The present study, which was carried out in collaboration with GKN Fokker, focuses on incorporating bird strike crashworthiness requirements within a multidisciplinary optimization (MDO) framework. During the preceding three-month internship in the same company, a pivotal contribution to this project was the development of an Abaqus interface for the Multidisciplinary Modeller, MDM, created within the Center of Competence in Design department. MDM is a Python/ParaPy-based automated generator of wings, moveables and flaps, starting from a set of user-specified parameters. The generation of ready-to-run input files thus lays the foundation for the subsequent optimization process, as any changes in materials or geometry can be easily accommodated.

The core objective of the research is to minimize the weight of an aircraft wing while taking into account additional requirements related to the extent of damage caused by bird strikes. Unfortunately, such events occur more frequently than one would be comfortable with, and stringent requirements are set in place to guarantee the safety of the passengers. Among these requirements, the aircraft must be capable of landing safely after such an event, being subject to loads associated with get-home conditions.

As a consequence, two critical constraints are formulated within the optimization framework, addressing the residual strength of the damaged front spar following a bird strike, coupled with a requirement based on a maximum penetration depth. The last constraint has also been included due to the rising popularity of the electric vertical take-off and landing aircraft, which not only fly at low altitudes, thus increasing the risk of bird strike, but may also contain battery packs in the leading edge, for instance, which can pose a significant risk if damaged. To tackle the complexity of this highly-dimensional optimization challenge, a methodology based on Bayesian optimization is proposed, employing surrogate models coupled with a preliminary variable ranking procedure.

The Kriging metamodel is identified as a suitable candidate, thanks to its error prediction capabilities, which are paramount in Bayesian optimization. A variance-based dimensionality reduction method is proposed, which makes use of an initial surrogate to estimate the main and interaction effects of the variables. The quantification of the significance of a variable is expressed as its percentage contribution to the total variance, thus allowing for an intuitive selection of the most important parameters. After the screening procedure is complete, the optimization procedure is carried out in the reduced design space, which uses the constrained expected improvement as an acquisition function. The proposed methodology is then applied on a case study problem, involving a five-bay metallic wing segment subject to the constraints aforementioned, involving 19 design variables representing the thicknesses of various components.

Remarkable weight savings have been achieved, the final result being 40\% lighter than the lightest feasible design among the initial data points. A significant dimensional reduction has also been attained for the maximum depth constraint, which is expected due to the local nature of the impact. Not only did the number of variables greatly decrease from 19 to just 3, but a considerable increase in the accuracy of the corresponding metamodel has also been registered, thanks to an increase in sampling density in the reduced space. However, the variable screening procedure revealed intricate interaction effects with respect to the residual strength of the front spar, emphasizing the nuanced complexity inherent in crashworthiness considerations. Nevertheless, a moderate dimensional reduction has been achieved for this constraint as well, reducing the number of variables to 8, thus proving the efficacy of the proposed variable screening procedure.

In conclusion, the utilization of Kriging models, variable ranking procedures, and Bayesian optimization collectively contributed to the success of achieving remarkable weight savings, proving the efficiency of the proposed methodology. Moreover, it has been shown that the integration of a residual strength requirement is necessary, as many cases were uncovered where no significant penetration occurred, although the application of the considered load case, which is not from critical to an undamaged wing, resulted in high stresses to the front spar of the damaged structure.

Decentralised drone swarms need real time collision avoidance, thus requiring efficient, real time relative localisation. This paper explores different data inputs for vision based relative localisation. It introduces a novel dataset generated in Blender, providing ground truth optic flow and depth. Comparisons to MPI Sintel, an industry/research standard optic flow dataset, show it to be a challenging and realistic dataset. Two Deep Neural Network (DNN) architectures (YOLOv3 & U-Net) were trained on this data, comparing optic flow to colour images for relative positioning. The results indicate that using optic flow provides a significant advantage in relative localisation. The flow based YOLOv3 had an mAP of 48%, 9% better than the RGB based YOLOv3, and 23% better than its equivalent U-Net. Its IoU_0.5 of 63% was also 14% better than the RGB based YOLOv3, and 51% than its equivalent U-Net. As an input, it generalises better than RGB, as test clips with variant drones show. For these variants, the optical flow based networks outperformed the RGB based networks by a factor of 10.

The electricity generation from wind has seen significant growth over the past two decades with a compound annual growth rate of over 21%, and this trend is projected to continue, aligning with the goals of the European Green Deal. In the context of offshore wind turbines, operational and maintenance costs can account for up to 30% of the total costs throughout the project's lifecycle. This highlights the need for the development of efficient condition monitoring methods aimed at mitigating maintenance expenses.
The gearbox is one of the most failure-prone components in wind turbines, leading to extended downtimes and substantial financial outlays. When the components of the gearbox degrade, the heat generated within the gearbox increases. This increase leads to higher oil temperatures, making the temperature signal a suitable indicator of the gearbox health condition. The main objective of this thesis is to design a physics-based normal behaviour model (NBM) for the estimation of the wind turbine gearbox oil temperature and to use field data to validate its effectiveness. The energy conservation principle is applied to the gearbox to formulate an equation for the calculation of the gearbox oil temperature. To account for some unidentified design specifications of the gearbox, the equation is fitted to available historical data from the Supervisory Control and Data Acquisition (SCADA) system to determine the unknown parameters. The model uses as input an array of operational measured signals available in the SCADA system, including rotor speed, power output, nacelle temperature and inlet oil temperature, which is the temperature of the oil after running through the cooling system.
A case study demonstrates the model's effectiveness in accurately predicting the gearbox oil temperature with a mean absolute percentage error of 0.75%. In order to investigate how well the physical characteristics of the gearbox are represented by the model, the coefficients of the energy balance equation derived from fitting it to the field SCADA data are compared to the parameters calculated using known characteristics of a reference gearbox. The comparison results indicate that the model's parameters are generally within the same order of magnitude as the reference values. To provide a benchmark, the performance of the physics-based model is also compared to that of two data-driven models using artificial neural networks (ANNs) for temperature prediction proposed in the literature. The results show that the proposed physics-based model outperforms both ANN models, achieving 14% and 59% lower root mean squared error (RMSE) and a reduction in time required for training of 70% and 95% when compared to the two ANN NBMs respectively.

Structural health monitoring of maritime and offshore structures can contribute to the reduction of the amount of uncertainty in assessment of the operational loads and the structural integrity. Such a survey can also reduce the large safety margins during the design and optimize the conservative inspection and maintenance schedules. One of the emerging variants of structural health monitoring are the so-called digital twin systems that not only allow the monitoring of the structures, but can also simulate the past or the future state of the structural integrity. In the context of this thesis, the feasibility of a digital twin system for the structural health assessment of marine structures based on machine learning algorithms and utilization of big data is examined. The main research question is if the digital twin technology can be enhanced with machine learning to accurately monitor the structural health of marine structures in terms of loading and fatigue damage accumulation. In order to answer this question four modules have been developed and tested. The first module uses artificial neural networks trained on operational data to predict the fatigue damage accumulation and the frequency that corresponds to the maximum stress power density. The error of the developed networks turned out to be negligible in the investigated cases, with standard deviation of less than 4% in the predictions. The peak frequency of the stress power density is predicted using a random forest regression algorithm. The optimized version of the algorithm has led to an accuracy of 93% with standard deviation of less than 1% in the predictions.
The second module is used to recalibrate the design response amplitude operators of the structure using the predictions of the first module or the operational data, if available. The design response amplitude operators are scaled and shifted in order to minimize the deviation between the spectral fatigue calculations and the predicted (or measured) data.The third module is a static/quasi-static load-reconstruction module. Using on-board strain measurements it is able to calibrate the loading of the structure based on a reformulation of the conventional finite element problem for static and quasi-static loading. A sensitivity analysis of the algorithm effectiveness have been performed using multiple load cases in which the reconstruction error turned out negligible. The fourth and final module relates to structural reliability analysis. This module is based on the estimation of the Hansofer-Lind reliability index as a minimization problem. The employed optimization engine is a variation of the particle swarm optimization algorithm using chaotic system behaviour to iteratively calculate the user-defined parameters. The limit state equation is formulated in a way that it takes the uncertainties related to the fatigue damage accumulation prediction model and Miner's Rule into consideration. Uncertainties related to the material and the fabrication process are not taken into account. The developed modules have been tested on structural details of a vessel monitored in the scope of Monitas Joint Industry Project (JIP). About a year's worth of data has been used to train the machine learning algorithms.

Type IV composite pressure vessels (CPVs) are used commercially for the gaseous storage of hydrogen in fuel cell electric vehicles (FCEVs). However, their economic implementation requires material optimization and a reliable prediction of the vessel strength. In this regard, their burst when loaded under internal pressure is impacted by the combined effect of the stacking sequence design and the variability of mechanical properties resulting from the manufacturing process. This work shows a framework for analysis that accounts for some of these manufacturing-induced characteristics in the mechanical response, namely the relation between the vessel stacking sequence, its final geometry, and the material properties. Furthermore, vessel burst pressures are alternatively estimated from the failure criteria evaluation in constitutively elastic analyses and the modeling of damage progression. Predictions are reasonably accurate when the collapse occurs in the cylinder (+2.2 %), although a more considerable discrepancy exists with experimental results when vessels fail in the dome transition region (+12.3 %).

Wind energy is one of the cleanest methods of producing electricity. It plays a major role in the transition towards green energy. It is estimated that the installed capacity of wind energy will cover more than 6 % of total global electricity demand by the end of 2019. As the demand grows the requirement for bigger wind turbine also grows. Each year larger wind turbine blades are being manufactured to satisfy the growing demand, resulting in high manufacturing cost and operation and maintenance cost. The main reasons for wind turbine blade failures are poor design, material failure, power regulator failure, extreme load buckling, lightning strikes, foreign object damage and human errors. The failure rate of the wind turbine blade is around 7 % compared to other components of wind turbine. So, an effective and proactive monitoring system for the wind turbine blades becomes necessary due to the large downtime associated with failure and high replacement cost. The main objective of this thesis is to use smeared crack modelling and integrating it with FEM tools (ANSYS 18.2) to identify the damage in the wind turbine blade. The concept that the modal parameters such as natural frequency, mode shape and damping will be affected when a structure is damaged, is used to identify and localize the damaged elements. First, spar cap - shear web assembly is modelled as simple "I" beam in ANSYS. The dimension of the blade is measured from the "NedWind 40 turbine" provided by LM Wind Power. The structure has been modelled using composite materials. The spar cap fibres are laid at 0 ̊ and the fibres of face sheet are laid at ± 45 ̊. One end of the structure is fixed to simulate the actual working conditions. Similarly, the material property of all parts was chosen to represent the actual wind turbine blade. Second, damage is introduced into the structure in the form of microscopic damping into the material property. To incorporate damping into the structure, the loss factor obtained from experiment is used. The damage modelling is done for different severity at different locations of the modelled structure. Third, the eigen value and eigen vector for different microscopic damping coefficient is extracted. Modal damping of the structure is calculated using the extracted eigen value. And eigen vector is used to calculate the phase angle of each elements. The phase angle of each element is plotted to localize the damaged elements. The results show that this method is suitable for identifying defects near the fixed end and in the midsection of a wind turbine blade. Also,1st flap wise and edge wise bending frequencies are suitable for identifying the defects near the fixed end and 2nd flap wise bending frequency is suitable for identifying the defect near the mid area.

The structural properties of a composite laminate can be tailored by allowing the fibres to vary their orientation within the individual plies. This class of composites is termed variable stiffness and they are able to enhance the structural performance over traditional laminates by redistributing the internal load of the structure. Therefore, it is of interest to pinpoint the structural properties that may benefit most from this concept and the extent thereof. To that end, a multi-objective optimisation of a variable stiffness cylinder was performed which identifies the Pareto front for objectives such as buckling, natural frequency, stiffness and strength. Manufacturing considerations were also incorporated by accounting for the allowable tow curvature and the resulting overlaps during the process. The stiffness variability was formed by means of a Lagrange polynomial, the non-linearity of which is determined by a specified amount of control points. An investigation on the suitable degree of non-linearity of the functional representation of the cylinder was thus carried out. In order to complete the optimisation in a practical time frame, machine learning algorithms were utilised to act as surrogate models and perform predictions on the optimisation objectives. Different algorithms were implemented for that task, namely Radial Basis Function, Kriging and Artificial Neural Network models. The Artificial Neural Network surrogate model outperformed the others when the dimensionality of the problem was high, although the opposite was observed for low dimensional problems. Cylinders with and without cut-outs were optimised separately and compared with their optimised constant stiffness counterparts. It is found that the performance gains for the cylinder without cut-outs are not significant when the additional mass due to the overlaps is accounted for. On the other hand, the cut-out cylinder greatly benefits from the stiffness variation for multiple objectives. Additional structural improvements were witnessed for higher degrees of non-linearity of the utilised Lagrange polynomial.

Airlines experience schedule disruptions on a daily basis. Poor weather conditions, unscheduled aircraft maintenance and congested air spaces are just a few of the causes that prevent airlines from operating their flight schedules as planned. In the third quarter of 2017 over 20% of all scheduled flights in Europe suffered from delays. Operation Research based decision support systems (DSS) help airlines with their disruption management processes and provide suggestions for recovery options.

For large hub-and-spoke carriers, with an extensive network and a large number of aircraft and flights, the computation time required to find the optimal recovery solution after a disruption increases rapidly. As airlines require fast recovery solutions when disruptions occur, there is an ongoing trade-off between computation time and system sophistication. In the majority of disruption cases, no or a limited number of undisrupted aircraft are required to find the optimal recovery solution. By selecting a limited number of aircraft, flights and airports used to find a recovery solution, the computation time is reduced exponentially. The challenge is determining which aircraft and flights should be selected.

This research aims to develop a decision support system for the schedule and aircraft recovery process that is able to present a feasible solution to a disruption in less than 120 seconds. An aircraft recovery model will be developed based on the integer linear programming model that was created by Vink et al. (2019) and Vos et al. (2015). Crew and passenger recovery are not considered. To recover disruptions, the optimization model can delay and cancel flights as well as perform tails swaps, where the flights from two aircraft are switched. The novelty of the work is that machine learning is used to predict which undisrupted aircraft will help recover a disruption. Based on those predictions a sub-network selection algorithm will select the subset of aircraft to be included in the optimization instead of the entire aircraft fleet.

The performance of the system is tested on a case study for the domestic hub-and-spoke network of Delta Airlines. The dataset for the study consists of 2200 daily flights, 147 airports and 827 aircraft in 8 aircraft families. The results of the system are compared with the optimal solution, where no aircraft selections were made. The case study shows that the system is able to make an aircraft selection where 50% of the fleet is discarded, while still finding the optimal solution in 98.9% of the 556 disruptions tested. Furthermore, the system reduces the computation time by 45%, resulting in an average time of 48 seconds. For the disruptions, the computation time varied between 9 and 180 seconds.

Induction welding is an effective technique for joining unidirectional carbon fiber reinforced thermoplastic composites and L-joints can be produced through quick and cost-effective processing steps. However, due to high localized stresses in the skin-stiffener interface, these L-joints are often avoided in primary aircraft structures. Also, no international testing standards have been developed for testing of such joints.

A method was developed for the implementation of a neat thermoplastic resin fillet between the L-joint skin and stiffener web using the induction welding process in an attempt to remove the high stress concentration at this location. A 35.4% increase in quasi-static pull-off strength was measured with a weight penalty of less than 0.5%. This result was compared with a similar autoclave co-consolidated joint, which showed an 80.9% improvement. An ANSYS Parametric Design Language finite element model was developed based on the virtual crack closure technique and it showed that the joint pull-off performance is strongly dependent on geometric parameters such as the skin and stiffener thickness. Also, a new test setup was developed, which reduced internal stresses created by the setup compared to those commonly used in literature.

By further improving the method through which the fillet is joined to the induction welded L-joint, a performance increase similar to that of the co-consolidated joint should be achievable. Test results have shown that the use of this type of fillet can lead to the skin-stiffener interface no longer being the critical failure point for realistic joints in primary aircraft structures.

Operations & Maintenance (O&M) contributes to about 25% of the total offshore wind farm costs which is rather large. It also influences the levelised production cost per kWh over the entire lifetime of the wind farm. Therefore, it is necessary to reduce the costs involved in O&M activities. One such activity is the logistics involved in O&M. The increasing logistical costs have been a challenge which can hamper the O&M market in the future. Therefore, it is important to take necessary steps to reduce the logistical costs. Wind farm operators have to take up logistical decisions to make optimal use of the available resources to maximise the wind farm availability. These decisions involve which turbines to repair, the route to follow considering constraints like weather conditions and vessel availability. A decision support tool aids in taking appropriate decisions to carry out the logistical activities. This calls for a testing tool which can assess the decisions taken by the planner or the decision support tool. Current research helps in developing a method to assess the quality of the decision taken by the decision maker or the decision support tool. This will help in making better decisions and reducing the costs involved in the logistics of O&M for Offshore Wind Farms (OWF). The objective of this graduation project is to set up guidelines for a test bench to examine the optimality of the decisions taken by the planner or the decision tool to carry out the logistical operations for a shorter duration.

An increase in apparent mode II fracture toughness in Fiber Reinforced Polymers has been found from modified Transverse Crack Tension (mTCT) tests with compressive stresses added orthogonal to the delamination plane. Numerical reproduction of these results has not yet led to a good understanding of the reasons for this behavior. Currently, an analytical approach is used for determining the mode II fracture toughness from an mTCT test. In this research, the validity of the analytical relation is investigated for the case of an added compressive stress. The behavior of the material in the
test is therefore numerically investigated in a qualitative manner. A mesh, using a material model for plasticity and a cohesive law for interface elements to form the crack, which have been proven to give good results, has been build. A refinement on mesh and geometrical dimensions, focused on the pressure zone of the specimen, ensures the degree of detail required to investigate the local behavior. A parameter study on the inlfuence of geometrical and material parameters is performed and linked to results from previous conducted experiments. A standard mTCT test without compressive stress shows a linear load displacement relation until crack initiation and then a constant load at which the crack grows through the specimen. This crack growth load is used for the analytical approach. The results from this research show a change from a constant crack growth load to a linear crack growth load range in the zone of the applied compressive stress. After crack initiation, the crack growth load through the compressed zone has an increasing, nearly linear relation, based on the compressive stress and length of the pressure zone. The analytical approach, in
which the highest reached load on the specimen before failure is used as input, no longer applies for this case. The failure load is now dependent on the length of the pressure zone and pressure magnitude.

Ever since the attacks on the World Trade Center, airport security has been a topic of interest. The United States was caught by surprise and the attacks triggered a renewed interest in aviation security. It was well recognized that airport security was one step behind on intelligent attackers and this created the need to developed better risk assessment methods.

Unfortunately, none of the risk assessment methods developed has the possibility to quantify the vulnerabilities in an airport checkpoint. One of the main challenges in developing a method which can do this, is to find a technique that can account for the complexity of the airport environment from which the vulnerabilities emerge. Furthermore empirical research has shown that security operators do not necessarily follow protocol, but behave as autonomous agents which regularly bend or break the rules.

A method which can potentially identify vulnerabilities in such a complex environment is agent based modeling. This modeling technique has proven to be very powerful in modeling complex systems that emerge from the behaviour of autonomous agents. Some work has been done in this area, but until now this technique has not been used to quantify vulnerabilities in an airport checkpoint. Therefore, the aim of this project is to develop an agent based model of an airport checkpoint, quantify the vulnerabilities emerging from this checkpoint and analyze the effect of employee behaviour on these vulnerabilities. To do this, the behaviour of the security operators is modeled using models that are rooted in behavioural psychology and have strong empirical backing. The employees decision making is modeled using Decision Field Theory and the employee performance is modeled using the Functional State Model.

With the model developed, a set of experiments is performed to calibrate the model and analyze the vulnerabilities. These experiments result in the quantification of vulnerabilities for a predefined set of threat scenarios.

The analysis of these threat scenarios shows that employee behaviour mainly impact threats scenarios in which a weapon is hidden in the carry-on baggage. The reason that security operators have a large influence on the outcome of this screening process is that it is the process in which employees have to perform multiple activities and make multiple decisions. It is found that the vulnerabilities in this screening process are mainly dependent on the speed/accuracy trade-off as made by the employees. The perceived risk of the detected prohibited items plays a limited role, since most prohibited items are only seen as a small risk. The performance of the employees played a less important role on the outcome of the screening process and differences in performance are mainly caused by the personality type of the agents. The personality type that put more effort into a task, outperformed the other agents. The skill level of the operators however, did not significantly effect the outcome the simulation. This suggests that the effort an operator puts in is more important than the skill level of the operator.

Finally it is found that it is beneficial to minimize the number of steps in the screening process. This benefits overall performance, since adding steps to the screening process means adding opportunities to make mistakes.

This report, which from now and onwards will be called the main body of the thesis, covers the core of the
problem that has to be solved. Generally, the aim of this thesis is to investigate a laminated composite’s
structural response, made of FRP, when its material properties are assumed stochastic instead of
deterministic. The materials shown in this thesis are investigated as part of a bigger project that concerns the
construction of several Mine Counter Measure Vessels.
In this document, the main focus is given on the results of experiments and/or analytical calculations whereas
the theories/methods used to extract these results are presented in the supporting document. As a result,
readers that are familiar with statistics, classical lamination theory, FEA modeling among others, can easily
read this report without coming across knowledge that they already have. In general, this document covers
the writer’s contribution to academics whereas the supporting document gives an extensive explanation on
what was used in this document.