Integrated CAE Solutions Multidisciplinary Structure Optimisation with CASSIDIAN LAGRANGE

Integrated CAE Solutions Multidisciplinary Structure Optimisation with CASSIDIAN LAGRANGE Cassidian Air Systems 2011 European HyperWorks Technology Conference Dr. Fernass Daoud, Head of Design Automation and Optimization

Contents Contents Introduction: Cassidian Air Systems Motivation: Challenges & Opportunities of the Airframe Design Process Multidisciplinary Airframe Design Optimization at Cassidian Air Systems Traditional Airframe Design vs. automated Multidisciplinary Design Optimization Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Applications Overview on past applications Application to the Unmanned Aerial Vehicle Talarion LAGRANGE profile in Hypermesh Benefits of the Automated Airframe Design Process 2010 CASSIDIAN - All rights reserved Page 2

Introduction Introduction: EADS and Cassidian Structure EADS Airbus Eurocopter Cassidian EADS Astrium Divsions Tom Enders (CEO) Fabrice Brégier (COO) Airbus Military Domingo Ureña-Raso Lutz Bertling (CEO) Stefan Zoller (CEO) François Auque (CEO) Cassidian Business MBDA Cassidian Systems Cassidian Electronics Cassidian Air Systems Units A. Bouvier H. Guillou B. Wenzler B. Gerwert 2010 CASSIDIAN - All rights reserved Page 3

Motivation Motivation Challenges of the Multidisciplinary Airframe Design Process The aircraft design process requires the combination of a broad spectrum of commercial as well as company specific analysis and sizing methods: a broad spectrum of A/C (company-) specific strength and stability analysis methods company specific aerodynamic and aero-elastic / loads analysis methods company specific composite analysis, design and manufacturing methods. The aircraft design is driven by a huge number of multidisciplinary design criteria (manoeuvre, gust and ground loads, aeroelastic efficiencies, flutter speeds, strength and stability criteria, manufacturing requirements etc.) resulting from different disciplines (loads, flight controls, dynamics, stress, design, etc.) The design process needs to consider and meet all these design driving criteria simultaneously, in order to determine an optimum compromise solution, i.e. all disciplines and multidisciplinary design criteria driving the airframe structural sizes and the composite lay-up need to be combined and have to interact within an integrated airframe design process. 2010 CASSIDIAN - All rights reserved Page 6

Traditional Airframe Design vs. automated Multidisciplinary Design Optimization Automation of the Global Airframe Development Process 1. Aerodynamic Analysis and Loft Optimization Aerodynamics Optimization 2. Loads Analysis and Aeroelastics Load Loops LAGRANGE: Automation of Loads and Sizing Loop Structures + Loads (2007-11) 3. Structural Analysis and Sizing 2010 CASSIDIAN - All rights reserved Page 9 Structural Optimization (until 2006) Aerodynamic Shape + Structural Sizes (2012+)

Traditional Airframe Design vs. automated Multidisciplinary Design Optimization Traditional Sizing Process Aerodynamic Loft, Inboard Profile Panel Model SDC Requirements, Parameter envelope Structural Concept Material Selection Key- Diagram Mass GFEM*) Property Update for follow up Loops Loads Loop modified for Dyn. Sizing Loop Loads Manual Sizing **) (global / local strength & stability checks) FCS Dynamic Model Dynamic Loads Flutter not fulfilled Aerodynamics, Preliminary Design Property Selection 1. Loop: (guess / engineering judgement) 2010 CASSIDIAN - All rights reserved Page 10 Check Stress *) No mass application => no correlation between stiffness & masses **) Dynamic landing and dynamic gust loads. NOT SYNCHRONISED with static loads Design Stress Loads Dynamics Mass

Traditional Airframe Design vs. automated Multidisciplinary Design Optimization Multidisciplinary Airframe Design Optimization Process Structural Concept Key- Diagram Material Selection Property Selection Mass Aerodynamic Loft, Inboard Profile Panel Model Dynamic Nodal loads GFEM**) Prelim. Dynamic Checks 1. Loop : (acc. to mass breakdown) Flight Physics Loads select design driving manoeuvres *) Within Block 0 dynamic nodal loads will be provided externally. Integration into process ongoing **) stiffness & masses in correlation ***) 1st Global Sizing Loop performed MDO (Loads, Stress, Dynamics) Balanced GFEM ***) 2010 CASSIDIAN - All rights reserved Page 11 Dynamic Check Property Update Manual Sizing (global / local for prob. follow up Loops Check Stress strength & stability checks) Loads Check Checks with tools from traditional Sizing Process Aerodynamics Design Stress Loads Dynamics Mass

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Eurofighter Composite Wing & Fin Large Spectrum of applications in Military and Civil Aircraft Design A380 Leading Edge Rib Multidisciplinary Analysis and Optimization Software Tool A400M Rear Fuselage & Cargo Door Talarion FEM Statics Aeroelasticity Steady Flutter Gust Dynamics Stability Aerodynamics Doublet Lattice Higher Order Skill Tools Strength Buckling Postbuckling Design for Manufacturing Composite Design Models & Manufact. Constraints Developed by CASSIDIAN Air Systems since 1984 More then 140 man-years of development A350 Wing A350 Fuselage X31 CFC Wing 2010 CASSIDIAN - All rights reserved Page 13

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Multidisciplinary Analyses within the Optimization Process Simultaneous consideration of aiframe design driving disciplines during analysis and optimisation: Stress Optimisation Optimisation model Criteria model Optimisation runs Stressing criteria (strength & stability) On basis of GFEM (with mass data) Loads Manufacturing Minimum & maximum thickness / dimensions Composite manuf. rules Thickness jumps, etc. Dynamics Frequency requirements Flutter speed DLM model for unsteady aeroelasticity 2010 CASSIDIAN - All rights reserved Page 14 Aeroelasticity Aeroelastic efficiencies Flutter speed Gust Aeroelastic requirements Selected design driving manoeuvres including flight conditions for each manoeuvre Aerodynamic HISSS model Coupling model (Beaming)

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Structural Components to be optimized h b b e c D Composite & Metallic Stringer Cross-sectional dimensions Ply thicknesses b e 1 Composite & Metallic Skin: Ply thicknesses / fibre orientation e.g. composite skin (wing, fuselage, taileron) Shear walls & longerons Cross-sectional dimensions Skin thicknesses Metallic frames: Cross-sectional dimensions Stringer-stiffened panels: Cross-sectional dimensions Skin thicknesses 2010 CASSIDIAN - All rights reserved Page 15

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Multidisciplinary Analysis Types Multidisciplinary structure optimisation (variable structure & variable loads) Analysis Model Criteria Model HISSS Linear Statics Linear Dynamics Steady Aeroelastics Unsteady Aeroelastics FEM Optimisation Model 2010 CASSIDIAN - All rights reserved Page 16

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Design Variables and Design Criteria Multidisciplinary structure optimisation (variable structure & variable loads) Analysis Model HISSS Linear Statics FEM Criteria Model Linear Dynamics Steady Aeroelastics Unsteady Aeroelastics c Optimisation Model Parametric model defining the design variables: h b b e D b e 1 Cross-sectional area of bars (sizing) Thickness of shell or membrane elements (sizing) Omega profile T profile Box profile C profile LZ profile Ply thickness of composites (sizing) Fibre orientation in composite stacks (angles) Coordinates of FE nodes (shape) Coordinates of control points (CAD, NURBS) Geometric Sizes + Composite-Lay-up Trimming variables (angles of attack) 2010 CASSIDIAN - All rights reserved Page 17

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Design Variables and Design Criteria Multidisciplinary structure optimisation (variable structure & variable loads) Analysis Model Criteria Model HISSS Linear Statics Linear Dynamics Steady Aeroelastics Unsteady Aeroelastics FEM Optimisation Model Parametric model defining the design variables: Cross-sectional area of bars (sizing) Thickness of shell or membrane elements (sizing) Ply thickness of composites (sizing) Fibre orientation in composite stacks (angles) Coordinates of FE nodes (shape) Coordinates of control points (CAD, NURBS) Trimming variables (angles of attack) 2010 CASSIDIAN - All rights reserved Page 18

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Design Variables and Design Criteria Multidisciplinary structure optimisation (variable structure & variable loads) Analysis Model Criteria Model HISSS Linear Statics Linear Dynamics Steady Aeroelastics Unsteady Aeroelastics FEM Optimisation Model Parametric model defining the design variables: Cross-sectional area of bars (sizing) Thickness of shell or membrane elements (sizing) Ply thickness of composites (sizing) Fibre orientation in composite stacks (angles) Coordinates of FE nodes (shape) Coordinates of control points (CAD, NURBS) Trimming variables (angles of attack) 2010 CASSIDIAN - All rights reserved Page 19

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Design Variables and Design Criteria Multidisciplinary structure optimisation (variable structure & variable loads) Analysis Model Criteria Model Strength (diverse models depending on failure criteria) Analytical buckling analysis (also theory 2nd order) Local stability analysis (for critical parts of cross-section), crippling Displacements (stiffness requirements) Constraints for natural frequencies aeroelastic requirements (steady, unsteady): efficiencies, flutter, etc. Manufacturing constraints HISSS Constraints for trimmed flight and landing manoeuvres Linear Statics Linear Dynamics Steady Aeroelastics Unsteady Aeroelastics comp allow tens allow xxxx s yyyy s FEM Damage Tolerance & Repairability Yamada-Sun Puck Tsai-Hill. Optimisation Model Parametric model defining the design variables: Cross-sectional area of bars (sizing) Thickness of shell or membrane elements (sizing) Ply thickness of composites (sizing) Fibre orientation in composite stacks (angles) Coordinates of FE nodes (shape) Coordinates of control points (CAD, NURBS) Trimming variables (angles of attack) 2010 CASSIDIAN - All rights reserved Page 20

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Design Variables and Design Criteria Multidisciplinary structure optimisation (variable structure & variable loads) Analysis Model Criteria Model Strength (diverse models depending on failure criteria) Analytical buckling analysis (also theory 2nd order) Local stability analysis (for critical parts of cross-section), crippling Displacements (stiffness requirements) Constraints for natural frequencies aeroelastic requirements (steady, unsteady): efficiencies, flutter, etc. Manufacturing constraints HISSS Constraints for trimmed flight and landing manoeuvres Linear Statics Linear Dynamics Steady Aeroelastics Unsteady Aeroelastics w eff Skin & Column Buckling for isotropic, orthotropic and anisotropic skins FEM Post-Buckling for isotropic and composite structures Optimisation Model Parametric model defining the design variables: Cross-sectional area of bars (sizing) Thickness of shell or membrane elements (sizing) Ply thickness of composites (sizing) Fibre orientation in composite stacks (angles) Coordinates of FE nodes (shape) Coordinates of control points (CAD, NURBS) Trimming variables (angles of attack) 2010 CASSIDIAN - All rights reserved Page 21

Multidisciplinary Airframe Design Optimization Procedure LAGRANGE Design Variables and Design Criteria Multidisciplinary structure optimisation (variable structure & variable loads) Analysis Model Criteria Model Strength (diverse models depending on failure criteria) Analytical buckling analysis (also theory 2nd order) Local stability analysis (for critical parts of cross-section), crippling Displacements (stiffness requirements) Constraints for natural frequencies aeroelastic requirements (steady, unsteady): efficiencies, flutter, etc. Manufacturing constraints HISSS Constraints for trimmed flight and landing manoeuvres Linear Statics Linear Dynamics Steady Aeroelastics Unsteady Aeroelastics FEM Optimisation Model Parametric model defining the design variables: Cross-sectional area of bars (sizing) Thickness of shell or membrane elements (sizing) Ply thickness of composites (sizing) Fibre orientation in composite stacks (angles) Coordinates of FE nodes (shape) Coordinates of control points (CAD, NURBS) Trimming variables (angles of attack) 2010 CASSIDIAN - All rights reserved Page 22

Overview on past applications Overview on past applications Eurofighter ( 1985) Composite Wing & Fin X-31A Wing (1990) Composite Wing Stealth Demonstrator (1995) Full A/C Design Trainer Wing (2000) Composite Wing& Fin A400M (2004-2006) Rear Fuselage Skin+Frames Advanced UAV (2006 + ) Composite Wing 2010 CASSIDIAN - All rights reserved Page 26

Application of MDO at Cassidian (2) Overview on selected Applications A350 XWB VTP Optimisation A350 XWB Wing Optimisation Optimum Composite Sizing Layout within 2 Month (MAS- Acquisition phase) ca. 3000 DV 250 000 Constraints Aeroelastics Optimum Composite Sizing of 40 Variants with 3 FTE * 6 Month for AI Toulouse ca. 20 % weight saving! A350 XWB Fuselage Optimisation Sec. 13-14 A30X Wing Optimisation Topology & Sizing Optimization of Sec. 19, A350 > 20 % weight reduction! Optimum Composite Sizing of several variants with 2 FTE * 12 Month (AI UK) > 35 % weight saving (compared to AL-design)! ca. 15000 DV 1000 000 Constraints A380 Leading Edge Rib Optimization Optimum Composite Sizing with 2 FTE * 5 Month Feasible Design without weight increase! (PAG) > 40 % weight reduction! Page 27 2010 CASSIDIAN - All rights reserved Page 27

Application to the Unmanned Aerial Vehicle Talarion Application to the Unmanned Aerial Vehicle Talarion Unmanned surveillance and reconnaissance aircraft Appr. Dimensions: Length: 14 m; Height: 4,5 m ; Span 26 m Take-off weight : 8000 kg class Performance Loiter Speed: >200 ktas Ceiling: > 43 kft Endurance: > 20 h class 2010 CASSIDIAN - All rights reserved Page 35

Application to the Unmanned Aerial Vehicle Talarion Talarion Rapid Fuselage Design Study Objective: Mass estimation In consideration of Flight-Physical constraints Strength Stability Manufacturing Automation of both loops: Loads & Sizing Loop Aerodynamic Model FE Model Analysis Model: Fully coupled structure-aerodynamic (Aeroelastic) 2010 CASSIDIAN - All rights reserved Page 36

Application to the Unmanned Aerial Vehicle Talarion Steady manoeuvre loads analysis within LAGRANGE Phase 1: Manoeuvre load simulation Requirements: (full aircraft) finite element model (full aircraft) aerodynamic model HISSS (Higher Order Sub- and Supersonic Singularity Method) or DLM (Doublet Lattice) panel model Coupling model: Beaming & Splining Methods used in order to transfer aerodynamic loads to the FE-Model and structural deflections to the Aero-Model i.e. fully coupled analysis models, without condensation FE model aerodynamic model 2010 CASSIDIAN - All rights reserved Page 37

Application to the Unmanned Aerial Vehicle Talarion Steady manoeuvre loads analysis within LAGRANGE Manoeuvre Load Simulation Based on the Mission and Structural Design Criteria the flight envelope is established and scanned (10 3-10 5 manoeuvres) in order to determine the design driving, steady manoeuvres with maximum loads Down selection of design driving steady manoeuvres (~10 2 manoeuvres). 2010 CASSIDIAN - All rights reserved Page 38

Application to the Unmanned Aerial Vehicle Talarion Steady manoeuvre loads analysis within LAGRANGE Manoeuvre Load Simulation Each design driving, steady manoeuvre can be described by the Mass configuration and CoG position Altitude Mach number Accelerations and rotational speeds (up to 9 values) Global Equilibrium of Forces and Moments has to be achieved for these pre-scribed manoeuvres of the elastic aircraft: Flight Parameter F y (inertia) + F y (aerodynamic) = 0 F z (inertia) + F z (aerodynamic) = 0 M x (inertia) + M x (aerodynamic) = 0 M y (inertia) + M y (aerodynamic) = 0 M z (inertia) + M z (aerodynamic) = 0 2010 CASSIDIAN - All rights reserved Page 39

Application to the Unmanned Aerial Vehicle Talarion Steady manoeuvre loads analysis within LAGRANGE Trimming Process For each steady manoeuvre (Mass, CoG, Altitude, Mach, Accelerations) the angles of attack (pitch angle, yaw angle and the AoA of the control surfaces are determined by minimizing the residual forces. β: mainly trim sideslip antimetric ailerons: mainly trim roll axis Flight Parameter sym. elevators: mainly trim pitch axis rudder: mainly trim yaw axis sym. elevators: mainly trim pitch axis antimetric ailerons: mainly trim roll axis F y (inertia) + F y (aerodynamic) = 0 F z (inertia) + F z (aerodynamic) = 0 M x (inertia) + M x (aerodynamic) = 0 M y (inertia) + M y (aerodynamic) = 0 M z (inertia) + M z (aerodynamic) = 0 alfa = - xx beta = - yy delta aileron = zz delta elevator = hh delta rudder = kk β α α: mainly trim lift 2010 CASSIDIAN - All rights reserved Page 40

Application to the Unmanned Aerial Vehicle Talarion Under Development Integration of Transient Gust into Optimisation A gust case is defined as a combination of: a) steady manoeuvre: flight condition and mass configuration (c.o.g. position!) (altitude & aircraft speed; usually 1g cruise) b) gust condition: wave-length and up- or down wind gust velocity and incidence angle (usually sinusoidal shaped) leading to huge amount of different gust cases (up to ~10000), which have to be considered! example: ~ evaluated time steps (approx. 1000) gust upwind profile flight condition (incl. mass configuration) gust wave length 2010 CASSIDIAN - All rights reserved Page 43

Application to the Unmanned Aerial Vehicle Talarion Under Development Gust Process Many gust blocks gust block for selected mass configuration Many gust cases gust case = incremental gust analysis specific mass configuration, specific incidence angle, specific wave length, specific speed, specific altitude + basic flight attitude trimmed aeroelastic steady manoeuvre for specific mass configuration, specific altitude, specific speed to be superimposed to an incremental gust Database (HDF5) evaluation of all gust cases for selected mass configuration Implementation of the Incremental Gust Response and the Sensitivities is completed. Implementation process for the fully automated determination of the design driving time steps and the superposition to manoeuvre load cases is ongoing. 2010 CASSIDIAN - All rights reserved Page 44

Application to the Unmanned Aerial Vehicle Talarion Currently under development Application to the Unmanned Aerial Vehicle Talarion Summary for Phase 1: Manoeuvre, Gust and Landing Loads Analysis The manoeuvre load simulation of elastic aircraft (fully coupled aerodynamicstructure model) is combined with a trimming process (optimisation task) in order to provide the distributed, elastic aircraft manoeuvre loads. The distributed aerodynamic and inertia loads are directly applied to the global, non-condensed FE model, providing the stresses and displacements for the subsequent strength and stability analysis. Gust loads are determined as incremental dynamic response. The time steps resulting in maximum local stresses are determined and the resulting deflections are superimposed to the corresponding steady manoeuvres. Landing Gear Loads are determined by an external Multi-Body-Analysis and then applied to the global full aircraft model in order consider them in the sizing process. By incorporating the loads analysis into the optimisation platform LAGRANGE both very time consuming loops (loads & sizing) are automated. 2010 CASSIDIAN - All rights reserved Page 45

Application to the Unmanned Aerial Vehicle Talarion Talarion Rapid Fuselage Design Study Phase 2: Multidisciplinary Sizing optimisation Analysis Model Criteria Model Aeroelastic analysis Manoeuvre Simulation trimming Process Optimisation Model Strength analysis: Damage tolerance Von Mises Skin buckling (for composite skin and metallic shear walls) Column buckling Flight-Physics: Force equilibrium in Y, Z Moment equilibrium around X, Y, Z Ply thickness of composite skin Thickness of metallic shear walls Cross-sectional areas of stringers Trimming variables 2010 CASSIDIAN - All rights reserved Page 46

Application to the Unmanned Aerial Vehicle Talarion Talarion Rapid Fuselage Design Study Optimisation Model: Composite skin Composite/Metallic Stringer Metallic Shear Walls 2312 elements linked to 43 patches 45-45 90 0 45-45 90 0 0 90-45 45 0 90-45 45 Stacking sequence: 16 layers linked to 3 design variables 4212 elements linked to 174 patches 174 design variables (1 * 174 patches) 8087 elements linked to 712 patches 712 design variables (1 * 712 patches) 129 design variables (3 * 43 patches) Total: 1015 design variables 2010 CASSIDIAN - All rights reserved Page 47

Analysis Model: Steady Aeroelasticity Optimisation Model: Talarion Rapid Fuselage Design Study antimetric ailerons: mainly trim roll axis sym. elevators: mainly trim pitch axis rudder: mainly trim yaw axis sym. elevators: mainly trim pitch axis antimetric ailerons: mainly trim roll axis β: mainly trim sideslip β α α: mainly trim lift 5 design variables for each load case ~ 50 Load cases ~250 Design variables 2010 CASSIDIAN - All rights reserved Page 48

Application to the Unmanned Aerial Vehicle Talarion Talarion Rapid Fuselage Design Study Criteria Model: Strength Stability Composites: Maximum strain (Damage Tolerance) 36992 constraints * 34 load cases Metallic: Von Mises stress 12299 constraints * 34 load cases 1.675.894 strength constraints Skin & shear wall buckling: Stringer Column Buckling: 1080 constr. * 34 load cases 419 constr. * 34 load cases 50966 buckling constraints Total: 1.726.860 constraints 2010 CASSIDIAN - All rights reserved Page 49

Optimisation at Cassidian Air Systems Criteria Model: Talarion Rapid Fuselage Design Study Flight-Physics Flutter Displacement U max max Manufacturing t tot t tot1 t tot2 c M M M x y z 0 0 0 5 constraints / load case ~ 50 load cases Y 0 Z 0 ~ 5 Displacement constraints / load case 50 load cases h b b e D Layer thickness fitting: 8layer * 47 steps = 376 b Minimum relative group thickness: 147 Minimum absolute stack thickness: 49 e 1 ~ 250 constraints 1 flutter constraint / mass configuration ~ 250 constraints 572 constraints 2010 CASSIDIAN - All rights reserved Page 50

Application to the Unmanned Aerial Vehicle Talarion Talarion Rapid Fuselage Design Study Skin Thickness & overall Reserve Factor: Total Thickness Overall R f Composite Skin 2010 CASSIDIAN - All rights reserved Page 51

Application to the Unmanned Aerial Vehicle Talarion Talarion Rapid Fuselage Design Study Thickness of Metallic Shear Walls, Frames, Floors & overall Reserve Factor: Total Thickness Overall R f All Results are available in different formats (colour plots, Excel Tables, Database etc.) Metallic Shear Walls, Frames, Floor 2010 CASSIDIAN - All rights reserved Page 52

Under Development LAGRANGE Input deck LAGRANGE Profile in HyperMesh Four distinct decks of the Lagrange input file to be considered Design Control Deck (control commands) NASTRAN- Deck Case Control Deck Optimization Data Deck (optimization data) Bulk Data Deck......... ENDCONTROL......... BEGIN BULK......... ENDDATA......... ENDE Chap. 3 Chap. 4 Chap. 5 to Chap. 8 HM user-page LAGRANGE Profile I/O loads reduced NASTRAN template uses standard NASTRAN I/O- procedures (CCD & BDD) provides extended I/O- features (DCD and ODD) 2010 CASSIDIAN - All rights reserved Page 54

Under Development Automatic Generation of DVs and BFs (defopt) External defopt- usage fully supported manage config-files execution of external binary defopt writes files to disk data to be imported generic process extendible for further external executables 2010 CASSIDIAN - All rights reserved Page 55

Summary Benefits of the Automated Airframe Design Process The optimization assisted airframe design process has been established and applied within all design phases of a broad range of A/C projects (civil and military applications; components, large assemblies & full A/C). The multidisciplinary design optimisation with LAGRANGE leads to a feasible airframe design which satisfies the requirements of all relevant disciplines with minimum weight. The automation of both loops: structural sizing and loads loop results in an drastic reduction of development time and effort. The strategic decision for an continued development the in-house MDO tool LAGRANGE is due to the specific aerospace design criteria on one hand (no Commercial Of The Shelf tool available) and the tremendous benefits and competitive advantages on the other hand. The in-house software availability allows the fast adaption to advanced analysis methods as well as to new technological product and customer requirements. Further Applications and Co-Operations are welcome! 2010 CASSIDIAN - All rights reserved Page 59

Thank you for your attention! The reproduction, distribution and utilization of this document as well as the communication of its contents to others without express authorization is prohibited. Offenders will be held liable for the payment of damages. All rights reserved in the event of the grant of a patent, utility model or design. 2010 CASSIDIAN - All rights reserved Page 60

Optimisation at Cassidian Air Systems Benefits of Design Automation Multidisciplinary optimisation with LAGRANGE provides a structural design which satisfies the requirements of all relevant disciplines with minimum weight Automation of both loops: structural sizing and load analysis (through full coupling to aerodynamic analysis tools) Huge multidisciplinary criteria model covering requirements of all structure design driving disciplines Drastic reduction of development time and effort (>50%) Reduced cost due to automation of the design process Estimated Saving for Talarion Development Process: Traditional Manual Process: e.g. 6 Load Loops (LL) * 1 J / LL * 40 FTE (Stress, Loads, Dynamics, Mass, Design, etc.) = 240 Man-Years With MDO Process: 2 Load Loops * 0,65 J / LL * 40 FTE = 52 Man-Years Estimated Saving: 188 Man-Years (37,6 Mill. )!!! Product/project specific requirements can be integrated (if required) in the platform easily (Property of CASSIDIAN Air Systems) Over 3 Million lines of code, 140 Man-Yeas development 2010 CASSIDIAN - All rights reserved Page 61