JAST. ol. 9, No., pp 4-8 Iranian Aerospace Society, Summer - Fall 1 J AST Journal of Aerospace Science and Technology The Utilization of High Fidelity Simulation in the Support of UA Launch Phase Design: Three Case Studies M. Mortazavi 1, A. Askari Improvement of the launch phase of a jet powered Unmanned Aerial ehicle (UA with Jet Assisted Take Off (JATO, has been the subject of attention in the - case studies of their application to UA launch phase problems are presented. X W,Y W, Z W U W, W,W W U,, W P, Q, R(or p, q, r Stall RE RE Stall U,, W P, Q, R F x, F y, F z l, m, n a b Aircraft location in inertial/world coordinates (m Aircraft velocity in inertial/world coordinates (m/sec Azimuth, Elevation, Roll in inertial/world coordinates (radians Linear velocity along X, Y and Z body axes (m/sec Angular velocity around X, Y and Z body axes (rad/sec Resultant velocity vector ( U W Stall speed (m/sec Reynolds number at Reynolds number at Stall Wind velocity across tail of aircraft Linear acceleration (m/sec Angular acceleration (rad/sec Forces acting on aircraft in body axes Moments about the X, Y and Z axes Angle of attack [tan -1 (W/U] Sideslip [tan -1 (/U] Air density aut.ac.ir ac.ir Nomenclauture µ e a r S L, D, S F b C l f Df X cg Z cg X pin Z pin W M UA M JATO M JATOCase X JATO nose Z JATO nose JATO MJT I xx, I yy, I zz Subscript W Air viscosity Surface area of wing (m Lift, drag and side force Wing span (m Chord lengt UA length UA fuselage section diameter Longitudinal distance of variable cg position from nose of the UA ertical distance of variable cg position from the lowest point of fuselage of the UA Longitudinal distance of pin position from nose of the UA ertical distance of pin position from the lowest point of fuselage of the UA Weight (N Mass of the UA Mass of JATO Mass of JATO without propellant Longitudinal distance of nose of JATO from nose of the UA ertical distance of nose of JATO from the lowest point of fuselage of the UA Angle between JATO and UA longitudinal axis in XZ-plane Angle between JATO and UA longitudinal axis in XY-plane Moment of inertia (kg-m Inertial/World coordinate system
46 M. Mortazavi, A. Askari C L C D C C C m C C Lq : C mq C L C m C C C lp C lr C C np C nr : C C C C Reference lift at zero angle of attack Reference drag at zero angle of attack Lift curve slope Drag curve slope Pitching moment at zero angle of attack Pitching moment due to angle of attack Lift due to pitch rate Pitch moment due to pitch rate Lift due to angle of attack rate Pitch moment due to angle of attack rate Side force due to sideslip Dihedral effect Roll damping Roll due to yaw rate Weather cocking stability Rudder adverse yaw Yaw damping Lift due to elevator Drag due to elevator Pitch due to elevator Roll due to aileron 1 Introduction Today, there is considerable interest in Unmanned Aerial ehicle (UA which is characterized by jet engine, zero length launcher and launching with JATO[1]. The best available synthesis and analysis techniques should, therefore, be used. The use of the simulation cost-effectiveness of exploring this high risk portion of Simulation analysis also proved valuable in developing phase. Recent developments in the practical application of simulation theory combined with the availability of computer-based synthesis and analysis tools offer the during launch and also in reduction of the associated The use of airplane simulation has an extensive history. Mathematical models used to describe these con- hand-adjusted by engineers and guided by pilots sub- controls, the need and importance of developing high Successful utilization of simulation in the launch phase, There have been many attempts to generate and improve modeling of the airplane s dynamic model, which ultimately improve the simulation s predictive capabili- Here, the implementation of launch phase simulation performed. The purpose of the test programs is to validate the system design and the simulation algorithm. UA launch phase is one of the most important critical phases of a UA and many failures occur in this phase. Some reasons are: 1. System instability in the launch phase,. The complexity of the aerodynamic behavior of phenomena such as Misalignment of JATO Thrust or Unsimultaneous Cut of Pins. Because of these reasons, it is not possible to use a simple simulation for the UA launch phase. Therefore, the objective of this paper is to develop the launch phase simulation of the There are some papers on the subject of UAs launch phase simulation. However, it should be noted that the launch phase simulation of a UA with this level of details such as Misalignment of JATO Thrust or time in the UA launch phase simulation in the present article. Therefore, the approach of launch phase simulation, in this article is unique and there is no similar work in the other studies. More importantly, it should be noted that three real case studies shows the simulation can accurately predict the system behavior in the launch phase. Actually, this simulation would be very turbances such as gust and turbulence. Hence, this article discusses the issues that are important in the launch phase of the UAs which have not been addressed pre-
The Utilization of High Fidelity Simulation in the Support of UA Launch Phase Design: Three Case Studies 47 viously. This is the second contribution of this paper. This article is also provided as an educational issue the subject of the UA launch phase simulation. First, the considerations that should be taken into account in the UA launch phase have been studied. Then, the test results are presented to achieve two goals: 1. verifying simulation and test results and also providing a stepwise logical analysis to explain these differences. Therefore, simulation in order to predict the safety factor of the UA launch phase. Flight Dynamic Modeling Coordinate System and Terminology In airplane simulations, coordinate systems fall into two broad classes, body coordinates and inertial or world and velocities are calculated in the body coordinate system and then converted to the inertial or world coordinate system prior to updating an airplane s position and attitude. Figure 1 shows the generally accepted convention for labeling the axes in the two coordinate systems. Body Because of its limited effect, the curvature of the earth fer to the geometric body axes. However, if a reference is made to the inertial or world coordinate system the subscript w is used (Figure 1. Figure. Mathematical Model Force equations: R WQ g sin FX m (1 U R WP UR g sin cos FY m ( WP W UQ P P Moment equations: g cos cos F I P I R I PQ( I I RQl (4 XX XZ XZ ZZ YY I Q I I PRI P R m Z m YY ( XX ZZ XZ ( I R I P ( I I PQI RQn (6 ZZ XZ YY XX XZ Kinematic equations: sin P (7 Q cos sin cos (8 R sin cos cos (9 Figure 1. body coordinate and inertial/world systems The mathematical model presented takes forces, con- generating linear and angular velocities in aircraft body coordinates as outputs (Figure. Navigation equations: X W cos sin Y W sin cos. h 1 W cos sin1 U 1 cos sin sin cos sin cos W
48 M. Mortazavi, A. Askari The forces and moments are: FX FX A Fengine FX JATO (11 FY FY A FY JATO (1 FZ FZ A FZ JATO l l A l JATO (14 m m A A m JATO engine m JATO nn n (16 Engine forces, torque and gyroscopic effect as well as environmental forces such as wind shear can have any- tions are made. Engine thrust is limited to the X-axis located above the aircraft center line that makes m Thrust and no calculations are made for the gyroscopic effect Aerodynamic Model The terms F AX, F AY and F AZ represent the resultant aerodynamic forces. (17 FXALsin Dcos SFsin (18 FYA SFcos F Lcos Dsin (19 ZA sideforce are calculated as follows: C C SF (C C c C q CL e rqs c L L L q L L Y C r Y qs ( (1 ( The aerodynamic moments represent the torque about the center of mass of the aircraft and are determined in the following equations: b Cl C l P P (4 la qsb b Cl R C R l acl r a r m n A A c Cmo C m Cmqq c C m C me qsc b Cn C n P P qsb b Cn R C R n ac a n r r UA Characteristics The UA characteristics are shown in table 1 [8]. Item Nomenclature Quantity 1 External Dimensions 1.1 Wing Span 1. Wing Aspect Ratio 4 Mean Aerodynamic Chord.8 m 1.4 Fuselage Length Fuselage Diameter.4 m 1.6 Wing Area 1.7 C.G. Position From Nose Weight.1 Gross Weight. Launch Weight Aerodynamic Wing Lift Curve Slope.68 /rad Horizontal L.C.S..6 /rad ertical L.C.S.6 /rad 4 Performance 4.1 ISA Condition 4. Stall Speed 7 m/s Engine Thrust Table 1- UA general characteristics (6
The Utilization of High Fidelity Simulation in the Support of UA Launch Phase Design: Three Case Studies 49 Aerodynamic Forces and Moments Correction The idea used in this research for the aerodynamic forces and moments in low a Reynolds number, is to consider a linear and quadratic relation between Reynolds and Aerodynamic forces and moments as below: l f Stalll f RE, RE Stall (7 Figure 4. JATO location on the UA for RE RE Stall : RE RE FXYZ,,, M XYZ,, FXYZ,,, M XYZ,, (8 Stall RE F X, Y, Z X, Y, Z X, Y, Z, X, Y, Z (9 RE Stall, M F M where F X,Y,Z and M X,Y,Z are the forces and moments in equations 11 to 16. JATO Model thrust is as follows: F T cos( X JATO JATO JATO F YJATO F T sin( Z JATO JATO JATO Misalignment of JATO Thrust Ideally, there is no misalignment in JATO thrust, but actually JATO thrust line is not on the JATO longitudinal axis. Asymmetric Forces and moments acting on the UA when JATO thrust is aligned to right side are: (MJT: Misalignment of JATO Thrust TLong. JATO TJATO cos( MJT T T sin( Lateral JATO JATO MJT FX MJT TLong. JATO cos( JATO F YMJT T LateralJATO FZ MJT TLong. JATO sin( JATO mmjt FX MJT ( Xcg XJATO nose F ( Z Z Z MJT cg JATO nose n F ( X X MJT Y MJT cg JATO nose l F ( Z Z (4 MJT Y MJT cg JATO nose Figure 3. JATO thrust model JATO Installation on the UA Figure 4 shows the schematic installation of the JATO JATO, causes an instability behavior and decreasing it, causes the UA not to reach a proper altitude. Obviously, the JATO thrust line must pass through the cg (UA+JATO. Unsimultaneous Cut of Pins - the UA separates from the launcher. It is necessary to check the UA behavior when pins do not cut simultaneously.
M. Mortazavi, A. Askari Center of Gravity and Moment of Inertia separated from UA. The variable cg position is: t 1 <Time<t : r cg ar M UA r cg UA M UA M M JATO cg JATO JATO r Time=t : Figure. UA connection to launcher (F z is perpendicular and toward the page Where the left pin cut, forces and moments acting on the UA are: (UCP: Unsimultaneous Cut of Pins FX UCP FX MJT Fengine (41 F YUCP FYMJT (4 r cgar M r M r M M Time>t : r cg r ar UA cgua JATOCase cgjato Case cg UA UA JATOCase Change of weight and moment of inertia of the system (UA+JATO during JATO burning are considered in the simulation. F ZUCP F ZMJT m F ( Z Z UCP X UCP cg pin F ( X X ZUCP cg pin (44 Because of changing the cg location, the correction on n F ( d F ( X X UCP X UCP f Y UCP cg pin l F ( d F ( Z Z (46 UCP Z UCP f Y UCP cg pin C C m ( UA JATO m UA C [( X X X ] L UA cg ( UA JATO cg UA ac Total JATO forces and moments acting on the UA are FXJATO FXMJT FXUCP (47 FYJATO FYMJT FYUCP (48 FZJATO FZMJT FZUCP (49 mjato mmjt mucp njato nmjt nucp ljato lmjt lucp 1 and cg position move back, causing a decrease in the stability margin that must be considered in the analysis. The simulation model should predict the UA attitude, position and velocity in launch phase before the test. The simulation results are compared with three launch/ ity analysis using simulation can be helpful to increase the insight of parameter effectiveness in the launch phase.
The Utilization of High Fidelity Simulation in the Support of UA Launch Phase Design: Three Case Studies Results and Discussions Case Studies Three case studies of launch simulation of the UA are presented. All of the case studies involve two control e a that may be changed by operator and/ test launch is without engine and the second and third tests are launched with engine. Measurement devices are vertical gyro, GPS receiver, barometric altimeter and telecommunication apparatus. All these devices were calibrated and have been used before on the other UAs. Measurement parameters are pitch angle, bank angle, altitude, pressure altitude, temperature, latitude, longitude, and ground speed. Note This fact causes some data loss and decreases the precision of the test data in comparison with the simulation result. Because of this, the simulation and test results can be compared in large scale. Case Study I: Launch Test without Engine The UA weight is 4 kg and JATO produces 4 KN in average during t -t 1 =. seconds. The simulations and - JATO =1.67 = e,t 1 =. To show the correction factor which is considered in equations 7 and 9, four scenarios are considered as below: Scenario 1: simulation of the launch phase considering the quadratic correction factor ( RE RE Stall in equation 9. Scenario : simulation of the launch phase considering the linear correction factor ( RE RE Stall in equation 8. sidering correction factor ( RE RE Stall in equation 8. Scenario 4: simulation of the launch phase considering zero forces and moments for RE RE Stall. There is a good agreement in altitude between scenario test. Therefore, it can be concluded that considering the quadratic form of the correction factor for forces and moments (equation 9 is more accurate than the linear model (equation 8. Figure 7 shows simulation cause pitch angle and altitude are not in their reasonable range. Although scenario 4 looks acceptable, there is an - (deg (deg (deg (deg Scenario 1. Scenario Flight Test 4 6 8 1 1 7 4 6 8 1 1 1 4 6 8 1 3 1-1 4 4 6 8 1 1 7 (deg a X a Z 4 1 4 6 8 1-1 4 6 8 1 4 1 4 6 8 1-1 4 6 8 1 Figure 6. Simulation of the UA Scenario 1, Scenario 3 Scenario 4 Flight Test 4 6 8 1 4 6 8 1 1 1 4 6 8 1 7 6 4-4 6 8 1 (deg a X a Z 7 4 6 8 1 1 4 6 8 1 4-4 6 8 1 1-1 - 4 6 8 1 Figure 7. Simulation of the UA Scenario 3,4
M. Mortazavi, A. Askari error in velocity that shows a good agreement. Differences in speed after t=7 seconds can be because of some er than is the one considered in the simulation. Another effect is because of lateral oscillation that causes loss of UA energy and reduces speed to lower than the point predicted in the simulation. simulation and test. It seems a pilot command after t=4 seconds in E Also, during JATO burning, the bank angle increases to. Using simulation, it is concluded that this occurs because of both JATO misalignment in xy-plane and non-simultaneous cut of pins ( MJT =.1, t UCP =.1 sec. Although speculating what the pilot may have thought 9 illustrates simulation of the test using an unstable au- gain of bank attitude hold mode is deliberately selected as K Comm. =. Figure 9 does not resemble what happened in the real test. The bank angle gradually increases during 1 seconds and the maximum ;two points which are different from the test. Another assumption is the pilot s reaction to hold the bank angle near zero. Figure 1 illustrates shows a similarity between the simulation and the test. Therefore, it is concluded that the pilot tried to stabilize the UA in longitudinal and lateral-directional modes. As a result, because the UA has a high acceleration during launch, pilot commands cause an instability system. Figure 11 shows the UA can be effectively controled with a stable gain autopilot which is shown in (deg a X 4 1 1 Simulation Flight Test 4 6 8 1 4 6 8 1-1 4 6 8 1 (deg (deg p (deg/sec 1 4 4 6 8 1 4 1 (deg (deg a Z 1 7 4 6 8 1 -. 4 6 8 1 4 6 8 1 1-1 4 6 8 1 Simulation Flight Test 4 6 8 1-1 4 6 8 1-1 -1 4 6 8 1-4 6 8 1 (deg/sec r (deg/sec 1 7 4 6 8 1 1 1 4 6 8 1 6 4-4 6 8 1 1 - -1 4 6 8 1 Figure 9. Simulation of the UA Using autopilot command with an unstable gain (K =.
The Utilization of High Fidelity Simulation in the Support of UA Launch Phase Design: Three Case Studies (deg Simulation Flght Test 4 3 1 4 6 8 1 1 7 4 6 8 1 [phi _Comm ] K_phi A demand A actual f(u From Gain Aileron Actuator UA Gyro f(u Figure 1. Roll Attitude Hold/Select Mode autopilot (deg p (deg/sec (deg (deg p (deg/sec 1-1 4 6 8 1 - - 4 1 4 6 8 1 4 6 8 1 (deg/sec r (deg/sec 1 1 4 6 8 1 1 1-4 6 8 1 1 Figure 1. Simulation of the UA Using operator command 4 6 8 1-1 4 6 8 1-1 -1 4 6 8 1 4 6 8 1 (deg/sec r (deg/sec - -1 4 6 8 1 1 7 Simulation Flight Test 4 6 8 1 1 1 4 6 8 1 6 4-4 6 8 1-4 6 8 1 Case Study II: Launch Test with Engine In this test, the prototype has an engine that produces a is the empty weight. The JATO thrust is 117 N during t -t 1 JATO =, e =1,t 1 =. test of the UA. Lateral-directional parameters are near zero and ignorable. The difference between t in simulation and test is due to the difference in the burning time of the JATO ideal and actual models used for simulation and test, respectively. The difference between t in the simulation and the test is due to difference in the burning time of JATO considered in ideal and actual models for simulation and test, respectively. Like case study I, due to pilot command during t= and t=4 seconds, cause pitch angle is lower than that in the simulation. Therefore, velocity in the simulation is lower than the which is acceptable. (deg 1 1 3 4 6 7 8 9 1 1 1 3 4 6 7 8 9 1 1 1 3 4 6 7 8 9 1 Figure 13. Simulation of the UA Simulation Flight Test Figure 11. Simulation of the UA Using autopilot command with a stable gain (K =.
M. Mortazavi, A. Askari Case Study III: Launch Test in The Presence of Headwind The UA weight is 4 kg and JATO averagely produces 4 KN during t -t 1 =. sec. In the test, there is a the UA in the presence of headwind. The simulation - JATO =11.6 = e =1, t 1 =. UA. Because of headwind, the pitch angle increases about 1 good agreement of the pitch angle in the simulation and the pitch angle equal to Command. (deg 4 3 Simulation Flight Test 1 1 3 4 6 7 8 9 1 1 1 3 4 6 7 8 9 1 3 1 1 3 4 6 7 8 9 1 of headwind=3 km/h As discussed in case study II, because the impulses in JATO ideal and actual models are equal and burning come equal after 1 seconds. In this test, by removing some problems existing in the last two tests like pilot command or an accurate prototype, there are no differences between simulation and test after 1 seconds. Sensitivity Analysis tion and, therefore, it is possible to use this simulation as a design tool and in sensitivity analysis in order to measure the sensitivity of the UA variables to parameter variation in the launch phase. Actually, it helps de- JATO,, JATO thrust and duration (, t JATO, maximum t UCP and maximum MJT cost. Table includes different JATOs with similar impulses, different launch angles (, JATO angles (, JATO misalignment and non-simultaneous cuts of pins. The initial data is: =. sec, =, =1, e [7]. Case 1 KN t JATO sec 1,., Imp. KN.s. 4 deg JATO deg JATO deg t Pins sec 1 1,, 1 4. 1,, 4. 1.1,.1,. 1 Table. Case studies for the launch - e =.1.1,.1, The case studies of launch phase simulation of differ- pulse used for launch. Figure 16 shows using JATO with causes decrease in the velocity that may be lower than the stall speed. Conversely, decreasing it causes a decrease in the altitude, JATO onds and the stability of the UA. On the contrary, low JATO decreases altitude, which is not enough for recov- and illustrate the maximum time difference between Pins Pins longitudinal and lateral-directional modes. Similarly, - Unsymmetric, otherwise there would be instability in the lateral-directional mode..1
The Utilization of High Fidelity Simulation in the Support of UA Launch Phase Design: Three Case Studies (JATO Thrust (N 6 x 14 4 3 1 = KN,t JATO =1 sec = KN,t JATO =. sec =13.7 KN,t JATO =4 sec. 1 1.. 3 3. 4 4. (deg (deg 1 1-4 6 8 1 1 1 4 6 8 1 4 3 1 4 6 8 1 4 3 1 4 6 8 1 Figure 1. Case 1 = KN,t JATO =1 sec = KN,t JATO =. sec =13.7 KN,t JATO =4 sec (deg a x a z 4 3 1 4 6 8 1 4-4 6 8 1 1 6 4 - Figure 16. Case 1 4 6 8 1 4 6 8 1 (deg (deg (deg (deg 6 4-4 6 8 1 1 4 6 8 1 4 6 8 1 6 4 6 8 1 1 - (deg a X a Z 6 4 6 8 1 1-1 4 6 8 1 4 Figure 17. Case -1 4 6 8 1 11 1 7 4 6 8 1 4 3 1 4 6 8 1 4 3 1 4 6 8 1 Launch =1 Launch = Launch =4 4 6 8 1-4 6 8 1 (deg a X a Z 4 3 1 4 6 8 1 4-4 6 8 1 4-1 4 6 8 1 - JATO = JATO =1 JATO = 4 6 8 1 Figure 18. Case 3
M. Mortazavi, A. Askari (deg (deg 1 - t=.1 sec t=.1 sec t=.1 sec -1 4 6 8 1 1 4 6 8 1 4 6 8 1 1-4 6 8 1 (deg Q (deg/sec a x a z 6 4 - - 4 6 8 1-4 6 8 1 6 4-1 -1 Figure 19. Case 4 4 6 8 1-4 6 8 1 (deg (deg 6 4 4 6 8 1 1 JATO =.1 deg JATO =.1 deg JATO =.1 deg 4 6 8 1 4 4 6 8 1 1-4 6 8 1 (deg Q (deg/sec a x a z 4 3 1 4 6 8 1-1 4 6 8 1 4 Figure 1. Case 4 6 8 1-4 6 8 1 (deg (deg - - -4 t=.1 sec t=.1 sec t=.1 sec 4 6 8 1 (deg (m/ses - -4 4 6 8 1 4 (deg (deg 1-1 4 6 8 1-1 - JATO =.1 deg JATO =.1 deg JATO =.1 deg (m/ses (deg - 4 6 8 1-1 - P (deg/sec Y (m -6 4 6 8 1 - -4 1 4 6 8 1-1 4 6 8 1 a y R (deg/sec - 4 6 8 1 4 3 1 Figure. Case 4 4 6 8 1-1 4 6 8 1 P (deg/sec Y (m -3 4 6 8 1 - -1 4 6 8 1 - -1-1 4 6 8 1 R (deg/sec a y -3 4 6 8 1 1-1 - -3 4 6 8 1 - Figure. Case 4 6 8 1
The Utilization of High Fidelity Simulation in the Support of UA Launch Phase Design: Three Case Studies Conclusions In this paper, the simulation of UA in the launch phase oped. Accurate formulation of symmetric and asymmetric forces and moments acting on the UA offered additional insight and understanding, which could not have been gained without simulation. Three case studies were modest examples of this simulation. Comparison of simulation and test in asymmetric condition, in conjunction with headwind and the UA with and without engine showed an acceptable agreement between would behave in these conditions and how to decide whether it is safe to launch the UA. The key in solving this problem was to recognize the contribution of aerodynamic forces and moments, JATO forces and moments in symmetric and asymmetric conditions, cg traveling during launch and moment Sensitivity analysis was shown as a useful tool to design launch parameters. It increased the insight of the launch performance as a result of effective parameters in the launch phase. For developing the research, some measures can be done such as extending and imple- timization algorithm to design the launch parameters which are quantitatively more precise. Acknowledge The authors hereby express their gratitude for the grant support from the HESA design bureau. References 1. Johnson, Eric N., Fontaine, Sebastiene, Use of Flight Simulation to Complement Flight Testing. Brian L. Stevens & Frank L. Lewis, Aircraft Control and Simulation, John Willey & Sons INC, (199. Roskam, Jan, Airplane Flight Dynamics and Automatic Flight Controls, Jan Roskam, Kansas, (1997. 4. Louise. Schmidt, Introduction to Aircraft Flight Dynamis, AIAA, 1998. Cooke, Joseph M., Zyda, Michael J., Pratt, David R., McGhee, Robert B., NPSNET: Flight Simulation Dynamic Modeling Using Quaternion, In Presence, ol. 1, No. 4, pp. 44-4, 1994. 6. Ralston, John, Kay, Jacob, The Utilization of High Fidelity Simulation in the Support of High Angle of Attack Flight Testing, Hampton, A 7. Mortazavi, Mahdi, Askari, Abdorreza, Simulation of RP s launch phase, The forth biennial conference of the Iranian aerospace society Conference, Amirkabir University of Technology, 8. aziri M.A., Mortazavi M., Seyyed-agha AR., ration for a Civil Medium Altitude Jet Powered Unmanned Air ehicle, The Fourth Iranian Aerospace Society Conference, Amirkabir Uni-