Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)

Permanent URI for this collection

https://hdl.handle.net/10889/47

Browse

Now showing 1 - 20 of 636

Open Access
A computational study of blood flow in the left atrium of patients with atrial fibrillation
Βάρσος, Παύλος; Varsos, Pavlos
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
The main object of this Diploma Thesis is the computational study of blood flow in the left atrium of patients with Atrial Fibrillation. Atrial Fibrillation is the most frequent cardiac arrhythmia worldwide, since it affects 1 to 2% of the general population, while it is accompanied by high mortality rates. Like any other arrhythmia, Atrial Fibrillation refers to cases in which the electrical activity of the heart involves abnormalities, which are often not completely treated with neither medication nor surgery. A critical treatment method, minimally invasive, is catheter ablation which is simultaneously conducted with the procedure of electrophysiological mapping of the left atrium. The scope of this project is to create a framework for the comparison of mappings of multiple wall shear stress indices (which according to experiments are associated with the cardiac electrical activity) with electrophysiological mappings, using patient-specific data provided by doctors. Possible correlations can supply doctors with crucial information in a direction of improving the ablation process outcomes. Reconstruction techniques are implemented to convert two-dimensional MRI images into three-dimensional models of the left atrium. Following the optimization of the 3D model, the mesh generation and the solution of the cardiac flow procedures take place. Both of these processes are performed in the open-source software, OpenFOAM ®, as we take advantage of the countless capabilities offered. After generating the mesh, a procedure of paramount importance for a computational study, the mesh independence study is carried out under steady state conditions using the simpleFoam solver, aiming to obtain the most suitable model for the present project. At the same time, various turbulence models and boundary conditions are tested in order to detect any possible differences in blood flow behavior. Finally, a transient flow simulation of a realistic cardiac cycle is conducted, utilizing the pisoFoam solver, since our main goal is to visualize the distribution of a variety of hemodynamic parameters, such as the Oscillatory Shear Index, as well as indices for indicating blood stasis in high-risk areas of the left atrium, like the Left Atrial Appendage.
Open Access
A S-N-based cohesive element for delamination propagation in composite materials
Μήτρου, Ανατολή; Mitrou, Anatoli
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
Fracture in composites is composed of complex micromechanical networks of cracks. These damage mechanisms, which include fiber bridging, transverse matrix cracking, splitting, and migration, can provide a degree of resistance by redistributing the loads away from the stress concentrations and blunting the crack tips. The resulting increase in the critical energy release rate with crack extension is referred-to as a resistance curve or R-curve. The role of R-curve effects on the rate of crack propagation in fatigue is the subject of current research. An experimental study was conducted by Yao to quantify the significance of fiber bridging in the delamination growth in multidirectional composite laminates by comparing quasi-static and fatigue R-curves. He observed that fatigue R-curves are less pronounced than quasi-static R-curves. Analysis methods capable of capturing R-curves are not yet mature, especially in fatigue. Yao’s results were used in an effort to evaluate a recently proposed cohesive zone model (CZM) for both quasi-static and fatigue. The quasi-static R-curve behavior is modeled using a cohesive layer superposition approach. The model utilizes the CZM developed by Turon et al., and assumes that the quasi-static cohesive law is also the envelope of the fatigue damage. The fatigue model also relies on intrinsic relationships between an S-N curve and its corresponding Paris law. Damage is assumed to accumulate at a rate dependent on the displacement jump and described by a damage accumulation function. The parameters of the model are obtained by fitting the integrated form of the damage accumulation function to an idealized S-N curve. The assessment of the numerical procedure is done by evaluating the method’s ability to capture Yao’s rates of crack propagation for unidirectional (UD) laminates. The analysis results indicate that the numerical model can predict slope and the offsets in the Paris law observed in the experiments for short cracks. However, as cracks extensions grow past 35 mm propagation rates are overestimated. Additional work is underway to understand this discrepancy.
Open Access
Adaptive control of 3D printing device based on digital twin concept
(2023-07-06) Βαρσαμής, Αντώνης; Varsamis, Antonis
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
In the realm of the fourth Industrial Revolution, digital technologies are constantly integrated in Manufacturing processes in order to improve system reliability, monitor processes accurately and ultimately optimize processes. In order to tackle these issues, advanced simulation techniques such as Digital Twins (DT) have to be developed. Further to that, modern manufacturing systems are based on the integration of non-conventional processes, such as Additive Manufacturing (AM). Therefore, the scope of this thesis is to design and develop a Digital Twin based framework for Fused Deposition Modeling (FDM). The purpose of the Digital Twin is to integrate quality assessment modules, and setup a database for monitoring all the experiment results. Additional modules will also be created and integrated, such as compensating the effects of printing, as well as immersive interfaces enabling the users to control and monitor the machine state remotely. Consequently, one of the key concepts of this thesis is the design of the information flow between the above-mentioned modules for optimizing the FDM process parameters, and thus minimize time and resources caused by human error. The implementation of the proposed framework is based on the integration of suitable communication channels.
Open Access
Adaptive refinement techniques for faster wave propagation simulations in rods and beams via the multiresolution finite wavelet domain method
(2023-10-19) Σαμαρά, Αικατερίνη; Samara, Aikaterini
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
The Multiresolution finite wavelet domain method (MR-FWD) is a numerical technique that employs Daubechies wavelet and scaling functions as basis functions for the approximation of the structure’s displacement field. The method provides a hierarchical representation scheme involving an initial, zero-scale approximation, while finer, higher-scale components are superimposed to improve the solution’s accuracy. In the present Diploma Thesis, adaptive refinement techniques are presented to accelerate wave propagation simulations in rods and beams via MR-FWD method. The fundamental idea lies in separating the solution into high- and low-wavenumber components and approximating each component with an appropriate level of resolution. Smooth low-wavenumber variations are accurately represented through low-scale approximations, whereas abrupt high-wavenumber oscillations require advanced computational effort and as such, they are handled with higher resolution levels. A heuristic indicator is employed to assess the quality of the solution and localize areas that need solution enrichment in the displacement field. Accordingly, a dynamic refinement domain is formed to define regions where increased precision is required. The efficiency of the proposed refinement process is evaluated through numerical results for the prediction of wave propagation phenomena in aluminium rods and composite beams. The results demonstrate substantial reduction in computational effort compared to the solution using uniform refinement, while maintaining a high level of accuracy. However, it is essential to meticulously define an appropriate threshold for the indicator to accurately locate the refinement region while ensuring computational efficiency.
Open Access
Adaptive scheduling in structural steel manufacturing enabled by production monitoring
Γάργαλλης, Αντώνιος; Μούρτζης, Δημήτριος; Μούρτζης, Δημήτριος; Σταυρόπουλος, Παναγιώτης; Δέντσορας, Αργύριος; Gargallis, Antonios
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
Due to the massive growth in available data to consider, decision making within the factory setting has become ever more complex. Also the increasing market demands concerning product quality and timely delivery have further hindered the ability of having critical judgement and being decisive when dealing with manufacturing problems. Thus using only human resources to form decisions and achieve manufacturing goals has become insufficient. New emerging technologies like the Internet of Things (IoT) and Cyber-Physical systems (CPS), which exist in the core of Industry 4.0 enabled factories, can help achieve decision making goals. The use of Industry 4.0 technologies in order to extract and handle useful data can be used to support human decision making. This can enable insight of the current and future status of a given production system, resulting in more accurate predictions and enhanced critical decisions. Production scheduling is one of the main problems production engineers have to tackle. The decisions taken considering production scheduling can greatly affect the whole production process. Industry 4.0 enabled decision support tools can help making production scheduling effective, while considering more data and parameters than ever before. As such this thesis proposes a methodology for production scheduling, based on historical and near real time data for checking resource and task status. The developed framework aims at providing useful insight to production engineers, assisting them in the decision-making process. Finally, the applicability of the framework was tested in a real-life case study of a structural steel manufacturing industry.
Open Access
Aeroelastic and structural similarity optimization of scaled wing models
(2023-02-13) Φιλίππου, Ευάγγελος; Filippou, Evangelos
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
The ongoing pursuit for more efficient transportation has led engineers to design lighter and blended wings in order to reduce fuel consumption and emissions. However, these designs introduce strong aeroelastic couplings that can result in failure. As a result, aeroelastic analysis and optimization have become crucial aspects of aircraft design in modern times. Additionally, aeroelastic testing of scaled models is another key phase in the development of aircrafts. The accurate prediction of aeroelastic response through the construction of models is vital for cost reduction and avoiding risky flight testing. This is achieved by maximizing the similarity of stiffness and mass distribution along with the flow field similarity during the creation of the scaled model. However, in today's lightweight aircrafts, the exact same geometry cannot be scaled down, requiring the use of a different configuration. This variation in configuration affects the aeroelastic response, making the use of computational aeroelasticity tools and optimization algorithms necessary. This computational technique constitutes a modern approach to aeroelastic scaling. This thesis focuses on developing an aeroelastic scaling framework using multidisciplinary optimization. Specifically, a parametric wing finite element model (FEM) is created that incorporates both thickness and geometry parameterization, primarily using shell elements. The aerodynamic loads are calculated using the DLM method along with twist and camber correction factors, and the coupling is established through the use of infinite plate splines. The aeroelastic model is then integrated with an Ant Colony Optimization (ACO) algorithm to achieve static and dynamic similarity between the scaled model and the reference wing. The key contribution of this thesis is the inclusion of internal geometry parameterization within the framework. The results of the implementation of a two-step optimization technique indicated high similarity in both the static aeroelastic and modal response of the two wings. The root mean square error (RMSE) of the in-flight shapes converged to 0.055m in the full scale and the mean MAC value of the first 5 modes was 0.934.
Open Access
AI-driven production scheduling based on multi-agent system
Σιάτρας, Βασίλειος; Siatras, Vasileios
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
Mass production industries and global firms depend on production management in order to compete in the modern volatile manufacturing marketplace. Achieving high manufacturing performance requires a compound of AI services and information technologies working in parallel across different business layers and ensure information transparency and precise decision-making. As research and innovation took off, an ecosystem of smart autonomous assets (so called agents) has been introduced in all manufacturing levels constructing the main idea behind Industry 4.0. With Multi-Agent Systems (MAS) appearing in the forefront of nowadays manufacturing research and innovation, the requirements of interoperability and granularity is the primary challenge for I4.0 stakeholders. The realization of this objective is found in the core components of I4.0, so called Asset Administration Shell (AAS), which provide abstraction in the production assets description and interactions. As a result, managing to design and develop AI services with increased abstraction in the applied environment enables the deployment of increased automation and flexibility in the business operations. This thesis focuses on exploiting AAS technology for the design and development of a scheduling meta-agent that uses AI in order to solve different production schedule optimization problems. For that reason there was proposed a toolbox of three independent AI agents, that act as a plugin to the meta-agent interface and allow decision-making in three different scheduling problems namely factory, logistics, and conveyor scheduling. (a) The development of the factory scheduler plugin a heuristic decision-making algorithm for scheduling was utilized; (b) for Logistics scheduler was based on a mathematical optimization problem that uses Fuzzy Logic and is solved by a Genetic Algorithm; (c) while for the Conveyor scheduler there were used both a Mixed Integer Programming model and two Machine Learning model (LSTM RNN, and a FFNN). The MAS was deployed in two different manufacturing cases i.e. automotive and bicycle manufacturing industries demonstrating improvement in the business KPIs. In overall, the MAS was proven efficient in the ability of managing the production environment with low human interference required.
Open Access
Analysis of structural damping effects on the dynamic response of a wind turbine blade
(2022-10-13) Παπανακλής, Γεώργιος; Papanaklis, George
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
The main aim of the present Diploma thesis is the development of a Reference Wind Turbine blade’s numerical model and the implementation of structural dynamic analyses in order to evaluate the structural damping effects and the structural integrity of the blade. During the last decades, the need for renewable energy that has lower environmental impact has increased exponentially, drawing the attention of the scientific and industrial community towards the wind energy sector and leading them to put some serious effort on the development of larger and more efficient wind turbines as also dedicated analysis and design tools. In this work, some theoretical background related to the wind turbine blades’ structure and aerodynamic loads is provided, along with a brief presentation of the modelling methods that are used for the simulation of the blade’s dynamic response. Dedicated wind turbine simulation tools are examined regarding their framework, as well as the aerodynamic and structural theories and approaches they employ for the simulation of the entire mechanism. The theoretical part of this thesis is completed with the review of composites’ loss factors and proportional damping, while the definition of Rayleigh factors that will be utilized in the numerical models is explained. Before the description of the blade’s model, a brief review of the two numerical tools (hGAST and DAMPBEAM) that provide data for the modeling and structural assessment of the component takes place. During the definition of the wind turbine’s blade model in Abaqus, the CAD model is created, material properties and composite layups are defined, available dynamic analysis methods and procedures are examined, load modeling is described, and the meshing strategy is explained. The implementation of modal analysis provides the eigenfrequencies and eigenmodes of the structure that are successfully compared with the respective ones from the DAMPBEAM code. For the damping assessment, various dynamic analyses are performed, where different aerodynamic load cases, damping properties and analysis methods are employed and their results are compared, extracting important conclusions for the effect of each parameter. Finally, blade’s structural integrity is evaluated using Tsai-Wu stress-based and maximum strain failure criteria, validating that structure’s capability of dealing with the load condition of the simulated operation when damping is taken into consideration.
Open Access
Application of genetic algorithms for the optimization of the vibrational behavior of mechanical systems
(2022-10-13) Σώλου, Άννα Βασιλική; Solou, Anna Vasiliki
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
In recent decades, there has been a demand for the development of practical, cost-effective control system applications that properly address the negative effects of vibration and protect various machines from natural or man-made hazards. For this reason, vibration control has received considerable research attention. Vibration control devices can be divided into three categories called passive, active, and semi-active control systems. A parallel development in the last two decades has been the emergence of the field of computational intelligence (CI) consisting mainly of the three fields of neural networks, evolutionary computing, and fuzzy logic. This study focused on the application of genetic algorithms (GAs) to optimize the vibration behavior of two types of mechanical systems, a single-degree-of-freedom (SDoF) vibration system, and a mechanical system with a dynamic vibration absorber (DVA) that are both widely used as vibration isolation devices. In addition, a passive vehicle suspension system was modeled, and the system was optimized to check the reliability of the genetic algorithm. Finally, suitable user-friendly Suitable graphical user interfaces (GUIs) were created by applying the above optimization process, to train users in both the theory of oscillations and the understanding of the theory of genetic algorithms. From a technical point of view, this research argues that the use of GA to find optimal solutions in oscillation control cases can be a reliable and well-performing tool that can contribute to the training of people interested in such educational fields.
Open Access
Assessment of novel deep learning methods for sea vessel condition monitoring
Θεοδωρόπουλος, Παναγιώτης; Theodoropoulos, Panayiotis
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
In the past decade, in the context of a rapidly changing and fiercely competitive global economy along with continuous efforts against climate change, maritime and marine engineering endeavors towards rendering ships more efficient and environmentally friendly are of vital essence. In this framework, the concept of condition monitoring and its modernization utilizing the latest digital technologies and implementing data-driven analyses is gaining more popularity within the marine industry. This study proposes the use of 2 types of convolutional neural networks (CNN), to recognize patterns that indicate early signs of defective behavior in the operational state of the vessel, benefiting from the capability of these models to represent features at high levels of abstraction without requiring to mine the features on which they shall be trained, which in numerous situations of condition monitoring is a challenging task requiring a lot of domain expertise. Specifically, the two methodologies examined in the present study are based on a 2D CNN, whose inputs are contour images extracted by the dataset, and on a 1D CNN with the time series being fed directly to it, in both cases, the dataset undergoes a preprocessing procedure. A parametric inquiry is presented in an attempt to deduce the parameters of the networks that ensure the best performance in the task of timely detecting defective patterns. Comparisons between the two proposed methodologies and benchmark classifiers are carried out to determine the strengths and weakness of the proposed methodology. The results provide an insightful picture of the applicability of CNN models in performing condition monitoring in ships, demonstrating the remarkable acquired in both study cases outperforming the benchmark classifiers.
Open Access
Automated task allocation competency-based model
Γιγής, Γεώργιος; Gigis, Georgios
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
Industry 4.0 has created several challenges for the human element in the working environment. With the Industry 5.0 emerging, the workforce receives an even more necessary role, of integrating the technology to the everyday operations of a production unit and ensure that the adoption of novel solutions is successful, while on the same time change their role, from performing tasks to innovating and solving complex problems. Taking into consideration that the human operator shift their focus to more creative and strategic jobs, the task allocation process will face new challenges and technological advancements should provide solutions for them. Now, jobs are assigned to people based on the supervisor’s intuition and judgement, wasting important time from defining the task till finding the most qualified employee for the specific job, and lacking documentation of the decision-making process. Although, competencies have been discussed thoroughly over the past few years, there is still room for improvement in measuring and comparing different competencies of human operators, engineers, technicians, and using the results to manage human capital. To uncover the potential of the task allocation, process the competencies will be used as decision-making criteria, aiming to achieve the best possible performance, through an automated system to assist the project manager find the optimal division of tasks to the available human resources, based on their competencies. The proposed framework can be applied both on a project and on a production line because it can be performed with real-time data and task allocation methodologies are the same for them, with the only difference being the task duration. In order to validate the applicability of the proposed framework and the generated algorithm, a case study derived from a maritime company has been examined, with the objective to allocate in the optimal way seafarers to vessels and maintain a balanced competency level among the fleet.
Open Access
Collision detection of a robotic arm using statistical time-series methods : a comparative study
Ζησόπουλος, Σπυρίδων; Zisopoulos, Spyridon
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
This study explores the detection of collisions on a collaborative robotic arm (cobot) under different Case Studies that correspond to distinct robotic tasks using statistical time-series methods. The collisions cause abrupt changes in the investigated torque and force signals acquired from the test rig, and two unsupervised and one supervised method are postulated for the effective and early detection of them. The postulated methods are three, namely the variance-based, the wavelet-based (supervised) and the RAR model-based, which are comparatively assessed with each other, as well as with two state-of-the-art methods, namely the residual-based and the FFT-based (supervised), in terms of True Positive Rate (TPR), False Positive Rate (FPR) and Detection Delay Time (DDT). The three case studies investigated are based on the measurements and the results of previous works that use the same KUKA LWR4+ robotic arm and they include either the autonomous motion of the robot in the first two Case Studies or the physical collaboration between the human and robot in the third Case Study. The collisions are of different nature along the different Case Studies, with varying magnitude and direction. The results indicate great TPR via all three postulated methods, that outperform the state-of-the-art ones, while all methods have zero FPR. Regarding detection time, each method has better performance in a different Case Study, with the RAR model-based being the fastest overall, followed by the wavelet-based and then the variance-based one.
Open Access
Computational study of the effect of dimples and bumps on an unmanned aerial vehicle wing, configuration optimization and comparison with experimental wind tunnel results
(2023-03-03) Δημητρόπουλος, Δημήτριος; Dimitropoulos, Dimitrios
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
The climate crisis and the environmental degradation that have emerged in recent years have made the reduction of the aviation industry’s carbon footprint, via the aerodynamic optimization of aircrafts, a very important research topic. The aim of the present work is to improve the aerodynamic performance of the ethERAS UAV wing, by delaying the separation of the boundary layer on its suction surface using dimples and bumps. These modifications are a means of passive boundary layer control, which have been mainly used in golf balls. Their function is based on introducing instabilities and turbulent fluctuations to the flow, which lead to larger momentum transfer to the boundary layer, enabling it to travel longer distances, even under the effect of large adverse pressure gradients. To determine the optimal size and position of the modifications, five configurations were designed for each modification type, resulting in a total of ten configurations. All CAD models were created using DASSAULT SYSTÈMES Catia. Afterwards, the models were prepared for the computational simulations using ANSYS Spaceclaim. All wing designs, including the original, non-modified wing, were studied at a low velocity (7m/s), to simulate landing conditions. From this case study, the best performing configuration for each modification type was chosen. The two chosen configurations and the original, non-modified wing design were studied at cruising speed (27m/s). The computational grids were created in ANSYS Fluent Meshing, using the Poly-Hexcore method of the ANSYS Mosaic Meshing technology. The computational simulations were implemented using the ANSYS Fluent pressure-based solver, the SST k-omega (SST k-ω) turbulence model and second order spatial discretization schemes. The numerical results of the 7m/s case study were validated using experimental, wind tunnel data. The two chosen configurations from the 7m/s case study and the original, non-modified wing were 3D printed on scale of 1:4 and were used for the experimental study. The comparison of the computational and experimental results showed good agreement. From the overall study it was concluded that the use of dimples aided in the enhancement of the aerodynamic performance of the wing, by improving its aerodynamic efficiency. The use of bumps proved to be inconsistent, leading to the overall degradation of the aerodynamic performance.
Open Access
Conceptual development of a knee simulating device for friction and wear testing of biomaterials
Μωραΐτης, Κωνσταντίνος; Moraitis, Konstantinos
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
Over the past decades, the field of knee joint simulating devices has been constantly evolving. Different kinds of experimental simulators offered valuable insight over kinematic and loading behavior of a natural or artificial knee joint. However, as human life expectancy constantly increases, people tend to experience more and more fatigue around their knee joint, resulting in intense wear and pain, thus experiments over the new prosthetic technologies introduced have to accommodate accelerated and more demanding testing regimes of variant daily and sport activities. The main objective of this study was to develop a novel dynamic knee simulator that not only abide to but surpassed the requirements of the ISO 14243 Standard [1, 2], including the operating frequency,specimen mounting, positioning & alignment, range of motion and in-vitro conditions such as the incorporation of lubrication, temperature, and pH control systems. The first development step included the conceptual design of the simulator and its optimization using CATIA (Dassault Systèmes, Vélizy-Villacoublay, France) designing software, where factors such as design for assembly (DFA), design for manufacturing (DFM), ease of maintenance, safety measures and modularity determined the final result. Some of the key characteristics of the apparatus were the 6 degrees of freedom permitted between the two articulating bones with five controlled axes, the option to mount and investigate the performance of specimens regardless of their size, as well as a custom fluid test medium enclosing capsule that did not interfere with the outgoing procedure, even at the maximum flexion angle of 120 degrees. The next step was to investigate the structural integrity of the conceptual design introduced, by imposing severe static loading conditions to replicate complex daily activities by using the ANSYS Static structural software. Finally, through a modal analysis, vibration characteristics were inspected to ensure that the machine operated at a frequency different from its natural ones, so resonance could be avoided. Throughout the simulations, a mesh independence study in conjunction with non-linearities were included, to obtain more realistic and accurate results. According to the findings of this study, the introduced conceptual design of the knee simulating device was functional and reliable.
Open Access
Creation of virtual environments for enabling the assessment of the ergonomics of manual operations
Τόγιας, Θεόδωρος; Μούρτζης, Δημήτριος; Σταυρόπουλος, Παναγιώτης; Ανυφαντής, Νικόλαος; Togias, Theodoros
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
The content of this thesis describes the design and development of a workplace evaluation and design software. The aim of this software is to simulate manual assembly processes, and provide tools to evaluate their efficiency and ergonomics, using Virtual Reality (VR) technology. The developed software is named Workplace Optimization Service (WOS), consisting of two modules: i) Shop floor Visualization, Simulation and Design Module(SVSDM) and ii) Virtual Workplace Analysis Module (VWAM). The SVSDM is the component of WOS that allows the immersion of the workplace engineer and the operator in a virtual shop floor. SVSDM has two main functionalities, the simulation mode where the operator may perform the simulation replicating real procedures and the design mode, where the actors are able to visualize and alter the layout of the workplace in a collaborative way. The VWAM is the component of WOS that provides the engineer all the necessary tools to assess the ergonomics of the layout, towards the improvement of the working conditions of the operators.
Open Access
Damage detection using phased array ultrasound and data analysis using artificial intelligence
Γιοβάνης, Γεώργιος; Giovanis, Geprgios
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
The present thesis aims to damage detection of composite aeronautical materials with the phased array ultrasonic testing (PAUT) and data analysis artificial intelligence (AI). First, there is an introduction to the source of the project's inspiration and its scientific significance. Secondly, there is a reference on composite materials, how they are used in aeronautical applications, and the damage mechanisms that take place. At that point, after briefly examining some standard non-destructive testing (NDT) methods, the selected technique and its capabilities are analyzed in depth, both at a theoretical and a scientific level. Next, the concept of Artificial Intelligence, Machine and Deep Learning is introduced, and the methods of "Supervised Learning" and "Unsupervised Learning" are analyzed in detail. More specifically, the clustering algorithm K-means and the Deep Neural Networks (DNNs) with emphasis on the convolutional neural networks (CNNs) are examined. Then, after a literature review of the scientific efforts made concerning the analysis of ultrasound data using artificial intelligence, the description of the project is presented, followed by the forming of the research questions and the methodology that was applied in the following research. Subsequently, the experimental setup (materials, equipment), the way of capturing the data, and the pre-processing and post-processing methods are described. Next, the unsupervised machine learning method used for binary classification for two different composites is analyzed, and the respective results are presented. Furthermore, the supervised deep learning method used for binary classification is analyzed, and the relevant results are presented. Finally, the conclusions that emerge from this analysis are presented, and suggestions for improvement and expansion of the work in the future are reported.
Open Access
Design and analysis of a 3U cubesat for material exposure experiments in space
(2023-06-26) Ντίνος, Ανδρέας; Dinos, Andreas
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
Space is an unfriendly environment, where the materials of the space shuttle mission are going to face significant challenging conditions. As a result, a big part of the ongoing research concerns the materials’ degradation under space conditions. However, the ideal way to achieve this is by sending an actual space shuttle into space, exposing the materials to the environment, and inspecting them in a non-destructive way. Still, the economic aspect is an obstacle for universities to apply their knowledge in this particular research field. On the other side, the participation of universities in aerospace has grown dramatically over the last few years, with more and more technical departments designing and sending nanosatellites into space. The aim of the present thesis is the design a 3U CubeSat, which will include a deployable part for conducting a material exposure experiment in space, recording the samples through a camera, and transferring the results to Earth for further analysis. After completing the necessary bibliographic review, in the first phase all the components of the satellite designed using CATIA are presented in detail. Then and after the required specifications related to the geometry of the satellite have been checked, the deployment part of the structure is kinematically analyzed in the ADAMS software, where three different torsional springs are compared to be used for the purposes of the experiment. Finally, the designed model is imported into ANSYS finite element software in order to perform the necessary analyses and verify its structural integrity, examining its behavior under quasi-static and dynamic loadings
Open Access
Design and analysis of a deployable antenna for nanosatellites
(2023-03-01) Μπέκας, Ιωάννης-Θεόδωρος; Bekas, Ioannis-Theodoros
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
One of the leading trends of the past decades regarding satellites has been the ever more reduction in their size, leading to the creation of a class characterized by their small dimensions and light weight, nanosatellites. These have found use cases in a plethora of applications, many of which require high gain telecommunication antennas. The present diploma thesis aims to design and analyze such an antenna suited for use in nanosatellites. At first, a literature overview is conducted, in which deployable antenna concepts designed for satellites and their applicability for nanosatellites are presented. Moving forward, an antenna model is designed in Catia environment and kinematically analyzed in MSC Adams environment. With the results and knowledge gained from this model and with the target of greater reduction in stowed volume a second antenna model is created. After its design, kinematic and structural (modal and quasi-static) analyses are performed, with the structural analysis being performed in a FEA software environment. Concluding, the two models are compared and the effect of the design choices on the results of their analyses are outlined. Summarizing, based on the conclusions drawn from the models, changes and optimizations to the design of the antenna as well as the parametrization and execution of the analyses are proposed. Thus, the next steps for the continuation of this work are put forward.
Open Access
Design and analysis of deployable antenna for satellites
Σταματόπουλος, Φίλιππος; Stamatopoulos, Filippos
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
This thesis examines the viability of an umbrella like deployable antenna with an aperture of 1m, stowed in a volume of 1.5U or 2U. Some antenna basic theory is referenced and a literature survey on antenna deploying mechanisms is performed. Following this, a concept is proposed, based on an existing design with an enhancement to meet the desired deployed aperture. To assess the stowage of the reflector in a given volume, simple tools are developed which are utilized during the design process. All the components from the reference design are considered of constant dimensions expect for the ribs which are the focus of the thesis. Through adjustments, the final three-stage rib design is reached. Having completed the design, the deployment sequence is described. A single-rib kinematic analysis is performed to extract the ideal motion profiles of the deployment. A single-rib dynamic analysis is performed to be used as a reference to multi-rib analysis and to examine the real motion and the forces produced. Multi-rib dynamic analyses are performed to simulate the deployment of the mechanism from a holistic perspective. Finally, a simple analysis of the antenna’s electromagnetic pattern is performed. Summarizing, it seems possible to deploy a 1m aperture antenna from a 2U stowage volume. The proposed mechanism can be studied in further detail and more analysis should be done, but the overall concept is viable.
Open Access
Design and development of product – service system based on augmented reality for machine maintenance
Αγγελόπουλος, Ιωάννης; Angelopoulos, Ioannis
Τμήμα Μηχανολόγων και Αεροναυπηγών Μηχανικ. (ΔΕ)
Maintenance is a demanding set of tasks performed on industrial equipment. Technicians are obliged to carry printed manuals, in order to perform all the necessary corrective actions. Moreover, there are cases where the technician needs to communicate with an engineer or even call specialized personnel in order to accomplish complex maintenance tasks. Technological advances in mobile technologies and mixed reality, have enabled engineers over the world to produce useful applications for providing all the needed instructions and communication with the use of any smart device, just by registering 3D content on the real environment. Augmented Reality is a cutting-edge digital technology facilitating engineers in reducing the cognitive load of technicians. Further to that, with the utilization of AR the threshold for minimum special skills and/or training is also reduced. Taking into consideration the advances in Information and Communication Technologies (ICT), expert engineers are capable of establishing sufficient communication channels with the shopfloor technicians, in an attempt to provide to provide remote guidance. As a result, the total maintenance time of industrial equipment can be further reduced and simplified. The purpose of this research work is the design and development of an AR based remote and smart maintenance framework for Engineered to Order (ETO) equipment. The framework can be realized as a mobile AR application, which is provided to the clients from the OEM as a Service, thus creating a Product Service System (PSS).