Submitted by isabela.moljf@hotmail.com (isabela.moljf@hotmail.com) on 2017-02-13T14:23:53Z No. of bitstreams: 1 evelynaparecidadeoliveira.pdf: 4026204 bytes, checksum: c1b720f3501ffcd12b38f9d4c877dfa9 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-02-13T16:49:50Z (GMT) No. of bitstreams: 1 evelynaparecidadeoliveira.pdf: 4026204 bytes, checksum: c1b720f3501ffcd12b38f9d4c877dfa9 (MD5) / Made available in DSpace on 2017-02-13T16:49:50Z (GMT). No. of bitstreams: 1 evelynaparecidadeoliveira.pdf: 4026204 bytes, checksum: c1b720f3501ffcd12b38f9d4c877dfa9 (MD5) Previous issue date: 2016-04-01 / O desgaste do dente, que se manifesta com uma grande variedade de graus ou níveis, é uma das anomalias dentárias comumente encontradas em diversas populações. A modelagem computacional do problema de contato oclusal pode ajudar na compreensão das interações entre os dentes, que gera concentrações de tensões. A abordagem utilizada para simular o contato entre as superfícies ásperas, dadas as funções de densidade de probabilidade das alturas das asperezas, consiste em discretizá-las em vários intervalos, de modo que cada um deles represente uma aspereza principal. As deformações das asperezas principais são analisadas e, utilizando técnicas de homogeneização, é possível estabelecer a relação entre as respostas que ocorrem em microescala e as respostas esperadas na macroescala. Neste trabalho, foi criada uma rotina para geração numérica de superfícies, um script acoplado ao programa de elementos finitos, Abaqus, para análise dos microcontatos, e, finalmente, um código para o procedimento de homogeneização e análise paramétrica das medidas de rugosidade. Na estimativa da dureza superficial, verificou-se que a curvatura média das asperezas exerce maior influência. Os parâmetros que mais influenciam nas forças e áreas de contato são o coeficiente de achatamento (kurtosis), cujo aumento leva ao decaimento dessas grandezas, seguido da rugosidade média quadrática, que, por sua vez, tem uma relação diretamente proporcional com a carga e consequentemente com a área de contato. / Tooth wear, which manifests with a great variety of degrees or levels, is one of the dental abnormalities commonly found in different populations. The computational modelling of the occlusal contact problem can help the comprehension of any interaction between teeth generating stress concentration. The approach used in this work to simulate contact between rough surfaces, given the probability density functions of asperities, consists in discretizing them in several intervals, so that each one represents a main asperity. The deformations of the main asperities are analyzed and, using homogenization techniques, it is possible to developed the relationship among the responses occurred in micro-scale and the predicted responses in macro-scale. In this work a routine has been established for numerical generation of surfaces, a script as well as has been coupled to the finite element program Abaqus, for analysis of the micro-contacts, and, finally, a code for homogenization and parametric analysis of the roughness measure. In the estimate of the surface hardness it was found that the mean curvature of the asperities exerts the largest influence. The parameters that influence the forces and contact areas are the flattening coefficient (kurtosis), which when increases leads to decay of these greatness, followed by the mean square roughness, which, in turn, has a direct proportional relationship with the strength and consequently with the contact area.

This master’s thesis deals with the development of materials for classwork using EPLAN software tools for "XEPL" course. The introductory part describes issues with design in electrical engineering, with an emphasis on the formal aspects of project documentation standards and software tools for design support. Further in the thesis, there is an introduction to the EPLAN software environment. The EPLAN environment modules are presented one by one and the EPLAN Education software is introduced. The next part of this thesis involves a model project in designing electric installation of car washer in the EPLAN Education. The last part of this thesis includes a set of exercises based on this model project, where basic functions of the EPLAN software and their uses are presented.

This thesis report aims to help the client developing their new product. The new product to be developed is a safety bracket for a safety system. The safety bracket connects different parts which create the safety system and it should be able to withstand impacts from moving objects. The client has a set of requirements that needs to be addressed during the product development process. One of the most important requirements that must be fulfilled is the given impact energy that the safety bracket must withstand. The methodology used during this thesis work is the product development processes (PDP). The product development process is used to find concepts that have the potential to answer the research questions and to fulfil the requirements. Some methods used in the product development process are brainstorming, brainwriting and combining working principles. The concepts were evaluated with a combination of Pugh´s matrix and weighting matrix. The three best concepts were selected for further development and tested with FEA simulation with Abaqus CAE. The impact simulation gave indications if the concepts could handle the impact energy and if they could fulfil the requirements. All three concepts could withstand the impact energy based on the simulations and most of the requirements could be fulfilled. The concepts with thinner profile walls had a reduction in stress and an increase in impact duration, where the kinetic energy is distributed throughout the impact. A protective shell helps with the reduction of stress and the energy absorption during the impact simulation.

Smart wearables, known as emerging paradigms for vital big data capturing, have been attracting intensive attentions. However, one crucial problem is their power-hungriness, i.e., the continuous data streaming consumes energy dramatically and requires devices to be frequently charged. Targeting this obstacle, we propose to investigate the biodynamic patterns in the data and design a data-driven approach for intelligent data compression. We leverage Deep Learning (DL), more specifically, Convolutional Autoencoder (CAE), to learn a sparse representation of the vital big data. The minimized energy need, even taking into consideration the CAE-induced overhead, is tremendously lower than the original energy need. Further, compared with state-of-the-art wavelet compression-based method, our method can compress the data with a dramatically lower error for a similar energy budget. Our experiments and the validated approach are expected to boost the energy efficiency of wearables, and thus greatly advance ubiquitous big data applications in era of smart health.<br><div>In recent years, there has also been a growing interest in edge intelligence for emerging instantaneous big data inference. However, the inference algorithms, especially deep learning, usually require heavy computation requirements, thereby greatly limiting their deployment on the edge. We take special interest in the smart health wearable big data mining and inference. <br></div><div><br></div><div>Targeting the deep learning’s high computational complexity and large memory and energy requirements, new approaches are urged to make the deep learning algorithms ultra-efficient for wearable big data analysis. We propose to leverage knowledge distillation to achieve an ultra-efficient edge-deployable deep learning model. More specifically, through transferring the knowledge from a teacher model to the on-edge student model, the soft target distribution of the teacher model can be effectively learned by the student model. Besides, we propose to further introduce adversarial robustness to the student model, by stimulating the student model to correctly identify inputs that have adversarial perturbation. Experiments demonstrate that the knowledge distillation student model has comparable performance to the heavy teacher model but owns a substantially smaller model size. With adversarial learning, the student model has effectively preserved its robustness. In such a way, we have demonstrated the framework with knowledge distillation and adversarial learning can, not only advance ultra-efficient edge inference, but also preserve the robustness facing the perturbed input.</div>

The objective of this project is to develop a driving test using a Digital Human Modeling tool (DHM), specifically IPS IMMA, which will allow the evaluation of the ergonomics of the interior of vehicles as currently demanded by the automotive companies. Thus, improving both the design and the design process. This will involve a study of the driving and the tasks carried out by a real person to end up programming them in the DHM software. Based on this study an interface is suggested that guides engineers or ergonomists to design their own driving tests and enable them to evaluate their own designs without a high specialization in DHM tools and software. Taking into account the already present autonomous cars and their future development, the conceptual design of a two positions steering wheel (autonomous/manual driving) will be introduced as an example to be added in the driving test. This example is intended to show how DHM tools can be used to evaluate different designs solutions in early stages of the product development process. This project will be a contribution to one of the sections of the ADOPTIVE project carried out at the University of Skövde and in collaboration with Swedish automotive companies.

Der hier vorliegende Vortrag beschreibt erste Untersuchungsergebnisse mit dem CAE-Programm "Creo Simulate" im Vergleich zum Programm "Ansys Workbench" zur Berechnung der Spannungsformzahlen sowie der örtlichen Kerbspannungen an gekerbten Konstruktionselementen. Der Vortrag beginnt mit einer Einleitung zur Geschichte der Kerbwirkungsforschung. Es werden Schadensfälle dargestellt und anschließend die Grundlagen zur Berechnung der Form- und Kerbwirkungszahlen kurz beschrieben. Es wird fortgesetzt mit der Durchführung von Konvergenz-Untersuchungen an Vollwellen mit SR-Nuten nach DIN 471 sowie an Vollwellen mit Absatz nach DIN 743 bei Zug- und Torsionsbelastung. Die Konvergenz-Berechnungsergebnisse wurden zusammengestellt und kurz kommentiert. Weiterhin wurden FE-Berechnungen zur Bestimmung der Spannungsformzahlen an Hohlwellen mit Absatz sowie an Vollwellen mit einer Kerbüberlagerung "Absatz und Querbohrung" bei Zug-, Biege, und Torsionsbelastung durchgeführt. Es wurden entsprechende Formzahldiagramme und Formzahlwerte sowie Spannungsverteilungsbilder je nach Belastungsart und Vergleichsspannungshypothese (GEH bzw. NSH) angegeben. Eine Möglichkeit zur Herabsetzung der Kerbwirkung an der kritischen Kerbstelle besteht in der absichtlichen Anwendung von Zusatzkerben, welche der Hauptkerbe benachbart sind. Derartige Entlastungskerben können zwar eine bedeutende Spannungsverminderung an der gefährdeten Stelle ergeben, an den Entlastungskerben entstehen jedoch neue Spannungsspitzen, die sich u. U. ungünstig auswirken. Zur Milderung von Kerbwirkungen durch konstruktive Entlastungskerben wurden exemplarische FEM-Untersuchungen an Vollwellen mit Absatz und an Kerbzahnwellen nach DIN 5481 bei Zug-, Biege- und Torsionsbelastung ausgeführt. Die Einflüsse der rechteckigen Form der Entlastungsnut auf die Formzahlen je Belastungsart wurden präsentiert und kurz diskutiert. Ebenfalls wurden auch die Kerbwirkungen an Evolventen-Zahnwellenverbindungen nach DIN 5480 mit freiem Auslauf und mit gebundenem Auslauf sowie mit Sicherungsringnuten nach DIN 471 untersucht, mit dem Ziel, den Ort der Spannungsmaxima und deren Verteilung, die Höhe der Kerbspannungen und den Einfluss der Sicherungsringnut auf die Spannungsüberhöhungen in den kritischen Bereichen zu erfassen. Es wurden zusätzlich weitere FE-Untersuchungen zu den Spannungsüberhöhungen an komplexen Gussbauteilen am Beispiel eines Planetenträgers für Planetengetriebe im Einatzbereich der Windkraftanlagen realisiert. Die Festlegung der Randbedingungen wurde kurz präsentiert und die sich daraus ergebenden Ergebnisse dargestellt. In diesem Vortrag wurde auch die Entwicklung eines neuen didaktischen Konzepts für die Konstruktionsausbildung zur Verbesserung der Präsentationskompetenz und Teamfähigkeit der Studierenden kurz beschrieben und über erste Erfahrungen aus der Umsetzung in die Lehrveranstaltung "CAD/CAE" berichtet. Die Studierenden erarbeiten in nach der Rundlitzenseilmethode strukturierten Gruppen unter Berücksichtigung der heterogenen Umgebung numerische Lösungen zu Variantenrechnungen von Aufgaben. Die Studierenden präsentieren ihre Ergebnisse in Form von 100-Sekunden-Vorträgen. Bei der Entwicklung dieser Methode lässt man sich durch den Karatesport inspirieren. Es wurden verschiedene Kriterien zur Bewertung der Micro-Präsentationen festgelegt und angewandt. Erste Erfahrungen mit der Umsetzung dieser Methodenkombination sind erfolgversprechend. Eine detaillierte statistisch-psychologische Evaluation dieses didaktischen Konzepts ist Ziel weiterführender Untersuchungen.

During the conceptual design stage several concepts are generated, and a few are selected for detailed analyses. CAD models from conceptual design often follow the "over-the-wall" approach for downstream analyses such as FEA, CFD, heat transfer, and vibrations. A CAD-centric approach will be applied to the CAD-to-CFD process to help industry in an ongoing quest to shorten the design cycle time. The CAD-centric approach consists of using the CAD model as a source of data for downstream applications such as mesh generation, and CFD setup. The CAD model used in the CAD-centric approach contains the geometry to be analyzed and non-geometric data required to solve the CFD problem in the form of attributes. Attributes can be associated to entities of the geometry such as the faces, edges, and volumes. Any operations changing geometry require the CAD-centric model be reworked. One class of topology alterations is the discrete feature problem that is encountered when an array of features change in number. A method is proposed, developed and reported on that adapts the CAD-centric approach to account for discrete feature changes that occur during preliminary design.

Design decisions have a large impact early in the design process. Optimization methods can help engineers improve their early decision making, however, design problems are often ill-posed for optimization at this early stage. This thesis develops engineering methods to use optimization during embodiment design, despite these difficulties. One common difficulty in designing mechanical systems is in handling the effects that design changes in one subsystem have on another. This is made more difficult in early engineering design, when design information is preliminary. Increased efforts have been made to use numerical optimization methods in early engineering design – because of the large impact early decisions have on subsequent development activities. One step toward executing meaningful optimizations in early design is the development of an optimization framework to be used when conditions are expected to change as the design progresses and new information is gained. This thesis presents a design framework that considers such change by subjecting the parametric updating of CAD models to optimization criteria specific to the problem at hand. Under the proposed framework, a part or subassembly is parametrically modeled in CAD; when changes are made to the subsystems that interact with the part or subassembly, it is then updated subject to design objectives and constraints. In this way, the updated part or subassembly satisfies system and subsystem level optimization criteria, reducing the need for the designer to react to design changes manually. It is used to reduce the weight of a Formula SAE suspension rocker by 18%, demonstrating the utility of this framework. Next, we develop methods to help engineers by giving them options and helping them explore during configuration generation. The design of multiple-bend, progressive-die-formed springs typically comprises four steps: (i) functional specification, (ii) configuration generation, (iii) configuration selection, and (iv) detailed shape and size optimization. Configuration generation fundamentally affects the success or failure of the design effort. This presents an important problem: by not generating potentially optimal configurations for further development in detailed design, the designer may unknowingly set the design on track for sub-optimal performance. In response, a method is developed that improves configuration generation. Specifically, an optimization-based spring configuration generator – without which, the generation would typically be based solely on designer creativity, experience, and knowledge. The proposed approach allows the designer to explore numerous optimization-generated spring configurations, which feasibly satisfy the functional specifications. The feasibility study is carried out before a final configuration is chosen for detailed development. Thus streamlining the designer's efforts to develop a design that avoids sub-optimality. We use the feasibleconfiguration generator to identify twenty-two electrical contact spring configurations. All twenty-two of the configurations satisfy the design's functional specifications. Two important concepts that improve decision making in early design were chosen. First, is the concept of a paremetric CAD based framework. Second is the concept of generating iso-performing design solutions. A numerical computer-based application is explained that takes advantage of these two ideas. A genetic algorithm topology optimization framework with the ability to converge to iso-performing solutions was integrated with CATIA V5. This application is demonstrated on a Formula SAE frame where it develops a pareto frontier of designs, expands upon one compromise design by producing iso-performing solutions, and automatically produces designs with the same performance after a parametric suspension change.

Computer Aided Design (CAD) models often need to be processed due to data translation issues and requirements of the downstream applications like computational field simulation, rapid rototyping, computer graphics,computational manufacturing, and real-time rendering before they can be used. Automatic CAD model processing tools can significantly reduce the amount of time and cost associated with the manual processing.In this dissertaion, automated topology generation and feature removal techniques are developed to prepare suitable models with mimunum user interaction. A topology generation algorithm, commonly known as CAD repairing/healing, is presented to detect commonly found geometrical and topological issues like cracks, gaps, overlaps, intersections, T-connections, and no/invalid topology in the model, process them and build correct topological information. The present algorithm is based on the iterative vertex pair contraction and expansion operations called stitching and filling respectively. The algorithm closes small gaps/overlaps via the stitching operation and fills larger gaps by adding faces through the filling operation to process the model accurately. Processed models are guaranteed to be free of intersecting faces or surfaces. Moreover, the topology generation algorithm can process manifold as well as non-manifold models, which makes the procedure more general and flexible. This algorithm uses an automatic and adaptive distance threshold that enhances reliability of the process and preserves small features in the model. In addition, a spatial data structure, the octree, is used for searching and neighbor finding to process large models efficiently. In this way, the combination of generality, accuracy, reliability, and efficiency of this algorithm seems to be a significant improvement over existing techniques. Results are presented showing the effectiveness of the algorithm to process two- and three-dimensional configurations. Feature detection and removal and feature collapse algorithms are presented to detect and remove small features from CAD models automatically. The feature detection and removal algorithm uses a feature size measure based on the surface area and perimeter to detect small features accurately and remove them from the model. Small feature removal may create holes in the model that are post-processed using the stitching and/or filling operations of the topology generation algorithm. The feature collapse algorithm is based on the iterative vertex pair contraction operation, which is a generalization of an edge-collapse operation, to collapse small features. Unlike previous efforts that use edge-collapse as a dimension reduction operator, the feature collapse algorithm can pair up any arbitrary vertices and perform iterative vertex pair contraction to collapse small features as well as glue unconnected regions. Results showing the automatic detection and removal of most commonly found small features like small edges/faces, fillets, chamfers, nuts, and bolts from real mechanical parts are presented.

El confort del calzado es uno de los aspectos más importantes para los fabricantes y diseñadores de calzado ya que está directamente relacionado con el éxito en la venta del producto y con la salud del usuario. A pesar de ello, y por diversas causas, el análisis del confort que es capaz de proveer un diseño de calzado sigue siendo una asignatura pendiente del desarrollo del producto de las empresas, si es que se lleva a cabo. La causa principal es la ausencia de una herramienta capaz de llevar a cado un test de confort de manera fiable, rápida y abordable en materia de conocimiento y costes para las empresas. El objetivo principal de esta tesis es el desarrollo de una aplicación informática, fundamentada en modelos matemáticos asistidos por ordenador alimentados mediante datos reales, capaz de simular ensayos funcionales en diseños virtuales de calzado, y relacionar dichos resultados con la percepción del confort de los usuarios. Dicha herramienta permite detectar y eliminar errores en el diseño del calzado (que influyen en el confort) desde fases tempranas de desarrollo del producto, aumentando así el confort percibido por los usuarios del calzado, así como la eficacia del proceso de diseño. El trabajo ha tenido lugar en tres grandes bloques. En primer lugar se estudió la relación entre la percepción de las propiedades funcionales asociadas a la interacción usuario-calzado y el confort global percibido. Para ello se llevaron a cabo ensayos con maquinaria para caracterizar las propiedades del calzado y ensayos con usuarios para caracterizar su percepción a partir del uso del calzado. Este análisis permitió obtener un modelo del confort global a partir de las propiedades funcionales, además de identificar las propiedades funcionales relevantes en el confort. A continación se desarrollaron y validaron modelos virtuales de los ensayos de cuantificación de las propiedades funcionales capaces de evaluar el comportamiento de un calzado a partir de su diseño CAD. / Olaso Melis, J. (2010). Predicción del confort a partir de prototipos virtuales de calzado y modelos de ingeniería asistida por ordenador [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/7730