a derivative data set consistent with the performance flight data. These reference values represent an averag
37、e flight center of gravity for each wing sweep. Geometric Models Analytical model derivatives were obtained from two large computer programs based on a grid determined by dividing the aircraft geometry into constant pressure panels. Integration of these pressures yielded the forces and moments. Both
38、 programs assume inviscid, incompressible (with Prandtl-Glauert corrections), attached flow. They differ in the way they represent the aircraft geometry and in the solution for the constant pressure panels. The first program, referred to as the WING-BODY program, models both lifting surfaces (wings)
39、 and a body. It is a potential-doublet panel program in which the wing-body combination is represented by a large number of distributed singularities which are used to satisfy the linearized potential equations. It can be used for both supersonic and subsonic analysis. Further information on the the
40、ory and use of the program is given in references 7 and 8, respectively. For the second program, referred to as the VORTEX-LATTICE program, the lifting surface planforms are represented with a lattice of horseshoe vortexes. By solving the flow boundary condition of each horseshoe vortex, the element
41、al lift of each panel can be determined. The program is used for subsonic analysis. It can be used to compute the potential as well as the vortex leading- and side- edge components of forces and moments; however, only the potential solution was used for these studies. Further information on the theo
42、ry and use of similar programs is given in references 9 to 11. 5 I Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-In using the above programs , an attempt was made to model the geometry in a basic , conceptual manner. This was done to best represent
43、 the manner in which these programs are mechanized to minimize computer time. A WING-BODY panel model that was used to obtain rolling tail control effectiveness is shown in figure 3. Only the right half of this model was used for the typical case where vehicle symmetry is assumed. The panel coordina
44、tes are the basic inputs to the program. A typical VORTEX-LATTICE planform model is shown in figure 4. The panels are not shown since paneling is an internal process of the program. Only the planform view is shown since, for this case the wings are coplanar. It was not necessary to model wing camber
45、 in either program. DATA PRESENTATION The majority of the flight data were obtained overthe airplanes Mach number and angle of attack ranges, predominantly at wing sweep angles of 26O, 35O, and 58O with the airplane in a clean configuration (that is, landing gear and flaps retracted) Additional flig
46、ht data were obtained at low speeds (Mach numbers from 0.25 to 0.45) with various combinations of landing gear and flap positions for wing sweep angles of 16O, 20, and 26O. All the data are presented as a function of angle of attack for specific Mach number ranges. Where possible, a recommended fair
47、ing is given to aid in the interpretation of the flight derivatives. In addition certain derivatives that are strong functions of Mach number as well as angle of attack are presented as functions of Mach number for a given angle of attack. Uncertainty levels are shown for all flight data. These unce
48、rtainty levels are proportional to the Cram the actual angle of attack, being a state variable, will typically vary ?2O from trim. The resulting linearized derivative (in this case, CNA) tends to acquire an average slope over the angle of attack range for the maneuver. This, in effect, filters any s
49、harp breaks in the coefficient. Thus, even though the first indications of a break in the data of figure 5(a) occur at an angle of attack of approximately 7O, the actual break occurs at an angle of attack between 8O and go, as shown in figure 8, which was taken from reference 12. Note that in this reference, the normal-force slope break is correlated with buffet intensity rise, whic
