Failure Analysis of a Nose Landing Gear Fork of a GROB G115 Aircraft

Luís Daniel Diogo Fernandes

Thesis to obtain the Master of Science Degree in

Mechanical Engineering

Supervisor: Prof. Dr. Virgínia Isabel Monteiro Nabais Infante

Examination Committee

Chairperson: Prof. Dr. Luís Manuel Varejão de Oliveira Faria

Supervisor: Prof. Dr. Virgínia Isabel Monteiro Nabais Infante

Members of the Committee: Prof. Dr. Luís Filipe Galrão dos Reis

November 2015

Agradecimentos

Gostaria de expressar o meu mais profundo agradecimento à minha orientadora, Dr. Virgínia Infante, pelo seu apoio e orientação durante os meus estudos e, em particular, durante a minha tese de mestrado no Instituto Superior Técnico. Gostaria também de agradecer ao Dr. Álvaro Neves, diretor do Gabinete de Prevenção e Investigação de Acidentes com Aeronaves – GPIAA, pelo seu apoio e ajuda fornecida. Agradecer também a todos os meus amigos e professores que me apoiaram ao longo da minha formação académica. Um agradecimento especial ao Engenheiro Diogo Rechena pelo apoio prestado na minha tese, principalmente no estudo e análise de elementos finitos. Por último agradecer à minha família pelo incentivo, apoio, dedicação e confiança ao longo dos meus anos de curso.

Acknowledgements

I would like to express my deepest appreciation for my advisor, Dr. Virginia Infante, for her support and guidance during my studies and particularly during my master thesis at the Instituto Superior Técnico. I would like to thank Dr. Álvaro Neves, director of Portuguese Civil Aviation Safety Investigation Authority – GPIAA, for his support, cooperation and help provided. I would also like to thank all my friends and teachers who supported me throughout my academic training. Special acknowledgements to Engineer Diogo Rechena for his support in my thesis, mainly in the finite element analysis. Finally, I would like to thank my family for their encouragement, support, dedication and confidence throughout my years of study.

iii

Abstract

In this dissertation the study and analysis of a case of failure with a landing gear was performed. The developed work comes following an accident occurred in Cascais’ aerodrome in which the nose of the landing gear’s fork, of a light aircraft, Grob 115, bent during landing. Nose gear failures are a high concern in the aviation industry. In average 55% of aircraft failures occur during takeoff and landing while 45% of failures occur during flight, according to the Federal Aviation Administration reports. In order to determine the causes of the accident, a material analysis was performed, followed by a detailed study of the fracture’s surface both visually and using optic and scanning electron microscopy. It was observed that the cracks developed in the vicinity of the bolted holes, which work as supporting connections, on the topside of the nose fork and, as such, it can be concluded that the referred area was subjected to cyclic stresses originating and propagating cracks inside the material. This cracking is characteristic of the existence of areas of stress concentration. Identified the crack initiation zone with beach marks near the origin of the crack, combined with the fact that the nose wheel fork is subject to cyclic loading, leads to the conclusion that the component failed due to fatigue. Finite element analysis were also performed on the nose fork taking into account service conditions in order to assess the structural integrity of the referred component. During the analysis it was observed that the critical areas are located in the vicinity of the connecting holes due to the fact that they are stress concentrating features. Finally, some modifications to the fork have been proposed with the goal of improving its performance during service. The suggested changes do not affect the assembly of the landing gear and relate to simple changes such as changing material, thickness, radius and distance between bolts.

Keywords: Aircraft; Nose wheel fork; Landing gear; Fatigue; Finite Elements; Fracture surface; GROB G115

Resumo

Nesta dissertação foi feito o estudo da análise a um caso de falha de um trem de aterragem. A elaboração desta tese vem no seguimento de um acidente ocorrido no aeródromo de Cascais, em que uma aeronave ligeira, Grob 115, ao aterrar partiu a forquilha do trem de nariz. A falha do trem de nariz é de grande preocupação para a indústria da aviação. Em média 55% das falhas com aeronaves ocorrem durante a descolagem e a aterragem, enquanto 45% das falhas ocorrem durante o voo, segundo relatórios da Administração de Aviação Federal De forma a determinar as causas do acidente, foi feita a análise ao material da forquilha, tendo sido posteriormente analisada com detalhe a superfície de fractura, quer visualmente, quer recorrendo a microscopia óptica e electrónica de varrimento. Foi observado que as fissuras se desenvolveram na zona dos furos de fixação, que funcionam como ligação, no topo da forquilha do trem de nariz, e como tal pode concluir-se que esta zona esteve sujeita a esforços cíclicos originando e propagando fissuras no interior do material. As fissuras observadas são características de zonas de concentração de tensões. Identificada a zona de início da fissura com linhas de paragem perto da origem da mesma, combinado com o facto de o carregamento ser cíclico na forquilha do trem, permite concluir que a falha do componente foi por fadiga do material. Foram também realizadas análises de elementos finitos ao componente, tendo em consideração os esforços a que a forquilha está sujeita. Durante a análise foram observadas as zonas críticas, as quais se situam perto dos furos da forquilha, devido ao facto destas serem zonas de concentração de tensões. Por último, foram propostas modificações à forquilha com o objectivo de melhorar o desempenho da mesma. As alterações sugeridas não comprometem a montagem do trem de aterragem e dizem respeito a alterações simples como alteração de material, espessura, raios e distância entre parafusos.

Palavras-Chave: Aeronave; Forquilha do trem de nariz; Trem de aterragem; Fadiga; Elementos Finitos; Superfície de fractura; GROB G115

vii

viii

Table of Contents 1 Introduction ...... 1

2 Literature Review ...... 3

2.1 Landing gear records ...... 3

2.2 Landing gear and airplane characteristics ...... 4

2.3 Shimmy Vibrations ...... 7

2.4 Loads in the landing gear ...... 10

3 Nose wheel fork fracture surfaces ...... 17

3.1 Material ...... 17

3.2 Visual observation ...... 18

3.3 Optical microscopy ...... 21

3.4 Scanning electron microscopy ...... 22

4 Finite Element Analysis ...... 30

4.1 CAD model and characteristics ...... 30

4.2 Loads calculation ...... 31

4.3 Mesh ...... 33

4.4 Results and Discussion ...... 36

5 Proposal of new designs ...... 42

6 Conclusions and Future Developments...... 54

6.1 Conclusions ...... 54

6.2 Future Developments ...... 54

7 References ...... 56

8 Annexes ...... 58

8.1 Annex I ...... 58

8.2 Annex II ...... 60

8.3 Annex III ...... 61

8.4 Annex IV ...... 62

List of Figures

Figure 1.1 - Crashed aircraft in Cascais, Grob 115. Source: GPIAA – Portuguese Civil Aviation Accident Investigation Branch (2014)...... 2 Figure 2.1 - Grob G115 aircraft nose ...... 5 Figure 2.2 - Nose landing gear assembly. Nose wheel fork localization ...... 6 Figure 2.3 - Grob G115 general dimensions [29] ...... 7 Figure 2.4 - Oil-filled shimmy damper and its location in an aircraft [25] ...... 8 Figure 2.5 - Grob G115A nose leg. Source: GPIAA ...... 9 Figure 2.6 - Dynamic phenomena [26] ...... 9 Figure 2.7 - Level landing with inclined reactions [28] ...... 10 Figure 2.8 - Aircraft CG position ...... 12 Figure 2.9 - Dimensional Schematic ...... 12 Figure 2.10 - Eye-Bar Loading [3] ...... 14 Figure 2.11 - Contact patch region...... 15 Figure 3.1 - Chemical elements present in the sample ...... 17 Figure 3.2 - Scheme with fractures location...... 18 Figure 3.3 - Fractured component ...... 19 Figure 3.4 - a) zone 1 of figure 3.3; b) zone 2 of figure 3.3 ...... 20 Figure 3.5 - Cracks in attachment holes [13]...... 21 Figure 3.6 - Initiation and propagation zone of fatigue rupture. Figure 3.4a) expanded in detail A...... 22 Figure 3.7 – Beach marks. Figure 3.4a) expanded in detail A ...... 22 Figure 3.8 - Mapping of different areas along the fracture surfaces subject to SEM. a) Zone 1 of figure 3.3; b) Zone 2 of figure 3.3 ...... 23 Figure 3.9 – Initial zone of fracture surface (detail “A1” of Figure 3.8a)) and enlargement on right ...... 24 Figure 3.10 – Zone of crack propagation (detail “A2” of Figure 3.8a)) ...... 24 Figure 3.11 – Fatigue stretch marks in detail “A3” of Figure 3.8a) and enlargement on right ...... 25 Figure 3.12 - Transition between the propagation zone of crack front and final unstable rupture region (detail "A4" in the figure 3.8a))...... 25 Figure 3.13 - Ratcheting mark in the nucleation zone of fatigue cracks (detail “B1” of Figure 3.8a)) ...... 26 Figure 3.14 - Fracture surface (detail "B2" of Figure 3.8a)) ...... 26 Figure 3.15 - Fracture surface (detail "C1" of Figure 3.8b)) ...... 26 Figure 3.16 - Edge between fracture surface and component surface (detail "C2" of Figure 3.8b)) ...... 27 Figure 3.17 - Final unstable rupture region (detail "C3" of Figure 3.8b)) ...... 27 Figure 3.18 - Final unstable rupture region (detail "D2" of Figure 3.8b)) ...... 28

Figure 3.19 – a) Initial zone of fracture surface (detail "D1" of Figure 3.8b)) b) enlargement ...... 28 Figure 3.20 - Fracture surface (detail "D1" of Figure 3.8b)) ...... 29 Figure 4.1 - Nose landing gear assembly in SolidWorks ...... 30 Figure 4.2 - Vertical and Drag forces applied in bead seat region ...... 32 Figure 4.3 - Constraints applied to nose landing gear assembly ...... 32 Figure 4.4 - Split line used to convergence study ...... 33 Figure 4.5 - Nodes used to do convergence study ...... 34 Figure 4.6 - Mesh convergence analysis ...... 35 Figure 4.7 - Stress variation during refining ...... 35 Figure 4.8 - Mesh with 5mm elements used in simulations ...... 36 Figure 4.9 - Von Mises stresses distribution in nose wheel assembly ...... 36 Figure 4.10 - Maximum stress observed in bolt tip ...... 37 Figure 4.11 - Stress distribution in the upper surface of the fork ...... 38 Figure 4.12 - Stress distribution in lower part of the fork ...... 38 Figure 4.13 - Stress field progress ...... 39 Figure 4.14 - Plasticity zones in the fork ...... 40 Figure 4.15 - Plasticity zones in the fork (top view) ...... 40 Figure 4.16 - Normal stresses in the fork ...... 41 Figure 4.17 - Sliding between parts ...... 41 Figure 5.1 - Some important dimensions of original fork ...... 42 Figure 5.2 - Proposal 1 fork dimensions ...... 43 Figure 5.3 - Plasticity in the fork (left) and stress field of proposal 1 (right) ...... 43 Figure 5.4 - Detail of regions in plastic deformation of proposal 1 ...... 44 Figure 5.5 - Proposal 2 fork dimensions ...... 44 Figure 5.6 - Plasticity in the fork (left) and stress field of proposal 2 (right) ...... 45 Figure 5.7 - Detail of regions in plastic deformation of proposal 2 ...... 45 Figure 5.8 - Proposal 3 fork dimensions ...... 46 Figure 5.9 - Plasticity in the fork (left) and stress field of proposal 3 (right) ...... 46 Figure 5.10 - Detail of regions in plastic deformation of proposal 3 ...... 47 Figure 5.11 - Proposal 4 fork dimensions ...... 47 Figure 5.12 - Plasticity in the fork (left) and stress field of proposal 4 (right) ...... 48 Figure 5.13 - Detail of regions in plastic deformation of proposal 4 ...... 48 Figure 5.14 - Representation of the effects of changing bolt distance in proposals 3 and 4 ...... 49 Figure 5.15 - Proposal 5 fork dimensions ...... 49 Figure 5.16 - Plasticity in the fork (left) and stress field of proposal 5 (right) ...... 50 Figure 5.17 - Detail of regions in plastic deformation of proposal 5 ...... 50 Figure 5.18 - Detail of regions in plastic deformation of proposal 6 ...... 51 Figure 5.19 - Stresses field in the fork of proposal 6 ...... 51

xii

Figure 5.20 – Detail of regions in plastic deformation of proposal 7 ...... 52 Figure 5.21 - Stresses field in the fork of proposal 7 ...... 53

xiii

List of Tables

Table 2.1 - Parameters to compute the CG location [29] ...... 11 Table 2.2 - Dimensions of Figure 2.9 ...... 13 Table 4.1 - Nose landing gear assembly components used in SolidWorks ...... 31

xiv

Abbreviations

AWA - Aeronautical Academy Web CAD – Computer Aided Design CAE – Computer Aided Engineering CAM – Computer Aided Manufacturing CASA - Civil Aviation Safety Authority CFD – Computational Fluid Dynamics EASA – European Aviation Safety Agency EDS - Energy Dispersive Spectrometer FAA - Federal Aviation Administration FAR - Federal Aviation Regulations FEA - Finite Element Analysis FEM – Finite Element Method FSF – Flight Safety Foundation GPIAA – Portuguese Civil Aviation Accident Investigation Branch KBE - Knowledge Based Engineering LAA - Light Aircraft Association MTOW – Maximum Take-Off Weight NLG – Nose Landing Gear NTSB - National Transportation Safety Board SEM – Scanning Electron Microscope

xvi

1 Introduction

The landing gear supports the entire weight of an aircraft during landing and ground operations. It is attached to primary structural members of the aircraft [1] and it is one of the most critical subsystems of an aircraft meaning that landing gear detail design is taken up early in the aircraft design cycle due to its long product development cycle. The need to design with minimum weight and volume, reduced life cycle cost and high performance poses many challenges to landing gear designers. The purpose of the landing gear is to provide a suspension system during taxi, take-off and landing. It is designed to absorb and dissipate the kinetic energy of landing impact, during which loads are very high. Landing gears are often retractable in order to reduce aerodynamic drag during flight although, the majority of light aircrafts have a fixed one. Aircrafts have several landing gear configurations, such as tricycle landing gear, in with the nose wheel in the front and two main wheels located just behind the center of gravity and the conventional landing gear with one tail wheel and two main wheels. Different configurations will result in the different load paths and stress behaviors. This dissertation is focused on the nose gear of the Grob G115, originally manufactured by Grob, which is an advanced general aviation fixed-wing aircraft, primarily used for flight training. The landing gear design takes into account various requirements that are stipulated by the Airworthiness Regulations to meet operational requirements of safety, strength, stability, stiffness, ground clearance, control and damping under all possible ground attitudes of the aircraft. Divakaran, Ravi and Srinivasa [2] present in their work an overview and challenges in landing gear design and development as well as how technologies help meeting each challenge. Many manufacturers do not use Finite Element Analysis (FEA) to test the nose landing gear because the Federal Aviation Administration (FAA) does not require FEM as part of the approval process [3]. Although, some commercial available CAD/CAM/CAE/CFD and Dynamic Simulation software tools are used in the design and development of landing gears, many Knowledge Based Engineering (KBE) tools and information intelligence tools are being developed and used by landing gear designers to automate many engineering processes. Several authors have studied landing gears, since the loads they are subjected to, design, optimization and finite element analysis. Horack [4] proposed an analysis of the landing gear structure. Yangchen [5] developed work on the light weight structural design and optimization of landing gears. The analysis of the landing gear using finite element method was proposed by Briscoe [6]. Krason and Malachowski [7] proposed the finite element modeling of the landing gear and drop test simulation. An investigation of a nose landing gear failure was performed by Lal et al. [8] on fatigue fracture of a nose landing gear in military transport aircraft. A study developed by Al-Bahkali [9] with two different landing gear configurations for a light aircraft have been analyzed and modeled under different landing conditions. This thesis follows the investigation of an accident that occurred on November 3, 2014, with a German manufacturing GROB aircraft, G115A model, registration D-EGXI, operated by

1 flight school AWA - Aeronautical Academy Web, during an instruction flight in Cascais aerodrome, Lisbon. According to GPIAA, the plane landed in the center of the runway, first with the main gear and keeping the nose up in order to protect the nose gear until the plane dissipate speed, and when the nose gear landed the pilot in command felt vibrations in the airplane, and moments later the nose landing gear collapsed causing the plane to turn left sliding down the runway, Figure 1.1. The mentioned aircraft has performed more than 17,000 landings, and being an instruction aircraft it has been subjected to high mechanical stresses and high vertical loads. The objective of this dissertation is to make an assessment of the component to determine the possible causes of failure. The adopted methodology involves a detailed study of previous incidents of the same type and obtaining technical information from the manufacturer and other aviation authorities on the fractured component. Afterwards, a visual analysis is presented, as well as an optical microscopy with low magnification of the fractured surface in order to characterize the type of fracture and identify areas of interest to perform a more detailed analysis through electronic scanning microscopy. Another step was to perform the analysis and numerical simulation using the finite element method in order to determine the stresses that the component is subject to and to understand the causes that may have led to the failure of the landing gear. This dissertation will help landing gear manufacturers answer some questions related to the nose landing gear during landing, and the stress to which it is subject, so these answers can be used in the early stage of future the designs and in maintenance operations.

Figure 1.1 - Crashed aircraft in Cascais, Grob 115. Source: GPIAA – Portuguese Civil Aviation Accident Investigation Branch (2014).

2 Literature Review

2.1 Landing gear records

Although considered the safest means of transportation, airplanes are liable to accident occurrence usually with dire consequences. Even if there are no physical injuries, repairs are often costly. Several accidents and incidents have occurred with Grob 115 due to landing gear failure. In January 1991 during the runway approach maneuver, the plane bounced and pitched forward; the second touchdown was on the nose landing gear with the aircraft yawed slightly to the left and the landing gear fork bent [10]. In August 2008, in the same airplane, during a normal landing, the right main wheel separated from the aircraft as a result of fatigue cracking [11]. In April 2011, in the Dundee airport, during a training flight the trainee was performing a set of touch- and-go landing exercises. Following a firm touch down on the main and nose wheels, the nose landing gear collapsed. The maintenance organization advised that the nose leg collapsed as a result of the lower attachment on the shock strut due to heavy landing on its nose wheel [12]. Some other incidents and accidents have occurred with Grob airplanes due to failures in the nose landing gear fork and the assembly in general. Most of them were the result of incorrect maintenance, and others from component fatigue. As a result of these accidents, the maintenance organization has added extra inspection stages to the procedures and Grob also published some advices and bulletins for aircraft owners. A Service Bulletin in October 2009 [13], with the denomination “mandatory”, contains technical and planning information about nose gear forks. Grob received several reports from fleet customers pertaining to cracks in the nose wheel fork P/N 115C- 5200.14, which is why the Service Bulletin’s category is mandatory. The cracks were found in the attachment holes during regular maintenance and were identified as fatigue cracks, caused purely by the loads applied during landings. A year earlier another bulletin had been published by CASA (Civil Aviation Safety Authority), applicable to all Grob 115 series aircraft, with the purpose of drawing attention to several wheel forks that had cracked near the nose landing gear supports. All Grob 115 operators were recommended to inspect nose wheel forks at each periodic inspection and report all damaged and cracked forks to CASA [14]. Problems with the nose wheel assembly are not exclusive to Grob aircrafts. In early November 2007 Van’s Aircraft issued a mandatory Service Bulletin, calling for all nose wheel aircrafts to be fitted with an updated design of nose wheel attachment yoke [15]. In May 2011 Zenith Aircraft Company also introduced a new nose wheel fork [16]. Diamond aircrafts manufacturer also have a service bulletin regarding fatigue in nose gear forks, due to cracks having appeared in the nose landing gear of some aircraft as a result of hard landings [17]. This bulletin from 1999 addresses two parts: an inspection of the Nose Gear Fork and modifications required to remove the Nose Gear Fork and replace it with an optional heavy duty fork. In December 2010, a single-engine Diamond Aircraft DA 40 lost the nose landing

3 wheel during a training tough-and-go maneuver. This incident occurred due to multiple crack initiation and propagation, similarly to other incidents involving nose gear forks on Diamond DA 20 airplanes (with fork design similar to DA 40) investigated by NTSB (National Transportation Safety Board). NTSB implemented a method to nondestructively inspect nose wheel forks for fatigue cracks on a periodic basis [18]. Also a CASA issued bulletin refers to Cessna 100 and 200 series airplanes advising operators that airplanes continue to be subject to fatigue cracking and in some cases, failure in the nose wheel fork may occur [19]. Cessna Service Letter 63-31 and Service Letter SE71-34 are in accordance with the circular and contain useful inspection criteria. In November 2012, LAA (Light Aircraft Association), following the discovery of several unreported incidents related to nose wheel forks being found severely cracked during routine inspections, issued an Information Leaflet recommending the inspection of several nose wheel forks for SportCruiser aircrafts operating on a Permit to Fly [20].

2.2 Landing gear and airplane characteristics

The landing gear is one of the most important airplane components used during aircraft takeoff, landing, taxiing, parking and steering on ground. Aircraft use landing gears ever since the first flight was made. In the beginning of the twentieth century landing gears were simplistic, had no wheels and were not retractable. During the Second World War, landing gears were improved; became retractable and the configuration changed. All landing gears serve the same purpose [21]:  During landing, the landing gear must absorb and mitigates the landing shock;  During taxi, the nose gear is steerable to maneuver on the ground;  In order to reduce drag, the landing gear is designed to be extendable and retractable, although in most light aircrafts the landing gear is fixed;  After touch the ground, the aircraft decelerates using brake systems;  All landing gear system components must meet several regulations to guarantee safety aircraft operations.

In terms of design procedure, the landing gear is the last aircraft major component to be designed. In other words, all major components (such as wings, tail, fuselage, and propulsion system) must be designed prior to the design of the landing gear. Furthermore, the aircraft most aft center of gravity (CG) and the most forward cg must be known for landing gear design. In some instances, the landing gear design may drive the aircraft designer to change the aircraft configuration to satisfy landing gear design requirements [22]. Depending on the type of aircraft and the function it serves, different landing gear configurations are used. Light aircrafts use simpler landing gear configurations, large commercial aircraft use more complex landing gear configurations. The main landing gear configurations are:  Conventional gear;

 Simple main wheel;  Bicycle gear;  Tricycle gear;  Quadricycle gear;  Multi-bogey gear.

The nose landing gear has as many functions, namely, to increase the maneuverability of the aircraft and to support the nose of the aircraft during ground operations. Furthermore, the nose landing gear includes a system that damps shimmy (explained in detail in subchapter 2.3), which is an effect that causes the nose gear to be misaligned with the moving-direction. The studied component in this dissertation is part of the nose of a Grob G 115, Figure 2.1, and it is assembled in the nose landing gear as shown in Figure 2.2. The Grob landing gear configuration is fix tricycle with a main gear and a nose gear. This component fractured during service.

Figure 2.1 - Grob G115 aircraft nose

Nose Wheel Fork

Figure 2.2 - Nose landing gear assembly. Nose wheel fork localization

The aircraft is manufactured in Germany by Grob Aircraft is constructed of carbon composite materials and is certified by EASA and FAA. The general dimensions are represented in Figure 2.3.

Figure 2.3 - Grob G115 general dimensions [29]

2.3 Shimmy Vibrations

One of the most important functions of the landing gear is the capability of absorbing energy during the landing impact. Due to the fact that landings result in high amounts of energy transfers and vibrations resulting from the elastic behaviour and due to dynamic loads that can lead to material fatigue and failure, a shock absorber with a large stroke is required in order to limit the forces occurring during this maneuver; typically the available travel is in the range of 0.3 to 0.6 meter. As landing gears are important and complex systems and are one of the most failure prone systems, as pointed out in a study performed by the Flight Safety Foundation to commercial aircrafts [23], they have some design requirements [24], and design and maintenance procedures may still require improvements to enhance flight safety. Shimmy is a consequence of the interaction between tyre and landing gear’s structure, leading to oscillation phenomena. The motion typically has a frequency in the range between 10 to 30 Hz. It should be an important consideration during the design of a landing gear due to the fact that the amplitude may be amplified to a level of vibrations that affects the comfort and visibility of the pilot, or even result in severe structural damage and landing gear collapse. The pilot may notice a landing gear vibration, which can develop from many sources and does not necessarily have to be caused by shimmy. Shimmy can occur on both the nose and main landing gears, although the latter case is less frequent, and this phenomena can lead to structural damage or torque link failure. Experience indicates that main landing gear shimmy is most likely to develop very shortly after landing impact, reducing friction in the yaw direction and asymmetrical spin-up of the wheels may serve as an initial disturbance [25].

Shimmy in the nose wheel is caused by excessive vibration of the wheel during motion. Many nose gear shock struts also have attachments for the installation of an external shimmy damper. The Grob 115D2 is equipped with an oil-filled shimmy damper, Figure 2.4, that is designed to dampen the vibration. However, shimmy can still occur if the damper is not correctly maintained, the runway surface is poor (rough surface, or with irregularities or with adherence issues) or a load is placed on the nose wheel while the aircraft is travelling along the runway at a relatively high speed. The latter effect can be reduced by landing on the main wheels and keeping the load off the nose wheel by applying a backward pressure on the control column [12]. In other Grob models, like the one under study (Grob G115A), there is no shimmy damper, as seen in Figure 2.5 from another crashed aircraft of the same model, meaning that the nose landing gear is strongly affected by vibrations and more subject to loads and material behavior that lead to material failure.

Figure 2.4 - Oil-filled shimmy damper and its location in an aircraft [25]

Figure 2.5 - Grob G115A nose leg. Source: GPIAA

Oscillations can occur in the longitudinal, lateral and yaw directions as shown in Figure 2.6. Longitudinal vibrations can be generated by landing impact or during braking and are usually referred to as “gear walk” and are normally caused by variations of vertical loads acting on the wheels. Shimmy vibrations on the other hand occur in the lateral and yaw directions and are generated by self-excitation forces which may also be induced by asymmetric conditions occurring at landings with prevailing crosswind. Sometimes the wheels can have a rolling motions about their longitudinal axis [26].

Figure 2.6 - Dynamic phenomena [26]

2.4 Loads in the landing gear

In commercial aviation, aircrafts are often divided into several classifications nominated as FARs – Federal Aviation Regulations, which are rules prescribed by FAA. A wide variety of activities are regulated and rules are designed to promote safe aviation, protecting pilots, flight attendants, passengers and the general public from unnecessary risk. Activities such as design and maintenance, pilot training activities, lightning, model aircraft operations and typical airline flights are presented in FARs which are organized into sections called Parts. For instance, Part 23, of the FARs, concerns aircrafts with a maximum gross landing weight of 12,500 lbs. In Part 25 which deals with airplanes in transport category the aircraft has the maximum landing weight beyond 12,500 lbs. Part 27 concerns rotorcraft, which has a maximum landing weight of no more than 7000 lbs while Part 29 is for rotorcraft in the transport category with more than 7,000 lbs. Several other classifications exist. The Grob G115 aircraft is classified under Part 23 Aircraft because its maximum take-off weight is 1,873 lbs (850kg). In order to properly analyse the aircraft and, in particular, the nose wheel fork, a variety of forces have to be determined. During landing, there are three main types of loads acting on landing gear that cause axial compression, bending and torsion on the landing gear strut [27]: 1. Vertical load; 2. Drag load; 3. Side load.

The reaction force between the ground and the aircraft is caused by contact between the tyre and the runway. Figure 2.7 shows the landing configuration in which the contact between the nose gear and the ground is impending.

Figure 2.7 - Level landing with inclined reactions [28]

In normal landing conditions, the airplane is assumed to be in the following attitudes in case of airplanes with nose wheels: 1. Three-point landing, in which the main wheels and nose wheel will be in contact with the ground simultaneously. 2. Two-point landing, in which only the main wheels will be in contact with the ground, with nose wheel just clear of ground.

The first attitude may be used in the analysis required under the second point. Level landing conditions can be found in Part 23, of the FARs [28].

Computing the centre of gravity location requires establishing the basic empty weight centre of gravity location by weighing. For this purpose the airplane is placed on 3 scales (2 under the main wheels, 1 under the nose wheel) so that the bottom edge of the canopy frame is in a horizontal position. The empty weight is determined from the sum of the single weights G2, Glri and Glle. The expression to compute the empty weight CG position aft of datum is determined in equation 2.1 in which the variables depend on the characteristics presented in Table 2.1

퐺푙푙푒. 푏푙푒 + 퐺푙푟푖 . 푏푟푖 푥푠 = − 푎 [29] (2.1) 퐺퐸

Symbol in expression Characteristic Units

퐺2 Weight at nose wheel kg (lds)

퐺푙푟푖 Weight at RH main wheel kg (lds)

퐺푙푙푒 Weight at LH main wheel kg (lds)

퐺퐸 = 퐺푙푟푖 + 퐺푙푙푒 + 퐺2 Empty weight kg (lds)

푎 Distance nose wheel – datum level mm (in.)

푏푙푒 Distance nose wheel – LH main wheel mm (in.)

푏푟푖 Distance nose wheel – RH main wheel mm (in.)

Table 2.1 - Parameters to compute the CG location [29]

푥푠 can also be calculated dividing the total moment (135532 kg.mm), obtained by the aircraft operator, by the total weight of the aircraft (639 kg) resulting in the distance between the aircraft’s CG and the reference datum, see Annex I. The obtained position for the centre of gravity is of 212 mm. The aircraft’s CG position is indicated with respect to the reference datum in Figure 2.8. Also, the remaining dimensions are shown in Figure 2.9 with the objective of computing external loads. The CG height is not available due to the lack of available data, therefore it was considered the height from the ground to a longitudinal line passing through the tip of the propeller spinner, with aircraft level in flight line (1,16 m).

Figure 2.8 - Aircraft CG position

Figure 2.9 - Dimensional Schematic

The dimensions in the case with inclined reactions were computed using trigonometry as shown below

푎′ = (푎 − 푐. tan 훽). cos 훽 (2.2) 푑′ = 푑. cos 훽 (2.3) ′ ′ 푏 = 푑 − 푎′ (2.4)

Table 2.2 presents the values for the equations (2.2), (2.3) and (2.4).

a 1285

b 323

c 1160

d 1608

a’ 965

b’ 595

d’ 1560

Table 2.2 - Dimensions of Figure 2.9

An aircraft may not be able to sustain flight and can lead to an unavoidable crash of the aircraft if the centre of gravity is outside of an acceptable range. It can be seen in Annex II that for the maximum weight conditions of that airplane, the CG position is within the acceptable range. The procedure to make sure the CG is acceptable is to first draw a horizontal line from the aircraft weight and a line vertically from the fuselage station on which the CG is located. There are two enclosed areas: the larger is the CG range when operating in the Normal category only, and the smaller range is for operating in both the Normal and Utility categories [30]. The same procedure is applicable in weight moment limits, the obtained values can be reviewed in Annex III.

Appendix C Part 23, of the FARs, states that the vertical force at the nose gear is determined by equation (2.5), see Annex IV:

푏′ 푉 = (푛 − 퐿). 푊. ( ) 푓 푑′ (2.5)

W represents the maximum weight of the aircraft during landing; parameter n represents the ratio of external applied vertical forces to the weight; L is the ratio of the assumed wing lift to the airplane weight; b’ and d’ are the dimensions defined in Figure 2.9.

Figure 2.10 - Eye-Bar Loading [3]

This vertical force should be converted into pressure using the eye bar theory, which is applied in determining how to distribute the ground reaction force on the wheel, presented below, to apply at the bead seat location. To understand this theory, there is a representation in Figure 2.10 with the eye-bar under loading and this theory is used to study the method of applying the load directly on the wheel without analysing the tire.

Figure 2.10 expresses the maximum unit load 푞푚푎푥 obtained in the eye-bar theory which can be calculated from equation (2.6). Using the applied force W, r which is the radius of the pin and integrating, equation (2.7) is obtained

휋 2 푊 = ∫ 2. 푟. 푞. cos 휃 푑휃 (2.6) 0 휋. 푞 . 푟 푊 = 푚푎푥 2 (2.7)

Eye-bar theory has been applied to tire/wheel interfaces before. For example, Stearns [31] derived the applied pressure, W at the bead seat region as expressed in equation 2.8:

휃0 푊 = ∫ 푏. 푊푟. 푟푏푑휃 (2.8) −휃0 Stearns used the eye-bar analogy similar to Blake [32], although the first’s work deals with automotive wheel, it can also be applied for aircraft wheels, to determine pressure distribution in the contact area between the wheel and the bead seat region.

휋 휃 Equations (2.9) and (2.10) are obtained replacing 푊푟 with 푊표 cos ( . ) in equation (2.8) 2 휃0

푊 .4. 푏. 푟 . 휃 푊 = 표 푏 0 휋 (2.9) 푊. 휋 푊표 = (2.10) 4. 푏. 푟푏. 휃0

In which Wo is the maximum pressure, Wr is the distributed pressure, rb is the wheel radius, b is the width of the bead seat and 휃0 is the contact patch angle. The contact patch, sometimes referred to as footprint, corresponds to the region where the tire is in contact with the road region. The pressure at the tyre/wheel interface varies during landing as the reaction force between ground and tire changes and is distributed on the bottom part of the wheel according to the contact patch region theory. Based in Figure 2.11 it is possible to determine the central angle α with equation 2.11 therefore defining the contact patch region, in which z is the tire deflection and r is the radius of the tyre. The used values can be obtained from the Goodyear tire technical manual, since the tyre brand was considered to be Goodyear [33]. The z value corresponds to the tyre deflection which in case of aircrafts is typically of 32%, sometimes being 35% [34].

Figure 2.11 - Contact patch region

푧 훼 = 2. cos−1 (1 − ) 푟 (2.11) The tyre model of the crashed Grob is a 380x150-5/15x6.00-5 with a radius of 6,5in. Using Goodyear it was determined that the flat radius is of 4,3in which corresponds to a 34% of deflection.

After determine the value for the central angle in which the pressure caused by contact force during landing should be applied, the load can be applied on the bead seat region and equation

(2.12) can be written as a function of the vertical force Vf:

푉푓. 휋 푊표 = (2.12) 4. 푏. 푟푏. 훼

The drag force, Df, at the instant following the touchdown is also exerted onto the nose gear of the aircraft and at this moment the wheel does not yet spin. The drag force is defined in equation (2.13)

푏′ 퐷 = 푘. 푛. 푊. ( ) 푓 푑′ (2.13)

In this expression k is the linear variation constant, n is the inertia load factor, b’ and d’ are the dimensions defined in Figure 2.9 and W represents the maximum weight of the aircraft during landing. The used values can be obtained in the aircraft and FAA database.

The results of the applied loads are presented in chapter 4.

3 Nose wheel fork fracture surfaces

3.1 Material

Appropriate material selection during the design stage of the landing gear is very important. A material guideline was created by Curry [1] to assist designers and engineers when selecting materials for several aircraft components. His guidelines include the inspection method, strength requirement, material samples, hardness, and surface finishing. Aluminums and alloy steels are the most common materials used in the nose gear assembly. Aluminum provides a high corrosion resistance property while maintaining the structural integrity to support the aircraft. Some of the parts in the nose gear assembly are forged before the final products are machined. Aluminum alloys such as 2014 and 7075 are commonly used in nose gear parts. When compared with 2014, aluminum 7075 although both materials have the same density, the latter one has better mechanical properties and stress corrosion resistance. Through analysis of chemical elements, using an energy spectrometer (EDS - Energy Dispersive Spectrometer), it was confirmed the presence of the essential elements of a series 5000 aluminum alloy in which aluminum is the principal element and the Magnesium the secondary one, Figure 3.1.

Figure 3.1 - Chemical elements present in the sample

In order to identify the nose fork’s material, the CES and matweb were used as well as chemical composition and mechanical properties tables. Taking into account the Magnesium concentrations in the analyzed specimen coupled with an expected high yield strength the material was identified as being the 5182-H19 aluminum alloy. It is noteworthy that the material identification is based on the comparison between the characteristics of existing alloys which present similar characteristics to the ones observed during EDS. As there is no manufacturer information concerning the material of the landing gear fork there may be some discrepancies between the selected material and the real one.

3.2 Visual observation

The fracture surface on the nose wheel fork was inspected visually and through optical and scanning electron microscopy in order to characterize the fracture surface and identify the initiation and propagation of any cracks that would lead to the failure of the component. Figure 3.2 depicts a representation of the location of the fracture surfaces on the component. In Figure 3.2 4 points are identified (A, B, C and D) corresponding to the crack initiation locations and hence where more detailed observations were made. Figure 3.3 shows an overview of the fractures occurred wholly in the component under consideration, clearly indicating the orientation of the fracture surface, which follows a plane perpendicular to the axis of the nose gear. Fractures have developed in the area of the connecting holes, which are stress concentration zones.

Figure 3.2 - Scheme with fractures location.

Figure 3.3 - Fractured component

In Figure 3.4 there are smooth flat areas indicating crack initiation which subsequently led to propagation and fracture in zones 1 and 2 of Figure 3.3. It should be noted that since zones 1 and 2 are crack initiation areas, tension stresses should be dominant, a topic that shall be addressed further in detail. A more detailed view of the fracture surfaces indicates that there is evidence of a crack surface propagation, consistent with cyclic loading dependent failure mechanisms, originating in the stress concentration zone in the vicinity of the holes. The marks on the fracture surface near the starting area are characteristics from fatigue propagation, commonly referred to as beach marks. Similarly to the present case study there is a case in which fatigue cracks were detected in the attachment holes of nose forks (see Figure 3.5) in several Grob aircrafts which led to the issue of the Service Bulletin No. MSB1078-165 directing operators to reduce the time between inspections.

Figure 3.4 - a) zone 1 of figure 3.3; b) zone 2 of figure 3.3

Figure 3.5 - Cracks in attachment holes [13].

The present study is focused on the analysis of the fracture surfaces indicated in Figure 3.4. The fractured component was cut in order to allow for a more detailed analysis through optical and scanning electron microscopes. During microscopy analysis there was evidence of existence of lines on the fracture surface, consistent with the occurrence of a process of material fatigue. These lines result of plastic deformation of the material by the effect of successive loadings by the component service.

3.3 Optical microscopy

Through observation, using optical microscopy with low magnification, of the fracture surface it was found that there is a regular surface (as shown in Figure 3.6), with two very different areas: one in which the fracture approaches the stress concentration area, indicating initiation of cracks that propagate slowly along the surface causing fracture of the component, and a more distant zone in which the propagation is faster until the final fracture. A large part of the fracture surface is occupied by a fracture facies, which is typical characteristic of a fast fracture due to overstress. As the fracture area due to overstress is clearly lower than that of fatigue propagation, it indicates that the material has a relatively moderate fracture toughness and the loads that led to the fracture were low. In the fracture surface there is a relatively small area with a facies substantially different from the remaining surface and which contains characteristics of the material fatigue inside which one or several cracks propagated until a final break. Figure 3.6 shows one of the crack initiation areas located in the corner of the holes. As already mentioned, this is a critical area in terms of stress concentration. The existence of stress concentrations promotes the formation of micro cracks during the initial stage of nucleation of fatigue cracks, which eventually collide forming "ratcheting" marks. After thorough analysis of the nose wheel fork fracture surface, it was not possible to detect any evidence of impact on the component side surface that can be identified as the origin of fracture.

Beach marks are visible but not numerous (see Figure 3.7). There are also radial lines, which can be identified through optical microscopy, although they are few in number and small in size, from the fracture’s initiation point. These radial lines correspond to the intersection of different planes in which different cracks propagate independently until the moment that, by propagation to the interior of the component, they intersect and form a single crack which spreads rapidly until the final break. These radial lines almost disappear before the macroscopic beach marks, indicating the existence of a single crack propagating macroscopically.

Figure 3.6 - Initiation and propagation zone of fatigue rupture. Figure 3.4a) expanded in detail A.

Figure 3.7 – Beach marks. Figure 3.4a) expanded in detail A

3.4 Scanning electron microscopy

In a more detailed examination of the different morphological characteristics of the fracture surfaces described above, different zones were considered along the surfaces indicated in Figure

3.4 a) and b), and in accordance with the mapping presented in Figure 3.8, for observation under scanning electron microscopy (SEM). The beach marks were observed in both surfaces covering approximately two thirds of the extension (Figure 3.4). The remaining portion, however, presents a distinct morphological pattern, characterized by a clear upper surface roughness caused by fragile fracture process, which occurs suddenly and precedes the full separation of the component. This fracture occurs due to resistance failure of the section not affected by the evolution of the crack front due to fatigue, since this causes a decrease in advancing the cross-sectional area that supports the mechanical loads acting there.

Figure 3.8 - Mapping of different areas along the fracture surfaces subject to SEM. a) Zone 1 of figure 3.3; b) Zone 2 of figure 3.3

The detail "A1" of Figure 3.9 corresponds to the initiation zone of one of the fatigue cracks in which the shape of the fatigue surface indicates a propagation of fatigue crack with a curve front, which can be attributed to the existence, in the central zone of the specimen of an approximate state of plane strain, causing the existence of stress triaxiality and further advance breach in the referred area. Figure 3.9 is an amplification of detail "A1" of Figure 3.8 indicating that the crack initiation does not occur due to any metallurgical defects or mechanical defects (e.g., machining, coating). Another characteristic that can be pointed out, through the observation of the propagation of the fatigue crack region is the presence of beach marks (Figure 3.10). These lines display a pattern of uniform progression, without significant evidence of change in the direction of the crack

23 front as a result of torsional effects. This evidence suggests that the fatigue process is mainly due to the cyclic bending stresses caused by the aircraft landings.

Figure 3.9 – Initial zone of fracture surface (detail “A1” of Figure 3.8a)) and enlargement on right

Figure 3.10 – Zone of crack propagation (detail “A2” of Figure 3.8a))

Figure 3.11 (magnification of detail "A3") confirms the existence of streaks on the fatigue fracture surface of the component. Striations are the result of a plastic deformation occurring in the crack front when the latter moves due to the influence of each loading cycle, which is why striations are parallel between themselves and perpendicular to the crack propagation direction. The spacing between two successive marks allows to account for the velocity of propagation of the crack front and is therefore a determining analysis parameter for the correct characterization of a fatigue process. The transition between the stable crack propagation (fatigue) and the unstable propagation area (giving rise to total failure of the material) is visible in Figure 3.12 (detail “A4” of Figure 3.8).

Figure 3.11 – Fatigue stretch marks in detail “A3” of Figure 3.8a) and enlargement on right

Figure 3.12 - Transition between the propagation zone of crack front and final unstable rupture region (detail "A4" in the figure 3.8a)).

In this study, the existence of "ratcheting" marks located in the crack propagation area was observed, as exhibited in Figure 3.13. These marks are the result of the merger of two micro crack propagation planes originating in the maximum stress region, resulting from the high stress state of the referred region.

Figure 3.13 - Ratcheting mark in the nucleation zone of fatigue cracks (detail “B1” of Figure 3.8a))

Figure 3.14 and Figure 3.15 reveal that there is no evidence of any defect responsible for the fatigue crack initiation.

Figure 3.14 - Fracture surface (detail "B2" of Figure 3.8a))

Figure 3.15 - Fracture surface (detail "C1" of Figure 3.8b))

In Figure 3.16 the edge between the surface and fracture surface of the component can be observed in greater detail. There is also no evidence whatsoever of any defects that could be the cracking or final failure source.

Figure 3.16 - Edge between fracture surface and component surface (detail "C2" of Figure 3.8b))

Figure 3.17 and Figure 3.18 exhibit in detail the surface roughness increase in the final rupture area.

Figure 3.17 - Final unstable rupture region (detail "C3" of Figure 3.8b))

Figure 3.18 - Final unstable rupture region (detail "D2" of Figure 3.8b))

Detail a) of Figure 3.19 is a photograph of the observation made on the fatigue initiation zone. The facies represented in detail b) is an enlargement of detail a), displaying fatigue cracking with low cracking rates. With this magnification some crack propagation plans, become more visible. These radial lines are of a very reduced size.

Figure 3.19 – a) Initial zone of fracture surface (detail "D1" of Figure 3.8b)) b) enlargement

In an attempt to visualize the distribution of precipitates SEM was used with retrodiffusion electrons to observe the component surface. It can be observed on Figure 3.20 that the precipitates are uniformly distributed over the entire surface of the material, including the border area with the fracture surface.

Figure 3.20 - Fracture surface (detail "D1" of Figure 3.8b))

4 Finite Element Analysis

4.1 CAD model and characteristics

Before a structural integrity assessment could be performed, the nose wheel fork, as well as some of the adjacent components, needed to be modeled. To do so, the Solidworks software was used. The nose gear assembly is composed of several components. However, most of them are used during ground operations and the performance during landing is not affected by those components, therefore they are not included in the present analysis. Figure 4.1 depicts all components used in the analysis (the nose gear assembly can be seen with all components in Figure 2.2) and Table 4.1 lists them. In the present dissertation a CAD the tyre was not modelled. Still the eye-bar theory was employed in order to model the interaction between the tyre and the wheel’s rim and to best approximate this behavior the bead seat was incorporated into the tyre/wheel interface.

Figure 4.1 - Nose landing gear assembly in SolidWorks

Part Number Component name Quantity 1 Fork 1 2 Shaft 1 3 Rim 1 4 Fork support connector 1 5 Bolt M8 4 Table 4.1 - Nose landing gear assembly components used in SolidWorks

4.2 Loads calculation

The vertical load in the aircraft, Figure 4.2, can be obtained with information based in airplane and FAA database and the equation (2.5) can be computed as shown below.

595 푉 = (2.67 − 0.67) × 850 × ( ) = 648.4 푘푔 푓 1560 (4.1)

This load is then converted into a pressure and calculated based on equation (2.12). In the present case the pressure was not determined due to the fact that the load can be inserted directly in the bead seat region in the CAD model and the CAE software is able to convert it into a pressure. It represents an approximation since the software does not use eye bar theory to convert force into pressure. The contact patch region is obtained computing the central angle with equation (4.2)

2.2 훼 = 2. cos−1 (1 − ) = 97° 6.5 (4.2)

The drag force, (see Figure 4.2), can be obtained using equation (4.3) 푏′ 595 퐷 = 푘. 푛. 푊. ( ) = 0.25 × 2.67 × 850 × ( ) = 216.4 푘푔 푓 푑′ 1560 (4.3)

Figure 4.2 - Vertical and Drag forces applied in bead seat region

In this study the nose landing gear was constrained using a fix type boundary condition on the base of each as depicted in Figure 4.3. This is a rough approach since the bolts are not fixed but connect the nose wheel fork to the upper part of the landing gear assembly which fixes them to the entire aircraft. Still, the objective in this case is to model the interaction between the bolt’s threads, the fork and the support connector.

Figure 4.3 - Constraints applied to nose landing gear assembly

4.3 Mesh

SolidWorks Simulation uses the displacement formulation of the finite element method to calculate component displacements, strains, and stresses under loads. The mesh should be as regular as possible, so the maximum numerical accuracy could be achieved and the mesh size can be defined by the user. A convergence check has to be performed by meshing the models with different size elements until the stress converges. In order to find a balance between numerical accuracy and computational time, a convergence study was performed. To do so, a control line (see Figure 4.4) was created linking the edges of two holes in the nose wheel fork. In this line, five points were selected, as depicted in Figure 4.5, to control for stresses in each mesh refinement iteration. The control points are all in the same x and y coordinate with the z coordinate being the only one that changes. This information is what allowed for result probing during post processing.

Figure 4.4 - Split line used to convergence study

Figure 4.5 - Nodes used to do convergence study

When performing a convergence study, the finer the mesh, the longer it will take to compute the results but their accuracy will increase. If increasing the number of elements does not change the stress values significantly, it is considered that the results have converged. The convergence analysis was performed comparing meshes with element sizes of 20 mm, followed by 10, 5, 2.5 and finally 1.25 mm. The convergence was performed using Figure 4.6 and Figure 4.7. The plot in Figure 4.6 presents the stress results for the control points indicated in Figure 4.5 while Figure 4.7 displays stress variations between mesh refinement iterations in the referred points. It is noteworthy that when changing the mesh size from 2.5mm to 1.25mm the results do not change significantly. In fact, the maximum stress variation in the control points is of 1.28% with a high computational time for both meshes making them undesirable for other analysis that were performed during this study. The meshes of 20mm and 10mm are also undesirable due to the fact that there are large result variations between each mesh and its respective refinement. When refining from 20mm to 10mm, variations are as high as 9.19% in the control points and when refining from 10mm to 5mm, variations reach values of 15.28%. Finally, since the maximum stress variation between the mesh size of 5mm and 2.5mm is of 3.84% it is considered that the results are acceptable and, therefore, the picked mesh is of 5mm.

50 20 40 10

30 5

Von Mises (MPa) Mises Von 2,5 20 1,25 10

0 24 12 0 -12 -24 Z coordinate

Figure 4.6 - Mesh convergence analysis

18,00% 16,00% 14,00% 12,00% 10,00% 20 to 10 8,00% 10 to 5

Variation (%) Variation 6,00% 5 to 2.5 4,00% 2.5 to 1.25 2,00% 0,00% 24 12 0 -12 -24 Z coordinate

Figure 4.7 - Stress variation during refining

Figure 4.8 - Mesh with 5mm elements used in simulations

4.4 Results and Discussion

After finite element analysis the stress fields were obtained, as is depicted in Figure 4.9. The maximum Von Mises stress occurs on the tip of the bolts with a value of, approximately, 1GPa. The high stress values are explained due to the fact that there is a fixed geometry boundary condition in a surface with a singularity (the chamfer between the threaded surface and the bolt’s tip), therefore the results on the bolts are going to be ignored.

Figure 4.9 - Von Mises stresses distribution in nose wheel assembly

The maximum stress occurs in the rear bolts (see Figure 4.10) of the nose wheel fork. This occurs due to the fact that the rear bolts are under compressive normal stresses due to axial forces, caused by the impact load, and bending moments, caused by the drag force. The frontal bolts, on the other hand, are subjected to compression due to axial loads and tension due to bending moments.

Figure 4.10 - Maximum stress observed in bolt tip

The main component to analyze in this thesis is the fork. Figure 4.11 and Figure 4.12 depict the nose wheel fork isolated from the remaining parts. It can be observed that the higher stresses occur near the attachment holes. In the upper surface higher stresses appear on the line that defines the border of the contact surface between the fork and the connector. The lower surface is also under high stress values in the region around the holes. This is a zone of stress concentration, and therefore more susceptible to failure.

Figure 4.11 - Stress distribution in the upper surface of the fork

Figure 4.12 - Stress distribution in lower part of the fork

Figure 4.13 depicts the progression of von Mises stresses inside the material. This type of plot helps locate areas under higher stress values and, from its analysis, it can be concluded that the critical areas are near the bolted connections. Also, it is noteworthy that higher stress zones are in the vicinity of the actual fracture surfaces meaning that it is likely that this stress state promotes crack propagation in that direction leading to component failure.

Figure 4.13 - Stress field progress

In order to identify the regions under plastic deformation, it was necessary to plot von Mises stresses above the allowable stress only. Aluminum alloy 5182-H19 has a yield strength of 395 MPa. The allowable stress depends of the safety factor which, for aircrafts, varies between 1.2 and 3, depending on the application and material. For main landing gear structures it is often 1.25 [35]. The allowable stress was computed using equation (4.4) and then used to plot the material under plasticity in Figure 4.14 and Figure 4.15. 휎 395 휎 = 푦 = = 316 푀푃푎 푎푑푚 1.25 1.25 (4.4)

Figure 4.14 - Plasticity zones in the fork

Figure 4.15 - Plasticity zones in the fork (top view)

The zones depicted in Figure 4.15 are the areas under the highest stresses in the studied landing conditions. During flights, the stresses in the nose landing fork are practically inexistent, reaching their maximum when the contact between the tyre and the ground occurs. So the areas represented as being under plasticity are the ones in which the stress amplitude is higher and therefore these zones of stress concentration are more likely to create cracks that propagate through fatigue.

Figure 4.16 - Normal stresses in the fork

Through Figure 4.16 - Normal stresses in the forkFigure 4.16 we can see that the highest normal stresses are near the attachment holes. This fact is in accordance with the hypothesis that the cracks nucleate in this zone. In order to try and increase the model’s accuracy, the contact between the fork and the connector part was also modeled. Figure 4.17 shows the sliding between both components. After finite element analysis, it was concluded that modeling and simulating the contact between components yields no significant changes in the stress results and significantly increases computation time.

Figure 4.17 - Sliding between parts

5 Proposal of new designs

In this chapter different possible configurations for the nose landing gear fork, of the studied aircraft, will be analyzed. The objective in this chapter is to improve component performance under cyclic loading without compromising its performance during service. To do so, changes in material, thickness, size, shape, among others are going to be proposed. Since the convergence analysis was performed in the previous chapter, used element size was of 5mm. The used methodology’s objective is to present, describe, analyze and discuss each of the fork configurations and subsequently accept or dismiss those configurations depending on the obtained results. To better understand the modifications there’s a schematic of the original nose gear fork displayed in Figure 5.1.

Figure 5.1 - Some important dimensions of original fork

Proposal 1

The first proposed modification was decrease in radius of the fork’s bending angle as shown in Figure 5.2.

Figure 5.2 - Proposal 1 fork dimensions

Figure 5.3 - Plasticity in the fork (left) and stress field of proposal 1 (right)

Figure 5.4 - Detail of regions in plastic deformation of proposal 1

The results indicate a decrease in the amount of material under plastic deformation (compare with Figure 4.15). The radius decrease does not lead to modifications in the nose wheel assembly while making the fork more resistant.

Proposal 2

In this case, in addition to the radius decrease, the bolts were moved further apart in the wheel axis direction as shown in Figure 5.5.

Figure 5.5 - Proposal 2 fork dimensions

Figure 5.6 - Plasticity in the fork (left) and stress field of proposal 2 (right)

Figure 5.7 - Detail of regions in plastic deformation of proposal 2

In this design the stresses above the allowable value are very similar to the ones presented in previous proposal. The stress field seems to be more reduced, but did not show significant differences.

Proposal 3

The third configuration represents a smaller radius than previous models and bolts are farther from the center, Figure 5.8. The stress results for this new configuration can be seen in Figure 5.9 and Figure 5.10.

Figure 5.8 - Proposal 3 fork dimensions

Figure 5.9 - Plasticity in the fork (left) and stress field of proposal 3 (right)

Figure 5.10 - Detail of regions in plastic deformation of proposal 3

The new configuration presents a slightly smaller amount of material under plasticity than in previous proposals but the changes do not seem to be significant.

Proposal 4

In this case, the change proposal is to increase the radius (see Figure 5.11) and reduce the distance between the bolts in the wheel axis direction. The results are displayed in Figure 5.12 and Figure 5.13. Stresses in this new design proposal are slightly higher than in proposals 2 and 3 where the bolts are further apart, the reason for this phenomenon in depicted in Figure 5.14. Although the stress results for the current proposal are lower than the original ones, presented in Figure 4.14, this design proposal does not perform as well as the previous ones.

Figure 5.11 - Proposal 4 fork dimensions

Figure 5.12 - Plasticity in the fork (left) and stress field of proposal 4 (right)

Figure 5.13 - Detail of regions in plastic deformation of proposal 4

Considering distances a and b (in which a>b) and the landing force F, the resulting moment M = F. d, (in which d is the distance) since d can be a or b, the bending moment is larger when d=a, which is when the bolts are closer to each other (see Figure 5.14). This increase in bending moment results in higher bending stresses, which explains the higher stress values.

Figure 5.14 - Representation of the effects of changing bolt distance in proposals 3 and 4

Proposal 5

In addition to changes in bolts distance and fork radius, thickness changes were also studied. As such, a 2mm increase in fork thickness was proposed (see Figure 5.15). The obtained results are plotted in Figure 5.16 and Figure 5.17. In this design proposal, the plasticity region is very localized. In fact, this region is localized in a singularity area, which is the 0mm radius fillet in the bolted hole, therefore, these high stress values can be attributed to the singularity and not on design characteristics.

Figure 5.15 - Proposal 5 fork dimensions

Figure 5.16 - Plasticity in the fork (left) and stress field of proposal 5 (right)

Figure 5.17 - Detail of regions in plastic deformation of proposal 5

Proposal 6

Material changes were also proposed. In this case, the aluminum alloy was changed from a 5182 alloy to a 2014 one, which is much more common in the aerospace industry. This alloys has a yield strength of 415 MPa which means that the allowable stress of 332 MPa.

Figure 5.18 - Detail of regions in plastic deformation of proposal 6

Figure 5.19 - Stresses field in the fork of proposal 6

Although there are no significant changes in the stress field, since the yield strength of the 2014 aluminum alloy is higher than the original one, the amount of material under plasticity is reduced.

Proposal 7

Another common alloy in the aerospace industry is the 7075 aluminum due to its high yield strength. In this case, the material’s yield strength is even higher than the 2014 and 5182 alloys with a value of 505MPa. The stress results are presented in Figure 5.18. It is noteworthy that, like in previous examples, the regions under plasticity are residual and caused due to the singularities in the bolted holes. This proposal represents a significant improvement for the nose wheel fork resistance without significantly changing weight (1.5kg).

Figure 5.20 – Detail of regions in plastic deformation of proposal 7

Figure 5.21 - Stresses field in the fork of proposal 7

6 Conclusions and Future Developments

6.1 Conclusions

In this dissertation, a nose landing gear fork was studied with the goal of finding the causes that lead to the failure of this component. Using SolidWorks, the fork was analyzed using the finite element method. The stress field on the fork has been determined and it was concluded that the region around the attachment holes is a zone of stress concentration which is critical for crack propagation. It is noteworthy that before the structural analysis was made, the material had to be identified and that the selected material may not correspond exactly to the real one and, therefore, the obtained results do not correspond exactly to the reality. In the present case study, cracks developed near the fixing holes, which are stress concentration zones, and propagated from the upper surface of the fork. As such, it can be concluded that this area has been subject to cyclic tensile loads, which rise to fatigue cracks. The fracture was initiated by cracking in holes of the fork structure. This cracking is characteristic of the existence of areas of stress concentration. Evidences of a fracture surface are visible in the component. Through optical microscopy, with low magnification, it is possible to identify a crack initiation zone with beach marks near the origin of the crack which, combined with the fact that the nose wheel fork is subject to cyclic loading, leads to the conclusion that the component failed due to fatigue. Taking into account that this accident occurred, according to the GPIAA, after landing 17000, it is recommended a detailed examination of the area of the fork in shorter periods of time. According to service bulletin NO MSB1078-165, October 2009, in which Grob received a report of a customer fleet pertaining to cracks in the nose wheel fork, the service life was limited to 8,000 landings before the nose wheel forks had been non-destructively tested. These cracks as the ones studied in this thesis were identified the fatigue cracks, caused purely by the number of landings. Also AWA was advised to inspect visually the nose wheel fork every overall inspection. One of the main goals of this dissertation was to present new proposals for the nose landing gear fork. Some new configurations (with simple changes) and materials were proposed with the goal of, reducing the material under plasticity without compromising the fork’s function.

6.2 Future Developments

One of the topics that can be further explored is the fatigue analysis of this component since it is concluded that the fork’s failure occurred due to fatigue. Another near future development has to do with the external load modelling. The forces used were the drag and contact forces with the ground. It has to be taken into account that the component under analysis belongs to a trainer aircraft for training pilots. For this reason, during landing, it is expected that loads are higher than usual. This causes changes in stresses. This factor was not taken into account but is an important subject to be treated. In the literature there

54 is no reference to this fact. Finally, experimental validation should be performed in order to assess the real stress/strain values during landing maneuvers.

7 References

1. Currey, N.S., Aircraft Landing Gear Design: Principles and Practices. Aiaa Education Series. 1988: Amer Inst of Aeronautics & (October 1988). 2. N., D.V., R.K.G.V. V., and S. Rao, Aircraft Landing Gear Design & Development - How Advance Technologies are helping to meet the challenges? White Paper, 2015. 3. Nguyen, T.D., Finite Element Analysis of a Nose Gear During Landing, in Engineering. 2010, University of North Florida. 4. Horak, I.V. Advanced Landing Gear Fatigue Test Method. in LMS Conference Europe. 2006. Munich. 5. Deng, Y., Application Of Shape Optimization In Landing-Gear Structural Design Of Small Aircraft. Mechanics in Engineering, 2008. 30(1): p. 47-51. 6. Briscoe, D., ME 548 Aerostructures Final Project ANSYS Analysis Landing Gear. 7. Malachowski, J. and W. Krasoń, Effort analysis of the landing gear with possible flow during touchdown. International Journal of Mechanics, 2008. 2(1): p. 16-23. 8. Franco, L.A.L., N.J. Lourenço, M.L.A. Graça, O.M.M. Silva, P.P. de Campos, and C.F.A. von Dollinger, Fatigue fracture of a nose landing gear in a military transport aircraft. Engineering Failure Analysis, 2006. 13(3): p. 474-479. 9. Al-Bahkali, E.A., Analysis of Different Designed Landing Gears for a Light Aircraft. International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering, 2013. 79: p. 406 - 409. 10. AAIB Bulletin 3/1991 Ref: EW/G91/01/11 11. AAIB Bulletin 9/2009 Ref: EW/G2008/09/01. 12. AAIB Bulletin 8/2011 Ref: EW/G2011/04/25. 13. Service Bulletin NO. MSB1078-165, 2009. 14. Airworthiness Bulletin 32-012 Cracking in nose landing gear fork on Grob 115 series aircraft, 2008. 15. Van’s Aircraft Service Bulletin 07-11-09 dated November 9, 2007. 16. New stol Nosewheel Fork and Towbar, STOL CH 750. 17. Service Bulletin NO. DA-20-32-02 Nose gear fork, Fatigue, 1999. 18. NTSB Identification: CEN11IA114. ntsb.gov. 19. Airworthiness Bulletin 32-016 Cessna single engine nose landing gear forks, 2008. 20. Airworthiness Information Leaflet MOD/338/016 Inspection for cracks in nosewheel fork, 2012. 21. Gowda, A.C. and N.B. S, Linear Static and Fatigue Analysis of Nose Landing Gear for Trainer Aircraft. Trends In Mechanical Engineering & Technology, 2014. 4. 22. Sadraey, M.H., Landing Gear Design, in Aircraft Design: A Systems Engineering Approach, Wiley, Editor. 2012. 23. Foundation, F.S., Landing Gear Topped List Of Aircraft Systems Involved In Accidents During 35-Year Period. Flight Safety Digest, 1994. 13: p. 13-16.

24. KrÜGer, W., I. Besselink, D. Cowling, D.B. Doan, W. KortÜM, and W. Krabacher, Aircraft Landing Gear Dynamics: Simulation and Control. Vehicle System Dynamics, 1997. 28(2-3): p. 119-158. 25. Besselink, I.J.M., Shimmy of Aircraft Main Landing Gears. 2000, Delf University of Technology. 26. Lernbeiss, R., Simulation of the Dynamic Behavior of Aircraft Landing Gear Systems. 2010, Vienna University of Technology: Vienna. 27. KUMAR, R.R., P.K. DASH, and S.R. BASAVARADDI, Design And Analysis Of Main Landing Gear Structure Of A Transport Aircraft And Fatigue Life Estimation For The Critical Lug. Design And Analysis Of Main Landing Gear Structure Of A Transport Aircraft And Fatigue Life Estimation For The Critical Lug, 2013. 1(4). 28. The Code of Federal Regulations of the United States of America. 1997. 29. Airplane Flight Manual GROB G 115. 1989. 30. Administration, F.A., Aircraft Weight and Balance Handbook. 2007. 31. Stearns, J.C., An Investigation Of Stress And Displacement Distribution In A Aluminum Alloy Automobile Rim, in Graduate Faculty of The University of Akron. 2000, University of Akron. 32. Huston, R. and H. Josephs, Practical Stress Analysis in Engineering Design. 3 ed. Mechanical Engineering. 2008: CRC Press. 33. Aviation, G., Aircraft Tire Data Book. 2002: Akron. 34. Aviation, G., Aircraft Tire Care and Maintenance. 2004: Akron. 35. Burr, A.H. and J.B. Cheatham, Mechanical Analysis and Design. 2 ed. 1995: Prentice Hall.

8 Annexes

8.1 Annex I

8.2 Annex II

8.3 Annex III

8.4 Annex IV