UAV Sky-Y flight loads: a Multi-Disciplinary approach

Transcription

1 UAV Sky-Y flight loads: a Multi-Disciplinary approach Daniele Catelani (MSC) G.M. Carossa, E. Baldassin, R. Digo, E. Marinone, O. Valtingojer (Alenia Aeronautica) 2010 Alenia Aeronautica S.p.A. The contents of this document are the intellectual property of Alenia Aeronautica S.p.A.. Apart from those contractually-agreed user rights, any copying or communication of this document in any form is forbidden without the written authorisation of Alenia Aeronautica S.p.A..

2 2 Target of the study - Integration of various disciplines (flight mechanics, aerodynamics, structures, loads) simulation models - Benchmark on flight simulation using Multi-Disciplinary approach - Air Vehicle Methodology improvement: the goal is to achieve the result through an unique simulation

3 3 Unmanned Air Vehicle (UAV) From surveillance to reconnaissance, from patrol to the most risky defence operations, the new frontier of aeronautics lies in the UAV (Unmanned Air Vehicle) domain development. Alenia Aeronautica, thanks to its effort in dealing with the various aspects of this new environment, is today the European leader in the UAV technology. Alenia Aeronautica major achievements in UAV domain: Sky-X (1 flight May 05), the first in its c lass to have flown in Europe Sky-Y (1 flight June 07), operating demonst rator to gain experience both in civil and military domains, aimed to Medium Altitude Long Endurance (MALE) missions.

4 4 Sky-Y Conceived as technology demonstrator for a MALE surveillance UAV. Its purpose is to develop enabling technologies that increase the aircraft s autonomous flight and data collection and distribution capabilities. Sky-Y has an all-composite structure, all-electric systems and a diesel engine of automotive derivation giving it up to 12 hours endurance. Dimensions Dimensions Length Length Span Span Weights Weights MTOW MTOW OEW OEW Fuel Fuel Payload Payload m m kg 1200 kg 850 kg 850 kg 200 kg 200 kg kg kg Performances Performances LOS LOS Radius Radius nm nm Range Range 500 nm 500 nm Altitude Altitude ft ft Endurance Endurance 12 h 12 h Engine Engine 1 1 Diesel Diesel 160 HP 160 HP

5 5 Sky-Y - SMAT The SMAT (Sistema di Monitoraggio Avanzato del Territorio Advanced Monitoring System of Territory), lead by Alenia Aeronautica and Selex Galileo, is a research project mainly based on Sky-Y UAV. It is financed by Regione Piemonte and Distretto Aerospaziale Piemontese. Sky-Y Aim of SMAT is to study and demonstrate the feasibility of an automated surveillance system for environmental and civil security, traffic surveillance, pollution control, etc It will integrate three different UAV platforms with the existing ground infrastructures. Ground Control Station Ground Network Sensors Coordination center

6 6 Multi-Disciplinary (MD) simulation Dynamic model of UAV flight: coupling aerodynamic data set, structural models, FCS (Flight Control System) flexible structural model (including control surfaces) control surfaces actuator models aerodynamic and inertial loads due to flight manoeuvres Aerodynamic data set MB Data (Control surfaces) FEM Model with Inertial data Pilot input ADAMS FCS model Flight Loads Time History

7 7 MD process Requirements Pilot Input FCS model Aero data set Matching aerod/fem Action 1 FEM + pressure distribution Create modal data Action 2 Modal data Create MB model Action 3 MB controlled model Manoeuvre analysis Action 4 CAD & FEM & MB & CFD Pre processing Action 0 Extracting specific responses Action 5 Generic manoeuvre response Data Flight Tecnologies Domain Manoeuvre response data FEM + inertia data Structure Technologies Domain Loads on structure MB data: control surfaces System Technologies Domain Actuators loads and position

8 8 Aerodynamic data set The aerodynamic analysis is accomplished by using various analytical techniques such as Vortex Lattice or Euler formulation methods (CFD - Computational Fluid Dynamics). These analytical techniques allow to generate an aerodynamic model starting from the external shape of the air vehicle. Panels Macropanel Macropanel For a set of flight parameters (α, β, δ, Mach), relevant aero-coefficients on predefined grids of the aerodynamic mesh are calculated. These coefficients are assembled into the aerodynamic data set.

9 9 FEM model - Original FEM model: full A/C NASTRAN model used for aeroelastic analyis - Modification of FEM model: modifications have been applied to take into account static and dynamic analysis differences and MultiBody needs: - separation of FlexBodies from assembled FEM model - master nodes definition - number of modes definition - ADAMS MNF cards Aileron Airframe Rudder Elevator

10 10 Structural/Aerodynamic coupling CFD aero coefficent data set are transferred on the structural mesh (not coincident with the aerodynamic one) using RBE3 element Aerodynamic model mesh Structural model (FEM) Models coupling

11 11 Structural/Aerodynamic coupling CFD aero coefficent data set are transferred on the structural mesh (not coincident with the aerodynamic one) using RBE3 element Aerodynamic model mesh Structural model (FEM) Models coupling

12 12 Flight Control System (FCS) model A Flight Control System has been developed, using General State Equation (GSE) element (which allows discrete integration time step) Input data time history of pilot commands (aileron, elevator, rudder, throttle) are transferred through FCS model (Alenia Fortran routine black box) and GSE to control surfaces Output are controlled surfaces rotations

13 13 Assembly of MD model 1 Aerodynamic Forces in Adams CFD / Aerodynamic data From aero mesh to structural mesh FORCE FORCE* * FORCE* * FORCE* * FORCE* * FORCE* * FORCE FORCE* * FORCE* * From structural mesh to Adams input: Fortran routine Nodal forces VFORCE Grid Force Loads Modal forces MFORCE Case: L_

14 14 Assembly of MD model 1 Aerodynamic Forces in Adams CFD / Aerodynamic data From aero mesh to structural mesh FORCE FORCE* * FORCE* * FORCE* * FORCE* * FORCE* * FORCE FORCE* * FORCE* * From structural mesh to Adams input: Fortran routine Nodal forces VFORCE Grid Force Loads Modal forces MFORCE Case: L_

15 15 Assembly of MD model 2 Structural elements in Adams Separated FE bodies Master nodes MNF files Aeroelastic model MNF statements Flex Bodies FULL MultiBody model Connections

16 16 Assembly of MD model 2 Structural elements in Adams Separated FE bodies Master nodes MNF files Aeroelastic model MNF statements Flex Bodies FULL MultiBody model Connections

17 17 Assembly of MD model 3 FCS in Adams GSE subroutine GSE array state variables Pilot Input read pilot input link FCS to pilot input at sampled time get FCS output Aileron rotation Rudder rotation Elevator rotation Throttle link output to actuators mobile surfaces rotation thrust

18 18 Assembly of MD model 4 33 rigid bodies 7 Forces flex bodies 2 vector torques Simulation 1 single command component force 34 requests Solver parameters Full Adams model Connections aerodynamic 47 joints fixed thrust rev 15 drag bushings sph 1550 damping aerodynamic moment forces (vector forces) bushing or 27 modal forces disp Data splines matrices array vars Measures vel acc loads Graphics

19 19 Manoeuvres description The manoeuvre simulation is splitted in two phases: 1. Trim analysis defining the first instanct of the dynamic response. At the required speed, altitude, load factor (Nz), the equilibrium aircraft condition is found with angular acceleration = 0. Three kind of trimming have been considered: pull out (Nz 1 ) steady turn (Nz > 1) inverted flight (Nz = -1) 2. Dynamic analysis starting from trimmed position imposing pilot commands

20 20 MD simulation Trim Analysis Implemented Strategy for trimming the aircraft: pin the aircraft to the aggregate mass node (C++ feature) through revolute and bushing allowing pitch rotation define the required A/C condition in term of speed, altitude, load factor and trim type (GUI) apply a sequence of static analysis reducing bushing Kt to 0 apply methods to obtain robust trim solution: run rigid model first, then flexible one increment number of iterations and KT values find intermediate (not trimmed but closer) configuration and then restart GUI reduce error tolerance

21 21 MD simulation Trim Analysis Implemented Strategy for trimming the aircraft: pin the aircraft to the aggregate mass node (C++ feature) through revolute and bushing allowing pitch rotation define the required A/C condition in term of speed, altitude, load factor and trim type (GUI) apply a sequence of static analysis reducing bushing Kt to 0 apply methods to obtain robust trim solution: run rigid model first, then flexible one increment number of iterations and KT values find intermediate (not trimmed but closer) configuration and then restart GUI reduce error tolerance

22 22 MD simulation Trim Analysis results Results in tabular form and in GUI

23 23 MD simulation Dynamic Analysis Implemented strategy for Dynamic Analysis: start from trimmed analysis results or from position imposed by user (GUI) define pilot inputs in terms of longitudinal, lateral, directional and handle commands in the bulk data file define time for balancing and simulation run rigid or flexible model GUI

24 24 MD simulation Dynamic Analysis Implemented strategy for Dynamic Analysis: start from trimmed analysis results or from position imposed by user (GUI) define pilot inputs in terms of longitudinal, lateral, directional and handle commands in the bulk data file define time for balancing and simulation run rigid or flexible model GUI

25 25 MD simulation - results 1 PULL OUT

26 26 MD simulation - results 2 Alpha Drag PUSH DOWN Elevator

27 27 MD simulation - results 3 STEADY TURN

28 28 From Adams to FE: Nodal Loads (GPFORCES) Adams/Durability module exports Adams analysis as a modal deformation file (mdf) file Nastran Restart analysis is performed including GPFORCE/ALL The Nastran output (f06 file) contains GPFORCES informations on each structural node for each time step Patran is used for graphical visualisation A reading/writing Fortran (C++) routine could be developed for extracting hotspot informations (max forces, monitoring stations loads, etc.)

29 29 Nodal Loads: GPFORCES Example of output TIME = E+00 G R I D P O I N T F O R C E B A L A N C E POINT-ID ELEMENT-ID SOURCE T1 T2 T3 R1 R2 R BEAM E E E E E E BEAM E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E ROD E E *TOTALS* E E E E E E BEAM E E E E E E BEAM E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E *TOTALS* E E E E E E-11

30 30 Nodal Loads: GPFORCES Example of output TIME = E+00 G R I D P O I N T F O R C E B A L A N C E POINT-ID ELEMENT-ID SOURCE T1 T2 T3 R1 R2 R BEAM E E E E E E BEAM E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E ROD E E *TOTALS* E E E E E E BEAM E E E E E E BEAM E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E QUAD E E E E E E *TOTALS* E E E E E E-11

31 31 Conclusions Performed test simulating various types of manoeuvres evidence: ADVANTAGES: - the results are in agreement with the outputs of currently used method - integration of simulation models of different disciplines - easy I/O management by use of GUI DISADVANTAGES: - aerodynamics/structural modal simulation quite complex - requirements of high computing power (possibly HPC)

32 32 Way forward - extension of the module to generic airframe: generic number of mobile surfaces, different aerodynamic data set, different FCS - Introduction of aeroelastcity - implementation of hydraulic/electric systems (co-simulation) - simulation of landing / taxing introducing landing gear model in the airframe multibody model - improvement of output results

33 33 Questions?

34 Thank You