Landing Gear Design for Blended Wing Body Flight Test Demonstrators Aerospace Engineering

Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Fabrizio Rizzi

Thesis to obtain the Master of Science Degree in Aerospace Engineering

Supervisor: Prof. Afzal Suleman

Examination Committee Chairperson: Prof. Fernando José Parracho Lau Supervisor: Prof. Afzal Suleman Member of the Committee: Dr. Frederico José Prata Rente Reis Afonso

The Blended Wing Body (BWB) configuration is a hybrid shape with unique features capable to take benefits from both flying wing and conventional aircraft. The poof of an unconventional concept configuration requires flight tests and validation on small Flight Test Demonstrators, avoiding cost and risk related to the use of a full scale model. A first 7% scaled Unmanned Aerial Vehicle (UAV) for a new generation BWB aircraft, has been designed and flight tested at the Center for Aerospace Research (CfAR), operating on unprepared grass runways (GRS). The need to evaluate the UAV, fully controlled with an autopilot, even in the critical phases of take-off and landing on concrete runways (CON), has required the design and integration of an undercarriage system. One of the main goals of the present thesis work is to carry out the design and development of a landing gear for FTV7%, including the integration of the system into the aircraft, ground testing and mock-ups preparation for a wheeled flight test campaign. In the process of scaling towards the faithfully representation of all the aspect that concern a full scale aircraft, the design of the landing gear has been started in parallel, even for a new 16.5% scale, involving some design similarities and the sizing of additional mechanical subsystems, such as dedicated suspension and braking systems. At last, the basic take-off and landing performance evaluation of the landing gears, designed for both FTV7% and FTV16.5%, is presented highlighting the influence of some design parameters.

Keywords: BWB, Undercarriages, Design, Scaling, Performance evaluation.

i ii Resumo

A Blended Wing Body (BWB) é uma configuração com forma híbrida e com características únicas que lhe permitem beneficiar das vantagens da configuração convencional e em asa voadora. A prova do conceito de uma configuração não convencional requer ensaios em voo e validação através de Flight Test Demonstrators, evitando assim custos adicionais e riscos associados a ensaios à escala real. Um primeiro modelo Unmanned Aerial Vehicle (UAV) à escala de 7% de uma aeronave com confi- guração BWB tem sido desenvolvido e submetido a ensaios de voo pelo Centre for Aerospace Research (CfAR), os quais têm sido feitos em pistas de aterragem indevidamente preparadas. A necessidade de avaliar um UAV controlado com um piloto automático em cenários críticos de descolagem e aterragem em pistas cimentadas requer o projecto e integração de um trem de aterragem. Um dos objectivos principais deste trabalho consiste no projecto do trem de aterragem do FTV7%, incluindo a integração do sistema na aeronave, ensaios no solo e a preparação duma maquete para a campanha de ensaios de voo. Paralelamente, durante o processo de redução de escala, foi desenvolvido o projecto do trem de aterragem de um modelo à escala de 16.5%. Este contem semelhanças com anterior mas com subsistemas mecânicos adicionais, tais como uma suspensão e um sistema de travagem. Por fim, o desempenho do trem de aterragem durante a descolagem e aterragem são avaliados e pro- jectados para o FTV7% e FTV16.5%, são apresentados com foco na influência de alguns parâmetros de projecto.

Palavras-chave: BWB, trem de aterragem, projecto, modelos à escala, avaliação de desempenho.

iii iv List of Figures

2.1 Landing gear for different aircraft applications. [6] ...... 5 2.2 Ground clearance requirement during take-off rotation. [10]...... 7 2.3 Take-off rotation and tip-back parameters. [10]...... 7 2.4 Landing gear application for several UAVs...... 10

3.1 Interface of the StatData feature of the Flight Test Data Processing software ...... 12 3.2 CAD assembly of the designed main landing gear for FTV7...... 14 3.3 Option 1: Skymaster wheels...... 15 3.4 Option 2: Turnigy HK wheels...... 15 3.5 SW structural simulations of the Al2024-1/8" leaf strut ...... 16 3.6 Drop Test of the Al2024 1/8" performed at CfAR...... 16 3.7 Geometric parameters for the leaf strut design ...... 16 3.8 Design taper parameters for the leaf strut. [8]...... 17 3.9 Integration of the main gear with the mid bay of the aircraft ...... 17 3.10 Crack on the lower bending radius of the leaf strut, after the ﬁrst bending machine process 18 3.11 Final bending hot process of the main gear struts...... 18 3.12 Heat treatment of the leaf strut. [27]...... 18 3.13 Displacement of the main gar strut with a static load corresponding to the MTOW of FTV 7 19 3.14 Test rig with no loading applied...... 20 3.15 Test rig with static loading corresponding to the MTOW of the aircraft ...... 20 3.16 Load displacement curve for the main gear leaf strut...... 21 3.17 CAD assembly of the designed nose gear for FTV7...... 22 3.18 Nose strut with static load corresponding to the aircraft MTOW...... 23 3.19 Nose strut with bottoming force, corresponding to the impact in hard landings ...... 23 3.20 CAD design of the custom components for the nose gear of FTV7...... 24 3.21 Geometry and kinematics of the steering system ...... 25 3.22 Turning radius for the steering system...... 25 3.23 Steering servo components...... 26 3.24 Mechanical steering system components ...... 26 3.25 Static test for the characterization of the nose gear strut: the test rig is represented on the left, the resulting load/deformation curve on the right...... 27 3.26 External and internal integration of the main landing gear with the airframe ...... 28 3.27 Nose landing gear external and internal integration with the airframe...... 29 3.28 Weight and Cost distributions for the production of one landing gear set for FTV7 . . . . 30

4.1 Tuning of the steering system control parameters for the autopilot ...... 32 4.2 Component list and mounting drawing for the Nose Gear strut assembly 1...... 33 4.3 Landing overview using the Flight test data processing software ...... 35

5.1 Work break-down structure of the landing gear design...... 38 5.2 Multidisciplinary process of the landing gear design...... 39 5.3 Basic requirements for the 16.5% Landing Gear design...... 40 5.4 Equilibrium about the main gear contact point at take-off rotation...... 41 5.5 Sketch of the swivel component, necessary to rotate the main gear position ...... 42 5.6 Take-off rotation ground clearance requirement...... 43 5.7 Geometric parameters involved in the determination of the turnover angle...... 43 v 5.8 Design procedures for impact loading condition: on the left the scaling design procedure with influence of the parachute rate of descent; on the right the classical design procedure with influence of the design wheel travel on landing...... 47 5.9 Sketches of the concepts for the internal layout configuration of the main landing gear. . . 48 5.10 Off-the-shelf tires from the UAV and ultralight aircraft market ...... 49 5.11 Off-the-shelf wheels from the UAV and ultralight aircraft market...... 50 5.12 Off-the-shelf brake solutions for FTV16.5 applications ...... 50 5.13 Off-the-shelf shock absorbers for FTV16.5 applications...... 51 5.14 Conceptual geometry of the main gear strut leg...... 51 5.15 Initial structural evaluation of different custom main gear struts ...... 52 5.16 Conceptual sketches of the carbon fiber strut ...... 52 5.17 Sketches of the concepts for the nose landing gear for FTV16.5...... 53 5.18 Wheel assembly solutions for the nose gear of FTV16.5...... 54 5.19 Concepts for the steering system of the landing gear for FTV16.5: Rack and Pinion mechanical transmission on the left and Pulley-Belt mechanism on the right ...... 54 5.20 Test rig for shock absorbers: it shows all the components needed for the test rig assembly. 55 5.21 Test rig concepts: on the left the Speed Rating Test rig, on the right the Drop Test rig. . . 56 5.22 Feature importance for the Main gear concept selection ...... 57 5.23 Feature importance for the Nose gear concept selection...... 57 5.24 Concepts selected for the main gear (on the left) and for the nose gear (on the right) . . 58 5.25 Procurement of the main gear wheel assembly...... 59 5.26 Procurement of the nose gear wheel assembly...... 59 5.27 Off-the-shelf shock absorber for the nose and main landing gear...... 60 5.28 Internal chambers of the shock absorber [38]...... 60 5.29 Shock load for the FOX 3 DPS 6.5"×1.5", selected for the Main gear application . . . . . 61 5.30 Shock load for the FOX 3 DPS 5.5"×1", selected for the Nose gear application ...... 61 5.31 Design space for the main landing gear inside the airframe of FTV16.5...... 62 5.32 Preliminary design geometry and load distribution of the main gear selected concept. . . . 62 5.33 Braking system for ultralight applications: internal caliper plus brake disk. [43] ...... 64 5.34 Braking system from bike applications applied to the X-48 UAV. [17]...... 65 5.35 Design space for the nose landing gear inside the airframe of FTV16.5...... 65 5.36 Preliminary design geometry and load distribution for the nose gear selected concept. . . 66 5.37 Pulley-belt steering system design parameters. [37]...... 66 5.38 Weight and Cost estimation for the production of one landing gear set for FTV16.5 . . . . 68

6.1 Aircraft taxiing operation over a 1-cosine modeled runway...... 70 6.2 Vertical acceleration during taxiing over a 1-cos dip runway for the FTV7 and FTV16.5 . 71 6.3 Heave response during taxiing over a 1-cos dip runway for the FTV7 and FTV16.5 . . . . 71 6.4 Aircraft model during landing...... 72 6.5 Vertical impact deceleration on the main gear of FTV7 and FTV16.5...... 73 6.6 Vertical displacements on landing on the main landing gear of FTV7 and FTV16.5 . . . . 73

B.1 Main Gear assembly drawing 1: leaf strut and attachment with the airframe ...... 86 B.2 Main Gear assembly drawing 2: wheel assembly connection with the leaf strut ...... 86 B.3 Nose gear assembly drawing 2: wheel assembly and piston strut...... 87 B.4 Nose gear assembly drawing 3: steering system ...... 87 B.5 Shock absorber Test Rig assembly drawing 1 ...... 88 B.6 Shock absorber Test Rig assembly drawing 2 ...... 88

C.1 Rigid aircraft with equivalent model of undercarriages during taxiing over the design runway.[39] ...... 90 C.2 Simulink model for the heave response on time domain during landing...... 91

vi vii List of Tables

2.1 Pros and Cons of the most used landing gear types [10] ...... 6

3.1 General requirements for the landing gear design for FTV7 ...... 12 3.2 Aircraft and performance requirement for FTV7 ...... 13 3.3 Geometry requirements for FTV7...... 13 3.4 Specific requirements for the main and nose gear of FTV7...... 14 3.5 Analytical Hierarchical table for tire and wheel selection of FTV7 main landing gear. . . . 15 3.6 Geometric parameter determination of the custom leaf strut design...... 17 3.7 Resulting displacements from Static Tests with increasing weight: (0 : 3 : 30kg) ...... 20 3.8 Linearized stiffness from experimental, analytical and computational tests...... 21 3.9 Drop test setting and results for different simulated landing cases...... 22 3.10 Linearized stiffness coefficients for the nose strut during soft and hard performances . . . 27 3.11 Main landing gear position with respect to the Fuselage Station...... 28

5.1 Set of landing gear configuration for different position of the c.g...... 42 5.2 Track and turnover angle for each main gear configuration...... 45 5.3 Static loading cases on the nose and main gear for each landing gear configuration . . . . 45 5.4 Total loading cases on the nose and main gear for each landing gear configuration . . . . . 45 5.5 Scaling design procedure: impact accelerations developed on nominal and parachute landings 46 5.6 Classic design procedure: impact accelerations developed on nominal and parachute landings 46 5.7 Analytical hierarchical process table for the Main Gear concept selection...... 57 5.8 Analytical hierarchical process table for the Nose Gear concept selection...... 58 5.9 Braking distance calculation according to FAR 23 regulations...... 63 5.10 Calculation of the parameters for the preliminary sizing of the braking system ...... 64 5.11 Preliminary selection of the pulley-belt steering system...... 67

6.1 Input for the taxiing dynamic model...... 70

viii ix Contents

List of Figures v

List of Tables viii

1 Introduction 1 1.1 Background and Motivations ...... 1 1.2 Scope of the Thesis Work ...... 2 1.3 Contributions at the Center for Aerospace Research ...... 2 1.4 Collaborations ...... 2 1.5 Layout of the Thesis document...... 3

2 Guidelines for Landing Gear Design5 2.1 Design keys and variables for the Landing Gear...... 5 2.2 Regulations for Landing Gear Design...... 8 2.2.1 Design limit parameters...... 8 2.2.2 Testing procedures...... 9 2.3 State of the art of Landing Gear for similar aircraft applications ...... 10 2.3.1 References for FTV7%...... 10 2.3.2 References for FTV16.5%...... 10

3 Landing Gear Design and Development for FTV7% 11 3.1 Requirement list ...... 11 3.1.1 General requirements ...... 12 3.1.2 Aircraft and Performance requirements ...... 12 3.1.3 Geometry requirements ...... 13 3.1.4 Main and Nose gear requirements...... 13 3.2 Design and development of the main gear assembly...... 14 3.2.1 Main gear features...... 14 3.2.2 Wheels and Tires for the main gear ...... 15 3.2.3 Design of the main gear strut...... 15 3.2.4 Testing of the main gear...... 18 3.3 Design and development of the nose gear assembly...... 22 3.3.1 Nose gear features ...... 22 3.3.2 Component oﬀ-the-shelf for the nose gear ...... 23 3.3.3 Design of additional custom components...... 23 3.3.4 Design of the steering system...... 24 3.3.5 Testing of the nose gear...... 26 3.4 Integration of the landing gear with FTV7...... 27 3.4.1 Integration of the main gear...... 27 3.4.2 Integration of the nose gear...... 28 3.4.3 Belly Pan design...... 29 3.5 Weight, cost and conclusions ...... 29

4 Ground mock-ups and flight test planning for FTV7% 31 4.1 Preparation of the aircraft for ground and flight testing ...... 31 4.1.1 Ground mock-ups ...... 31 4.1.2 Landing gear toolkit...... 33 4.2 Ground testing...... 33 x 4.2.1 Taxiing of the FTVs...... 33 4.2.2 Take off run testing of FTV 2B...... 34 4.2.3 Improvements on the landing gear ...... 34 4.3 Flight test planning ...... 34 4.3.1 Flight test data processing ...... 34

5 Landing Gear design for FTV16.5% 37 5.1 Design process for the 16.5% Landing Gear system...... 37 5.2 Basic features of the 16.5% Landing Gear ...... 39 5.2.1 Landing Gear position and aircraft center of gravity...... 40 5.2.2 Landing Gear height...... 42 5.2.3 Track and Turnover angle...... 43 5.2.4 Design loading conditions for the nose and main gear...... 45 5.2.5 Impact loading condition for the landing gear...... 45 5.3 Conceptual Design...... 47 5.3.1 Main gear concepts...... 47 5.3.2 Nose gear concepts...... 53 5.3.3 Test rig concepts...... 55 5.4 Preliminary Design...... 56 5.4.1 Selection of the concepts...... 57 5.4.2 Procurement and testing procedures for the oﬀ-the-shelf components ...... 59 5.4.3 Preliminary design of the main gear ...... 62 5.4.4 Preliminary design of the nose gear...... 65 5.5 Weight, cost estimation and conclusions...... 67

6 Basic performance evaluation of aircrafts with a landing gear system 69 6.1 Response of the aircraft with landing gear during the typical ground maneuvers ...... 69 6.1.1 Taxiing dynamic model ...... 69 6.1.2 Taxiing performance evaluation and comparison for FTV7 and FTV16.5 ...... 70 6.2 Response of the aircraft with landing gear during a typical design landing...... 72 6.2.1 2-points landing dynamic model ...... 72 6.2.2 Landing performance evaluation and comparison for FTV7 and FTV16.5 . . . . . 73

7 Conclusions and Future developments 75 7.1 Conclusions...... 75 7.2 Future developments...... 76

A Landing gear Matlab parameter calculator 77 A.1 Input parameters...... 77 A.2 Output parameters...... 77

B CAD assembly drawings 85 B.1 Assembly drawings and Bill of components for the Landing Gear for FTV7% ...... 85 B.2 Assembly drawings for the Shock Absorber Test Rig ...... 85

C Performance evaluation 89 C.1 Taxiing Model ...... 89 C.2 Landing simulink model...... 91

Bibliography 93

xi Chapter 1

Introduction

1.1 Background and Motivations

The design and development of Unmanned Aerial Vehicles (UAVs) is the first step towards the proof of feasibility of new generation aircrafts. Instead of building an expensive full scale flight test demonstrator, the general approach is to scale the model and demonstrate that the configuration and associated technologies warrant the development of a full-scale, certifiable aircraft. Experimental data for scaled models are used to define and review the basic characteristics of full-scale aircrafts, verify theoretically predicted behaviour and provide support for making decisions in low time, cost and risk. [1] In this framework, the Canadian Aircraft Companies Bombardier and Quaternion Aerospace are moving towards the development of an advanced, unconventional, Blended Wing Body (BWB) business jet with an estimated entry into commercial service in 2035. [2] The scaling process is a step-by-step progress that requires the development of small flight test vehicles (FTVs) in different scales and configurations, in order to test the effectiveness of each system and the affected behaviour of the aircraft during flight. The first scaled representation of the new generation aircraft was built and flight tested since 2016 at Center of Aerospace Research (CfAR), with the collaboration of the University of Victoria (UvIC) in British Columbia - Canada. It is a 7% of the full scale aircraft, designed and developed having in mind that the scaling of the physics of such a complex system goes far beyond merely scaling down the size. [3] The objective of the first configurations of FTV 7% was to collect aerodynamic data from flight tests in order to validate and improve the control laws, used from the autopilot to maneuver the aircraft. At this time the phases of take-off and landing were not important to evaluate: the aircraft was catapulted into the air using a shoot launcher and the landing was performed using the belly pan of the aircraft. The next generation has required a lower risk to damage the vital components of the aircraft in terms of airframe and systems and the necessity to test the aircraft’s behaviour and control even during the critical phases of take-off and landing. In order to accomplish these new needs towards an improves faithfullness of the represented scaled aircraft, a landing gear system has been needed and the ground operation of the aircraft has moved from Grass airstrip to paved runways. 1 In parallel with the flight tests and demonstrations of expected performance for a wheeled configuration of FTV 7%, the Center for Aerospace research has been working on the design and development of a larger scale flight test demonstrator that represents the 16.5% of the full aircraft size. The new scale requires take-off and landing on paved concrete runways, so the design of a detailed landing gear system is one of the hard-points that conditions the final layout and behaviour of the aircraft. The design and sizing of a landing gear system for such a large scale of UAV will provide an additional value towards the development of the new generation regional jet aircraft, now enabled to handle and dissipate impact shock forces with an adequate suspension system, to be stably maneuvered on the ground thanks to a reliable steering system and to be stopped in a specific distance by using a certified braking system. [4]

1The aviation code for a grass runway is GRS,for a paved concrete runway is CON. 1 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

1.2 Scope of the Thesis Work

The main purpose of the present thesis work is to carry out the design and development of both undercarriages for two different scales of UAVs, that represent the new blended wing body business jet. For the smaller flight test vehicle application, it is required that the landing gear enable the aircraft to be tested in a wheeled configuration with steering and braking capabilities linked to the operational conditions of the selected airstrip for testing. The landing gear has been designed, manufactured and integrated with the airframe, including many aspects of aircraft design, fabrication, testing and analysis. The landing gear for the larger scale demonstrator, instead, follows the general time-line and schedule of the other design teams of the aircraft and a constant interchange of data and information is necessary for the development of a feasible aircraft concept with all the required systems and subsystems stated in the flying demo spec document [2]. The undercarriage system for FTV 16.5% is demanded to have a dedicated suspension in order to manage and dissipate the impact loads developed on take-off and landing, and a detailed braking system to stop the aircraft within a certain distance. The project also focuses on the design and development of custom test rigs in order to test the properties of the components used, as well as prove the integrity and functionality of the designed landing gear in both static and impact load case scenarios. All the phases of the work have been accompanied with the necessary documentation containing all the technical information of the landing gear, including component list definition, CAD drawings, updated requirement spreadsheets and support material for simulations and testing. At the end of the present thesis work, the reader should be aware of all the unique challenges that the design of a landing gear offers, due to the multidisciplinary nature of the design process and the necessity to evaluate structural behaviours and general performances since the first phases of preliminary design. He will be even conscious of all the ground mock-ups operations required after the installation of a landing gear, that enable the wheeled aircraft to be fully controlled, in all the phases of flight, by an autopilot.

1.3 Contributions at the Center for Aerospace Research

The contributions resulting from the present thesis work are attributable to the area of design, integration and performance evaluation of unmanned air vehicles equipped with landing gear. In particular the Center for Aerospace Research is now provided of the following main outcomes, that are the result of seven months of design work performed from March until September 2018:

• Design Excel spreadsheet: It contains all the critical requirements and key variables, necessary for the whole design process of a landing gear system.

• MatLab® Sizing tool: The software receives some input parameters stated in the Excel spreadsheet, and automatically calculates the resulting preliminary valid geometry that respects all the requirement veriﬁcations. The tool includes also sections to execute preliminary calculations for the mechanical subsystems required to a landing gear.

• MatLab® & Simulink performance evaluation tool: The software can be used as a valuable tool for the heave performance evaluation of the landing gear during take-oﬀ and landing since the beginning of the preliminary design and so, it can guide and demonstrate the feasibility of some design decisions.

• Checklist and hardware toolkit for the landing gear of FTV7%: All the landing gear fasteners and components, designed and manufactured for FTV7%, have been organized in a dedicated toolkit box, with part number descriptions and all the necessary mounting drawings. A checklist for the landing gear spare parts has been prepared as well, in order to facilitate the operations of pre-ﬂight checking and repairing of damaged components.

1.4 Collaborations

All the design and realization phases of the project have been supported by the collaboration with several entities. First of all, the Center for Aerospace Research, located in an hangar of the International Victoria Airport in Sidney and built in 2012 by the Professor Afzal Suleman in collaboration with the University of Victoria, has represented the physical location where all the design process and ground 2 1 – Introduction testing phases have been performed. The Center is specialized for UAV design in Western Canada and its shop mechanical machines have been extremely helpful to characterize the experimental nature of design systems for Unmanned Aerial Vehicles. The Center’s human resources are organized in several teams and a close collaboration has been needed with each of them that affects the landing gear design, such as the Airframe design, Recovery system and Control system teams. Ground mock-ups and flight test preparations for the wheeled configuration of FTV 7% have been performed in collaboration with the CfAR Flight Test Crew composed by Stephen Warwick, Jenner Richards, Sean Bazzocchi and Max Rukosuyev. A constant reporting of the major design outcomes and progress on the production and manufacture side of the project, has been presented to the Bombardier stakeholders. All the activity for the landing gear project have been resumed in weekly progress reports, support technical documentation (quarterly reports and coordination memos) and meetings every two weeks in order to verify and double check the feasibility of the designed landing gear, ensuring that the design decisions made were not drastically changing the desired behaviour of the full scale aircraft. The manufacturing phase of the landing gear for FTV 7% has been supported by several local companies that have provided the necessary mechanical instrumentation. The production of the leaf strut for the main landing gear has been supported by the facilities of the companies Stark CNC and Western Edison, specialized for waterjet cutting and bending metal components, and Pyrotek Aerospace, specialized for the heat processes of metal parts for aircraft applications. The manufacturing of needed design components for the nose landing gear, as well as modifications for off-the-shelf components, have been achieved using the manual lathe and mill machines provided at the Mechanical Laboratory of University of Victoria. The design of the landing gear for FTV 16.5% has constantly been characterized by back and forward exchange of landing gear information with the major companies identified as suppliers for feasible off-the-shelf components, such as Matco Mfg and Aircraft Spruce Canada specialized for wheel assembly components and Marc-Ingegno Italy ,Vorsprung and Fox, leaders of different suspension system applications.

1.5 Layout of the Thesis document

The organization of the thesis work in the present document tries to recall the chronology and evolution of the design process for landing gear applications from small UAVs to larger scale demonstrators. In some cases the work has been performed in parallel and so necessary rearrangements have been done in order to preserve the reading ﬂow. The resulting structure is described in the following itemize:

• Chapter 2: The general guidelines for the landing gear design for UAV applications, used for both 7% and 16.5% FTVs, are presented. The chapter begins with an initial description of the design keys and variables and speciﬁcations used to limit some design parameters. Then a state of the art of landing gear design applications, currently used for aircraft comparable with the two ﬂight test demonstrators, is described.

• Chapter 3: The third chapter shows all the design and development processes performed for the landing gear of FTV7%. The main important requirements, that have conditioned all the design, are described and lead to the deﬁnition of the preliminary feasible layout of the landing gear. Then the design and production phases for both main and nose gear are explained in detail, including the required integration phase with the airframe of two FTV7% provided at the Center.

• Chapter 4: This chapter describes all the ground mock-ups and ground testing necessary to validate the aircraft with landing gear before a flight test. In addition, it includes the flight test planning and the successive phase of analysis required to evaluate the performance of the landing gear and the influence of some design choices.

• Chapter 5: It illustrates the basic features and all the design decisions made towards the preliminary design of the landing gear for FTV16.5 application. It includes all the major outcomes related to the conceptual design phases and the initial calculations necessary to deﬁne the layout and the structure of the landing gear, including the design of the required ground test rigs.

• Chapter 6: The basic performance evaluation in terms of heave response of the aircraft with the designed landing gear is illustrated in chapter 6. The most critical phases (take-oﬀ run and landing), 3 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

that aﬀects the landing gear design, are considered and compared for both landing gear of FTV7% and FTV16.5%. • Chapter 7: The last chapter of the document highlights the most important conclusion of the thesis work and the planning of the future works for the design of the landing gear for FTV16.5, expected to be ﬂight tested by the end of 2019. • Appendix: The appendixes contain the support documentation to understand the features of the MatLab®software for sizing calculations, the CAD assembly drawings and bill of all the components used for the 7% landing gear, the theoretical background for the mathematical models used for the performance evaluation described in chapter6.

4 Chapter 2

Guidelines for Landing Gear Design

The landing gear can be defined as the essential intermediary that prevents the airplane from the cata- strophe, so particular attention and engineering effort to premeditate possible failure modes, are required since the beginning of the design. [5] Since the functions fulfilled by the aircraft are extremely different, it becomes clear that each landing gear represents an individual case, designed with specific considerations and decisions regarding its own application. Several types of landing gear for different aircraft applications are shown in figure 2.1.

Figure 2.1. Landing gear for diﬀerent aircraft applications. [6]

In general the following functions are required to the most variety of landing gear system: [7]

• Allow Take-oﬀ and landing operations;

• Provide stability for ground maneuvering taxiing and take-oﬀ;

• Transfer the ground loads to the airframe;

• Convert the longitudinal kinetic energy in heat thanks to a braking system;

• Damp the vibrations and bouncing caused by the kinetic energy developed upon impact and take-oﬀ run operations;

The undercarriages have essentially to convert the aircraft from its natural airborne environment into a lumbering ground vehicle on the ground. The general approach for the design of a landing gear, follows the normative established by the FAR regulations and typical considerations explained in the pillars of the landing gear design like the references Roskam, Currey and Niu. [7][8][9] Anyway in most cases this approach is not directly applicable for small scaled aircraft, where speciﬁc requirements and compromises between scaling process and UAV considerations are necessary. The three following sections describe the major keys and variables for the landing gear design and some examples of landing gear designed for airplanes with similarities to FTV 7% and FTV 16.5%, that have been considered as a reference for some design decisions.

2.1 Design keys and variables for the Landing Gear

The design of the landing gear is an iterative process which involves parameters that strongly influence the aircraft configuration design and aerodynamics performances. [10] All the keys and variables involved in the landing gear design from the beginning through the whole iterative process, have been evaluated in the first research phase and are explained schematically as follows. 5 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

• Scale of the aircraft: It is the first parameter that conditions from the beginning all the design process. Essentially the aircraft scale can vary among Radio commanded small planes, Unmanned Aerial Vehicles, Ultralight aircrafts, Commercial and Cargo aircrafts, Military aircrafts and for each of them the specific operation condition of the aircraft highly influences all the design decisions. [11]

• Type and complexity: They represent the two major decisions that must be made before the landing gear design process can start. The typical possible conﬁgurations of a landing gear are tricycle, bicycle, tailwheel or unconventional gear and its complexity is highly aﬀected by the need or not of a retraction system. The optimal landing gear layout can be decided after an analytical hierarchical process (AHP) method through the assignment of weighting factors to each of the feature described in the pros and cons table 2.1.

Table 2.1. Pros and Cons of the most used landing gear types [10]

Landing gear type Characteristic Tricycle Bicycle Tailwheel Weight Medium High Low Ground stability High Undetermined Low Steering ability High Medium Low Leveled attitude High Medium Low Take-oﬀ rotation High Low Medium

The decision to use or not a retractable system is guided by the aircraft cruise speed and weight/cost budget: the state of the arts shows that airplanes with cruise Mach number less then 0.85 tend to have fixed gears.[7] For light and not fast aircraft applications, the extra weight and the cost that accompanies a retractable landing gear are usually more disadvantageous than the parasite drag caused by the friction of the air flowing over the fixed gears. Lightweight airfoil-shaped fairings and wheel pants can be eventually used to streamline the airflow as aerodynamically as possible reducing a great amount of parasite drag. [12]

• Landing gear attachment: there are two possible attachment structures for the landing gear, represented by wing and fuselage. Usually the fuselage attachment is preferred for small aircrafts with a fuselage wide enough to allow the desired wheel track. In fact, at equal wheel track, the height of the landing gear for a wing attachment is bigger (resulting in more weight) and additional components are needed to transfer the loads from the wing ribs to the fuselage structure. [13]

• Center of gravity position: the center of gravity (C.G.) envelope inﬂuences the overall geometry of the system including the horizontal and vertical location of the undercarriages. If the horizontal position of the C.G. can vary between wide limits, the worst loading case scenario should be considered for the landing gear design. [14]

• Vertical Load ratio: The vertical load ratio between the nose and main landing gear aﬀects the position of the undercarriages with respect to the center of gravity as well as the structural components needed to manage the resulting load distribution. The normal force on the nose gear should be limited, but not less than 8% of the aircraft landing weight for an adequate steering [7]. The usual practice is to design the landing gear in order to distribute the vertical load between 8 ÷ 10% and 90 ÷ 92% respectively for the nose and main1 gear. The vertical static load on the nose PN and main gear PM can be calculated according to the system of equations 2.1, where lM and lN are respectively the main and nose arm ratio with respect to the c.g. and ns is the number of main gear wheel assemblies. [15]

( W lM PN = (lM +lN ) (2.1) W lN PM = ns(lN +lM )

1The "main" gear is so called because it carries the larger amount of load. 6 2 – Guidelines for Landing Gear Design

The nose and main landing gear are even subjected to dynamic loading condition respectively during landing with brakes applied and take-oﬀ rotation due to acceleration forces developed. The total nose gear load is therefore obtained adding the dynamic load to the static one, as shown in equation 2.2, where ax is the deceleration with brakes applied that is typically 35% of the gravity for a dry concrete runway. [10] ax hC.G.W PNTOT = PN + (2.2) g (lN + lM ) In a similar way the total load on the main gear is obtained adding to the static load, the dynamic load caused by the longitudinal acceleration during rotation, as shown in equation 2.3, where aT is the average acceleration imposed by the thrust.

aT hC.G.W PMTOT = PM + (2.3) g (lN + lM ) • Ground clearance: the height of the landing gear should ensures a reasonable clearance between the runway and all other parts of the aircraft in compressed position. The ground clearance requirement at take-off, that ensure the prevention of a fuselage or tail hit, is respected if the maximum take-off rotation angle αTO is less than the clearance angle αc, defined as equation 2.4. −1 Hf αc = tan (2.4) Lf

The dimensions Hf and Lf , shown in ﬁgure 2.2, are respectively the fuselage clearance in leveled position and the distance aft of the main gear to the beginning of the unsweep angle of rotation.

Figure 2.2. Ground clearance requirement during take-oﬀ rotation. [10]

• Take-off rotation and tip-back prevention: the geometry of the landing gear is also affected by the necessity to have a regular take-off rotation and an adequate tip-back prevention respectively during take-off and landing. The two parameters (αTO and clearance) involved in this design variable are shown in figure 2.3.

Figure 2.3. Take-oﬀ rotation and tip-back parameters. [10]

A regular take-off rotation is ensured if the A angle between the center of gravity position c.g. and vertical line on the ground contact is at least equal to the tip clearance angle αC and higher than 15◦. The A angle should not be too much different from the tip clearance angle, otherwise a great amount of load is needed on the tail in order to rotate the aircraft at take-off. [8] • Overturn prevention: The overturn of the aircraft on ground is the rolling over of the aircraft that can happen during ground turning and cross-wind conditions. The phenomena is prevented if the moment generated by the aircraft weight about one of the main gear contact point, is higher than the moment generated by the centrifugal force in ground turning maneuvers and the moment generated by acting force in cross-wind conditions. [7] The respect of the requirement passes through an analysis of ground turning controllability and stability during cross-wind conditions. [10] 7 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

• Structural integrity: The landing and taxiing loads over rough runways should be absorbed and dissipated through proper struts, possibly with either stiffness and damping properties, capable to maintain as minimal as possible the structural deflection of the landing gear. The maximum deflection of the landing gear, during impact loading, represents a limit for the maximum track of the main landing gear.

• Safety: It is important to consider from the beginning of the design that a failure of any landing gear parts does not represent a risk to critical damage the airframe and system components of the aircraft.[16]

• Low cost and low weight: The use of components oﬀ-the-shelf (COTS) is highly encouraged in order to avoid the cost of custom design whenever possible. If necessary, the landing gear designer should evaluate the possibility to use low cost and low weight components, designed for applications that goes beyond the aeronautical ones.[17]

2.2 Regulations for Landing Gear Design

The Unmanned Aerial Vehicle applications for the Landing Gear do not always allow the applicability of standard regulations. Depending on the scale, the UAV landing gear design can be considered in between the design of the system for Radio command small planes and ultralight aircrafts. The nature of these aircrafts implies, in some cases, a custom design depending on the specific mission and runway, where the plane is going to operate. Anyway, since the scaled FTVs considered are a representation of an aircraft that will necessitate a Federal Administration Regulation (FAR) certification, some design choices and parameter definition can be done referring to FAR 23 regulation, valid for normal, utility and aerobatic aircraft with maximum take-off weight less than 12500lbs.[12] The next subsections highlight how to select some important design parameter and the needed procedures for testing.

2.2.1 Design limit parameters The most important parameters, that the landing gear designer has to select in a phase of deﬁnition of all the requirements, regard the design landing to which the landing gear structure is expected to respond in elastic ﬁeld. They are related to the vertical and longitudinal behaviour of the structure, as described in the following itemize:

• Rate of descent: The rate of descent2 of the aircraft for the landing gear design should be in between 7 and 10 fps and can be determined using the equation 2.5, derived from FAR 23.725 normative. The values W [lbs] and S [ft2] refers to the aircraft weight and wing reference surface in landing condition. [18] 1 W 4 wTD = 4.4 (2.5) S L

• Ground reaction factor: The vertical dynamic loads developed at the ground contact point on landing are treated in the ﬁrst phase of the design as quasi-static loads and obtained by multiplying the static load times the ground reaction factor NG, as shown in equation 2.6. A typical value for 3 NG for ultralight aircraft is 3, simulating an impact acceleration of 3G .[7]

Dynamic load N = (2.6) G Static load

• Spin-up/ spring-back loads: In absence of speciﬁc tests for determining spin-up and spring-back loads, developed on landing when the wheels pass instantaneously from null speed to the aircraft horizontal landing speed, the appendix D of FAR 23 regulation can be used for an initial estimation.

2The rate of descent is often called sink speed or touchdown rate and is deﬁned as the vertical speed of the aircraft before touching down. It is dependent on the ﬂaring speed and angle of the aircraft in a landing attitude of two points contact on the two main landing gears. 3The impact acceleration is often measured as function of the gravity acceleration, so 3G corresponds to 29.43m/s2 8 2 – Guidelines for Landing Gear Design

The maximum value for the horizontal force (in pounds) acting on the wheel is determined through the equation 2.7, p 1 2Iw (VH − Vc) nFVmax FHmax = (2.7) re ts

where re is the effective rolling radius (in ft) of the wheel under impact, Iw is the rotational mass moment of inertia of rolling assembly (in slug ft), VH and VC are respectively the linear horizontal landing speed of the airplane and the peripheral speed of tire (in fps) if pre-rotation is used, n is 4 5 the effective friction coefficient , FVmax is the maximum vertical force on the wheel (in lbs), ts is the time interval between ground contact and reaching of the maximum vertical force on the wheel (in seconds). The dynamic spring-back effect can be estimated, in a level landing condition, assuming the load in equation 2.7 to be reversed.

2.2.2 Testing procedures The FAR regulation contains also the deﬁnition of some testing procedures needed to certify the design of the landing gear. The most relevant test peculiarities are presented as follows: • Aircraft attitude: For a leveled landing the attitude of a tricycle aircraft should be one of the following: – Simultaneous nose and man wheels contact on the ground; – Nose wheel clear of the ground when main gear touches down; • Limit Drop Tests: The full airplane or equivalent assemblies consisting of wheel, tire and shock absorber should be drop tested from free drop height in inches not less than 9.2 inches and not more than 18.47 according to the equation 2.8, derived from FAR 23.725 regulation. [18]

1 W 2 h = 3.6 (2.8) DT S The drop weight to use in equivalent drop tests should be determined by the equation 2.9.

[hDT + (1 − L) d] We = W (2.9) (hDT + d)

where hDT is the drop height calculated with equation 2.8, d is the deﬂection under impact of the tire plus the vertical component of the axle travel relative to the drop mass, L is the Lift to weight 6 ratio , W is the aircraft landing weight or the static load on the main gear WM or the static load on the nose gear WN if the drop tests are done considering assembly units. The limit inertia load factor n, applied to the center of gravity c.g., should be determined by the equation 2.10. W L n = n e + (2.10) j W W

where We and W are respectively the equivalent weight of drop test and the aircraft landing weight (or the static weight on the main gear or nose gear, if the drop test is done with equivalent assembly units), and nj is the load factor recorder in the drop test ((dv/dt)/g) plus 1. • Shock absorption Tests: the limit inertia load factor in 2.10 selected for the design should not be exceeded in energy absorption tests. The test must demonstrate the landing gear not to fail in a simulated descent velocity equal to 1.2 the selected rate of descent in equation 2.5, assuming the wing lift equal to the aircraft weight before the impact. • Tire rating Tests: Each tire should have a tire rating not exceeded by corresponding static ground reaction under the design maximum weight and critical position of the center of gravity.

4A typical value of the friction coeﬃcient on landing is 0.80. [18] 5A typical value of the time delay from the contact point on landing and the attainment of the maximum vertical load is 0.2s. [19] 6The Lift to weight ratio should be less than 0.667 according to FAR 23.725.[18] 9 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

2.3 State of the art of Landing Gear for similar aircraft applications

The present section illustrates the state of the art of landing gear for different aircraft applications, from small radio commanded planes and Unmanned Aerial Vehicles to ultralight aircraft certified with FAR regulations. Some of these aircrafts, with weight and performance similarities to FTV 7 and FTV 16.5, have been considered as a reference to select the off-the-shelf components and to make design choices. Figure 2.4 shows the weight and speed range of the aircraft applications considered, and where the two blended wing body FTVs are located among four of representative aircrafts for each category.

Figure 2.4. Landing gear application for several UAVs

2.3.1 References for FTV7% The 7% scale of the new generation blended wing body aircraft is located in between RC commanded hobby airplanes and small jet UAVs. The Skymaster RC jets represent the framework of tires and wheels oﬀ-the-shelf solutions that can be considered for aircraft with gross weight and speeds similar to FTV7. The tires for this aircraft application, are designed to withstand 40 m/s landing and take-oﬀ speed. [20] The tricycle landing gear of the Penguin C has been considered as a reference for designing the custom leaf strut and the nose gear strut for FTV7. The suspension of the landing gear is essentially attributed to the elastic behaviour of the bended aluminum main strut and to the simple spring elements inside the nose gear strut. The steering system is implemented using a servo linkage mechanism with self-centered caster nose gear strut. [21] The aircraft, provided at the Center for Aerospace Research, has been considered even as a reference for the ground mock-ups necessary to tune the control parameters for a fully operative autopilot on an aircraft with undercarriages.

2.3.2 References for FTV16.5% The larger flight test scale demonstrator can be considered part of the experimental UAV category, where the use of specific custom components, designed for the required mission, becomes highly discouraged due to the high cost and time required. For big aircraft scales, it is necessary to add some form of shock absorber to better manage dynamic load conditions during take-off and landing, and provide the aircraft with adequate steering and braking system. The general guideline for this category of aircraft is to search for low cost components off the shelf, as implemented on the landing gear of the X-48 Boeing Nasa, characterized by shock absorbers and brake system from bike applications. [17] As for tire and wheel solutions, the choice among hobby aircraft applications gets rapidly limited above 150mm diameter. Therefore a reference is represented by nose or tailwheel applications for small ultralight aircraft, such as the Thatcher CX4, even if in most of the cases they are not rated for high speed due to the low speed requirement for their application.

10 Chapter 3

Landing Gear Design and Development for FTV7%

The present chapter shows the detailed experimental design and development of the landing gear system for the 7% scale of the blended wing body new generation aircraft. Since the beginning it has been fundamental to consider the behaviour of the aircraft, already flight tested without undercarriages at the Center for Aerospace Research, during the critical phases that influence the landing gear performances. An initial description of requirement list, useful to define the framework in which the landing gear designer should operate, is followed by the explanation of all the fundamental design steps for both main and nose undercarriages. All the peculiarities and challenges, linked to the production phase and integration of the full system with the aircraft, are presented as well as the ground testing, necessary to validate the design itself.

3.1 Requirement list

The first step required to start the design of a landing gear for an already built aircraft is to define a list of core requirements, including the constraints imposed by the systems that cannot be significantly modified. The landing gear requirements for FTV7 have been determined considering all the general guidelines described in chapter 2 and considerations resulted from continuous discussions with the stakeholders. Five categories of essential requirements have been selected to guide the iterative design process: general requirements for the landing gear, aircraft performance requirements, scaled-geometry requirements and specific requirements for the main gear and nose gear. Due to the interdisciplinary and iterative nature of the design for aircraft systems, most of the requirements are interdependent and for this reason they have been organized in an Excel sheet, then imported in Matlab, using a parametric design approach. [22] In this phase, it has been useful to consider the real behaviour of the aircraft, already flight tested without undercarriages several times at the Center for Aerospace research, during landing approach and touch-down1, in order to evaluate in which ranges of impact the aircraft is used to operate. For this purpose, the StatData feature of the software Flight test data processing, developed by Sean Bazzocchi and Jenner Richards at the Center for Aerospace research, shown in figure 3.1, has been helpful to compute the average and maximum values of the parameters related to the impact, that are the rate of descent, landing speed and vertical deceleration.

1The phase of take-oﬀ did not inﬂuence the design of the landing gear since the aircraft was catapulted into the air, and so the resulting accelerations and oscillations were related to the shooting of the catapult. 11 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 3.1. Interface of the StatData feature of the Flight Test Data Processing software

3.1.1 General requirements

The most important general requirements of the landing gear design for FTV7 are summarized in table 3.1.

Table 3.1. General requirements for the landing gear design for FTV7

# General Requirements Value/Type 1 The type of landing gear should be as cheap and simple as possible Tricycle, Fixed 2 The aircraft should be able to perform in concrete not well prepared Merrit airstrip runways 3 The aircraft should have an adequate system to facilitate ground ope- Steering system rations 4 The braking system is not required but can be implemented to test the Electro/magnetic braking control parameters with autopilot brakes 5 The landing gear should be the same for diﬀerent aircraft conﬁgurations: FTV7-tailed and FTV7-no-tailed

As described in section 2.1, one of the most important factors that determines the type of the landing gear is the configuration of the aircraft. The flight test vehicle considered, representation of a new blended wing body concept aircraft, is developed in two different configurations, with tail and without tail. The main goal for the landing gear designer, in this specific case, is to guarantee the same landing gear between the two configurations in order to reduce the cost and complexity of the design.

3.1.2 Aircraft and Performance requirements

The requirement related to the aircraft and performance are shown in table 3.2. The reason of the large excursion of the center of gravity (C.G.) for the take-off and landing operations comes from consideration of the full scale aircraft: the blended wing body, with and without tail, is supposed to fly in non stable regimes and this implies the necessity of shifting the position of the C.G. to trim and balance the aircraft in all the flight phases. The data values of performance used for design considerations are the one obtained from the previous flight tests of FTV7 without landing gear, using the statistical software described in section 3.1. The 12 3 – Landing Gear Design and Development for FTV7% maximum values, especially as for the accelerations registered by the 50 Hz log file2, are referred to the maximum of the peak values for each flight, that in most of the cases are developed in a short timespan like 0.1 seconds. For this reason, they are useful to have a full picture of all the possible worst case scenarios experienced by the aircraft on landing, but nor directly applicable as design parameters, better represented by average values.

Table 3.2. Aircraft and performance requirement for FTV7

# Aircraft performance requirements Value/Type 1 Maximum Take-Off weight of FTV7 with tail 13.6 kg 2 Percentage of the center of gravity position for take-off and landing with 56 ÷ 66% respect to the mean aerodynamic chord (M.A.C.) 3 Average Landing Speed from previous flight tests 26.80 m/s 4 Maximum Landing Speed from previous flight tests 30 m/s 5 Average Rate of descent from previous flight tests 1.05 m/s 6 Maximum Rate of descent from previous flight tests 1.4 m/s 7 Average landing deceleration from previous flight tests 1.1G 8 Maximum landing deceleration from previous flight tests 4.3G

3.1.3 Geometry requirements The geometry of the landing gear should replicate as close as possible the limits imposed by the full scale aircraft. The most relevant geometry requirements are shown in table 3.3. The position of the nose gear is fixed to the scaled value from the full scale aircraft, whilst the location of the main gear can be moved depending on the center of gravity position. The design choice for the main gear position has been done in order to have a load distribution of 90% on the main gear and 10%, verifying that each location respects the take-off rotation and tip-back criteria introduced in section 2.1. The vertical behaviour of the landing gear is influenced by the definition of the height of the landing gear in all the phases that involves the landing: fully extended, static, fully compressed position where respectively no loads, MTOW corresponding loads and impact loads are applied. The static position is determined by the ground clearance criteria while the fully compressed position should replicate possibly the scaled geometry.

Table 3.3. Geometry requirements for FTV7

# Geometry Requirements Value/Type 1 Position of the nose landing gear with respect to the front fuselage 275 mm station 2 Load ratio on the landing gear (nose, main) due to the C.G. position 10%-90% 3 Main gear track measured at the outboard wheel position, in static 589 mm position of the landing gear 4 Vertical height of the gear in extended position with respect to the water 174.89 mm line (W.L.) reference 5 Scaled compression of the landing gear from its extended position 27.26 mm

3.1.4 Main and Nose gear requirements Speciﬁc requirements for main and nose gear are presented in table 3.4. In a tricycle conﬁguration the main and nose gears have the same height, so the aircraft is leveled on ground even if the main gear tends to have bigger wheels. The maximum allowed size with respect to the scaled diameter, has been selected to be 25% larger, after some considerations that consider the wheel’s market among RC planes

2The log ﬁle at 50 Hz is installed on the aircraft and records all the ﬂight data used from the autopilot every 0.02 seconds. 13 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators for comparable applications to FTV7 and maximum allowed drag penalties due to the increased frontal area, in comparison with the values expected for the full scale aircraft. [3]

Table 3.4. Speciﬁc requirements for the main and nose gear of FTV7

# Requirement Value/Type 1 Scaled tire size for the main landing gear 2.5 inches 2 Scaled tire size for the nose landing gear 1.5 inches 3 Maximum larger percentage for tire sizes 25% 4 Number of wheel assemblies per each strut 1

All the stated requirements have been imported in a developed Matlab® parameter calculator software, described in Appendix A, in order to deﬁne the resulting feasible geometry and preliminary layout of the aircraft with undercarriages.

3.2 Design and development of the main gear assembly

The present section describes the process of design and integration of the main landing gear for the 7% scale of flight test vehicles, in both tailed and no-tail configurations. The design has moved rapidly from scratch to solid CAD modeling in SolidWorks and the final modeled layout, with all the needed components, is shown in figure 3.2. The choice of the off-the-shelf components and the custom design of the leaf strut, as well as the features of the main landing gear, are explained in the following subsections.

Figure 3.2. CAD assembly of the designed main landing gear for FTV7

3.2.1 Main gear features

The main landing gear has been designed in order to support all the maximum take-off weight of FTV7, that simulates the case of not perfect three points landing. The initial approach has been to possibly search for off-the-shelf components in order not to affect drastically the cost and weight budget of the Bombardier-Quaternion project. The final assembly of the main landing gear, shown in figure 3.2 consists of wheels derived from the Skymaster radio commanded jets, introduced in section 2.3 and custom components for the leaf strut and the parts necessary for the integration with the airframe. The most interesting feature of the main landing gear is the possibility to move easily its location depending on the center of gravity position. This capability has been required in order to test the working operation of the autopilot during take-off and landing, varying the longitudinal C.G. envelope from 56% to 66% of the mean aerodynamic chord (M.A.C.). [2] The main landing gear is attached to the flat part of the fuselage, using apposite shifting slots, and it is in part covered by the redesigned internal foam of the belly pan. The structural stiffness of the main landing gear is carried by the elastic behaviour of the strut leaf, whilst the damping of the system is developed in the most part by the pure inertial motion of the aircraft and is slightly increased by using additional rubber cushion sheets in between the landing gear attachments and the skin of the airframe. 14 3 – Landing Gear Design and Development for FTV7%

3.2.2 Wheels and Tires for the main gear The wheel assembly is the component responsible to allow the aircraft to have contact with the ground. For small aircraft applications it is usually a nylon or metal alloy wheel, that houses the bead seat for a foam or rubber tire. Several options have been considered in a preliminary market research, all suitable for the size and weight of FTV7. The most interesting solutions, shown in ﬁgures 3.3 and 3.4, have been purchased and characterized.

Figure 3.3. Option 1: Sky- Figure 3.4. Option 2: Turnigy master wheels HK wheels The final decision has been done based on an analytical hierarchical approach with weighted features decided in relation to the potential application to FTV7, that is summarized in table 3.5. Since the aircraft is supposed to operate in runways with a considerable length, it is expected to be stopped without the need of a braking system. In addition, no fairing have been designed and so the parasite drag caused by the rolling frontal area of the wheel is expected to provide a backward force that can help to stop the aircraft. The option 1 has been selected because it privileges the weight saving and simplicity. However if the airplane operation will move to short and unprepared airstrip, a braking system should be essential, and so the electro-brakes of the option 2 would be preferable. The electro-magnetic brakes are easy to control and to integrate with the overall system: when the current is passed through the coil, the magnets into the outer rim of the wheel react with the magnetic field generated and slow the wheel down, thus causing the aircraft to brake. As there are no friction parts in the wheels, they cannot wear out so reliability and longevity should never be an issue. This solution could potentially be used even to brake progressively, avoiding the wheels to lock-up, and differentially between the left and right wheel, helping the aircraft to steer. [23]

Table 3.5. Analytical Hierarchical table for tire and wheel selection of FTV7 main landing gear.

Option 1 Option 2 Features Value Rank Value Rank Weight (50%) 83 gr 10 192 gr 5 Size (20%) 2.7"x0.8" 8 3"x0.8" 6 Brakes (5%) Extra 2 Electro-Magnetic 8 Application (25%) Model: 15 ÷ 20 kg 8 Model: 10 ÷ 15 kg 6 Total 100% 8.7 5.6

3.2.3 Design of the main gear strut The design and development of the main gear strut started from the evaluation of possible oﬀ the shelf components and moved towards a custom solution, including CAD design, production and testing. This subsection describes in detail the design and production processes performed, as well as the simulation and ground tests done to characterize the component and verify the design decisions made.

Off-the-shelf solutions The impact loads for such a small scale of airplane can easily be manageable by a strut leaf in Aluminum or Carbon fiber composite. Several off the shelf components, designed for radio commanded small aircraft, in stock at the Center for Aerospace Research, have been tested and evaluated. For the sake of brevity only the structural tests of the Aluminum 2024 leaf with thickness 1/8” are presented, since it has the most suitable geometry in comparison with the requirements stated in subsection 3.1.3. The landing scenario 15 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators simulated by structural simulations and ground drop test, shown in figures 3.5 and 3.6, represents an impact of 3G and all the equivalent MTOW weight acting on the strut. The landing condition has been simulated applying the procedures described by FAR regulation, introduced in subsection 2.2.2, and considerations from the previous flight tests without undercarriages. As result from both simulations and ground tests, the component has been deformed in plastic field, therefore it has not resulted stiff enough to be used for the FTV7 case design.

Figure 3.5. SW structural simulations of the Figure 3.6. Drop Test of the Al2024 1/8" Al2024-1/8" leaf strut performed at CfAR

Due to the difficulty to find a leaf strut among radio commanded applications, with the exact geometry required for the fully extended and fully compressed positions, a custom solution has been planned to be designed and developed. A solid model with the same geometry as the off-the-shelf component, has been evaluated in SolidWorks with different Aluminum grades and thickness, in order to guide the choice of material for the custom design. The simulations suggested to improve the grade of Aluminum to ERGAL 7075 and the thickness of almost 1.5 mm, without increase considerably the weight of the component.

CAD design of the custom leaf strut The design phase of the leaf strut has been a trade-oﬀ among the Coward methodology described in reference [8], the requirements stated in section 3.1 and the necessity to integrate the strut with the airframe of FTV7 without internal and external interference. The most important design parameters are illustrated in ﬁgure 3.7, where L is the equivalent geometric arm between the force on the wheel and the top strut bending radius, t is the thickness of the strut, αC , αs and θ are respectively the camber, sweep and bending angles of the strut, WR and WB are the evaluation of the strut width at the two bending radius locations.

Figure 3.7. Geometric parameters for the leaf strut design

The values of WR and WB can be calculated from the graph used in the Coward procedure [8], shown in ﬁgure 3.8. The root width dimension can be estimated entering in the graph with the beam width parameter b, calculated according to the experimental equation 3.1, valid for Aluminum alloy struts, 16 3 – Landing Gear Design and Development for FTV7%

3 where W is the aircraft weight and Sref is the reference wing area. The width at the bottom bending radius of the strut is assumed to be half of WR, and the thickness should be about t = WR/8.

Figure 3.8. Design taper para- Figure 3.9. Integration of the main gear with the meters for the leaf strut. [8] mid bay of the aircraft

0.01373W 1.5 b = 0.5 (3.1) LSref The θ angle is defined by the bending of the strut, necessary to guarantee the required vertical position of the landing gear. The wheel travel of the landing gear is then influenced by the camber angle αC , determined in an iterative process in order to have the wheel front section aligned with the vertical line when the strut is loaded in a 3G impact landing with maximum take-off weight. The sweep angles at the front an rear location of the strut leaf leg have been dictated by the integration of the component with the airframe, since the connection is supposed to be inside the mid bay of the aircraft, as shown in figure 3.9. In addition, a swept leaf strut is expected to have a better stress distribution during impact, since in most of the two points landing on the main gear with the nose gear clear from the runway, the ground reaction results to be transferred vertically to the strut leaf. The resulting design geometric strut parameters are organized in table 3.6.

Table 3.6. Geometric parameter determination of the custom leaf strut design

t θ αC αs−f αs−r L b WR WB 3/16" 31.33◦ 7◦ 9.4◦ 2.1◦ 6.55" 0.075" 3" 1.5"

Material procurement and production of the customized strut The first step of the production process has been the procurement of the material sheet in Aluminum 7075-T6, with thickness 3/16" and size 24"x24". The size has been selected in order to produce two sets of main gear strut for each aircraft FTV 7tailed and FTV 7no−tail and one extra spare part for testing. The flat pattern of the component has been obtained from the material sheet using abrasive waterjet cutting technique, selected because it is a cold process that makes no impact on the material being machined. At its most basics, a slurry of water, abrasive and air, flows from a pump with pressure between 60000 and 94000 psi, through plumbing with a cutting head of minimum kerf diameter around 0.035 inches. At the cutting head, a high speed air valve, allows the water to pass through the jewel orifice creating a supersonic waterjet stream, typically at Mach 3.0, able to cut the material without create heat affected zones (AHZ) typically caused by other machine processes that require for instance warping and clamping.[24]

The next step of production has been to obtain the ﬁnal geometry through the bending process of the water-cut parts. Appropriate bending brakes have been used to lock the part and load force around

3All the data in the beam width parameter equation are espressed in British imperial units. 17 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators the bending lines. The initial attempt to bend the part has been done as usual for the bending of lower Aluminum grades, with a cold process machine. As a result the component has started to crack at the location of the bending line, as shown in ﬁgure 3.10, due to the hardness of the 7075 Aluminum, the large thickness and sharp bending radius 1/2". For this reason, it has become necessary to anneal the bending lines by heating at temperature4 around 400 − 450◦C in order to bend properly the parts.[25] The result of the hot bended leaf struts is shown in ﬁgure 3.11.

Figure 3.10. Crack on the lower bending radius of the leaf strut, after the first bending Figure 3.11. Final bending hot process of machine process the main gear struts After machine processes it is important that the material recover ideally the original properties and so a typical heat treatment for the Aluminum 7075 has been applied, following the specifications ASM 2770- 2658 [26]. The leaf strut has been first heat treated at temperature 870◦F for 1 hour and 11 minutes, then it has been quenched for 7 seconds and finally the precipitation age for curing the material has been processed at the temperature of 250◦F for 23 hours and 10 minutes. The process, in terms of temperature with respect to the time intervals, is shown in figure 3.12.

Figure 3.12. Heat treatment of the leaf strut. [27]

3.2.4 Testing of the main gear

The Main Gear leaf strut has been tested ﬁrstly with ﬁnite element (FE) simulations in SolidWorks and then ground tested by means of Static tests to estimate the structural properties and Drop tests to verify the integrity of the component when all the aircraft weight is acting on the main gear in case of 3G impact landing.

4Aluminum 7075 should be formed at temperature not too close to the melting point that is 477 − 635◦C, since it would start to get brittle. 18 3 – Landing Gear Design and Development for FTV7%

Static tests for the main gear strut The theoretical linear stiﬀness can be calculated using the equation 3.2, described in the Coward methodology in [8], considering the Young Modulus of Aluminum 7075 equal to 10400 ksi.

3 E · WR N Kth = = 32.067 (3.2) 2 L 3 mm 96b · cos θ t The corresponding analytical displacement can by estimated considering one of the leg of the strut leaf to be approximated by a rectangular beam, using the equation 3.3.

3 3 FM · cos (θ) · Leq 4FM · cos (θ) · Leq zana = = 3 (3.3) 3EIeq E · beq · t

The beam approximation implies to consider equivalent geometric parameters to the ones of the real tapered strut described in figure 3.7. The equivalent beam parameter beq is defined as the average of the strut width at the root WR and bottom WB locations, whilst the equivalent length of the beam Leq has been derived assuming to have the same surface area between the real strut and the approximation. The other parameters are the same described in table 3.6. The same structural properties have been estimated using structural FE simulations in SolidWorks, with increasing weight, from 0 kg to 30 kg with 3 kg of each step. The boundary conditions used for the finite element simulations, that try to represent the reality as close as possible, are bearing constraints at the attachment points with the airframe and remote loads at the axle location of each leg. The geometry is modeled with 10557 solid tria mesh elements with overall size length 4.56 mm, capable to predict the bending behaviour of the material.

Figure 3.13. Displacement of the main gar strut with a static load corresponding to the MTOW of FTV 7

For sake of brevity, only the loading condition corresponding to the static load distribution of the maximum take-off weight of FTV7 is presented as follows. The corresponding maximum computational displacement, shown in figure 3.13, is 3.4 mm registered at the same location of application of the remote loads. The tests, with the same increasing weight, have been replicated in reality at the shop of the Center for Aerospace Research, by using a simple drop leverage test rig and digital caliper measurement for the displacement. The measurements have been taken between the ground reference and a reference on the Test rig, in both loading and unloading case. The material behaviour has been proved to be elastic, since 19 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators the outboard track between the axles at the end of the tests has remained the same of the unloaded case. The experimental maximum displacement, under equivalent maximum take off weight load, is proved to be similar to the computational one (3 mm) , as shown in figures 3.14 and 3.15.

Figure 3.15. Test rig with static loading cor- Figure 3.14. Test rig with no loading applied responding to the MTOW of the aircraft

The experimental (zexp), computational (zcomp) and analytical (zana) displacements resulting from the increasing static weight up to 30 kg are summarized in table 3.7.

Table 3.7. Resulting displacements from Static Tests with increasing weight: (0 : 3 : 30kg)

# Test Load [kg] zexp[mm] zcomp[mm] zana[mm] Loading Unloading 1 3 0.31 0.2 0.76 0.92 2 6 0.97 1.38 1.53 1.84 3 9 1.76 2.01 2.29 2.76 4 12 2.55 2.69 3.06 3.68 5 15 3.09 3.50 3.82 4.59 6 18 4.45 4.57 4.58 5.52 7 21 5.64 5.75 5.35 6.44 8 24 6.20 6.40 6.12 7.36 9 27 7 7.2 6.88 8.28 10 30 7.75 7.90 7.64 9.20

The difference in some values between loading and unloading condition can be justified due to the difficulty to take the measurements each time between the same points and to the not exactly sand equivalent weight used for each test. Due to the measurement uncertainty an average between the two loading and unloading cases can represent a better estimation of the real behaviour.

The results of table 3.7 are shown graphically in ﬁgure 3.16, where the displacements are approximated with polynomial trend for the experimental case and linear one for the analytical and computational cases. In the reality the strut stiﬀness, represented by the slope of the curve, progressively decreases with the increasing of the load demonstrating that the real behaviour is not linear.

A double least squares approximation can be used to linearize the real stiffness properties of the strut, in two different ranges of loads, denominated 1st slope (with load less than 170 N) and 2nd slope (with load more than 170 N). In the first load range the real linearized stiffness results to be well estimated by the computational analysis and much more stiff respect to the analytical model: this is probably due to the assumption used to approximate the tapered strut leaf as a single piece of rectangular beam. 20 3 – Landing Gear Design and Development for FTV7%

Figure 3.16. Load displacement curve for the main gear leaf strut

In the second range of loads, the strut progressively loses stiﬀness due the geometric non linearities of the component: in this case the structural properties are better estimated by the analytical approach with respect to the computational one, were inconsistencies are registered, probably due to the not uniform distribution of vertical load on the wheel axle when the load increases. The resulting values of linearized stiﬀness in all the cases analyzed, with the respective error eR respect to the real linearized experimental behaviour, are described in table 3.8.

Table 3.8. Linearized stiﬀness from experimental, analytical and computational tests

Exp 1st slope Exp 2nd slope Analytical Computational K [N/mm] 40 30 32.067 38.58 eR 1st slope X X −20% −4% eR 2nd slope X X +6.5% +22.2%

Impact drop test of the main gear strut

The Drop tests for the main gear have been performed according to FAR 23 regulations, described in subsection 2.2.2, and considering a total weight corresponding to the MTOW of the aircraft in tailed configuration, the worst case scenario of a two points landing on the main gear. The tests have been performed in the shop of the Center for Aerospace Research using the same test rig of the static tests, shown in figure 3.14. Different landing cases have been simulated, as shown in table 3.9, considering as variable input the vertical rate of descent and the drop height calculated according to equation 2.8. The first three cases reproduce typical landings experienced by the previous flight tests of FTV7 without undercarriages, with an average impact acceleration that is in between 1.8 and 2.4 the gravity. The last landing case, is the design landing corresponding to an average impact acceleration of 3G: the resulting displacement (27 mm) is quite similar to the scaled down compression of the main landing gear, stated in the requirement of subsection 3.1.3, demonstrating that the main landing gear has been designed in structural similitude with the full scale aircraft. 21 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Table 3.9. Drop test setting and results for diﬀerent simulated landing cases

Sink Drop Compression Time to Impact acce- Speed height stop leration Landing 1 m/s 51 mm 14 mm 28 ms 1.8G Case 1 Landing 1.16 m/s 69 mm 16.6 mm 28.6 ms 2G Case 2 Landing 1.5 m/s 114 mm 24 mm 32 ms 2.4G Case 3 Landing 1.81 m/s 167 mm 27 mm 29.8 ms 3G Case 4

3.3 Design and development of the nose gear assembly

The second, but not less important, component of a tricycle landing gear is the nose (often referred as front) undercarriage assembly. The design, procurement and production of all the nose gear components, has followed the same general approach of the main gear, evaluating and testing oﬀ the shelf components whenever possible and developing custom parts to enable the operation and integration with the aircraft. The CAD of the front gear system is shown in ﬁgure 3.17. All the relevant components shown, are discussed in detail in the following subsections.

Figure 3.17. CAD assembly of the designed nose gear for FTV7

3.3.1 Nose gear features

The nose gear functions for FTV7 are basically accomplished by an assembly of aluminum parts, helical springs inside the nose strut and rubber tire, providing the required support, stiffness and damping properties. The steering ground maneuverability of the aircraft is carry out by an electric servo that transfer the electric power in motion to the front wheel thanks to a shaft-coupler, horns and mechanical linkage. Custom made components have been designed and manufactured to attach the nose landing gear to the airframe and provide the needed kinematics and mechanics for the steering maneuvers. The fixed location of the front wheel is defined by the document for requirement specifications in reference [2] and, as a consequence, the nose landing gear results to be loaded vertically by 10% of the the total aircraft weight. 22 3 – Landing Gear Design and Development for FTV7%

3.3.2 Component oﬀ-the-shelf for the nose gear

The oﬀ-the-shelf components used for the nose landing gear are essentially the nose strut and wheel assembly, both selected after evaluation and testing of several solutions from RC jet applications. Due to the impossibility to fully integrate the components, designed for other speciﬁc small planes, in most of the cases it has been necessary to edit the parts in order to increase the reliability and the integration with the airframe.

Strut

The selected nose gear strut, in Aluminum 6061, has been designed as main gear leg for scale models as the Warbird jets. It consists of three different parts: piston, cylinder to house the compression springs and a double scissor to connect the two structural parts and allow the motion in loading condition. To demonstrate the possibility of using that solution for the nose landing gear of FTV7, it has been drop tested according to the regulations already used for the main gear testing, stated in section 2.2.2, with an equivalent weight corresponding to 10% of the aircraft maximum take-off weight. The stiffness properties are attribute to the internal springs, adjusted in order to have a double compression behaviour depending on soft or hard landings. The static position, with static load applied on the strut, and bottoming out of the strut are shown respectively in figures 3.18 and 3.19.

Figure 3.18. Nose strut with sta- Figure 3.19. Nose strut with botto- tic load corresponding to the aircraft ming force, corresponding to the im- MTOW pact in hard landings Due to the necessity to respect the height constraints stated in subsection 3.1.3 and to integrate the component with the internal front bay’s space of the aircraft, it has been necessary to edit the cylinder of the strut reducing the length of the leg by 10 mm. An additional aluminum bushing has been manufactured and insert to reinforce the strut where there is the accommodation for the wheel axle.

Wheel assembly

Aluminum wheels and rubber tires, designed for mid-sized turbine RC jets, have been selected to be applied for the nose landing gear. The wheel axles have been provided with stainless steel ﬂanged ball bearings with rating ABEC-4 and capable to withstand easily the load acting on the tire.

3.3.3 Design of additional custom components

Additional components for the nose gear integration and for the implementation of the steering system have been designed and manufactured, using the mechanical machines provided in the shop of the Center for Aerospace Research and in the mechanical laboratory of the University of Victoria. The following paragraphs show the CAD design and production process of the needed custom components, illustrated in ﬁgure 3.20. 23 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 3.20. CAD design of the custom components for the nose gear of FTV7

Attachment plates The Aluminum 7075 attachment plate has been designed in order to easily integrate the nose gear with the airframe and provide the needed alignment for the shaft of the steering system. The design goal has been to install the nose landing gear at the same location where the launcher hook for the catapult system was bolted, in order to reduce the amount of extra connections to the aircraft. The component, as shown in ﬁgure 3.20, has four counter-sunk clearance hole for the connection 6-32 screws with the airframe, two tap screw holes for the spring screws 2-56, and four tap screw holes 6-32 for the attachment of the strut cylinder. Two additional buttonholes have been placed in the central area of the component in order to release material where it is not for structural support. The component has been manufactured using the same material of the main gear leaf strut and the same waterjet technique, already described in subsection 3.2.3. An extra carbon ﬁber plate with thickness 1/16", has been used in between the skin of the aircraft and the Aluminum plate, with the scope to mate the attachment plate to the aircraft as uniformly as possible, in the same location where the launcher hook was installed and so where the skin surface was not perfectly smooth.

Horns for the steering system The steering motion is transferred from an electric servo, positioned inside the front bay, to the nose strut, through a servo horn and a strut horn, shown in figure 3.20, respectively clamped to the servo shaft and to the nose gear piston strut. The design of the two parts has been iterative, in order to facilitate the mounting procedure in a limited space as the front bay of the aircraft and to reduce the amount of torsion load, needed on the screws to clamp the horn to the shaft and strut cylinder: the clamping procedure is simplified thanks to the design of rectangular houses with edges parallel to the faces of the hex nut, resulting in no rolling tendency of the nut, when subject to torsion. The final design of the strut horn has a slot in the internal diameter, in order to allow the screw that holds the nose strut piston to pass through when the load on the wheel increases. Another feature of this part is the possibility to be connected, with extension springs, directly to the aluminum plate, maintaining the wheel self-centered in the mid-line axis even in case of servo or linkage failures.

Cylinder for the nose gear strut integration The nose gear strut is connected to the airframe by using an Aluminum 2024 cylinder that passes through the skin of the aircraft and is bolted to the attachment plate, as shown on the right of the ﬁgure 3.20. The translational and rotational motion of the strut, along longitudinal and transversal direction, is prevented thanks to the use of delrin sleeves around the strut, while the vertical location is hold by a lock-tided screw on the top of the cylinder, rotating on a delrin washer when the strut is steered.

3.3.4 Design of the steering system The steering system has been designed to transfer the motion from an electric servo to the front wheel assembly thanks to a mechanical linkage. The steering ability of an aircraft is measured in terms of the 24 3 – Landing Gear Design and Development for FTV7% maximum steering angle and consequent turning radius, necessary to perform a 180◦ turn within the width limit of the runway. [28] This requirement is less restrictive for small UAVs, where usually the runway of application is much more wide than the actual wing span of the aircraft. The maximum steering angle of the nose gear for FTV7 has been assumed to be ψ = ±25◦, the same limits of the rudder control already tuned in the autopilot. The basic geometry and kinematics of the steering system is shown in ﬁgure 3.21. The turning capability of the aircraft is evaluated based on the turning radius Rturn corresponding to the maximum steering angle ψ, as described in equation 3.4 and shown in ﬁgure 3.22.

Figure 3.21. Geometry and kinematics of the Figure 3.22. Turning radius for the steering system steering system

The parameter lM−N is the wheelbase and ψ is the maximum steering angle, measured with respect to the longitudinal centerline of the airplane.

l R = M−N (3.4) turn cos(90◦ − ψ)

The three bar mechanism of the steering system has been designed in such a way to ensure the maximum steering angle when the servo arm has an inclination of β = ±30◦, in order to not have interferences with other internal components. The resulting arm ratio is deﬁned as in the equation 3.5.

h tan (ψ) strut = (3.5) hservo tan (β)

All the components used in the steering system, including standard parts (electronic servo and mechanical bar) and custom parts are shown in figures 3.23 and 3.24. Due to internal space constraints, only the servo is allocated inside the nose bay of the airframe, whilst the other components of the steering system are positioned under the flat belly of the aircraft. The servo HV6130-MKS has been selected among RC airplane series applications and it results to be the best compromise among size, weight and torque required to steer the wheel. It has an aluminum case and metal gear, capable to transfer electric power in mechanical motion with very low slug and adjustable speed. [29] The attachment support for the servo has been designed to be produced with 3D printing technique in versatile plastic. The part has an inclined surface with branched ribs in order to reinforce the support and facilitate the bonding with the front wall of the nose bay of the airframe. On the other side, it allows the mating with the servo replicating its detailed shape. An adequate clearance hole allows the servo gear to rotate and move the shaft without interferences with the support. The motion is transferred to the other components under the flat belly thanks to a shaft connected to the servo with a coupler-screw mechanism and passing inside a delrin bushing through the airframe. The mechanical transmission is granted by connecting the linkage to both servo shaft horn and nose strut horn. 25 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 3.23. Steering servo components Figure 3.24. Mechanical steering system components Two additional extension springs are used as backup system in case of failure of the servo mechanism, that can occur during hard landing or high irregularities of the runway. The springs connect the strut horn to the attachment plate and they have been selected in order to allow the maximum design steering angle ψ. The two springs work oppositely in extension and compression, maintaining the front wheel in the neutral position when no motion is applied form the steering system.

The spring stiffness has been selected considering an equilibrium state about the center of the wheel, illustrated in figure 3.21, when the limit positions for the steering system are reached with steering angles ◦ ◦ equal to ψ = 25 and β = 30 . The resulting stiffness Kspring required can be calculated according to the moment equilibrium equation 3.6.

M · hstrut = Fspringhspring cos (γ) (3.6) hservo

The stiﬀness required is dependent on the torque applied on the servo M, the angle γ between the spring axis and vertical position, the spring arm hspring about the center of rotation and the maximum available spring travel sspring. Applying the equation 3.5 to the moment equation 3.6, yields the equation 3.7, that allows to select the extension springs with a feasible combination of stiﬀness and stroke.

tan (ψ) 1 Ks = M · (3.7) tan (β) hspring cos (γ) · sspring

3.3.5 Testing of the nose gear

The nose gear strut, already drop tested as discussed in subsection 3.3.2, has been characterized in terms of structural properties, through static tests with increasing weight between zero and the maximum value that corresponds to the bottoming out of the internal spring system.

The tests have been performed using a test rig mounted on the optical table provided in the shop of the Center for Aerospace Research, as shown on the left of ﬁgure 3.25. The loads on the strut have been applied with a screw mechanism and regulated by using a digital force gauge, while the resulting displacements on the strut have been registered using a digital caliper.

The stiﬀness of the nose gear strut is essentially carried by the internal spring system, regulated with a double acting spring in order to perform a progressive compression, as shown in the graph load/deformation shown on the right of ﬁgure 3.25. 26 3 – Landing Gear Design and Development for FTV7%

Figure 3.25. Static test for the characterization of the nose gear strut: the test rig is represented on the left, the resulting load/deformation curve on the right.

As expected, the stiffness property of the nose gear is represented by two different linearized slopes, corresponding to the action of a single spring or double springs. The spring lengths have been selected in order to attribute the load on the first spring until the acceleration reaches values of three times the gravity. As a consequence the nose gear can have a soft performance, to smooth the vibrations induced by taxiing over unprepared runways, and provide additional stiffness during hard impact landing. The actual stiffness coefficients of the nose gear for the two different soft and hard behaviour, are summarized in table 3.10.

Table 3.10. Linearized stiﬀness coeﬃcients for the nose strut during soft and hard performances

Soft performance Hard performance KN 3.9 N/mm 6.5 N/mm

3.4 Integration of the landing gear with FTV7

The present section illustrates the process of integration of the landing gear subsystems with two airframes of FTV7, respectively built for tailed and no-tail configurations. Both nose gear and main gear systems are attached to the flat belly of the aircraft through specific structural components inside the front and middle bay of the airframe.

3.4.1 Integration of the main gear Main Gear location The main landing gear has been placed at different locations with respect to the center of gravity position, as already discussed in subsection 3.3.2. Every location of the main gear meets the tip-back and take-off rotation requirements and ensure a load distribution of 90% on the main gear and 10% on the nose gear. The skin of the airframe has been provided with slots allowing an easy change of the main gear position before each flight test, whenever required. The main landing gear positions with respect to the fuselage front station (F.S.), depending on the center of gravity variation with respect to the reference mean aerodynamic chord length, are shown in table 3.11. 27 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Table 3.11. Main landing gear position with respect to the Fuselage Station

C.G. % w.r.t. MAC C.G. position [mm] Main gear position 56% 749 800.17 57% 755.66 807.57 58% 762.32 814.97 59% 768.98 822.37 60% 775.64 829.77 61% 782.30 837.17 62% 788.96 844.17 63% 795.62 851.97 64% 802.28 859.37

Main landing gear attachment to the airframe The flat part of the leaf has been designed in order to increase the capability of the strut to withstand the stress concentration, induced by the bending moments that are developed by the vertical loads on the wheels. In fact, the bending capabilities of a structure increase if the connection points are distanced, since the resulting stress distribution is more uniform, as demonstrated from the structural simulations performed and discussed in subsection 3.2.4. Six attachment points for bolts or rivet nuts connections are used in a three plate junction, that includes the leaf strut, the sandwich skin of the airframe and two internal attachment plates inside the middle bay of the aircraft, shown on the right of figure 3.26. The internal plates, in Aluminum 2024 with 1/16" thickness, are used not to overload the skin of the aircraft. Their function is fulfilled where the washers and bolts are acting and so a symmetric shaped cut-out has been designed to save weight and reduce possible interferences with internal components. The final design of the flat surface includes three buttonholes in the areas where the material is low stressed in order to reduce the weight of the landing gear and also facilitate the integration with the airframe and the electronic speed controllers (ESC), shown on the left of figure 3.26. The latter cannot be moved from their initial location because they need to be ventilated for all the duration of the flight, and for this reason they have been mounted over the main leaf strut. The damping capabilities to attenuate impact loads on main landing gear, have been improved thanks to rubber cushions, 60A in the durometer scale, located in between all the structural plates.

Figure 3.26. External and internal integration of the main landing gear with the airframe

3.4.2 Integration of the nose gear The nose gear assembly is fixed to the flat belly of the airframe, through the same bolt holes used for the launcher hook catapulting system of the aircraft without undercarriages. The first step has been to smooth the flat belly with sand paper P100 and sterilize the surface with alcool, in order to ensure a mating as large as possible. Then the carbon fiber plate has been applied and the attachment plate in Aluminum 7075 has been bonded on, using epoxy adhesive EA-608. After a full curing time of 24 hours, the attachment plate has been bolted in the four screw holes and all the assembly has finally been mounted, as shown on the left of figure 3.27. The same epoxy has been used to attach the servo 3D support to the inclined front wall of the nose bay. The nose gear cylinder passes through the skin of the aircraft and is located between the battery 28 3 – Landing Gear Design and Development for FTV7% packs that need to be there for trim and balance purpose. The final integration of the nose gear with the internal front bay is shown on the right of figure 3.27.

Figure 3.27. Nose landing gear external and internal integration with the airframe

3.4.3 Belly Pan design

The landing gear parts, including all the fasteners used to attach the component to the airframe, have been covered by a foam belly pan that has basically two functions: streamline the ﬂow on the bottom part of the fuselage and provide additional safety in case of emergency belly landing consequent to a failure of the landing gear. The foam belly pan has been provided with suﬃcient internal room to allow the shifting of the main landing gear and not to interfere with the steering system mechanism.

3.5 Weight, cost and conclusions

The weight and cost distributions, relative to one manufactured landing gear set for FTV7, are shown in the charts in figure 3.28. As already largely discussed, the main goal of the landing gear design, has been to keep as low as possible the weight and cost increasing respect to the original budget of the aircraft. The weight distribution results from the mass evaluation of all the components used for the landing gear. The total weight of the two undercarriages results 5% of the maximum take-off weight for the tailed configuration of FTV7, in line with the general mass distribution of landing gear for comparable aircraft. [30] The weight is distributed as 74% for the main gear, mostly due to the beefy leaf strut, and 26% for the nose gear, in majority represented by the nose gear strut and wheel. In order not to highly affect the mass properties and inertias of the aircraft, a weight reduction analysis that considers all the bill of materials and components used into the aircraft, has been performed in conjunction with the other responsible CfAR teams of the aircraft. The most candidate components to be revised in order to allow an integration of the landing gear, without altering significantly the weight of the aircraft, has resulted to be relative to the batteries and parts used for the previous catapult take-off launches. The reduced capacity of the batteries and the removal of all the components correlated to the catapult launch system, has allowed to integrate the landing gear without change the previous MTOW estimations. The cost distribution results from an economic analysis using the PO pipeline orders in reference [31]. The cost of the off-the-shelf components is relatively low in comparison with the production of custom components, even though extra time has been required to edit the standard parts and facilitate the integration with the overall landing gear design. 29 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 3.28. Weight and Cost distributions for the production of one landing gear set for FTV7

Conclusions The aircraft FTV7 in both tailed and tailess configurations, with undercarriages installed, are now ready to take-off and land for the first time on a dedicated airstrip. Before a flight test campaign, it is necessary to prepare all the settings and parameters for the autopilot, which is expected to have control of the aircraft from the initial taxiing until landing. The next chapter will introduce all the required ground mock-ups to prepare the aircraft with undercarriages for a flight test campaign, including the planning of all the activities connected to the landing gear analysis during the most critical phases of flight.

30 Chapter 4

Ground mock-ups and ﬂight test planning for FTV7%

Once the undercarriages have been integrated with the airframe, the FTVs are ready to be subjected to ground mock-ups in order to estimate the expected control parameters to tune the autopilot. The first part of the present chapter describes all the necessary steps to prepare the aircraft for ground and flight tests, with full commands imposed by the autopilot. Then the ground testing (taxiing and take-off run), performed in the airstrip where the aircraft is suppose to take-off for the flight tests, is presented mentioning to the necessary improvements suggested by the conditions of the runway. The last section of the chapter describes the flight test plan and introduces the process of flight test data analysis using the Flight Data Post-Processing software developed by Sean Bazzocchi and Jenner Richards at the Center for Aerospace Research.

4.1 Preparation of the aircraft for ground and ﬂight testing

Since all the phases of ground and flight testing are supposed to be controlled by the autopilot, specific ground mock-ups are required in order to estimate mass and inertia properties of the aircraft and the necessary control parameters for take-off run and steering.

4.1.1 Ground mock-ups Center of gravity trimming The first operation is to trim the aircraft for each position of the main landing gear and so for center of gravity in the range 56% ÷ 64% of the mean aerodynamic chord. The trimming has been done with different disposition of the movable components inside the front and middle bay of the airframe. To test the effective mass and balance at the specific center of gravity position, the aircraft has been lifted and pivoted about a steel rod support attached to the middle bay at the corresponding c.g. position. The procedure has been done for each landing gear position and for the two configurations with tail and tailess, with weight respectively equal to 13.6 kg and 12.9 kg, including undercarriages.

Bifilar pendulum test The Bifilar pendulum test is an approximate inertia calculator that uses a ground test rig and interaction among different softwares including data collecting and processing to estimate the inertias of the aircraft in the three representative directions (yaw, pitch, roll). The procedure can be accurate only if all the components, including the landing gear, are right in place as established from the center of gravity trimming. The aircraft is attached to a specific jig that is pivoted to different hooks, depending on the inertia property to test (pitch, yaw and roll configuration). The sinusoidal input is tuned on the autopilot with a specific frequency and the response collected on the inertia measurement unit is transferred to the VectorNav software that creates the corresponding graphs. The set-up data for the BFP test, including the aircraft assembly and jig mass, the jig fiber’s spacing and length and notes for the configuration tested, are collected in an Excel BFP Commissioning Card, ready to be used for post-processing. 31 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

The resulting inertias are calculated using the post-processor MatLab BFP software developed by Jenner Richards at the Center for Aerospace Research. The software needs inertia inputs from the BFP Jig SolidWorks model calculated about the experimental coordinate system, that is the center of gravity coordinate system of the full assembly, aligned with the CfAR aircraft coordinate system and located at the front fuselage station. The inertias of the aircraft are found removing the inertia of the jig about the experimental coordinate system from the total inertia of the assembly aircraft/jig. The latter is calculated, about the same point, from inputs of the experimental pendulum test according to the relation expressed in equation 4.1, where m is the overall mass, D is the distance between the jig ﬁbers, h is the length of the ﬁbers and T is the period of the input signal. [32]

mgD2T 2 I = (4.1) 16hπ2 The moment of inertias about the front reference system are then calculated applying the parallel axis theorem. Note that the inertias for FTV7 have been calculated with landing gear located at 56%, while the inertia for the other conﬁgurations can be calculated applying the parallel axis theorem, knowing the inertia of the main landing gear from SolidWorks mass properties.

Mapping of the servo steering system The ground maneuverability of the aircraft, controlled directly from the autopilot, is guaranteed by a careful mapping of the electric servo dedicated for the steering system. The operation has been accomplished using a goniometer to check the angular position of the nose gear respect to the neutral position and sending specific pulses from the ground station that is connected to the autopilot. It has been decided to use a total steering capability of 28◦, due to the large width of the airstrip and so the not required classical steering capability of bigger aircraft operating in airports. The next step has been to find the autopilot commands to maximize and minimize the steering ability selected. The corresponding pulse has been divided in ten mapping points and the resulting pulse spacings, measured in µs, have been tuned in the autopilot command in order to register the effective angular position of the nose gear. The data of the servo mapping are collected in the table and graph shown in figure 4.1. From the resulting control mapping it is possible to notice that the response of the steering system to the autopilot pulse input is almost linear in the selected range of steering ability.

Figure 4.1. Tuning of the steering system control parameters for the autopilot

Mapping of the thrust control parameters The thrust control parameters for the autopilot have been mapped through a thrust test mock-up. The aircraft with undercarriages has been placed on ground and connected through a rope to a load cell. The throttle command has been mapped with respect to the throttle percentage needed for all the phase of taxiing, take-off and flight, following a similar approach described for the steering servo mapping. The minimum (0% throttle) and maximum thrust (100% throttle) defined by the propulsion system, have been 32 4 – Ground mock-ups and flight test planning for FTV7% used to read the corresponding values of pulses on the autopilot, then the command has been equally spaced in ten mapping points and used as input in the autopilot. Finally the thrust mapping has been completed reading the corresponding values of thrust on the load cell.

4.1.2 Landing gear toolkit

All the landing gear spare part hardwares have been organized in a specific toolkit with the relative checklist annexed on the top, in order to facilitate the pre-flight preparation and the necessity to quickly repair or replace the components in case of relevant damages. The toolkit contains even all the landing gear assembly drawings with the description of the component list and relative part number. A sample of drawing, relative to the nose gear strut, is shown in figure 4.2. All the other mounting drawings of the landing gear for FTV7 are illustrated in Appendix B.

Figure 4.2. Component list and mounting drawing for the Nose Gear strut assembly 1

4.2 Ground testing

The FTV2B and FTV2C with undercarriages have been ground tested on the runway where the aircraft is supposed to take-oﬀ and land, in Merrit - Douglas Lake (Canada). The goal of the tests has been to verify the performance of the aircraft with landing gear in all the phases that precede the ﬂight, with the intention to make improvements if necessary.

4.2.1 Taxiing of the FTVs

During the preliminary taxiing tests, the aircraft has resulted not to respond as expected from the autopilot nose gear inputs. Manual testing has revealed compromised yaw authority due to signiﬁcant play in the coupling between the steering shaft and linkage horn, probably due to the quite uneven runway conditions. As temporary solution it has been decided to add a drop of cyano-acrylate glue in order to reinforce the clamp ﬁt. The aircraft has so regained yaw authority and several taxiing tests have been performed in autopilot and manual mode. 33 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

4.2.2 Take off run testing of FTV 2B The take-off run tests are the next step towards a fully operating aircraft, ready to be flight tested. The aircraft has been attempted to be controlled in the centerline of the airstrip in a take-off run sequence. The initial response has showed out of control of the aircraft when high speeds were achieved, due to the airstrip crack induced bouncing. Tuning PID loops on the autopilot control have been necessary to try and get the aircraft running true in the take-off run even if the runway conditions were not as expected. After several attempts it has been agreed that larger wheels on the main gear could have helped to let the aircraft controllable with reasonable values of PID parameters.

4.2.3 Improvements on the landing gear The problems encountered in taxiing and take-off run testing, due to the bumpy conditions of the airstrip, have led to the necessity of making improvements in the design of some components of the lading gear. Due to the steering fit slopping, it has been decided to improve the design of the horns used in the steering system, increasing the contact area in between the servo shaft and the horn. In addition the two horns have been provided with a full open clamping slot, in order to increase the effectiveness of the fitting with an higher torsion clamp. The difficulties encountered during take-off run over an high irregular runway, have indicated the necessity to swap for bigger size of wheel assemblies for the main landing gear. Tires with 25% bigger diameter respect to the previous solution adopted, shown in figure 3.3, have been selected. The aircraft attitude has been maintained the same of the previous design, compensating the increasing of the height (12.5%) on the main gear contact point by adopting a nose gear with the original strut length (as from manufacturing), previously described in subsection 3.3.2.

4.3 Flight test planning

The FTV2B and FTV2C aircrafts, with the improvements on the landing gear adopted from the ground testing results, are now ready to be ﬂight tested even in a rough and bumpy airstrip, like the one in Merrit - Douglas Lake (Canada). The new inertias of the aircraft with a new set of gears and nose strut, can be estimated removing the contribution of the previous components and adding the inertia of the new components to the aircraft inertia about the front fuselage station, as described in subsection 4.1.1. The planning of the ﬂight tests, in particular regarding the implications on the landing gear, is described through the following steps:

1. Full take-oﬀ operation of the wheeled FTV2B and FTV2C in autopilot mode, with ground station oﬃcer (GSO) ready to control the aircraft if something does not go as expected.

2. Flight of the aircrafts in autopilot mode, with a ﬂight path based on the endurance and ﬂight requirements agreed with the stakeholders.

3. Landing of the aircrafts on the same airstrip used for take-oﬀ.

4. Check the aircraft conditions after the full operational ﬂight.

5. Analyze all the critical ﬂight test phases for the landing gear using the Flight Data Post-Processing software.

6. Use all the processed data to redesign components, if necessary, and to make decisions on the landing gear design for the larger ﬂight test demonstrator 16.5%.

4.3.1 Flight test data processing The response of the aircraft with landing gear during all the flight phases of interests (take-off and landing) can be evaluated using the Flight test data processing software provided at the Center for Aerospace research and developed by Sean Bazzocchi and Jenner Richards. The general graphical interface of the software, for a typical landing overview of a previous flight test of FTV2B without landing gear, is shown in figure 4.3. 34 4 – Ground mock-ups and flight test planning for FTV7%

Figure 4.3. Landing overview using the Flight test data processing software

The landing gear designer can obtain all the telemetry data of the aircraft during take-off and landing, including the variation of altitude, vertical rate of descent, take-off and landing horizontal speed, attitude of the aircraft in terms of roll, pitch and yaw, and even information regarding the accelerations developed on take-off (horizontal acceleration ax) and landing (vertical acceleration az) and so the effective load on the wheels developed at the ground contact. Specific points on the flight phase (for example the instant of touch-down or take-off) can be evaluated using a data cursor extrapolation tool. All these data can be visualized on multiple plots: for instance, the center body of the software interface shown in figure 4.3 is representing the trend of the altitude and true air speed in a landing of FTV2B without undercarriages, and the g developed in all the directions. The analysis of flight test data is followed by the development of flight test statistics for important design parameter, such as the g developed on landing and the rate of descent of the aircraft before touching-down, that can be compared with the values used for the design. The data can be analyzed for each configuration of the main landing gear and for the determination of the optimal center of gravity location for take-off and landing, from a landing gear point of view, that can be useful to guide the design of the landing gear for a larger flight test vehicle demonstrator. The obtained time-history plots can be even useful to prove the reliability of design structural simulations and prove the development of a Simulink model of the aircraft response with undercarriages.

35 36 Chapter 5

Landing Gear design for FTV16.5%

The design of the new generation blended wing body aircraft has been started for a bigger 16.5% FTV scale, in parallel with the 7%, in the process of scaling towards to a flight test vehicle that is more representative of the full scale aircraft. The large scale demonstrator is expected to represent as close as possible the same behavior of the full-scale in all the flight phases, including take-off and landing operations that require the design of a reliable undercarriage system. The landing gear system for an aircraft with size and weight of FTV 16.5 is not readily available form radio commanded aircraft or ultralight applications, so it has to be designed and produced from scratch. The main goal of the designer of the landing gear for big UAV scales, is to search at the same time for a simple landing gear in terms of weight and cost, and a complex machine capable to support and transfer the largest local loads from the ground to the airframe. This chapter highlights all the progress done for the design of the landing gear towards to the preliminary definition of the working laws involved in its functionality. An initial brief presentation of all the aspect concerned in the design of the fully operative system, leads to the determination of all the basic features of the undercarriages for FTV16.5, before entering in detail of the conceptual design, followed by the selection of the most suitable concepts and preliminary design, of both main gear and nose gear system with the relative mechanical subsystems.

5.1 Design process for the 16.5% Landing Gear system

The design of the landing gear system for a new big scale aircraft is a self-contained project that passes through all the typical design phases, but at the same time it requires close collaboration among all the different design team, including airframe, sizing, recovery system, propulsion, flight control and aerodynamics. The work break down structure (WBS), that defines the guideline framework for the design progress of the landing gear system for FTV16.5, is shown in figure 5.1. The design of the landing gear follows the same schedule and deliverables time of the airframe design and is accompanied with all the necessary documentation and reports, that represent the basis of discus- sion with the stakeholders. The present thesis work covers the first three phases of the project, performed in the period in between May and September 2018. The first approach has been to define the project charter that tracks all the function required for the final product and some basic requirements, agreed with the stakeholder, in order to guide all the design process. In addition to the all aspect involved in the design of the landing gear for FTV7, discussed in chapter3, the design for a bigger scale UAV requires the development of a suspension system, dedicated for the shock load management, and braking system to stop the aircraft within the length of the runway. The project is also focused on the development of specific Test rigs that allow the testing of the integrity and functionality of the designed landing gear. Specific tests are required to verify the response of shock absorbers, the speed rating of the tires, the energy dissipation upon braking and the structural response of the aircraft on the landing gear during static and drop tests. [4] The design starts with the evaluation of all the possibilities and concepts for the main systems that define the basic operation of the undercarriage and then moves, through a market analysis and analytical hierarchical prioritization of the proposed solutions, to the selection of concepts. The successive phase regards the definition of preliminary design of the basic components needed, including the evaluation of the basics of braking and steering system. Off-the-shelf components are evaluated, adjusted and fully integrated into the design. At this level a preliminary CAD model is prepared and it defines the boundary 37 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators space needed for the following detailed sizing.

Once all the components needed have been characterized, the design moves to the detailed definition of each system and subsystem, including the clarification of all the test rig procedures needed to certify the design. At this point detailed structural and aerodynamic finite element based analyses are required to optimize the modeled components. The final objective is to generate the complete component list and assembly drawings that lead to the manufacture process and integration within the airframe.

Figure 5.1. Work break-down structure of the landing gear design

The undercarriage design is a multidisciplinary process that requires the use and interaction of different softwares, as illustrated in figure 5.2. The basic features of the landing gear are defined using Excel spreadsheets and Matlab sizing tools. The next step is to translate the initial layout of the landing gear in solid CAD modeling. Once the design is refined, structural and aerodynamic simulation tools can be used to evaluate the responses of the landing gear to simulated loading cases that represent some specific phases of the flight. The design is then optimized and evaluated in the contest of the full aircraft, by using Simulink models, ground testing and integration checks. 38 5 – Landing Gear design for FTV16.5%

Figure 5.2. Multidisciplinary process of the landing gear design

5.2 Basic features of the 16.5% Landing Gear

The basic characteristics of the landing gear for FTV16.5 have been determined after a scaling process from the 7% and 100%, guidelines from FAR 23 regulations and typical design steps described in reference [10] and already implemented for the design of the landing gear for the 7% aircraft. The landing gear for the large scale demonstrator is assumed to have the same configuration (tricycle and fixed) and geometric similarities with respect to the 7% landing gear and the design load factor used to size the structural elements is decided to be the same suggested by certification of landing gear for FAR 23 regulations, as expressed in subsection 2.2.1.

The first step of the design process has been to identify and list all the design requirements in an Excel spreadsheet file. The requirements, as already done for the FTV7, are divided by category depending on the type of the feature they affect: aircraft and performance, geometry, main and nose gear peculiarities. The list can be filtered by the origin of the requirement (scaling of Bombardier aircrafts, sizing design team, airframe design team, parachute design team, FAR and MIL specifications, design book guidelines), status of the requirement (open, completed, deferred, no longer required), priority (high, medium, low) and date. An example of the Excel format sheet for the basic requirements is shown in figure 5.3.

The next stage of the design process has been to import all the requirements from Excel to MatLab in order to calculate the basic features of the landing gear, using the developed Matlab® parameter calculator software, described in Appendix A and already used for the FTV7 landing gear. At this point the design of the landing gear becomes iterative and all the changes in the landing gear variables should satisfy the listed core requirements. 39 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 5.3. Basic requirements for the 16.5% Landing Gear design

The next subsections describe all the basic aspects that have been considered to characterize the initial layout of the landing gear for FTV16.5.

5.2.1 Landing Gear position and aircraft center of gravity The importance to relate the landing gear design to the center of gravity envelope is to make sure, in a parametric approach, that the major geometric landing gear variables satisfy all the requirements. The design approach in this case has been to scale and fix the nose gear position and determine from the C.G. envelope, contained in reference [2], the position of the main gear that better implies the respect of all the design requirements. The final position of the main landing gear has been determined after a series of iterations and revision of all the landing gear parameters. In a tricycle configuration the optimal location of the main gear relative to the forward position of the C.G. is governed by the take-off rotation requirement. The consequent tip-back angle, that is the angle between the farther and lowest point of the fuselage or tail and the ground, should respect the tip-back angle requirement that prevents the aircraft from tipping back on its tail during rotation, as already introduced in section 2.1.

Take-off rotation requirement The distance between the main gear and the most forward position of the c.g. is defined by the take-off rotation requirement, that allows the aircraft to rotate around the main gear in order to achieve the angle of attack required for take-off. An important parameter that defines the take-off rotation requirement is the pitch acceleration θ¨, when the aircraft begins to rotate. The rate of change of the pitch rotational speed depends upon several parameters including tail geometric parameters, elevator control power, aircraft weight, rotational speed and distance between the main gear and the aircraft c.g. A typical ¨ 2 take-off pitch angular acceleration for the FTV16.5 high maneuverable aircraft is θTO = 15[deg/s ].[10] During the rotation the speed VR of the aircraft can be assumed to be a function of the stall speed Vs, as expressed in equation 5.1.

VR = 1.1 ÷ 1.3VS (5.1) The forces and moments involved during take-off rotation are represented in the body force diagram in figure 5.4. The represented vectors refer to the aerodynamic forces at the aerodynamic center of the wing and tail (if present), propulsion thrust, aircraft mass forces and friction force at the main gear contact point. The equilibrium state represented refers to the instant following the nose gear lift-off from the ground, so no forces on the nose landing gear are considered. 40 5 – Landing Gear design for FTV16.5%

Figure 5.4. Equilibrium about the main gear contact point at take-oﬀ rotation

The maximum main gear position depending on the take-oﬀ rotation requirement can be calculated applying a moment equivalence about the MG contact point, as shown in equation 5.2, where the pitching inertia (Iyy−mg) about the main gear contact point can be calculated with the parallel axis theorem from the inertia of the aircraft about center of gravity (Iyy), as in the equation 5.3. ¨ Iyymg θ − W xcg − mazcg + Lxac−w + M0w − Lhxac−h − D · zD + T · zT xm = (5.2) L − Lh − W W I = I + (x − x )2 (5.3) yymg yy g m c.g.

Tip-back requirement

The tip-back angle αtb is the maximum aircraft nose-up altitude with the tail touching the ground and the gear in fully extended position. It should be at least 5◦ larger than the take-oﬀ rotation angle and needs to be equal or less than the angle A, measured from the vertical at the main gear location to the aircraft most aft center of gravity, as expressed in equation 5.4.[7] ( α ≥ α + 5◦ tb TO (5.4) A ≥ αtb

The A and αtb can be calculated according to equations 5.5 and 5.6, where zcg considered is the vertical position of the c.g. in the worst case scenario (higher position), xfp and zfp are respectively the longitudinal and vertical location of the farther and lowest point in the worst case scenario (no tail conﬁguration).

x − x A = tan−1 m c.g. (5.5) zg −1 zfp αtb = tan (5.6) xfp − xm

Position of the main gear The longitudinal C.G. envelope of the aircraft is quite wide (from 56% − 66% of the mean aerodynamic chord) because, as already discussed for the 7% aircraft, the new generation flight demonstrator needs to be tested with a movable mass and C.G. inside the aircraft, allowing the flight in both stable and unstable configurations. If the position of the main gear is wanted to be fixed, it needs to be located 41 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators at the forward minimum location that results from the aft location of the C.G (66%). This position respects the tip-back requirement, expressed in equation 5.4, but the resulting angles for the forward ◦ ◦ limit of the C.G. would be too much different, respectively A=34 and αtb = 18.6 , and this implies an huge amount of load needed on the elevon to trim the aircraft and allow the take-off operation, that is not easy manageable in a no tail configuration. The decision made at this point has been to privilege the design of a "semi-movable" landing gear, which ensure similar load distribution between the nose and main gear along the C.G. envelope. Moreover this configuration implies at the same time the easier take-off rotation (A and αtb angle similar), even though it adds complexity and extra structural components respect to the classic fixed gear. The main landing gear strut leg can be fixed before each flight test choosing from 4 position inside a swivel component, directly attached to the pivot point on the fuselage. The basic concept of the swivel component which allows different settings of the main gear is shown in figure 5.5.

Figure 5.5. Sketch of the swivel component, necessary to rotate the main gear position

As the landing gear configuration changes, with a rotation of the strut leg about the vertical axis on the pivot point of the component, the resulting wheel track dimension changes, maintaining the same height of landing gear. The variation of the track between the two main gear assemblies will be discussed in subsection 5.2.3. The main landing gear different configurations, depending on the C.G. envelope that results in stable or instable positions, involve the main gear contact points,1 shown in table 5.1.

Table 5.1. Set of landing gear conﬁguration for diﬀerent position of the c.g.

Conﬁguration Aircraft stability C.G. envelope Contact point [mm] A Yes 56 − 58% 1940 B Yes 59 − 61% 1987.1 C No 62 − 64% 2034.2 D No 65 − 67% 2093

5.2.2 Landing Gear height The landing gear height is deﬁned as the distance between the ground contact point and a reference in the aircraft, in this case the front fuselage water line. As done for FTV 7, the initial approach has been to scale the dimensions of the full scale aircraft with the landing gear in fully extended, static and fully compressed positions and its determination needs to be validated after checking the take oﬀ-rotation ground clearance criteria expressed in section 2.1. The clearance angle αC is calculated according to

1The dimension of the main gear contact point are relative to the front fuselage station (F.S.) 42 5 – Landing Gear design for FTV16.5%

equation 2.4 and is proved to be less than αTO. The height and take-oﬀ ground clearance requirement imply a ground clearance of Hc = 95.80mm, as shown in ﬁgure 5.6.

Figure 5.6. Take-oﬀ rotation ground clearance requirement

5.2.3 Track and Turnover angle The wheel track T is defined as the distance between the most right and most left main landing gear on the ground, in a frontal plane view. Its determination for the FTV16.5 aircraft has been finalized after scaling considerations and the necessity to move the main landing gear for different position of the center of gravity. The minimum allowable value for the wheel track must satisfy the turnover angle requirements (turning controllability and ground stability), whilst the maximum position should meet the structural integrity requirement. The turnover angle φ is determined from a drawing sketch procedure as shown on the left of figure 5.7, referred to the scaled down track.

Figure 5.7. Geometric parameters involved in the determination of the turnover angle

Turning controllability on ground The wheel track T plays an important role in the ground controllability of the aircraft and its determination passes through the moment equilibrium about the gear contact, as illustrated in equation 5.7, that 43 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators refers to ﬁgure 5.7. T F · z = W · (5.7) c cg 2 The turnover prevention requirement imposes that the moment generated by the aircraft weight W should be higher than the one caused by the centrifugal force Fc. The consequent minimum track dimension and turnover angle are expressed in the system of equations 5.8, where Rturn is calculated according to equation 3.4 stated in chapter3.

 2 T > 2Fc·zc.g. = 2V zc.g. mg gRturn 2 (5.8) φ > tan−1 Fc = tan−1 V = tan−1 T  mg gRturn 2zc.g.

Ground stability

The other case that can roll over the aircraft is a cross-wind taxiing, assuming in the worst case scenario that the wind is perpendicular to the centerline of the aircraft, ass represented on the right of ﬁgure 5.7. To prevent the aircraft from overturning during cross-wind conditions, the moment generated by the weight about the main gear contact point should be higher than the one generated by the cross wind force. The requirement is critical for the lowest mass of the aircraft (zero fuel) and it is shown in equation 5.9 2F · z T > W c.g. (5.9) mg

The aerodynamic cross wind force FW is estimated according to equation 5.10.

1 F = ρV 2A C (5.10) W 2 w S DS

The aerodynamic parameters that determine the cross-wind force are the wind speed VW (assumed to be 50knots), the aircraft cross side area AS invested by the cross-wind, and aircraft side drag coeﬃcient

CDS (the value is in between 0.3 and 0.8). [10]

Structural integrity requirement

The maximum value of the wheel track is determined from a structural point of view that interests the deﬂection of the strut leg, attached to the pivot point on the fuselage. In this analysis it is assumed that there is no suspension system and the strut leg is directly loaded when the ground force is applied on the wheel. The most critical condition is achieved when the static ground reaction Fmmax , corresponding to the aft limit position of the c.g., is applied on the wheel and the deﬂection is the maximum value of wheel travel with 2 inches of clearance to the wing. [33] The corresponding maximum track can be estimated by using equation 5.11, that comes from the approximation of the strut leg as a beam, applying the displacement equation 3.3 stated in chapter3.

s 1 3 3EIeq · ymax 2 Tmax = 2 + ymax (5.11) Pmmax

Final determination of the wheel track and turnover angle

The final wheel track of the main landing gear is expected to change among the four configurations described in subsection 5.2.1, as the main strut rotates about the vertical axis on the pivot point. It is assumed that configuration B (at 60% location of the center of gravity) has the scaled down track dimension from the full scale aircraft, shown in figure 5.7, in order to keep as minimal as possible the resulting variation of the track dimension for the other configurations. The resulting track and turnover angle for each configuration, shown in table 5.2, respect the limit values stated by the turnover angle and structural integrity requirements. As observed, the track offset respect to the scaled-down nominal dimension is quite low:+1% for configuration A, -2% for configuration C and -4.3% for configuration D. 44 5 – Landing Gear design for FTV16.5%

Table 5.2. Track and turnover angle for each main gear conﬁguration

Conﬁguration Wheelbase Rturn Tmin Tmax T φ A 1292 2582 750 1470 1344 40.76 B 1339 2676 724 1520 1337 41.80 C 1386 2770 699 1570 1323 42.65 D 1445 2888 670 1637 1293 43.83

5.2.4 Design loading conditions for the nose and main gear The loading distribution between main and nose gear is largely affected by the wheelbase, defined as the distance between the nose and main landing gear and in this case it is determined by the fixed scaled- down position of the nose gear and the resulting position of the main gear for each of the 4 configurations described in subsection 5.2.1. The static load distribution can be determined from the arm ratio of the nose and main landing gear with respect to the most critical position of the center of gravity, that is respectively the forward limit for the nose gear loading and the aft limit for the main gear, as expressed in equation 2.1 in section 2.3. The resulting combination of loads for all the main landing gear configurations and for the most critical case of C.G. for each configuration are summarized in table 5.3.

Table 5.3. Static loading cases on the nose and main gear for each landing gear conﬁguration

Conﬁguration Wheelbase [mm] C.G. shifting PNmax [kg] PMmax [kg] Load Ratio A 1292 56-57-58 % 16.8 77.83 9.6%- 89% B 1339 59-60-61 % 16.86 78.17 9.6%-89.3% C 1386 62-63-64 % 17 78.50 9.7%-89.7% D 1445 65-66-67 % 16.94 78.14 9.7%-89.3%

The vertical loads on the nose and main gear are increased during braking and take-oﬀ, according to the equations 2.2 and 2.3 stated in section 2.1. In this case, the highest value of the vertical C.G. envelope of the aircraft is used because it represents the most critical situation with higher load. The resulting maximum static and dynamic loadings for the nose and main gear, summarized in table 5.4, are utilized for the wheel and tire design as well as for the shock absorber selection.

Table 5.4. Total loading cases on the nose and main gear for each landing gear conﬁguration

Conﬁguration PNdyn [kg] PNTOT [kg] PMdyn [kg] PMtot [kg] A 22.23 39.03 25.40 103.23 B 21.45 38.31 24.51 102.68 C 20.73 37.73 23.69 102.19 D 19.88 36.82 22.72 100.86

5.2.5 Impact loading condition for the landing gear The landing cases that the designer should consider are nominal landing and parachute emergency landing. The design of impact behaviour on landing for the speciﬁc case of FTV16.5 can be outlined following either a scaling procedure from the full scale aircraft or applying the typical design procedure described in references [7] or [8].

Scaling procedure The input for the scaling procedure is to assume that the performance of the landing gear for FTV16.5 is the same in terms of vertical deflection of the full scale model. The second parameter essential to 45 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators determine the required impact acceleration on landing is the vertical rate of descent, calculated in the case of nominal landing with maximum take-off weight according to the equation 2.5, introduced in chapter 2.3 and valid for aircraft certified according to FAR 23. In case of parachute emergency landing, the typical rate of descent is selected in a trade-off with the parachute design team. The resulting average impact load factor, developed on landing, can be calculated as shown in equation 5.12, obtained simply applying the basics of kinematics and dynamics principles.

V 2 N = z (5.12) G 2 · d · g

The resulting normalized accelerations developed on landing for the two design case scenarios is shown in table 5.5.

Table 5.5. Scaling design procedure: impact accelerations developed on nominal and parachute landings

Compression [d] Sink speed[Vz] Impact acceleration Nominal 64.11 mm 7.33 fps 4g Parachute 64.11 mm 15 fps 16.6g

Classic design procedure

The classic design procedure for the landing gear design, already used for the design of the undercarriages for FTV7, is to select from the beginning the design ground load factor and then calculate the required wheel travel to perform the consequent impact acceleration. A typical ground load factor for small aircraft compliance with FAR 23 regulations is 3, as stated in the section 2.2. The wheel travel, corresponding to the 3g nominal landing, is obtained using the equation 5.12 and results to be 25% more than the nominal scaled down compression. In the emergency landing case, it is assumed that there is the bottoming out of the internal suspension system and that the wheel travel is higher than the nominal value. This procedure allows to lower the impact normalized acceleration on parachute landing from 16.6g, calculated according to the scaling procedure, to 11.5g. The load factor for the two nominal and emergency cases, using the described design procedure is represented in table 5.6.

Table 5.6. Classic design procedure: impact accelerations developed on nominal and parachute landings

Compression [d] Sink speed[Vz] Impact acceleration Nominal 84.8 mm 7.33 fps 3g Parachute 92.9 mm 15 fps 11.5g

Selected design procedure for the landing gear of FTV16.5

Due to the higher impact acceleration to dissipate in a scaling design procedure, it has been agreed with the stakeholder to design the landing gear in order to have the wheel travel for nominal and parachute landing that results in the lower impact acceleration. The graphs in ﬁgure 5.8 show the trends of the impact average acceleration developed on landing in function of the compression and rate of descent. 46 5 – Landing Gear design for FTV16.5%

Figure 5.8. Design procedures for impact loading condition: on the left the scaling design procedure with inﬂuence of the parachute rate of descent; on the right the classical design procedure with inﬂuence of the design wheel travel on landing

The design impact acceleration is lower if the admitted rate of descent decreases or the compression of the landing gear increases. The classical design procedure has been chosen, since the rate of descent used for the design of the parachute system has been resulted hard to change signiﬁcantly and also because the suspension system is expected to manage the impact loads with the maximum possible wheel travel and the minimum shock stroke.

5.3 Conceptual Design

In the first conceptual phase of the project the designer is faced with a variety of possible configurations for the most important components of the main and nose gear and their attachments with the airframe. As done for FTV7, the general design guideline has been to use, where possible, off-the-shelf components (COTS) in order to reduce the complexity and cost of the design work. At this stage the landing gear designer is aware of the ground tests needed to prove the structural integrity and functionality of the selected components and if necessary, redesign parts and adjustments.

5.3.1 Main gear concepts The present section summarizes all the concepts and the guidelines used for the conceptual design of the tricycle landing gear. The concepts are shown at level of conﬁguration (external and internal) and components needed, analyzing standard components whenever possible.

External layout configurations The external layout of the main landing gear is guided by the type of attachment with the airframe. The two primary options are the fuselage and the wing and the final choice influences the take-off and landing performance as well as the ground stability. The full scale aircraft has the main landing gear attached vertically to the wing: this configuration on FTV16.5 would imply an optimal shock load distribution inside the suspension system and a constant value of the wheel track between fully extended and fully compressed position of the gear, but on the other side, it would require additional structural components to transfer the loads from the wing ribs to the bulkheads of the fuselage, that means additional load path complexity and weight. A fuselage attachment represents a valid alternative since the fuselage is wide enough to provide the required wheel tracks and to carry directly the loads developed on the landing gear. Although the resulting load distribution in this configuration on the shock absorber is higher, it has been preferred because it can allow the variation of the ground contact point on the main gear, as discussed in subsection 5.2.1. 47 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Internal layout configurations The internal configuration of the main landing gear is essentially guided by the suspension system and structural supports for the attachment on the bulkhead. Different configurations have been compared and evaluated in terms of the resulting kinematics and required components, and the most interesting ones are sketched in figure 5.9.

Figure 5.9. Sketches of the concepts for the internal layout conﬁguration of the main landing gear.

The sketches show, for each conﬁguration, the kinematic motion from fully extended to fully compressed position of the landing gear, the pivot points on the bulkhead, the wheel travel and the stroke on the shock absorber. The main features for each conﬁguration are summarized as follows:

• Configuration 1: The first solution presents a pivot point on the fuselage frame with an internal shock absorber that is disposed with a certain angle with respect to the strut leg. The motion ratio, between the wheel travel and the stroke of the shock, is completely governed by the geometric arm ratio between the internal and external strut leg. This results in an efficient and simple kinematics with the possibility to easily integrate a common attachment plate for the shock absorbers of the two main gears.

• Configuration 2: The second configuration is similar to the first one in terms of pivot point on the fuselage frame, but it presents an additional component that can be used to increase the motion ratio and so have the maximum wheel travel, with the minimum stroke on the shock. On the other side this adds complexity to the design, due to the necessity to integrate the motion of the shock absorber with a third linkage component and an additional hard-point on the fuselage.

• Conﬁguration 3: The last solution presents an internal pivot point connected to the fuselage bulkhead, with an internal shock absorber disposed in the upper-left part. This results in a low required internal space and easy integration with the internal layout of the fuselage. On the other side, the available stroke is lower and cut-outs on the fuselage are needed to allow the strut leg to rotate in loading conditions.

The choice of the most suitable configuration depends even on the type and size of the off-the-shelf components needed for the suspension system. The final selection will be the first step to start the preliminary design of the landing gear.

Components off the shelf (COTS) As already largely discussed in chapter2 and3, the simplicity and cost of the design of the landing gear is guided by the availability of standard components from UAV and ultralight aircraft applications. A detailed market research has been done for each basic component needed for the main landing gear: tires, wheels, brakes and shock absorbers. The tire has essentially the function to support the aircraft structure off the ground, help to absorb the shock loads depending on the inside pressure, transmit the accelerations from the ground to the landing gear structure and braking forces to the runway surface, help to maintain or change the direction 48 5 – Landing Gear design for FTV16.5% of motion. The main parameters, that guide the selection of the proper tire, are the geometric outside diameter and width, and the total static loads that the landing gear is expected to experience, stated in table 5.4. It has been established that the maximum geometric size2 of the tires should have not exceeded the 25% more of the scaled down dimensions from the full scale, in order to keep minimum the difference of behavior, between the FTV and full scale aircraft, during the phases of take-off acceleration and ground effect on landing. [33]

The scaled down size from the full scale aircraft is 6"x2"-1.5" and such a type of tire has been difficult to find in off-the-shelf market, since it is in between radio commanded and ultralight aircraft applications. The loads that the tires should withstand are easily manageable, but the speed target for FTV16.5 is quite high for almost all the tire in that application range, only rated for low speed tail-wheel applications. These tires are applicable to the main gear concept only if they can be ground tested and certified for the speed range of interest, otherwise the decision should go towards tires for other UAV applications (such as the Sonex Jet and the Boeing experimental drone X-36, introduced in section 2.3), where the rated speed values are more comparable, even if they have bigger size. The main results from the tire market research are summarized in figure 5.10.

Figure 5.10. Oﬀ-the-shelf tires from the UAV and ultralight aircraft market

The wheel size required for the main gear of FTV16.5 is guided by the bead seat diameter of the selected tire. Different solutions have been evaluated and the main feasible ones are illustrated in figure 5.11. The classical solutions are made from aluminum forged alloy, usually in dural 2024 or ergal 7075 and they are produced in two halves then joined together by a number of tiebolts. The hub is designed to house the wheel bearings, that are taper roller type and sealed to ensure their grease is not ejected at high speed. Each of the solutions proposed are rated for high static capacity and load limit, with respect to the needs for applications to FTV16.5. An innovative solution is offered by the tubeless STS wheel with integrated brakes machined in ergal 7075. A tubeless solution has essentially three advantages in the aeronautical field: safety and less probability of punctures, lightness avoiding the use of a tube/air chamber, easy affordability. However the minimum tubeless off-the-shelf option found in the market has a rime diameter of 4", as shown in figure 5.11, that does not allow the integration with a 6" tire. [34]

2 The geometric size of tire is represented as D0 × W − DW , where D0 is the outside tire diameter, W is the width and DW is the wheel diameter, or tire bead seat diameter. 49 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 5.11. Oﬀ-the-shelf wheels from the UAV and ultralight aircraft market

The brakes can be either off the shelf solutions from ultralight applications for the wheels described before, or they need to be integrated from other applications, such as bikes or quad karts. The brakes can be distinguished for type of actuation (pneumatic, electro-magnetic, hydraulic) and type of brake contact (integrated disk brake, external or internal caliper3, drum internal brakes). Most of all the applications for aircraft in the range of FTV16.5 use hydraulic brakes with external or internal caliper, so the market research has been addressed to this type. The feasible solutions for the wheels described in figure 5.11, are presented in figure 5.12. The smaller wheel solution can be integrated with disk brake and caliper for bicycle applications, as already implemented for the X-36 UAV aircraft, if no specific braking distance requirement are needed and the length of the runway is much more higher than the scale of the aircraft. If the aircraft needs to be stopped within a certain distance and an higher stop kinetic energy is required, the applications should move towards calipers designed and sized for ultralight applications, in compliance with FAR 23 regulations. In alternative, the STS tubeless wheel described, presents the possibility to be integrated with a brake system that works as a clutch, with pistons that generate an axial force and the disks packed alternatively with friction material and steel. This solution is extremely light weight and it results highly performing, because the friction material has a larger surface than a classic braking system, and so the same braking torque is obtained with half hydraulic pressure. [34]

Figure 5.12. Oﬀ-the-shelf brake solutions for FTV16.5 applications

The internal suspension kinematics and structural response is mostly dictated by the shock absorber

3External or internal refer to the position of the caliper with respect to the wheel diameter. 50 5 – Landing Gear design for FTV16.5% selection. In general, for fixed landing gear, a solid strut leaf or a rubber bungee is a suitable options, but in this case due to the scale of the aircraft and to the applications in not perfect smooth runways, it has been decided to rely all the suspension properties (stiffness and damping) to apposite shock absorbers. The first approach has been to search for components among bicycle/bike and ultralight aircraft applications, in order to alleviate the impact cost of a custom solution. The main interesting solutions from the preliminary market research are shown in figure 5.13. The simpler solutions is a coil spring with separate elements for the compression (spring) and damping (oil, piston and orifices), even though it needs a relative wide amount of space. A more compact solution is represented by an air spring with an internal chamber, design in order to use the air for both compression and damping properties. The air spring shock absorber shown, has ten settings of rebound control and three position for compression adjustment with apposite rotating valves. The best solutions in terms of rebound and compression is an oleo-pneumatic shock absorber with separate chambers and fluid for the two properties, in order to maximize and optimize the fluid application for the specific task required. However, this is not an easy solution to integrate with small airplanes, especially if the shock absorber needs to be allocated inside the aircraft, due to the large size of the shock and to the high cost and weight.

Figure 5.13. Oﬀ-the-shelf shock absorbers for FTV16.5 applications

Custom components The off the shelf components described are expected to be integrated with a custom strut leg that can rotate about a swivel bracket before fixing the position for take-off and landing. The conceptual design of the strut has started from the evaluation of typical material that can potentially be used for the custom component. The geometry has been modeled considering the nominal outboard track at the position of the c.g. at 60% of the M.A.C. and following the classical procedure already used for the 7% and described in subsection 3.2.3. The part, modeled in SolidWorks, is shown in figure 5.14 with all the typical design elements that include pivot points, internal and external arm, house for the wheel axle.

Figure 5.14. Conceptual geometry of the main gear strut leg

Different materials, typically used for landing gear strut components (Steel, Aluminum, Carbon fiber 51 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators laminate), and thickness have been used to evaluate the best conceptual application with the same overall geometry. The structure has been modeled with TRIA solid mesh elements and the boundary conditions applied are hinge constraints for the two pivot points, that connect respectively the strut to the fuselage and to the shock absorber, and remote vertical load at the location of the wheel, corresponding to the design impact load. The material and thickness have been changed in an iterative process in order to keep the structural safety factor equal to 1.5, commonly used for sizing the aircraft structure. The analysis results, illustrated in figure 5.15, show that the best combination for a strut leg for the FTV16.5 is a carbon fiber laminate, in this case modeled with 6 layers4. This solution, at equal geometry, results to be lighter respectively 80% and 60% of the steel and ergal solutions.

Figure 5.15. Initial structural evaluation of diﬀerent custom main gear struts

Although the laminate solution analyzed is suitable in terms of performance, it results expensive to manufacture because it requires the design of carbon fibre molds specific for the designed geometry. Two valid alternatives are sketched in figure 5.16. The concept 1 is made by two central bodies producible with 3D printing technique and then bond together through a bath in different layers of carbon fibre. Instead, the concept 2 results from the integration of an off-the-shelf rod in carbon fibre with a 3D printed strut cover. Either concepts 1 and 2, can be easily attached to the swivel component, described in section 5.2.1, and to a similar custom swing component on the wheel axle, capable to modify the camber and tow angle of the wheel assembly, since the track dimension is changed at different c.g. locations.

Figure 5.16. Conceptual sketches of the carbon ﬁber strut

The parasite drag induced by the frontal area of the wheels can be reduced by designing 3D printable fairing covers and wheel pants, aerodynamically shaped. At last, the landing gear should preserve the integrity of airframe and systems in case of parachute emergency landing. As already introduced in subsection 5.2.5, the landing gear design is inﬂuenced by the design of the recovery system of the aircraft in order to decide which is the trade-oﬀ in terms of rate of

4The layer combination used is [0/ + 45/90/90/ − 45/0] 52 5 – Landing Gear design for FTV16.5% descent to which the landing gear and the recovery system should be designed. The designed emergency situation corresponds to a vertical sink speed of 15 fps or equivalent impact acceleration around 11.5g. In order to not largely oversize the structure and weight of the landing gear system for a parachute landing, the extra kinetic energy, associated with the emergency, is thought to be absorbed by an internal crashworthy component, located in the middle-low part of the fuselage and connected to the bottom pivot points of the two shock absorber.

5.3.2 Nose gear concepts The concepts for the nose gear assembly, including the conﬁguration and oﬀ-the-shelf possibilities are illustrated in the present section.

Types of nose gear The first design variable for the nose gear design is the type of configuration. The concepts considered in this stage are shown in figure 5.17.

Figure 5.17. Sketches of the concepts for the nose landing gear for FTV16.5

The first possibility is a traditional nose landing gear with a single vertical strut mounted directly on the wheel axle, a setup that guarantees a direct load path to the airframe supports. This solution is the simplest and cheapest, but the rebound and compression control is exclusively attributed to the shock absorber, that for this reason needs to have enough stiffness and damping properties to manage all the vibrations and dissipate the related energy. A valid alternative to the classical straight-vertical nose gear is represented by a trailing link solution. It is made essentially by an L-shaped flexible arm, ahead the wheel, with the shock absorber disposed with a certain angle with respect to the vertical. In a trailing link connection, the link itself represents an extra shock absorber and contributes to smooth the vibrations caused by hard landing and rough runway surface, whilst the shock absorber is responsible to react to the first vibrations induced by landing. An other advantage of the trailing link is the easy maintainability respect to the straight vertical nose gear: repairing the shock absorber from a trailing link gear is quite easy because it does not require to disassembly all the system, but only the shock absorber needs to be replaced. On the other side, a trailing link means also more moving parts to lubricate and complicates the design of the retraction system, if required, so its choice depends of the type of landing gear for the specific application. [35] An other important decision, that conditions the design of the nose gear assembly for FTV16.5, is the number of tires per strut, choosing from one or two. A double wheel means better ground handling, lower rolling resistance helping the take-off and redundancy, in case of failure of one assembly, but at the same time it implies more weight and drag penalties to the aircraft design. Due to the relatively small size of the aircraft and low load condition expected on the front of the aircraft, a single wheel unit, with bigger diameter respect to the scaled down size, can represent a valid alternative capable to have an adequate ability to deal with less than perfect runways.

Components oﬀ the shelf (COTS) The initial market research has been done for the wheel and tire applications among big radio commanded airplanes and UAV applications. The feasible options for the concepts described are shown in ﬁgure 5.18. 53 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Robart tire is a custom goodyear application for small airplane (like scaled Spitﬁre and P-38) and it has a tough outer skin, softer inner core and design for low bouncing. [36] The other solution, used for RC big jet airplanes (like Skymaster jets), has the same scaled-down outside diameter than the full scale aircraft and so it is a good candidate to be applied, in a double version, for the nose gear of FTV16.5. The tire is rubber type and the sidewalls are slightly thicker than the tread, so it is designed to hold up to side-load ground rolls during steering. [20]

Figure 5.18. Wheel assembly solutions for the nose gear of FTV16.5

The suspension properties are accomplished by a shock absorber to be selected from similar options to the ones in figure 5.13, but smaller size and stroke (indicatively 5.5"× 1"). For a straight vertical configuration, an oleo shock absorber should be required to mitigate the vibrations induced by bouncing on uneven runways, whilst in a trailing link configuration a simple air-damper spring should be enough to work in conjunction with the trailing link arm itself. The nose gear is designed to give to the aircraft the necessary steering ability during ground maneuvers. The steering system on the nose gear can be implemented using a rack and pinion mechanical transmission or simpler pulleys moved by a belt, as shown schematically in figure 5.19. In both cases, the motion is transfered by a dedicated servo actuator that can be pulsed modulated in order to be commanded directly from the autopilot. In the first concept the rotational motion of the nose gear is driven by a linear movement of the rack commanded by an electric motor, while in the second concept the motion is transfered from a smaller pulley with a specific drive ratio. [37]

Figure 5.19. Concepts for the steering system of the landing gear for FTV16.5: Rack and Pinion mechanical transmission on the left and Pulley-Belt mechanism on the right

Custom components

The nose gear is supposed to be positioned at the scaled down longitudinal position respect to the fuselage front reference and attached to the front bulkhead of the aircraft using a custom bracket in metal. The strut and additional scissors to prevents shimmy instability (in case of straight vertical conﬁguration) are addressed to be designed in the preliminary phase of the landing gear design, once the size of the selected components are established. 54 5 – Landing Gear design for FTV16.5%

5.3.3 Test rig concepts The design of a full landing gear system from scratch requires even the definition of the testing procedures in order to prove the feasibility of the selected concept and demonstrate the structural integrity upon the typical take-off and landing conditions, in which the aircraft should operate. Most of the test rig procedures follows the indications stated in FAR 23 regulations. [12]. The present subsection describes the concepts and ideas to test the off-the-shelf components, with take-off and landing loading simulated scenarios, as well as the full assemblies during impact loading.

Shock absorber test rig

Both the selected shock absorbers for the main and nose gear assembly, are needed to be tested in order to determine the stiffness and damping properties along the full stroke of the piston. Due to the impossibility to have a shock dynamometer in CfAR shop and to the high cost of using the instrumentation from other shock specialized companies, the first approach to verify the feasibility of the shocks has been to design a custom test rig with a simple leverage mechanism, to be mounted on the precision optical table, provided in CfAR shop. The outlined design sketch and the relative main components are shown in figure 5.20.

Figure 5.20. Test rig for shock absorbers: it shows all the components needed for the test rig assembly.

The system has been designed with a leverage arm ratio of 1 : 8, sufficient to reproduce in the shock absorber the worst case scenario of impact for design nominal landing. The leverage is realized with a double rectangular cold steel tube distanced with enough clearance to allow the shock absorber to be positioned in between and to be moved, when loaded, without interferences. The rectangular cross section has been selected in order to increase the moment of inertia about the longitudinal axis and so decrease the stress distribution inside the long leverage. The rigidity on the contact area with the shock absorber, pivot point and load input, is increased using specific steel inserts with the same internal shape of the rectangular tubes. All the insert in the rectangular bar are fixed with standard screws and hex nuts. The shock absorber is connected to the two bars with a bolt connection and located in the center using two circular steel spacers. The connection of the shock absorber is vertically offset from the top plane on the centerline in order to have the leverage horizontal when the travel of the shock piston is half stroke: in this way it is valid the approximation of small angles when the bars move. On the other side, respect to the longitudinal centerline, there is house for a counter balance weight used to pre-load and balance the system, before applying the test loads. The other end of the shock absorber is fixed between two angle brackets designed to be bolted on the optical table provided in CfAR. The pivot hinge of the leverage system is realized by using a rectangular steel bar in the center plane of the system: it is connected with two pins to additional custom brackets, bolted to the optical table, in order to remove the oscillations during the motion of the leverage, and different holes in the bottom part are used to create the pivot point of the overall system depending on the height of the shock absorber to test. Even for the pivot steel bar, the central positioning is maintained thanks to round spacers in between the two rectangular tubes. The testing load is applied at the end of the larger side of the leverage using a simple hook, force transducer, and a bucket to insert a variable equivalent weight. Two steel spacers are added in between the two bars, in proximity of the hook application in order to increase the bending rigidity of the leverage where the load is applied. 55 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Speed rating test rig A truck test can be designed to test the speed rating of both main and nose gear selected tires, with equivalent take-oﬀ speed of FTV16.5. The test rig can be designed as a trailer for the truck provided in CfAR in order to be easily installable and removable. The basics of the speed rating test rig are shown on the left of ﬁgure 5.21. The wheel to test is linked to a fork pivoted on the standard rectangular tube trailer, used for the connection with the truck. On the same wheel axle, a support plate is pivoted in order to apply the testing static load on the wheel. The oscillation of the plate is prevented with a double arm connection with the fork. The equivalent static weight is applied symmetrically on the support plate around the wheel thanks to high density lead rectangular blocks. Safety rubber pads are disposed around the perimeter of the bottom part of the support plate in order to prevent damages on the wheel if the rubber tire blows up during the test and to avoid the destruction of the test rig at high speed on the concrete ground. If the static test is proved to be successful, the tire can be tested under the real longitudinal spin-up loads, developed on landing when the speed of the tire passes from zero to the aircraft landing speed. For this purpose some steel ropes can be used to drop the test rig from a limited height. The ropes should be attached on the wheel axle and commanded from a pulley system inside the truck.

Static drop test rig A static drop test rig, in compliance with the normative FAR 23 for the height and equivalent mass test parameters, is needed to be designed for testing the main gear and nose gear full assemblies. The test rig for the main gear, sketched on the right of ﬁgure 5.21, mainly requires a beefy support for sliding the double adjustable plate, that represents the support to attach the shock absorbers and to apply the distributed mass. The bottom plate has two cut-outs in order to allow the vertical motion of the struts during impact loading. The double plate before being dropped is connected to the test rig support with hooks and steel ropes with adjustable length according to the required drop height. The sketch in ﬁgure shows the overall boundary dimensions needed for a similar test rig concept.

Figure 5.21. Test rig concepts: on the left the Speed Rating Test rig, on the right the Drop Test rig.

5.4 Preliminary Design

In the preliminary design phase, the landing gear concept is chosen and the design activity becomes more analytical and detailed. The designer starts to procure and evaluate the off-the-shelf components available 56 5 – Landing Gear design for FTV16.5% in the market and to integrate them into the design. This phase of the project is highly influenced by the design of other systems of the aircraft and so it is important to define a design space for the landing gear inside the airframe, in order to avoid all the possible interferences with other components. This phase of the design includes also the definition of the mechanical subsystems (braking and steering systems), the preliminary design of all the needed structural components and the design of the preliminary testing procedures in order to determine the feasibility of the system. At the end of the preliminary design stage it is expected that all the landing gear components, with the basic features defined, are placed in CAD and integrated with the airframe design. The present section shows the initial phase of the preliminary design and the planning of the next steps needed before entering in the critical design phase of the landing gear, in conjunction with the development of the airframe design.

5.4.1 Selection of the concepts The concepts for the main and nose gear have been selected through an analytical hierarchical approach that compares the possible solutions described in subsections 5.3.1 and 5.3.2. The most important features, identiﬁed to select the concepts, have been rated with an importance percentage that comes from engineering criteria and evaluation of the application and integration of the landing gear for FTV16.5. The discriminatory features used for both main gear and nose gear concept evaluation are shown in the charts in ﬁgures 5.22 and 5.23.

Figure 5.22. Feature importance for the Main Figure 5.23. Feature importance for the Nose gear concept selection gear concept selection

The final selection for the main gear is mostly guided by cost, design simplicity and the needed space inside the fuselage, quite limited due to the presence of the center of gravity shifting mechanism and parachute components. The AHP summary table used for the main gear internal concept selection is shown in table 5.7. The selected configuration, with pivot point of the leverage strut on the fuselage bulkhead and the shock absorber directly connected to the strut, has been proved to have the best features regarding weight, availability and estimated cost, even though the rebound control and stiffness properties are all attributed to the shock absorber, unlike the other configurations where the linkage permits to add an additional kinematic suspension.

Table 5.7. Analytical hierarchical process table for the Main Gear concept selection

Configuration 1 Configuration 2 Configuration 3 Features % Rank [1-10] Rank [1-10] Rank [1-10] Space needed 20 10 7 7 Weight 15 8 5 6 Simplicity 20 10 7 5 Rebound Control 10 6 8 7 Availability 10 10 8 5 Cost 25 8 6 5 Tot 100 8.8 6.65 5.85

As for the nose gear concept selection, the final decision is guided by the need to have a good directionality of system and controllable rebound to keep minimum the vibration induced from taxiing 57 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators and take-off run on rough runways. The AHP summary table used for the nose gear concept selection is shown in table 5.8. The trailing link solution, has been preferred to be developed, because it allows to use a simpler air spring shock absorber, since the stiffness and damping are controlled also by the trailing structure itself.

Table 5.8. Analytical hierarchical process table for the Nose Gear concept selection

Trailing link Classic conﬁguration Features % Rank [1-10] Rank [1-10] Space needed 5 8 6 Weight 10 5 8 Directionality 30 9 6 Simplicity 5 5 9 Rebound Control 15 8 6 Maintainability 10 7 5 Cost 25 9 7 Tot 100 8 6.5

The next step, after the concept selection, is to select the off-the-shelf components and integrate them into the design process. The preliminary CAD models for both main and nose undercarriages are shown in figure 5.24. The model of the main gear assembly includes the most important components encountered at the cut section, that corresponds to the main gear positioned in the configuration B (with nominal scaled-down track dimension). The basic leverage suspension includes FOX Float DPS with size 6.5” × 1.5”, selected because of its high availability, low cost and the possibility to change the set up of stiffness and damping before each flight. The wheel assembly comprehends the solution 1 for tires, wheels and brakes, described in subsection 5.3.1, chosen because it fulfills the scaled requirements of FTV16.5 and it’s the simplest and cheapest solution to integrate into the design. The solid model for the nose gear assembly contains the preliminary brackets that delineate the trailing link suspension including the same type of shock absorber selected for the main gear, but smaller in size (5.5” × 1”). The wheel assembly solution 1 has been selected, due to the high quality of rubber plus foam of the tire and because it has a 25% more of the scaled down size, that can help to achieve firstly the take-off run aircraft speed conditions and absorb the oscillation induced by the runway thanks to the inner core designed for low bouncing. The steering system represented is a pulley-belt mechanism, selected for its adjustability and relatively ease to be integrated with standard components.

Figure 5.24. Concepts selected for the main gear (on the left) and for the nose gear (on the right)

58 5 – Landing Gear design for FTV16.5%

5.4.2 Procurement and testing procedures for the oﬀ-the-shelf components The present subsection describes the testing procedures necessary to prove the feasibility of the selected oﬀ-the-shelf components for FTV16.5, even though the tests have not been performed before October 2018 due to delays in the procurement of the components and in the manufacturing process of the designed test rigs.

Wheel assemblies The combination of tires and wheel selected for both main and nose gear are shown in ﬁgures 5.25 and 5.26.

Figure 5.25. Procurement of the Figure 5.26. Procurement of the main gear wheel assembly nose gear wheel assembly Since the main gear wheel has been designed for not high speed tailwheel applications, it requires speed testing and certification. Moreover, it doesn’t provide a readily integrated brake system so the wheel can be considered as starting point for the sizing of the braking system required. At the same time, the nose gear assembly necessitates speed testing, because it has been designed for big radio-commanded airplanes that usually have lower speed requirements in comparison with FTV16.5. Both Main gear and Nose gear wheel assemblies are ready to be speed tested, developing the speed rating test rig conceptualized in subsection 5.3.3. The testing procedures needed are summarized as follows: 1. Weight distribution: the weight distribution on the support plate should represent the maximum vertical loads on the nose and main gear, stated in table 5.4 in subsection 5.2.4. The loading condition can be represented by using several layers of four lead blocks, with approximately weight of 10 kg and thickness of 3 mm. The blocks are then locked in position by using steel ropes. 2. Speed-up the truck: the truck should reach gradually the equivalent take-off speed of the aircraft, stated in the list of requirements. If necessary, the tire to test, can be pre-rotated in the opposite sense of the forward speed, by using a simple electric motor, in order to lower the speed requirement on the truck. 3. Evaluation of the tire conditions: after the speed testing evaluate the tread condition of the rubber and if visible deformations have been induced from the testing high speed. 4. Dynamic speed test: if item 3 is successful, the next step can be an evaluation of dynamic drop performances of the tire in order to estimate the spin-up loads induced on the tire when it touches the ground at the landing speed of the aircraft. The support plate can be connected, through steel cables, directly to the back of the truck, and all the system can be dropped from a low defined ground clearance. The latter can be as minimal as possible in order to reduce the possibility of blow-up, since no suspension system is used in the speed rating test. The test should only replicate the phenomena of spin-up, where the tire passes from almost zero speed to the aircraft landing speed. An indicative value of ground clearance can be 1 inch. If the selected tires do not pass successfully the described test, bigger aircraft rated and certified solutions should be used, as the ones presented in figure 5.10 in subsection 5.3.1.

Shock absorbers The selected shock absorbers for the suspension system of the nose and main gear are shown in figure 5.27. Both solutions have been preferred to an oleo-pneumatic shock absorber because of the less cost, 59 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators weight and space needed inside the fuselage. On the other side, despite of a simple coil spring, they provide a manageable compression and damping thanks to the design of two internal chambers where the air passes through, during loading. Figure 5.28 illustrates the cut section view and the most interesting points involved in the functionality of the shock absorber: the air is pumped inside the upper chamber by a specific valve (number 4 in figure); the stiffness of the shock can be regulated with a compression valve (indicated by number 2 in figure) according to three different setting (firm, normal, soft), that can be selected depending on the type of runway; the damping is achieved thanks to small orifices like the one shown in item 3; the last item indicates a safety valve of depressurization in order to protect the internal chambers in case of high load. [38]

Figure 5.28. Internal Figure 5.27. Oﬀ-the-shelf shock absorber chambers of the shock for the nose and main landing gear absorber [38]

The shock absorbers are now ready for characterization and for the evaluation and proof of the feasibility of the selected concepts. The shock load curves have been ﬁrst determined by an analytical procedure that considers the internal geometry of the components shown in ﬁgure 5.28.

Although the stiﬀness and damping eﬀect are correlated in a shock absorber with not separate chambers, they have been estimated, via analytical method, as if they act separately. The analytical expression used to approximate the trend of the spring shock load Fs along the piston travel s, is shown in equation 5.13, where Ac is the area of the cylinder, Ap is the area of the piston, V0 is the volume of the upper chamber in static pre-load condition, P0 is the pre-load pressure and γ is the index of the polytropic compression, assumed to be 1.1 because the reaction is expected to be so fast that it can be assumed to be close to an isotherm. [39]

γ V0 Fs = P0Ac (5.13) V0 − Ap · s

On the other side, the damping force has been estimated as a quadratic function of the piston travel rate of change Vt, as shown in equation 5.14, where Ao is the orifice area and Cd is the orifice discharge coefficient estimated using reference [40].

3 2 ρAc Vt Fd = 2 (5.14) 2 (CdAo)

The resulting shock loads for the shock absorber selected for the main and nose landing gear, in function of the piston travel, are shown respectively in ﬁgures 5.29 and 5.30. 60 5 – Landing Gear design for FTV16.5%

Figure 5.29. Shock load for the FOX 3 DPS 6.5"×1.5", Figure 5.30. Shock load for the FOX 3 DPS selected for the Main gear application 5.5"×1", selected for the Nose gear application

The true trend of the shock load along the piston travel needs to be validated by experimental tests. Due to the difficulties to use a proper shock dyno test rig that can test the behaviour of the shock absorber in relation to different inputs of force and different setting of the inside pressure, it has been decided to test the shocks in a custom test rig, already described in subsection 5.3.3, with the goal to prove the feasibility of the shock within the expected impact loads, corresponding to different internal geometries. The following itemizes summarize all the testing procedures that should be used to characterize the internal disposition of the shock absorber in FTV16.5.

1. Setting the inside pressure: The inside pressure can be regulated with an hand pump to values in between 100 and 300 psi. Note that the shock absorber should be slowly compressed with the shock pump screwed on ten times by around 25% of its stroke, since it is needed equalization of the positive and negative air chambers. Typical values of inside pressure are based on the static load on the shock and are indicated in the manual of the shock absorber.[41] The minimum value of pressure is set at the beginning of the test in order to balance, with a counterweight, the test rig in the initial position.

2. Setting the SAG: the SAG (negative spring deﬂection) is the degree by which the shock compresses under the static load. Typically the SAG can be set in between 20% and 25% of the shock stroke for a ﬁrm compression set-up, while a value in between 25% and 35% should be imposed in case of expected higher loads, for instance due to the not regular runway. The inside pressure should be regulated until a reasonable value of SAG is achieved. [41]

3. Adjusting the compression: there are three compression settings, with an apposite valve, that can be imposed for each test. If the valve is in position "open" the compression is the most sensitive to unevennesses on the ground, if it is in "drive" it corresponds to the nominal setting of the compression, whilst in "firm" position the piston is locked and the suspension results more rigid. A blow-off valve in the bottom part opens the flow of fluid in the case of heavy impacts preventing damages to the shock.

4. Adjusting the rebound: the rebound can be regulated through a rotating rebound wheel that changes the velocity of the piston in loading condition. The rebound can be set depending on the type of the runway being traveled. If low bouncing is desired the setting wheel valve should be rotated in clock-wise direction.

5. Load and stroke data collecting: For each loading testing conditions corresponding to the diﬀerent inside pressure set, the load on the shock, that results from the leverage ratio and the corresponding stroke, can be read and collected in an Excel data sheet.

6. Data processing: the data collected can be used to create a curve load/deformation depending on the variation of the inside pressure.

Note that the testing procedure is applied for diﬀerent setting of inside pressure and increasing load until the SAG requirement is respected. 61 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

5.4.3 Preliminary design of the main gear The preliminary design of the main landing gear is developed in parallel with the airframe design and each component should be designed inside a defined volume space, in order not to interfere with the design of other systems inside the fuselage. The design volume is defined considering possible variations of the length of the internal strut leg, the location of the shock absorber and a certain clearance respect to the size of the shock absorber. It includes also the central space needed for the design of the crush-worthy component for emergency landing. The relative size needed for the preliminary design of the landing gear is shown in figure 5.31. The internal location of the main gear is supposed to be attached on the rear side of the aft bulkhead, since the front side is occupied by the wing attachments and fuel system. The internal layout is fixed among the main landing gear 4 configurations and so it needs to be defined for the worst case landing scenario.

Figure 5.31. Design space for the main landing gear inside the airframe of FTV16.5

The final needed space can be specified after the shock absorber characterization and definition of the needed leverage arm ratio between the external (dext) and internal (dint) strut leg. The geometric parameters of the main landing gear in the front plane are shown in figure 5.32. The geometry of the strut is characterized by the determination of the external dext and internal dint lengths, the camber angle γ of the wheel, the inclination α of the external strut with respect to a vertical plane, and the inclination β of the shock absorber with respect to the internal strut.

Figure 5.32. Preliminary design geometry and load distribution of the main gear selected concept.

One of the results of the shock absorber testing described in subsection 5.4.2 is to find the maximum load that the shock absorber can withstand with the full stroke, at different setting of inside pressure. Once the ultimate load Fs−MAX is determined, it is possible to calculate the required moment arm ratio illustrated in figure 5.32 and so determine the internal layout, when a static load FT −mg is applied to the wheel contact point. The arm ratio can be calculated from an equilibrium momentum about the pivot point O that leads to the expression in equation 5.15.

a F = s−MAX (5.15) b FT −mg 62 5 – Landing Gear design for FTV16.5%

The geometry of the strut should respect the moment arm ratio as described in equation 5.16, where T is the maximum main wheel track between the possible conﬁgurations introduced in subsection 5.2.1 and do is the distance of the pivot point from the vertical mid plane of the aircraft, deﬁned at this point from the height requirement of the landing gear and considerations with Airframe Design Team.

a T − d = 2 o (5.16) b dint · sin (β) Stated that the values of the track and position of the pivot point cannot be signiﬁcantly changed due to the constraint explained in section 3.1, the only chance to validate the use of the selected shock absorber is to move the length of the internal strut dint and the angle β considering the following design approach:

1. Change β: The initial value of dint is ﬁxed to the value used to model the initial geometry shown in subsection 5.3.1 and the resulting β is calculated using equation 5.16, checking whether the β is less of 90◦, condition that ensure a valid internal kinematics.

◦ 2. Change dint: If the resulting geometry is not valid, ﬁx the angle β to 45 and change the length of the internal strut, ensuring that the resulting geometry has the internal strut inside the delimited design space. If the geometry is not valid change also the value of the angle β. 3. Change the design space: If items 1 and 2 do not lead to a valid geometry, check if the design space can be enlarged without hardly interfering the design of other airframe components. 4. Change the shock absorber: If none of the items are respected, the landing gear designer should change the oﬀ-the-shelf component and opt for an oleo-pneumatic shock absorber. In that case all the procedure described should be reiterated for the new component until a valid combination of shock absorber and internal layout is achieved. Once the geometry and loads on the shock absorber, in both static and impact cases, are determined it is possible to size the pivot hard-points on the bulkhead using the static equilibrium system of equations 5.17. ( V = F + F · cos (θ) o T −mg s−mg (5.17) Ho = Fs−mg · sin (θ)

Preliminary design of the braking system for the main gear The classical preliminary sizing of the braking system, used to certify braking systems according to FAR 23 regulation, starts from the determination of the brake roll distance required to stop the aircraft, as expressed in equation 5.18. The total landing distance (sb) is determined by the ground roll distance with brakes applied plus a short "free roll", to account for the reaction time required to apply the brakes and engages the spoilers, that usually is on the order of 1 to 3 seconds. [42]   2 W/Sw ρVTD sb = ln 1 +  + VTD · tf (5.18) ρg (CD − µrCL) CD 2W/Sw − CL µr The landing distance has been calculated for both tailed and tailess configurations of FTV16.5 in order to consider the worst case scenario for the braking system. The aerodynamic lift and drag coefficients in landing clean configuration for both tailed and tailess have been scaled from the values of FTV7, according to the data provided by the aircraft aerodynamic team, and are reported, with all the other needed values, in table 5.9. The calculations have been done considering no wind and no thrust reversal, with the engine set to IDLE and the friction coefficient with brakes applied equal to µr=0.4. [7] All the quantities in table are expressed in Imperial units.

Table 5.9. Braking distance calculation according to FAR 23 regulations

W SW VTD z CD CL µR sTOT Tailed 386 50.19 83.45 4000 0.0448 0.2894 0.4 585 No-Tail 386 50.19 130.90 4000 0.0233 0.0785 0.4 815

63 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

The longitudinal force and the torque needed per each wheel assembly can be calculated assuming that all the aircraft longitudinal kinetic energy is ideally dissipated by an equivalent longitudinal braking force acting along the braking distance, as in the system of equations 5.19. ( 1 mV 2 = F s 2 TD h b (5.19) T = Fh · rw The values for the Kinetic energy and torque needed to size the braking system are calculated and summarized in table 5.10.

Table 5.10. Calculation of the parameters for the preliminary sizing of the braking system

sTOT Kineticenergy Fh rw T Fn Tailed 585 42671 ft lbs 71 lbs 1.75 in 124 in lbs 63.6 lbs No-Tail 815 93000 ft lbs 114 lbs 1.75 in 200 in lbs 102.5 lbs

The table contains even, in the last column, the normal force of the brake pads required to size the brake disk. In the calculation it has been assumed that the friction coefficient developed on the brake disk is 0.65 (typical for steel brakes) and the medium disk radius equal to 1.5 inches that guarantees the development of a brake system with an internal caliper within the wheel size selected, with diameter 3.7 inches. The values required for the braking system sizing have been used to select standard components among the solutions described in subsection 5.3.1. The results show that, if the aircraft is wanted to be stopped within a specific distance according to FAR regulations, the only chance is to use a braking system certified for ultralight aircraft and in this sense the solution with internal caliper and disk is the preferable because it is cheap and reliable. One of the possible solutions, with the relative maximum values of kinetic energy and torque applied, is shown in figure 5.33. However, to integrate this solution with the wheel presented in subsection 5.4.2, it is necessary to design a custom brake disk, since the standard coupled disk is designed for a wheel larger than the one selected.

Figure 5.33. Braking system for ultralight applications: internal caliper plus brake disk. [43]

In alternative, the complexity and cost of design and production of the braking system could be highly decreased if it is taken into account that the length of the aircraft is less than 4% of the the total length of the runway where it is suppose to fly the aircraft, Foremost (Alberta - Canada). If no requirements on the braking system are applied the aircraft could potentially be stopped in a distance much more longer than the braking distance sb that results from equation 5.18, and the resulting torque need to dissipate the kinetic energy is much more manageable, using hydraulic brakes for rear bike applications. The latter is the strategy often applied to brake UAVs when no specific stopping distances are required and the priority is just to brake the aircraft within a defined clearance from the end of the runway, using 64 5 – Landing Gear design for FTV16.5% the simplest, lightest and cheapest solution possible. An example is presented by the solution used for the experimental aircraft X-48 shown in figure 5.34.

Figure 5.34. Braking system from bike applications applied to the X-48 UAV. [17]

The next step for the design of the braking system is to select an off-the-shelf brake kit from bike applications, test it and find the maximum torque applicable, from which it would be possible to estimate the effective braking distance required and finalize the preliminary design of the braking system.

5.4.4 Preliminary design of the nose gear

A similar approach to the one used for the preliminary design of the main gear has been followed for the nose gear, starting from the deﬁnition of the design space, illustrated in ﬁgure 5.35. It has been determined from the expected needed space for the steering system and for the attachment to the front side of the nose bulkhead.

Figure 5.35. Design space for the nose landing gear inside the airframe of FTV16.5

The basic geometry, with the load distribution and the constraints on the pivot point and on the nose attachment point to the front bulkhead, are shown in ﬁgure 5.36. As already stated for the preliminary design of the main landing gear, the geometry of the nose gear can be ﬁnalized once the properties of the shock absorber will be known and tested. 65 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 5.36. Preliminary design geometry and load distribution for the nose gear selected concept.

The resulting geometric arm ratio, referred to figure 5.36, that defines the final disposition of the trailing link, can be calculated applying a moment equilibrium about the pivot point of the trailing link, as shown in equation 5.20 c (L + L ) · sin (δ) F = 1 2 = s−ng (5.20) d L1 sin (σ) FT −ng where FT −ng is known from the static load distribution between the nose and main gear and Fs−ng is the maximum shock load registered on the shock absorber test. Following an optimization procedure similar to the one described in subsection 5.4.3 for the main gear, the landing gear designer can now rearrange the inclination of the trailing link respect to the vertical (δ) and the slope of the shock respect to the trailing link bigger bracket (σ), as well as move the attachment point of the shock on the trailing link (L1 and L2), checking that the geometric requirement 5.20, defined by the load ratio, is respected. Once the geometry is defined, hard-points corresponding to the pivot on the trailing link and the attachment points on the front bulkhead can be sized using simple static equilibrium equations.

Preliminary design of the steering system for the nose gear The pulley and belt mechanism has been preferred for the steering system, since it is the most reliable and the tension applied can be even adjusted with apposite tensioner, in case of harder steering necessities caused by rough conditions of the runway. In addition, its maintainability is easier respect to the other classical mechanical transmission, since all the units (driven pulley, input pulley and belt) can be replaced singularly. The initial sizing for the preliminary design of a pulley-belt steering system, sketched in ﬁgure 5.37, is accomplished deﬁning the center distance of the pulleys and the relative drive ratio. The design approach, even in this case, has been to consider whether standard sizes of pulleys could satisfy the needed drive ratio inside a space limited by the volume design, arranged and established in conjunction with the airframe design team.

Figure 5.37. Pulley-belt steering system design parameters. [37]

66 5 – Landing Gear design for FTV16.5%

The calculation of the steering parameters has been done using drive ratio tables contained in reference [37] and the software Center Distance Designer provided by the SDP/SI manufacturer. Due to the size of the nose gear shaft to steer it has been preferred to have as much contact area as possible and so the larger pulley, connected to the nose gear and driven by a smaller pulley, whose motion is activated by a dedicated servo. Therefore, the drive ratio results less than one and the transmission is step-down drive. A value of 0.7 has been chosen as desired drive ratio, since for steering system applications on UAVs it is not required an high rate of change of steering. The resulting preliminary steering system has the parameters shown in table 5.11, where N1, N2 and NB are respectively the number of teeth for the larger pulley, smaller pulley and belt.

Table 5.11. Preliminary selection of the pulley-belt steering system

Drive Ratio Center distance Pitch (P) N1 N2 NB 0.7 5.0187 in 0.080 30 21 151

The resulting geometry of the steering system is inside the design space and so it results to be a valid candidate for the steering system of FTV16.5. The next step of the preliminary design should be the procurement and characterization of a servo actuator to connect to the smaller pulley and then the evaluation of forces and torque needed to steer the nose gear shaft within a speciﬁc angle ψ with respect to the neutral position.

5.5 Weight, cost estimation and conclusions

The weight of the landing gear is one of the most important factor that, in most of the cases, guides the design process and selection of the components. A target of the landing gear weight can be established depending on some basic parameters of the aircraft and of the undercarriage system, using the procedure described in reference [30]. The mass estimation of the system can be calculated according to the empirical equation 5.21.

H W = K · K · K · W · LG · (N )0.2 (5.21) LG L ret LG L b g where Kret is equal to 1 for fixed landing gear, KL takes in account the extra landing gear weight needed to land on a carrier aircraft and so in this case is 1, KLG is the landing gear weight factor typically equal to 0.62 for general aviation and home-built aircraft, WL is the landing weight in this case assumed to be equal to the maximum take-off-weight that represents the worst case scenario, HLG is the landing gear height, b is the wing span, Ng is the selected landing ground load factor. The resulting landing gear weight target is 8 kg that represents 4.5% of the aircraft maximum take-off weight. The effective weight distribution estimation, at the phase of preliminary design that includes some defined off-the-shelf components, is shown on the left of figure 5.38. The total estimated weight, based on the preliminary sizing of basics components and selected standard components, is 9.5 kg. It represents 5.4% of the MTOW and is distributed as 79% on the main gear, mostly due to the strut leg design and braking system, and 21% on the nose gear mostly due to the internal steering system and trailing link structure. As notable the suspension system, for both nose and main gear assembly, results to be less than 5% of all the landing gear weight, even though the shock absorbers result the principle responsible components to dissipate the kinetic energy developed on landing. This first estimation of weight distribution can tend better towards to the target with the optimization of the design, successive to the detailed definition of all the components needed. The cost estimation of the landing gear has been conducted through a market analysis of the typical costs of components needed and purchased. [31] The most huge amount of cost is expected to be reserved for the production of the main gear strut legs, especially if they require custom mold design. The second system that affects the cost budget is the suspension system. Braking system and steering system have similar cost estimation and the effective value depends upon the use of standard components. 67 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 5.38. Weight and Cost estimation for the production of one landing gear set for FTV16.5

Conclusions The undercarriage system, at this point, results defined in all its basic functionality and the detailed sizing of all the needed components can be started as soon as the off-the-shelf components are tested according to the described testing procedures. The process requires close collaboration with the airframe design team in order to delineate the hard-points inside fuselage bulkhead. From the first definition of the landing gear layout it is important to have an idea of the performance of the aircraft with undercarriage on the most critical phases for a landing gear point of view: take-off and landing. A preliminary evaluation of the basic performances on take-off and landing of a simplified model of landing gear, will be presented for both 7% and 16.5%, in the next chapter.

68 Chapter 6

Basic performance evaluation of aircrafts with a landing gear system

The behaviour of the aircraft on ground in most of the cases is not linear, mainly due to the complexity of the landing gear in order to absorb energy on landing and dampen vibrations during the ground maneuverability. Dynamic loads are developed and their determination is important to evaluate the performance of the landing gear and aircraft itself during all the phases of preparation and flight. The basic performance evaluation is treated in the present chapter, assuming the gear as a linear or non- linear black-box depending on the type of loads developed on the particular phase analyzed. The basic dynamic response is evaluated for both FTV7 and FTV16.5 equipped with the undercarriages designed and described in chapters3 and5. The first dynamic phase analyzed is the taxiing/take-off run over a typical runway surface model, used to certify, according to FAR regulations, the dynamic response of the aircraft with landing gear: the necessity to evaluate the response of the aircraft with landing gear over an uneven runway is highlighted by the results of ground mock-ups for FTV7, as described in chapter 4. The other dynamic response evaluated for aircrafts with landing gear is the landing, assumed to be a two-point touch-down, as usually performed from aircraft with conventional tricycle landing gear.

6.1 Response of the aircraft with landing gear during the typical ground maneuvers

The taxiing response considered in this thesis work, is concerned to the entire phase of ground movements towards to the straight line motion on ground before the final take-off, without including braking and turning that are considered as separate operations. The dynamic calculations and simplifications used follow the rational criteria that are common used for the design of landing gear for bigger aircraft, in order to meet the certification requirements. [?]

6.1.1 Taxiing dynamic model During taxiing the non linear response of the aircraft in terms of heave and pitch motion, is mostly due to the non linearity of the runway above which the landing gear is demanded to move the aircraft. In order to empathize the irregularities of the ground for the dynamic response, the landing gear has been modeled as a simple linear spring/damper system. The reaction on the undercarriages has been evaluated considering the aircraft as a rigid body, the worst case scenario in terms of induced vibrations and bouncing on ground. Since every runway has its own profile, in a design phase, the aircraft response with landing gear is evaluated over a typical non-smooth runway, the San Francisco 28R1, according to the CS-25.491. [?] The runway is modeled as a series of different dips and bumps with 1-cosine shape, as shown in figure 6.1.

1San Francisco 28R is often used to evaluate the performance of the landing gear still in a design phase, so often it is called "Design runway". 69 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure 6.1. Aircraft taxiing operation over a 1-cosine modeled runway

The modeling of the runway depends upon the depth h (xr) of dips/bumps at the location xr respect to the reference coordinate system, usually located where the irregularity starts. The detailed procedure to model the runway is described in Appendix C. The dynamic model of the aircraft response is evaluated by the two degrees of freedom most involved in this phase, heave and pitch motion, applying the Lagrange method, as discussed in the Appendix C. The system of second order differential equations, shown in 6.1, has been reduced to first order differential equations and solved using the time-space domain in Matlab® & Simulink. The degree of freedoms are evaluated at the center of gravity position and the equivalent matrices that describe the system are dependent on the position of the nose and main gear with respect to the center of gravity and equivalent properties of linear stiffness and damping for the main and nose gear. ˙ ˜ z¨ ˜ z˙ ˜ z ˜ hN ˜ hN M ¨ + C ˙ + K = CR ˙ + KR (6.1) θ θ θ hM hM Note that the non linearity on the response is imposed by the right-hand side terms in the system of dynamic equations due to the variation in elevation of the runway surface.

6.1.2 Taxiing performance evaluation and comparison for FTV7 and FTV16.5 The time-history response of the aircraft during taxiing over a typically bumpy runway has been evaluated for both FTV7 and FTV16.5 and the major results, in terms of vertical acceleration and heave response, are illustrated and commented in the present subsection. The mass, geometry, stiﬀness and damping properties of the FTV7 and FTV16.5, used to solve the dynamic model are explicated in table 6.1.

Table 6.1. Input for the taxiing dynamic model

m Iy lN lM KN KM CN CM [kg] kgm2 [m] [m] [N/m] [N/m] [N s/m] [N s/m] FTV7 13.6 2 0.474 0.0526 6500 32067 50.726 229.25 FTV16.5 175 75 1.117 0.124 27450 82138 555 2508

The mass (m) and location of nose (lN ) and main gear (lM ) for both FTVs results from the scaling process of the full scale aircraft, with the center of gravity located at 56% of the main aerodynamic chord. The pitch inertia for FTV7 has been estimated using the approximate inertia process described in subsection 4.1.1, whilst for FTV16.5 the value has been obtained from the inertia load cases stated in the flying demo spec document in reference [2]. The linearized stiffness for the main and nose landing gear of FTV7 are the ones obtained from ground testing described respectively in subsections 3.2.4 and 3.3.5. For FTV16.5 instead, the stiffness properties result from the linearization of the polytropic theoretical relation used for a preliminary characterization of the shock absorbers, as described in subsection 5.4.2. The damping properties for FTV16.5 have been estimated using a similar linearization, and the resulting values have been scaled for FTV7 using the MTOW and stall speed formal factors. The two aircrafts have been evaluated when they encounter a 1-cosine bump, with depth and length respectively equal to ∆ (h) = 15mm and LR = 8m for FTV7, and ∆ (h) = 30mm and length LR = 15mm for FTV16.5, when the aircrafts are at 75% of the take-off speed. The values for the geometric modeling 70 6 – Basic performance evaluation of aircrafts with a landing gear system of the runway, have been estimated from the real condition of the airstrip in Merrit - Douglas Lake for FTV7, while for FTV16.5 typical values used for ultralight aircraft have been found in reference [39]. The heave acceleration and displacement response at the nose and main landing gear’s location, for the two flight test demonstrators, are respectively shown in the graphs a and b of figures 6.2 and 6.3.

Figure 6.2. Vertical acceleration during taxiing over a 1-cos dip runway for the FTV7 and FTV16.5

Figure 6.3. Heave response during taxiing over a 1-cos dip runway for the FTV7 and FTV16.5

The most relevant observations for the taxiing performance evaluation when the aircraft encounters a bump/dip, are underlined in the following itemize:

• Rebound induced by the bump: the negative acceleration and displacement values in the ﬁgures are caused by the rebound of the nose and main landing gear that occurs when they encounter the ground, that can be considered an inﬁnite rigid target.

• Maximum value of heave acceleration: for FTV7 the maximum value (0.55g) is registered on the nose gear when it encounters the bump, while for FTV16.5 the maximum (0.63g) is observable on the ﬁrst rebound of the nose gear, probably due to the nose gear shock absorber recoil during rebound. As observable the maximum acceleration transmitted to the landing gear structure and then transferred to the airframe, is almost the same for the two FTVs, since the two landing gears have been designed with structural similarities.

• Maximum value of heave displacement: the heave response follows the motion induced by the bouncing acceleration caused by the encounter of a bump inside the runway. The maximum peak value (18mm for FTV7 and 40mm for FTV16.5) is determined by the depth of the bump and the compression when the rigid target is encountered.

• Peaks and oscillations: the peaks are a measure of the rebound control performed by the landing gear. The peaks of the main gear follow the ones of the nose gear with a lag that is dependent on the location of the main gear with respect to the nose gear. For FTV16.5 the heave motion is 71 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

completely damped in 5 peaks within 1.5 seconds after the bump encounter, while the motion is not properly damped for the FTV7 since no specific damping components are used. Note that the bouncing effect on FTV7, could be significantly reduced by adopting tires with larger diameter and so higher surface to damp vibrations. In last measure, the analysis of the results is voted to understand the applicability of the adopted models and the reliability of the assumptions made. The linearization assumptions are fully true only for the nose gear of FTV7 since it is characterized by a linear progressive spring. The real behaviour of the main landing gear for FTV7 and both undercarriages of FTV16.5 is not linear since the suspension system is respectively made by a leaf strut, with behaviour similar to a parabola, and air/damper shock absorber with exponential behavior weighted by the index of gas compression inside the shock, similar to an isotherm reaction. On the other side the damping properties of the FTV16.5 shock absorber are dependent on the square of rate of position change of the piston, whilst for FTV7 the only damping is provided by the rubbers on the main gear and by inertial effect. Anyway, the simplified model explained in this section highlights the basic taxiing response of the aircraft empathizing the importance of the non linear geometry of the runway for the performance evaluation.

6.2 Response of the aircraft with landing gear during a typical design landing

Landing is the most critical phase for both landing gear and airframe design, because a significant amount of energy needs to be dissipate in order to reconduce safely the aircraft on ground. The landing phase analyzed in this section is the operation typically performed with the aircraft flaring and the nose gear still airborne at the moment of landing. The rates of descent used for both FTV7 and FTV16.5 performance evaluation are the same used for the design, that come from previous flights for FTV7 and FAR regulations for FTV16.5.

6.2.1 2-points landing dynamic model A simple representation of the aircraft during touch-down on the main landing gear, is to consider the wheel/tire assembly and the suspension system in series, distinguishing the aircraft mass m from the unsprung mass mT (that consists of wheels, tires, brake assemblies on the two wheels). The multi- system model analyzed is sketched in figure 6.4. The dynamic behavior of the landing gear is described by the equivalent stiffness KM and damping cM of the suspension system and an equivalent stiffness KT of the tire/wheel assembly. In this analysis, the main gear strut has been considered rigid, the suspension system aligned with the aircraft center of gravity and the tire without damping properties. These assumptions have been done since, for a nominal landing of FTV 16.5, it is assumed that all the impact load is transferred to the internal shock absorbers, the center of gravity is quite close to the ground contact point on the main landing gear and the damping forces on the tires are expected to be small in comparison with the damping properties of the shock. [39]

Figure 6.4. Aircraft model during landing

The aircraft displacement and tire compression are described by the quantities zM and zT , while the compression on the suspension system is expressed as zSA = zM − zT . The forces acting on the aircraft at the moment of the touch-down are the weight, residual lift and the non linear load on the shock absorber induced by the ground reaction inside the landing gear system, and for simplicity in this model they have been assumed to be applied on the aircraft center of gravity. The landing gear, on 72 6 – Basic performance evaluation of aircrafts with a landing gear system a landing dynamic analysis, needs to be considered non linear due to the non linear nature of impact. The dynamic behaviour is governed by the system of equation 6.2, where the last equation represent the initial condition of vertical speed of the mass and tire, assumed to be equal to the design vertical rate of descent VTD at the instant before the touch-down.  mz¨ + f (z , z˙ ) + mg − L = 0  M NL SA SA mz¨T − fNL (zSA, z˙SA) + KT zT = 0 (6.2)  z˙M (0) =z ˙T (0) = VTD The system of dynamic equation are modeled and solved in Simulink, whose block diagram is shown in Appendix C.

6.2.2 Landing performance evaluation and comparison for FTV7 and FTV16.5 The landing performance evaluation is measured in terms of impact acceleration developed and consequent wheel travel and shock absorber piston’s motion. The non linear properties of the landing gear have been introduced in the model thanks to look-up tables: the stiffness and damping properties for FTV16.5 are taken into account thanks to shock load graphs introduced in subsection 5.4.2; as for FTV7, the experimental non linear load-displacement curve, described in subsection 3.2.4, is imported in order to represent the structural response of the leaf strut. The stiffness effect on the tire has been assumed to be linear and estimated considering the maximum rolling radius upon MTOW loading distribution. The other necessary inputs (weight properties and geometry) are the same used for the taxiing model, introduced in table 6.1. The impact acceleration and vertical displacement time-history responses are shown in figures 6.5 and 6.6

Figure 6.5. Vertical impact deceleration on the main gear of FTV7 and FTV16.5

Figure 6.6. Vertical displacements on landing on the main landing gear of FTV7 and FTV16.5

73 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

The most valuable results deduced from the graphs are itemized as follows:

• Tire effect: Note that adding a simple tire model for both FTV7 and FTV16.5 the acceleration does not tend to infinite values at the moment of impact. • Peak of impact acceleration: the impact acceleration is a function of landing weight and of the vertical rate of descent before the aircraft touches down. The maximum peak value for both FTVs is comparable with the design load factor assumed to be 3 for both landing gear design cases. In terms of structure response it is important the values above the design load factor, if present, are limited to a small amount of time so that they cannot induce permanent and catastrophic deformations: as notable on the graphs, this is respected for both FTV7 and FTV16.5. • Vertical bouncing: The vertical bouncing on the main gear in this analysis does not consider the effect of the nose gear that rotates and touches the ground. • Displacements: the wheel travel (25mm) consequent to impact landing is quite similar to the one obtained in the ground drop test (27.4mm) shown in subsection 3.3.5. In the case of FTV16.5, the displacement consequent to impact landing, is distinguished between the wheel travel and piston’s motion of the shock absorber. The large wheel travel (the maximum value is 130 mm) is due to the nature of the suspension system finalized to have the maximum gear displacement controlled by the shock absorber’s stroke. The shock absorber in the model, has been considered in order to have the full stroke when the impact acceleration presents a peak. Although the dynamic landing model is a good mean to verify the aircraft response with landing gear in a phase of preliminary design, it needs to be improved including the real geometry of the landing gear, that in the case of FTV16.5 can be finalized only when the internal location of the shock absorbers is determine. For the best accurate response of the aircraft with landing gear, separated main gear models and nose gear effect should be considered in the analysis, and the model should be integrated with the complete Simulink model of the aircraft: this approach can lead to the performance evaluation of all the 9 degrees of freedom involved on the aircraft during landing.

74 Chapter 7

Conclusions and Future developments

7.1 Conclusions

The design and development from scratch of a fully operative landing gear, applied in this dissertation for two different scales of aircraft, lead to several causes for reflection that highlight the complexity related to experimental design. The nature of Unmanned Aerial Vehicles and their applications require iterations, ground testing and eventual redesign of components in order to improve and optimize the final layout and behaviour of the landing gear. Both FTV7% and FTV16.5% applications, evidence that the landing gear design for flight test demonstrators of bigger aircrafts, goes far beyond merely scaling down the size of the full scale solution. Although the designer has not wide freedom to change some parameters that define the final geometry and disposition of the undercarriages, he needs to have in mind that the resulting behaviour on impact landing and the complexity of the suspension system depends on the specific scale of the aircraft and on the operational conditions of the runway. The small scale of FTV7% does not allow a geometric scaling of wheel assemblies and so an additional requirement of maximum 25 % of extra size, as stated in subsection 3.1.4, has been established with the stakeholders. Furthermore the need to test the feasibility of take-off and landing operations, within a large range of center of gravity, has not allowed the positioning of the main landing gear at the scaled down value, because the take-off requirement, stated in section 2.1, and a reasonable load ratio distribution between the nose and main landing gear, should be respected for each c.g. position. The design for both aircrafts has been conditioned by cost, weight, availability and manufacturability constraints. As introduced in the current state of the art of landing gear for similar aircraft applications in section 2.3, the market research for aircraft off-the-shelf components that alleviate the cost of a custom solution, has been significantly complicated passing from the small scale FTV7% to FTV16.5% and solutions from other field of applications have represented the best alternative, even though appropriate tests are required to define the feasibility and integration with the overall design. The specific design of the landing gear for the small scale FTV has underlined the difficulties to integrate a new system into an already built airframe, especially because the available space inside the bays of the aircraft has not been designed to house the internal components of the nose steering system and attachment plates of the main gear leaf strut and so design trade-off and redistribution of internal components have been constantly necessary. Once the landing gear has been installed into the airframes of the two tailed and no-tail configurations of FTV7%, new ground mock-ups have been needed in order to validate the behaviour of the aircraft with undercarriages on the runway, and improve the design of some components (tire size and horns for the steering system), before proceeding straight to a flight test. The design of the undercarriages for FTV16.5% has shown up that the landing gear is not a "Cinderella subject" but, in some cases, it becomes a significant design driver, influencing the external and internal layout of the airframe since the initial phases of conceptual and preliminary design of a new aircraft. Therefore close collaboration with the other design teams of the aircraft (airframe, recovery system, system controls) has been maintained along all the crucial design decisions, avoiding possible interferences and inconsistencies. The two designed landing gears have been finally evaluated in terms of heave response during the most critical phases of take-off over a bumpy runway and landing. The irregularities on the two runways 75 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators chosen as model for the ground application of FTV7 and FTV16.5, as discussed in chapter6, produce similar average heave acceleration in the two landing gear structures, even if the induced oscillations result almost no damped for the 7% and well damped for the 16.5%, due to the different suspension system implemented. As for the heave response transmitted after touching the ground in a design landing, the two landing gears receive a similar normalized acceleration because both of them are designed to sustain nominally a 3G landing, but the resulting motion on the two applications, although they have a similar trend, is characterized by a different travel distribution depending on the weight of the aircraft and on the suspension system used.

7.2 Future developments

The FTV7%, in both tailed and no-tail configurations, is awaiting for its first flight test campaign with landing gear. The flight tests with variable position of the main landing gear and the successive data analysis process are already planned, as discussed in subsection 4.3.1. The flight test results can validate the structural response of the landing gear, estimated in the initial phase of the design, using structural simulation tools, and then verified by ground tests. Furthermore they can underline what is the performance of the aircraft with landing gear during take-off and landing at different position of center of gravity: the response evaluation can lead to the validation of the developed model described in chapter6 and to the definition of the best position of c.g. with the preferred control setting for the autopilot, that can support further design decisions for the landing gear of FTV16.5%. As for the landing gear of FTV16.5% the progress in the design will follow the development of the airframe of the aircraft, expected to be ready for production in 2019. The preliminary design can continue following the procedures described in subsections 5.4.3 and 5.4.4, once the shock absorbers for both main and nose gear will be characterized after testing, according to the steps defined in subsection 5.4.2. The preliminary phase of the project will be ultimated once all the aspect introduced in chapter5 will be finalized, including the definition of all the components and design parameters of the needed mechanical subsystems (steering and braking system) and the definition of the required test rigs and procedures to verify the behaviour of the landing gear assemblies. In the next stages the design will be refined and finalized in every detail so that all the parts will be geometrically and structurally fully determined and ready to be manufactured. In this phase, all the possible failure modes during parachute emergency landing and the resulting risk management will be considered, in conjunction with the Recovery System Team. If no redesign will be required, the component list definition and assembly CAD drawings will be prepared and the project will move to the procurement and readiness phase with the landing gear manufactured and assembled. After that the assembled landing gear will be ready to be ground tested and integrated with the first developments of the airframe. The whole aircraft, with landing gear, is then scheduled to be ready to fly in summer 2019.

76 Appendix A

Landing gear Matlab parameter calculator

The MatLab® code, attached to the present appendix, computes the basic parameters and features of the landing gear depending to some input that can be deﬁned by the user at the beginning of the program. It is an useful tool that can be used for the initial sizing of the landing gear, where some parameters are not fully deﬁned and iterations are necessary.

A.1 Input parameters

The input parameters, that the landing gear designer has to deﬁne according to a list of requirements list, are:

• Aircraft and performance data:

– Maximum Take of Weight MTOW; – Mean aerodynamic chord M.A.C. length and location; – Wing reference area; – Rate of descent; – Aircraft conﬁguration: tailed or no tail;

• Geometry input:

– Location of the nose gear; – Center of gravity position; – Wheel track for the main gear; – Size of the wheel assemblies;

• Design choices:

1 – Ground reaction load factor Ng ; – Vertical load ratio between nose gear and main gear;

A.2 Output parameters

The output data computed by the program are:

• Geometry output:

– Location of the main gear;

1It is the ratio between dynamic and static loads, used to estimate the ground reaction loads during impact 77 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

– Geometry angle: ∗ Angle A between the C.G. and the vertical line on the main gear; ∗ Angle C between the C.G. and the vertical line on the nose gear;

∗ Tip-back angle αtb;

∗ Angle αM−N between the nose gear and main gear outboard wheel; ∗ Turnover angle ψ;

• Impact parameters:

– Wheel travel of the landing gear; – Normalized acceleration developed on landing, depending on the rate of descent and compression of the strut; – Longitudinal spin-up and spring-back loads developed on landing;

• Other design parameters:

– Take-off Rotation requirement verification; – Braking system sizing: ∗ Brake kinetic energy; ∗ Torque required per each wheel; ∗ Horizontal and vertical load during braking; – Shock absorber sizing: ∗ Stiffness load; ∗ Damping load; – Drop Test parameter;

The code, applied for FTV16.5, is listed as follows. 1 clc 2 clear all 3 4 5 %%FIXEDINPUT 6 7 MTOW=175.24;%%[kg] 8 S_ref=50.19;%%[ft^2] 9 MAC=1.5697e+03;%%[mm] Mean aerodynamic chord from Horizon-16.5%BA document 10 x_MAC=886.46;%%[mm] Leading edge of theMAC 11 x_nose=647.70;%%[mm] Location fromBBA- scaled down dimension 12 track_main=1388.71;%% Scaled outboard wheel distance for the Main Gear calculated fromBACAD 13 V_land=39.93;%%[m/s] 14 g =9.81; 15 16 %%VARIABLEINPUT 17 18 %%1PARAMETER:CGPOSITION 19 percentage_cg_MAC=58; 20 21 %%2PARAMETER:LOADRATIO 22 percentage_load_main=90; 23 percentage_load_nose=10; 24 load_ratio=percentage_load_nose/percentage_load_main;%% Nose gear Load over Main Gear Load 25 26 %%3PARAMETER:WEIGHT 78 A – Landing gear Matlab parameter calculator

27 W= MTOW ; 28 29 %%4PARAMETER:TAILORNOTAIL 30 tail =1;%% 1=tail configuration, 0=no tail configuration 31 32 %%5COMPRESSION 33 d_scaled=64.11/1000;%% Scaled down compression[m]; The actual compression is 92.29 mm 34 d_design=92.29/1000; 35 d=0.0300:0.0001:0.100; 36 37 %%6RATEOFDESCENT 38 V_z_FAR=7.33*0.3048;%%[m/s] fromFAR 23. 15 fps from Parachute landing . 39 V_z_parachute=15*0.3048; 40 41 %%7 Design Ground load Factor 42 N_g =3;%% Ground Reaction landing factor to estimate vertical dynamic loads, suggested byFAR 23. 43 44 45 %%GEOMETRYCALCULATOR 46 47 alpha=asin((V_z_FAR/V_land))*180/pi; 48 V_horizontal=V_land*cos(alpha*pi/180); 49 50 x_cg=(percentage_cg_MAC/100*MAC)+x_MAC;%%[mm] Longitudinal position of thec.g. 51 z_nose=346.55;%%[mm] Vertical location of the nose reference to the ground reference line 52 z_cg=z_nose+56.90;%%[mm] Vertical loacation of thec.g. with respect to the ground reference line. Max value, from documentBOMBARDIERSPEC REVD 53 54 l_n=x_cg-x_nose;%%[mm]. Position of the nose gear with respect toC.G. 55 l_m=l_n*(load_ratio);%%[mm]. Position of the main gear with respect to C.G. 56 x_main=l_m+x_cg;%%[mm] Longitudinal position of them.g. 57 58 wheelbase=l_n+l_m; 59 60 A=(atan(l_m/z_cg))*180/pi;%% Angle between the vertical line on the main gear and the line to theC.G. 61 C=(atan(l_n/z_cg))*180/pi;%% Angle between the vertical line on the nose gear and the line to theC.G. 62 if(tail==1) 63 x_tip=3496.84;%%[mm] Location of the tip and farther point of the aircraft. Data from Solidworks model 64 x_tip_contact=x_tip-x_main;%%[mm] distance between the contact point and tip of the tail 65 y_tip_contact=433.45;%%[mm] vertical distance between the contact point and the tip of the tail 66 else 67 x_tip=3038.87;%%[mm] Location of the tip and farther point of the aircraft. Data from Solidworks model 68 x_tip_contact=x_tip-x_main;%%[mm] distance between the contact point and tip of the tail

79 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

69 y_tip_contact=373.39;%%[mm] vertical distance between the contact point and the tip of the tail 70 end 71 72 alpha_tipback=(atan(y_tip_contact/x_tip_contact))*180/pi;%% Tipback angle, calculated using dimensions form the Solidworks model 73 alpha_main_nose=(asin((track_main)/(2*(wheelbase))))*180/pi;%% In plane angle between the Nose gear and the outboard wheel 74 alpha_turnover=atan(z_cg/(l_n*sin(alpha_main_nose*pi/180)))*180/pi;%% Turnover angle in the cross plane. the maximum Turnover angle according toFAR 23 is 55 degree 75 76 %%LOADCALCULATOR 77 78 %Static Loads 79 P_m=(W*l_n)/(2*(wheelbase));%%[kg] Static Load on each wheel of the Main Gear 80 P_n=W*l_m/(wheelbase);%%[kg] Static Load on the Nose Gear 81 P_m_max=N_g*P_m;%%[kg] Peak vertical load on each wheel at landing 82 P_n_max=N_g*P_n;%%[kg] Peak vertical load on the nose gear at landing 83 84 static_loads = [P_m P_n]; 85 maximum_loads = [P_m_max P_n_max]; 86 87 %Impact Loads 88 a_g_FAR_design=V_z_FAR^2/(2*d_design*g);%average acceleration developed during landing, consideringa compression of 92.29 mm 89 a_g_parachute_design=V_z_parachute^2/(2*d_design*g);%average acceleration developed during parachute landing vonsideringa compression of 92.29 mm 90 91 % Fixed Rate of descent 92 a_g_parachute_scaled=V_z_parachute^2./(2*d*g); 93 a_g_FAR_scaled=V_z_FAR^2./(2*d*g); 94 % 95 plot(d*1000, a_g_parachute_scaled) 96 hold on 97 grid MINOR 98 plot(d*1000, a_g_FAR_scaled) 99 xlabel (' Compression[mm] '); 100 ylabel (' Normalized acceleration[a/g] '); 101 title (' Normalized acceleration depending on the wheel travel '); 102 legend ('Rate of descent: 15 fps[Parachute Landing] ','Rate of descent: 7.33 fps[FAR Calculation] '); 103 104 % Fixed Compression 105 d_scaled=64.11;%[mm] 106 d_CAD=92.29;%[mm] 107 d_FAR=84.8;%[mm] 108 109 V_z=0:0.1:15;%[fps]. 15 fps is the parachute landing 110 a_g_scaled=(V_z.*0.3048).^2./(2*(d_scaled/1000)*g); 111 a_g_CAD=(V_z.*0.3048).^2./(2*(d_CAD/1000)*g); 112 a_g_FAR=(V_z.*0.3048).^2./(2*(d_FAR/1000)*g); 113 114 figure 115 plot(V_z, a_g_scaled) 116 hold on

80 A – Landing gear Matlab parameter calculator

117 plot(V_z, a_g_FAR) 118 plot(V_z, a_g_CAD) 119 grid MINOR 120 xlabel ('Rate of descent[fps] '); 121 ylabel (' Normalized acceleration[a/g] '); 122 title (' Normalized acceleration depending on the Rate of Descent '); 123 legend ('Scaled down compression: 64.11 mm ', 'Nominal Landing Compression : 84.8 mm ', 'Parachute Landing Compression: 92.29 mm '); 124 125 %%TAKE-OFFROTATIONREQUIREMENT 126 %Altitudes 127 T_0 =288;%% Temperature atz 128 h =6.5; 129 z =0.887;%% Foremost airportaltitude in[km] 130 z_c =z +0.3;%% cruise altitude at 200m above theRWY 131 rho_0=1.225;%% Air density at sea level[kg/m^3] 132 rho=rho_0*((T_0-h*z)/T_0)^4.2561;%% Air density at the airport height 133 rho_c=1.0146;%% Air density at cruise height 134 135 % Speeds 136 137 V_S =32.46;%%[m/s] Stall speed at atMTOW clean configuration[From aerodynamic team] 138 V_R=1.2*V_S;%% Rotation speed atTO: 1.1 to 1.3V_S 139 V_c =52;%%[m/s] Cruise speed from spec documentBBA] 140 141 % Geometric data wing and tail 142 AR=b^2/S_ref;%% Aspect Ratio 143 S_h=0.67075;%% Tail surface: fromBBA documentation: 7.22 sqft 144 x_acwf=(0.25*MAC)+34.9*0.0254;%% aerodynamic center wing fuselage: 25% MAC, Leading edge at 34.9" of theF.S.[according to the document Horizon-16.5_Flying_Demo_Specs_RevB] 145 x_ach=3.1056;%% aerodynamic center tail:fromBBA documentation: 122.27 in 146 z_T =0.637;%% Arm of the Thrust with respect to the ground reference line[From Solidworks 16.5%]: center of the Nacelle 147 z_nose=0.412;%%177.22 is the height of the Nose station with respect theRWY 148 z_cg=0.46914;%% vertical position of thec.g. Maximum position from the document [0066-BBA 16.5%LANDINGGEARSYSTEM]. 149 z_D = z_cg ;%% Arm of the Drag with respect to the ground reference line. It has been assumed thatZ_D equalzc.g. 150 x_cg=0.56*MAC+34.9*0.0254;%% forward limit of thec.g. at 56%MAC 151 152 % Aerodnamic coefficients 153 C_D0TO=0.0668;%% C_D0 atTO with flaps andU-Tail 154 e =0.92;%% Ostwald factor 155 K=1/(e*pi*AR); 156 C_LC=(2*W)/(rho_c*V_c^2*S_ref);%%C_L cruise 157 C_Lflap =1;%% extra_lift coefficient atTO due to flap. No Flap are for the GEN2B and GEN2C 158 C_LTO=C_LC+C_Lflap;%%C_L atTO 159 C_D=C_D0TO+K*C_LTO^2;%%C_D atTO 160 C_macwf=-0.005;%% C_m0 wing-body 161 C_Lh = -1.1;%% Negative Lift coefficient Tail/elevon atTO 162 163 % Forces 164 D=0.5*rho*V_R^2*C_D*S_ref;%% Total Drag atTO

81 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

165 M_acwf=0.5*rho*V_R^2*C_macwf*S_ref*MAC;%% Pitching moment wing-body at thea.c. 166 L_TO=0.5*rho*V_R^2*S_ref*C_LTO;%% Total Lift atTO 167 L_h=0.5*rho*V_R^2*S_h*C_Lh;%% Lift Tail 168 L_wf=L_TO-L_h;%% Lift wing fuselage atTO 169 N=W- L_TO ;%% Normal force on the Main gear atTO 170 F_F =mu*N;%% Friction force 171 a_x=(T-D-F_F)*g/W;%% longitudinal acceleration[m/s2] 172 173 % Calculation of the longitudinal position of the main gear with respect to theFWD limit ofC.G.[Reference Mohammad, Sadraey- Aircraft Design& Systems Engineering Approach] 174 I_yy=90.718;%% Inertia of the aircraftw.r.t.C.G 175 I_yymg=I_yy;%% Inertia of the aircraft atTO with respect to the ground contact point at the main gear 176 theta_ddot=15*pi/180;%% required rotation acceleration atTO[deg/s^2]: Fora normal light general aviation aircraft it is 8-10 deg/s^2 177 x_mg_loop=(I_yymg*theta_ddot-D*z_D+T*z_T+M_acwf-(W/g)*a_x*(z_cg)-W*x_cg+ L_wf*x_acwf-L_h*x_ach)/(L_wf+L_h-W); 178 x_mg=1.905; 179 tail =1; 180 if(tail==1) 181 while(abs(x_mg-x_mg_loop)>1e-6) 182 I_yymg=I_yy+W/g*(x_mg-x_cg)^2; 183 x_mg_loop=x_mg; 184 x_mg=(I_yymg*theta_ddot-D*z_D+T*z_T+M_acwf-(W/g)*a_x*(z_cg)-W* x_cg+L_wf*x_acwf-L_h*x_ach)/(L_wf+L_h-W); 185 end 186 else 187 L_h =0; 188 while(abs(x_mg-x_mg_loop)>1e-6) 189 I_yymg=I_yy+W/g*(x_mg-x_cg)^2; 190 x_mg_loop=x_mg; 191 x_mg=(I_yymg*theta_ddot-D*z_D+T*z_T+M_acwf-(W/g)*a_x*(z_cg)-W* x_cg+L_wf*x_acwf-L_h*x_ach)/(L_wf+L_h-W); 192 end 193 end 194 195 %%SHOCKABSORBERSIZING 196 197 %Data 198 mh = m / 2;% Factor total mass for half aircraft 199 Wh = mh * 9.81; 200 Lh = Wh;% Lift balances weight 201 W_e = 2.23;% Vertical speed on landing(TAS) 202 203 % Single main landing gear parameters leading to non-linear behaviour 204 Pinf = 3.4e5;%Extended position pressure [50 psi] 205 PS = 5.4 e5;%Static position Pressure 206 PC = 2.4 e6;%Compressed position pressure 207 PA = 1e5;%Atmospheric pressure 208 zS = 0.0381;%Maximum Stroke[m] 209 Area = 1.6e-3%Wh/PS; 210 Vratio = PC / Pinf; 211 Vinf = Vratio * Area * zS / (Vratio - 1); 212 zinf = Vratio / (Vratio - 1) * zS;

82 A – Landing gear Matlab parameter calculator

213 ns = 1; 214 nd = 1.35; 215 216 dz = 0.001; z = [0: dz: zS]; [dummy, nz] = size(z); 217 for j=1:nz 218 Pd(j) = Pinf / (1 - z(j) / zinf)^nd; 219 Fd(j) = (Pd(j) - PA) * Area; 220 end 221 222 plot(z,Fd); 223 hold on 224 grid on 225 xlabel (' Displacement '); 226 ylabel ('Force '); 227 228 229 %%BRAKINGSYSTEMSIZING 230 distance_landing=250;%%[m]. Use Mohammed 231 t_stop=(2*distance_landing)/V_horizontal; 232 a_braking=V_horizontal/t_stop; 233 F_stop=MTOW*a_braking; 234 KE=0.5*(MTOW)*V_horizontal^2;% KJoule 235 KE_req_per_wheel=(KE/(0.737562))/2;% Ft lbs 236 237 tire_main_diameter=6;%% inches 238 Torque=((MTOW/2)*a_braking)*((tire_main_diameter*25.4/1000)/2);%%[Nm] Torque required per each wheel 239 Torque_req=Torque*8.851;%%[in lbs] Torque required per each wheel 240 241 F_N_braked=static_loads(2)+W*a_braking*(z_cg)/(wheelbase);%% Vertical loads on the Nose gear with brakes applied 242 F_M_braked=((static_loads(1))*2-W*a_braking*(z_cg)/(wheelbase))/2;%% Vertical loads on the Main gear with brakes applied 243 braked_loads=[F_N_braked F_M_braked]; 244 245 %%DROPTESTPARAMETERS 246 W_e=MTOW*0.454; 247 h=(3.6*(W_e/S_ref)^0.5);%FAR 23[mm]

83 84 Appendix B

CAD assembly drawings

B.1 Assembly drawings and Bill of components for the Landing Gear for FTV7%

This appendix shows all the assembly drawings that define the bill of components used for the landing gear for FTV7. The main gear drawings are illustrated in figures B.1 and B.2, while the nose assembly drawing are illustrated in figures 4.2, B.3 and B.4

B.2 Assembly drawings for the Shock Absorber Test Rig

The shock absorber Test Rig described in subsection 5.3.3 of chapter 5, is ready for manufacturing and its CAD assembly drawing with all the components needed is shown in ﬁgures B.5 and B.6.

85 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

8 7 6 5 4 3 2 1

F F 2 6

3 4

5 1 E 5 E

3 D D 7

C C

Item No. Part Number Description Quantity 1 0060-0206 Main gear strut 1 Locknut for 10-32 2 97231A440 screw 6 Washer for 10 3 91525A115 screw size 12 Support PROPRIETARY AND CONFIDENTIAL UNLESS OTHERWISE SPECIFIED 4 0060-0209 2 THE INFORMATION CONTAINED IN THIS DIMENSIONS ARE IN INCHES DO NOT SCALE DRAWING B attachment plate DRAWING IS THE SOLE PROPERTY OF B UVIC CENTRE FOR AEROSPACE RESEARCH MACHINING REQUIREMENTS: GENERAL TOLERANCES: (CfAR). ANY REPRODUCTION IN PART R63 5 1296N21 Rubber sheet 4 OR AS A WHOLE WITHOUT THE WRITTEN - BREAK SHARP CORNERS AND EDGES 0.005" MAX. +0.005" PERMISSION OF UVIC CfAR IS PROHIBITED. X.X ±0.02" HOLE ≤1": -0.002" N7 - INTERNAL CORNERS OF CUTOUT R0.010" MAX. X.XX ±0.005" 6 Aircraft Skin Aircraft Skin 1 CfAR OFFICE NUMBER BETWEEN HOLES: ±0.005" Socket head Screw - C'SINK INTERNAL THREAD 90 TO MAJOR DIA. X.XX5 ±0.005" 7 92200A347 6 778-351-1926 - CHAMFER EXTERNAL THREAD 45 TO MINOR DIA. X.XXX ±0.002" ANGLES: ±0.1° 10-32, 1" long NAME DATE

DRAWN Fabrizio Rizzi PARENT ASSEMBLY CHECKED This part is designed to be APPROVED made by A TITLE: Component List exploded view 4 A WEIGHT: [Kg]; [Lbs]

SIZE DWG NO. FILE REV DWG REV FINISH: B MATERIAL: SCALE: 1:2 DIMENSIONS ARE IN SHEET: 1 of 1

8 7 6 5 4 R:\0066-BOMBARDIER-16.5%3 LANDING GEAR\DESIGN\MECHANICAL\7 %\7%_CAD_LANDING_GEAR\SWEPT_VERSION\FULL

Figure B.1. Main Gear assembly drawing 1: leaf strut and attachment with the airframe

8 7 6 5 4 3 2 1

1 F 8 F

E E 10 9

D D

10 12

C C

Item No. Part Number Description Quantity 1 0060-0206 Main Gear Strut 1 Tire and wheel Tire and wheel 8 70mm 70mm 1 9 92981A042 Shoulder screw 1 PROPRIETARY AND CONFIDENTIAL UNLESS OTHERWISE SPECIFIED THE INFORMATION CONTAINED IN THIS DIMENSIONS ARE IN INCHES DO NOT SCALE DRAWING B 5mm, M4x0.7mm DRAWING IS THE SOLE PROPERTY OF B UVIC CENTRE FOR AEROSPACE RESEARCH MACHINING REQUIREMENTS: GENERAL TOLERANCES: Washer for M5 (CfAR). ANY REPRODUCTION IN PART R63 10 91116A140 2 OR AS A WHOLE WITHOUT THE WRITTEN - BREAK SHARP CORNERS AND EDGES 0.005" MAX. +0.005" screw PERMISSION OF UVIC CfAR IS PROHIBITED. X.X ±0.02" HOLE ≤1": -0.002" N7 - INTERNAL CORNERS OF CUTOUT R0.010" MAX. X.XX ±0.005" Spacer for M5 CfAR OFFICE NUMBER BETWEEN HOLES: ±0.005" 11 92871A042 1 - C'SINK INTERNAL THREAD 90 TO MAJOR DIA. X.XX5 ±0.005" screw, 6mm long 778-351-1926 - CHAMFER EXTERNAL THREAD 45 TO MINOR DIA. X.XXX ±0.002" ANGLES: ±0.1° 12 92871A031 Spacer for M5 1 NAME DATE screw, 2 mm long DRAWN Fabrizio Rizzi PARENT ASSEMBLY 13 94710A101 Locknut for M4x0.7 1 CHECKED This part is designed to be mm APPROVED made by A TITLE: Component List exploded view 5 A WEIGHT: [Kg]; [Lbs]

SIZE DWG NO. FILE REV DWG REV FINISH: B MATERIAL: SCALE: 1:1 DIMENSIONS ARE IN SHEET: 1 of 1

8 7 6 5 R:\0066-BOMBARDIER-16.5%4 LANDING GEAR\DESIGN\MECHANICAL\73 %\7%_CAD_LANDING_GEAR\SWEPT_VERSION\MAIN_GEAR_ASSEMBLY_WITH_SWEPT

Figure B.2. Main Gear assembly drawing 2: wheel assembly connection with the leaf strut

86 B – CAD assembly drawings

8 7 6 5 4 3 2 1

F F

E 4 Item No. Part Number Description Note E Shoulder screw 1 93985A209 3/16", 1-1/4" 2 90945A760 Washer 1/4" 3 is connected to 3 0060-0220 Bushing for the Rim the rim 4 Nose Gear Piston 1 strut D 5 92311A321 Set screw 4-48 D Bushing for the 6 is connected to 6 0060-0221 Nose gear strut the nose gear strut 8 7 Nose gear Tire and 2 Rim 8 90633A009 Locknut 8-32 3 3 C 5 C 7 6

PROPRIETARY AND CONFIDENTIAL UNLESS OTHERWISE SPECIFIED THE INFORMATION CONTAINED IN THIS DIMENSIONS ARE IN INCHES DO NOT SCALE DRAWING B DRAWING IS THE SOLE PROPERTY OF B UVIC CENTRE FOR AEROSPACE RESEARCH MACHINING REQUIREMENTS: GENERAL TOLERANCES: (CfAR). ANY REPRODUCTION IN PART R63 OR AS A WHOLE WITHOUT THE WRITTEN - BREAK SHARP CORNERS AND EDGES 0.005" MAX. +0.005" PERMISSION OF UVIC CfAR IS PROHIBITED. X.X ±0.02" HOLE ≤1": -0.002" N7 - INTERNAL CORNERS OF CUTOUT R0.010" MAX. X.XX ±0.005" BETWEEN HOLES: ±0.005" CfAR OFFICE NUMBER - C'SINK INTERNAL THREAD 90 TO MAJOR DIA. X.XX5 ±0.005" 778-351-1926 - CHAMFER EXTERNAL THREAD 45 TO MINOR DIA. X.XXX ±0.002" ANGLES: ±0.1° NAME DATE

DRAWN Fabrizio Rizzi PARENT ASSEMBLY CHECKED This part is designed to be APPROVED made by A TITLE: Component List exploded view 1 A WEIGHT: [Kg]; [Lbs]

SIZE DWG NO. FILE REV DWG REV FINISH: B MATERIAL: SCALE: 1:2 DIMENSIONS ARE IN SHEET: 1 of 1

8 7 6 5 4 R:\0066-BOMBARDIER-16.5%3 LANDING GEAR\DESIGN\MECHANICAL\7 %\7%_CAD_LANDING_GEAR\SWEPT_VERSION\FULL

Figure B.3. Nose gear assembly drawing 2: wheel assembly and piston strut

8 7 6 5 4 3 2 1

F F

Item No. Part Number Description Quantity Support Plate for 15 0060-0210 the Nose Gear 1 29 28 Linkage for the 24 2444 steering system 1 3D printed support E 28 0060-0225 for the servo 1 E Servo for the 29 HV6130 steering system 1 Coupler for the 30 30 20170919001 servo 1 91221A112/95868A Glass-Filled Nylon 31 15 31 screw/ nylon screw 1 132 2-56 D Bushing for the D 32 0060-0219 steering shaft 1 33 0060-0217 Steering shaft 1 Horn for the 34 0060-0212 steering shaft 1 Socket Head 35 92196A109 Screw 4-40, 7/16" 1 32 long 25 98019A216 Washer for N4 1 C 37 91834A102 Nut for 4-40 screw 1 C 34 33 37

35 PROPRIETARY AND CONFIDENTIAL UNLESS OTHERWISE SPECIFIED THE INFORMATION CONTAINED IN THIS DIMENSIONS ARE IN INCHES DO NOT SCALE DRAWING B DRAWING IS THE SOLE PROPERTY OF B 25 UVIC CENTRE FOR AEROSPACE RESEARCH MACHINING REQUIREMENTS: GENERAL TOLERANCES: (CfAR). ANY REPRODUCTION IN PART R63 OR AS A WHOLE WITHOUT THE WRITTEN - BREAK SHARP CORNERS AND EDGES 0.005" MAX. +0.005" PERMISSION OF UVIC CfAR IS PROHIBITED. X.X ±0.02" HOLE ≤1": -0.002" N7 - INTERNAL CORNERS OF CUTOUT R0.010" MAX. X.XX ±0.005" 24 BETWEEN HOLES: ±0.005" CfAR OFFICE NUMBER - C'SINK INTERNAL THREAD 90 TO MAJOR DIA. X.XX5 ±0.005" 778-351-1926 - CHAMFER EXTERNAL THREAD 45 TO MINOR DIA. X.XXX ±0.002" ANGLES: ±0.1° NAME DATE

DRAWN Fabrizio Rizzi PARENT ASSEMBLY CHECKED This part is designed to be APPROVED made by A TITLE: Component List exploded view 3 A WEIGHT: [Kg]; [Lbs]

SIZE DWG NO. FILE REV DWG REV FINISH: B MATERIAL: SCALE: 1:1 DIMENSIONS ARE IN SHEET: 1 of 1

8 7 6 5 4 R:\0066-BOMBARDIER-16.5%3 LANDING GEAR\DESIGN\MECHANICAL\7 %\7%_CAD_LANDING_GEAR\SWEPT_VERSION\FULL

Figure B.4. Nose gear assembly drawing 3: steering system

87 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

8 7 6 5 4 3 2 1

5 11 F F 6 12

E E 4 10

Item No. PN Description QTY D 1 Shock 6.5" 1 D 2 A0001-0001 Rect. Bar 2 3 A0001-0003 Insert 2 2 4 A0001-0004 Vertical Pivot Plate 1 5 A0001-0005 Angle for the bar 2 6 A0001-0006 Angle for the shock 2 7 1/2-20 screw 2 8 92196A389 1/2-20 screw 1 C C 9 90128A961 #10-32 screw 1 10 92390A908 Pin 3/8" 1 11 98306A809 Pin 3/8" 2 12 91864A062 1/4-20 screw 8 7

1 PROPRIETARY AND CONFIDENTIAL UNLESS OTHERWISE SPECIFIED THE INFORMATION CONTAINED IN THIS DIMENSIONS ARE IN INCHES DO NOT SCALE DRAWING B DRAWING IS THE SOLE PROPERTY OF B 3 2 UVIC CENTRE FOR AEROSPACE RESEARCH MACHINING REQUIREMENTS: GENERAL TOLERANCES: (CfAR). ANY REPRODUCTION IN PART R63 OR AS A WHOLE WITHOUT THE WRITTEN - BREAK SHARP CORNERS AND EDGES 0.005" MAX. +0.005" PERMISSION OF UVIC CfAR IS PROHIBITED. X.X ±0.02" HOLE ≤1": -0.002" N7 - INTERNAL CORNERS OF CUTOUT R0.010" MAX. X.XX ±0.005" 9 BETWEEN HOLES: ±0.005" CfAR OFFICE NUMBER - C'SINK INTERNAL THREAD 90 TO MAJOR DIA. X.XX5 ±0.005" 778-351-1926 - CHAMFER EXTERNAL THREAD 45 TO MINOR DIA. X.XXX ±0.002" ANGLES: ±0.1° NAME DATE

DRAWN PARENT ASSEMBLY CHECKED This part is designed to be APPROVED DR. AFZAL S. made by A TITLE: Exploded view shock Test rig A WEIGHT: [Kg]; [Lbs]

SIZE DWG NO. FILE REV DWG REV FINISH: B A0001_1 MATERIAL: SCALE: 1:20 DIMENSIONS ARE IN SHEET: 1 of 1

8 7 6 5 4 R:\0066-BOMBARDIER-16.5%3 LANDING GEAR\DESIGN\MECHANICAL\16.5 %\TEST RIGS\SHOCK_ABS_TEST\A0001

Figure B.5. Shock absorber Test Rig assembly drawing 1

8 7 6 5 4 3 2 1

F 7 F 2

E E

D D

Item No. PN Description QTY 2 A0001-0001 Rect Bar 2 C 7 1/2-20 screw 1 C 9 90128A961 #10-32 screw 1 13 A0001-0002 Insert 2 2

DRAWN PARENT ASSEMBLY CHECKED This part is designed to be APPROVED DR. AFZAL S. made by A TITLE: Exploded view shock test rig A WEIGHT: [Kg]; [Lbs]

SIZE DWG NO. FILE REV DWG REV FINISH: B A0001_2 MATERIAL: SCALE: DIMENSIONS ARE IN SHEET: 1 of 1

8 7 6 5 4 R:\0066-BOMBARDIER-16.5%3 LANDING GEAR\DESIGN\MECHANICAL\16.5 %\TEST RIGS\SHOCK_ABS_TEST\A0001

Figure B.6. Shock absorber Test Rig assembly drawing 2

88 Appendix C

Performance evaluation

The dynamic response of the landing gear is almost controlled by two dynamic components in series, the tire and shock absorber. The overall system should be properly modeled in order to have an accurate distribution of loads when the aircraft is in operation. The behaviour of tires under static and dynamic vertical and later loads is complex and depends upon material properties, pressure and temperature, wall ﬂexibility and runway conditions. [44] A simple representation of the tire can be an undamped model with stiﬀness proportional to tire pressure.

C.1 Taxiing Model

The landing gear, in the ﬁrst phases of preliminary evaluation of concepts and oﬀ-the-shelf components can be assumed to be a simple linear spring/damper, resulting in a set of linear equations. The taxiing response analyzed, refers to irregularities on the runway, that can be modeled as a series of dips and bumps, in a 1-cos shape.

Each dip/bump can be modeled according to the equations C.1, where ∆hr and Lr are respectively ˙ the depth and length of dip or bump, xr is the distance along the runway, h (xr) and h (xr) are the elevation proﬁle and rate of change at xr coordinate, assuming a true air speed V at the instant of time considered. [39]

∆hr 2πxr h (xr) = 1 − cos (C.1a) 2 Lr ˙ dh dxr dh h (xr) = = V (C.1b) dxr dt dxr

The nose and main gear pass at different times over the dip/bump causing different heave and pitch response of the aircraft. The motion of the aircraft, considered as a rigid body, over a one dip-cosine runway described in equation C.1 is regulated by the response of the landing gear, simply modeled with two mass-spring-damper1 respectively for the nose and main gear. The overall model, analytically analyzed and implemented in Matlab®/Simulink refers to the figure C.1.

1The spring and damping coeﬃcients are the ones resulting from a single model respectively for the nose and main gear. 89 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

Figure C.1. Rigid aircraft with equivalent model of undercarriages during taxiing over the design runway.[39]

The two degree of freedom that represents the response of the aircraft at the center of mass are the heave zc (downwards positive) and pitch θ (relative to any horizontal datum). The aircraft responses are evaluated with respect to a datum initial state, so they are incremental. The data that play an important role in the aircraft response are:

• Mass properties of the aircraft:

– Mass of the aircraft: m

– Inertia about the center of mass: Iy

• Stiﬀness properties of the landing gear:

– Equivalent stiﬀness on the nose gear: KN

– Equivalent stiﬀness on the main gear: KM

• Damping properties of the landing gear:

– Equivalent damping on the nose gear: cN

– Equivalent damping on the main gear: cM

• Geometry of the landing gear:

– Distance of the nose gear from the c.g.: lN

– Distance of the main gear from the c.g.: lM

The dynamic equation that describe the motion can be obtained from the Lagrange”s equations, shown in C.2, with the generalized coordinates zc and θ.

d ∂T ∂T ∂D ∂U ∂ (δW ) − + + = for j=1,2 (C.2) dt ∂x˙j ∂xj ∂x˙j ∂xj ∂ (δxj)

The kinetic energy, potential energy and dissipation function for the generalized model considered are expressed in equation C.3. 1 1 T = mz˙ 2 + I θ˙2 (C.3a) 2 c 2 y 1 1 U = K ∆ 2 + K ∆ 2 (C.3b) 2 N N 2 M M 1 2 1 2 D = C ∆˙ + C ∆˙ (C.3c) 2 N N 2 M M 90 C – Performance evaluation

The motion equations of the 2DoF system, shown in C.4, are obtained substituting the equation C.3 into the equation C.2. m 0 Z¨c CN + CM −lN CN + lM CM z˙C + 2 2 ˙ (C.4) 0 Iy θ¨ −lN CN + lM CM lN CN + lM CM θ KN + KM −lN KN + lM KM zC f(t) + 2 2 = −lN KN + lM KM lN KN + lM KM θ lT f(t) The system of equations can be applied for a dynamic system like the one described in chapter 6 in equation 6.1 and solved with time-space method in Matlab® & Simulink. with time-space method

C.2 Landing simulink model

Figure C.2. Simulink model for the heave response on time domain during landing

91 92 Bibliography

[1] C. H. Wolowicz, J. S. Bowman-Jr, and W. P. Gilbert, “Similitude requirements and scaling relati- onships as applied to model testing,” NASA Technical Paper, 1979.

[2] Bombardier and Q. Aerospace, “Horizon - Flying Demo Specs for the new generation FTV,” 2018.

[3] CfAR and Q. Aerospace, “BBA 7pcnt (Operational Documentation and Scaling process),” 2018.

[4] F. Rizzi, CfAR, Bombardier, and Q. Aerospace, “Landing Gear project charter for the 16.5% FTV,” 2018.

[5] H. G. Conway, “Landing Gear Design,” 1958.

[6] A. Argiolas, “Landing Gear Design and integration into aircraft,” 2016.

[7] J. Roskam, Airplane Design. 1985.

[8] N. S. Currey, Aircraft Landing Gear Design: Principles and Practices. 1988.

[9] M. C.-Y. Niu, Airframe Structural Design - Practical Design Information and Data on Aircraft Structure. 1988.

[10] Mohammad H Sadray, “Aircraft Design - A systems engineering approach,” 2013.

[11] V. N. Divakaren, G. V. V. Ravikumar, and S. R. Patnala, “Aircraft Landing Gear Design and Development,” 2017.

[12] Federal Aviation Authority, “Chapter 13 - Aircraft Landing Gear Systems,” Aviation Maintenance Technician Handbook - Airframe, 2014.

[13] Al-Hussaini, “Undercarriage Layout Design,” 2015.

[14] A. Jha, “Landing Gear Layout Design for Unmanned Aerial Vehicle,” 14th National Conference on Machines and Mechanisms (NaCoMM09), NIT, Durgapur, India, December 17-18, 2009, 2009.

[15] G. Roloﬀ, “Aircraft Landing Gear: The Evolution of a system,” in Airbus-Deutschland GmbH, no. April, 2002.

[16] T.-U. Kim, J. Shin, S. Kim, and I. Hwang, “Design of a crashworthy landing gear using composite tube,” ICCM International Conferences on Composite Materials, 2009.

[17] B. Management, “X-48B Blended Wing Body Flight Control Demonstrator,” 2009.

[18] Federal Aviation Administration, “Far Part 23 — Airworthiness Standards,” 2010.

[19] U. Commander John R. Brown, A critical study of Spin-up drag loads on aircraft landing gears. PhD thesis, 1949.

[20] Morne and Duke, “Skymaster jets ARF PLUS - Technical Manual,” 2015.

[21] Platform and U. Subsystems, “PENGUIN C - Unmanned Aircraft System,” 2017.

[22] R. F. Woodbury, Elements of Parametric Design. 2010.

[23] Hobbyking, “Aircraft RC parts catalog US,” 2018. 93 Fabrizio Rizzi: Landing Gear Design for Blended Wing Body Flight Test Demonstrators

[24] J. Moriarity and C. Gallagher, “The Ultimate Guide to Waterjet,” Flow Internal Corporation, 2001. [25] S. Benson, “Bending Basics: The fundamentals of heavy bending,” The Fabricator, 2017. [26] A. Handbook, “Heat treating of Aluminum Alloys,” ASM International, vol. 155, 1999. [27] Pyrotek Aerospace, “Certiﬁcate of conformance for Heat treatment of 7075-T6 Aluminum,” 2018. [28] S. T. Chai and W. H. Mason, “Chapter 3 Landing Gear Concept Selection,” in Landing Gear Integration in Aircraft Conceptual Design, 1997. [29] M. P. R. Products, “Airplane series Servo catalogue,” 2017. [30] Krauss and Wille, “Chapter 8 - Weight Estimation,” 2002. [31] CfAR and Quaternion, “Documentation of Purchase orders relative to the landing gear components for FTV7%,” 2018. [32] B. CfAR, “2018-08-08 FTV2B TAIL BFP Test Info,” 2018. [33] F. Rizzi, “BBA 16.5%FTV core requirements for the landing gear,” 2018. [34] Marc-Ingegno, “Analisi strutturale FEM – Cerchio scomponibile per carrello di velivolo,” iMex A, 2017. [35] R. Mark, “Basics of Trailing-Link landing gear,” 2017. [36] Robart, “Catalogue and speciﬁcations for tires and wheels for small airplane applications,” 2018. [37] SDP/SI, “Master inch catalog D820: drive components used in mechanical transmission,” 2018. [38] MTBR, “The Fox Float RP2(3) damper service thread: uncovering the secret,” 2017. [39] J. E. C. Wright, Han R., Introduction to aircraft Aeroelasticity and loads. 2014. [40] I. Neihouse, W. Pepoon, L. Aeronautical, and L. A. Force, “National advisory committee for aero- nautics,” 1950. [41] DT Swiss AG, “Dt Swiss - SHOCKS Technical Manual,” 2015. [42] Warren F. Phillips, Mechanics of Flight. Wiley, 2009. [43] MATCOmfg, Wheels and Brakes W600, W600XT, W600XLT. 2016. [44] H. B. Pacejka, Tyre and Vehicle Dynamics. 2006.