Zugl.: Magdeburg, Univ., Diss., 2009

Zugl.: Magdeburg, Universiẗat, Diss., 2009.

Actualmente existe un gran número de técnicas y herramientas para realizar estimaciones en los procesos de software, pero muchas de ellas requieren de gran volumen de información del proyecto que se está analizando, dificultando una estimación temprana del esfuerzo requerido para desarrollar dicho proyecto. Aquellos analistas que trabajan con el Léxico Extendido del Lenguaje, al contar con este modelo en etapas tempranas del software, pueden inferir ciertas características del proyecto, como pueden ser los Casos de Uso, las clases y entidades de base de datos que formaran parte del diseño del proyecto. Por otro lado, existen técnicas de estimación de esfuerzo ampliamente utilizadas y estandarizadas que se valen de estas características, como por ejemplo Puntos Caso de Uso, pero que en una etapa temprana de elicitación de requerimientos no son aplicables por falta de información. Este trabajo pretende brindar a los usuarios que utilizan Léxico Extendido del Lenguaje en su proceso de elicitación de requerimientos, una herramienta que, a partir de la información recabada en las etapas tempranas de dicho proceso, proporcione una estimación del esfuerzo necesario para realizar el proyecto, basada en un método ampliamente utilizado y estandarizado.

Applications involving high electrical conduction require complex components that are difficult to be manufactured by conventional processes. Laser sintering (LS) is an additive manufacturing technique that overcomes these drawbacks by offering design flexibility. This study focuses upon optimizing the process of laser sintering to manufacture functional prototypes of components used in high electrical conduction applications. Specifically, components for two systems – high current sliding electrical contacts and fuel cells – were designed, manufactured and tested. C-asperity rails were made by LS and tested in a high current sliding electrical setup. Corrugated flow field plates were created by LS and their performance in a direct methanol fuel cell (DMFC) was tested. This is the first experimental attempt at using laser sintering for manufacturing such complex components for use in high electrical conduction applications. The second part of this study involves optimization the laser sintering process. Towards this, efforts were made to improve the green strength of parts made by LS. Particle size of graphite/ phenolic resin and addition of nylon/11 and wax were tested for their effect upon green strength. Of these, significant improvement of green strength was observed by altering the particle size of the graphite/ phenolic resin system. New methods of improving green strength by employing fast cure phenolic resins with carbon fiber additions were successfully demonstrated. This study also identified a binder system and process parameters for indirect LS of stainless steel –for bipolar plate compression/ injection mold tooling. All the experimental results of this study lead us to believe that laser sintering can be developed as a robust and efficient process for the manufacture of specialized components used in advanced electrical conduction systems. / text

published_or_final_version / Industrial and Manufacturing Systems Engineering / Master / Master of Philosophy

Requirements engineers have to elicit, document, and validate how stakeholders act and interact to achieve their common goals in collaborative scenarios. Only after gathering all information concerning who interacts with whom to do what and why, can a software system be designed and realized which supports the stakeholders to do their work. To capture and structure requirements of different (groups of) stakeholders, scenario-based approaches have been widely used and investigated. Still, the elicitation and validation of requirements covering collaborative scenarios remains complicated, since the required information is highly intertwined, fragmented, and distributed over several stakeholders. Hence, it can only be elicited and validated collaboratively. In times of globally distributed companies, scheduling and conducting workshops with groups of stakeholders is usually not feasible due to budget and time constraints. Talking to individual stakeholders, on the other hand, is feasible but leads to fragmented and incomplete stakeholder scenarios. Going back and forth between different individual stakeholders to resolve this fragmentation and explore uncovered alternatives is an error-prone, time-consuming, and expensive task for the requirements engineers. While formal modeling methods can be employed to automatically check and ensure consistency of stakeholder scenarios, such methods introduce additional overhead since their formal notations have to be explained in each interaction between stakeholders and requirements engineers. Tangible prototypes as they are used in other disciplines such as design, on the other hand, allow designers to feasibly validate and iterate concepts and requirements with stakeholders. This thesis proposes a model-based approach for prototyping formal behavioral specifications of stakeholders who are involved in collaborative scenarios. By simulating and animating such specifications in a remote domain-specific visualization, stakeholders can experience and validate the scenarios captured so far, i.e., how other stakeholders act and react. This interactive scenario simulation is referred to as a model-based virtual prototype. Moreover, through observing how stakeholders interact with a virtual prototype of their collaborative scenarios, formal behavioral specifications can be automatically derived which complete the otherwise fragmented scenarios. This, in turn, enables requirements engineers to elicit and validate collaborative scenarios in individual stakeholder sessions – decoupled, since stakeholders can participate remotely and are not forced to be available for a joint session at the same time. This thesis discusses and evaluates the feasibility, understandability, and modifiability of model-based virtual prototypes. Similarly to how physical prototypes are perceived, the presented approach brings behavioral models closer to being tangible for stakeholders and, moreover, combines the advantages of joint stakeholder sessions and decoupled sessions. / Anforderungsingenieure erheben, dokumentieren und validieren wie Bedarfsträger in einzelnen und gemeinsamen Aktivitäten die Ziele ihrer kollaborativen Szenarios erreichen. Auf Grundlage von Angaben darüber, wer warum mit wem zusammen was erledigt, kann anschließend ein Softwaresystem spezifiziert und umgesetzt werden, welches die Bedarfsträger bei der Durchführung ihrer Abläufe unterstützt. Um Anforderungen verschiedener (Gruppen von) Bedarfsträger zu erfassen und zu strukturieren, werden szenariobasierte Ansätze genutzt und erforscht. Die Erhebung und Validierung von Anforderungen, die kollaborative Szenarios abdecken, ist dennoch kompliziert, da derartige Informationen hochgradig verknüpft, fragmentiert und über mehrere Bedarfsträger verteilt sind, wodurch sie nur in Gruppensitzungen effizient erhoben und validiert werden können. In Zeiten global verteilter Firmen ist die Planung und Durchführung solcher Workshops mit Gruppen von Bedarfsträgern nur selten praktikabel. Mit einzelnen Bedarfsträgern zu sprechen ist hingegen oft realisierbar, führt aber zu fragmentierten, unvollständigen Szenariobeschreibungen. Durch eine Vielzahl von Einzelgesprächen mit wechselnden Bedarfsträgern kann diese Fragmentierung aufgelöst werden – dies ist aber eine fehleranfällige und zeitaufwändige Aufgabe. Zwar bieten formale Modellierungsmethoden z.B. automatische Konsistenzchecks für Szenarios, doch führen derartige Methoden zu Mehraufwand in allen Gesprächen mit Bedarfsträgern, da diesen die verwendeten formalen Notationen jedes Mal erläutert werden müssen. Handfeste Prototypen, wie sie in anderen Disziplinen eingesetzt werden, ermöglichen es Designern, ihre Konzepte und erhobenen Anforderungen ohne viel Aufwand mit Bedarfsträgern zu validieren und zu iterieren. In dieser Dissertation wird ein modellbasierter Generierungsansatz vorgeschlagen, der kollaborative Szenarios prototypisch auf Grundlage von formalen Verhaltensmodellen für die beteiligten Bedarfsträger darstellt. Durch die Simulation dieses Verhaltens und dessen Animation innerhalb einer webbasierten, domänenspezifischen Visualisierung, können Bedarfsträger diese Modelle erleben und die bisher erfassten Szenarios validieren. Eine derartige interaktive Szenariosimulation wird als modellbasierter virtueller Prototyp bezeichnet. Basierend auf den Interaktionen zwischen Bedarfsträgern und einem virtuellen Prototypen ihrer Szenarios können zudem formale Verhaltensspezifikationen automatisch abgeleitet werden, die wiederum die fragmentierten kollaborativen Szenarios vervollständigen. Dies ermöglicht es den Anforderungsingenieuren, die kollaborativen Szenarios in individuellen Sitzungen mit einzelnen Bedarfsträgern zu erheben und zu validieren – entkoppelt voneinander, da Bedarfsträger webbasiert teilnehmen können und dabei nicht darauf angewiesen sind, dass andere Bedarfsträger ebenfalls in der gleichen Sitzung teilnehmen. Diese Dissertation diskutiert und evaluiert die Machbarkeit, Verständlichkeit sowie die Änderbarkeit der modellbasierten virtuellen Prototypen. Auf die gleiche Art wie physikalische Prototypen wahrgenommen werden, erlaubt es der vorgestellte Ansatz, Verhaltensmodelle für Bedarfsträger erlebbar zu machen und so die Vorteile von Gruppensitzungen mit denen entkoppelter Sitzungen zu verbinden.

Engineering educators the world over are being faced with the dilemma of combining traditional mathematically intensive courses, like Control Systems and Robotics with advances in computational hardware and software. This is because it is impractical to include both software engineering issues as well as conventional course content. A solution to the problem lies in Rapid Prototyping technology to develop and design software, for application on PC's and embedded systems. Rapid Prototyping, based on automatic code generation, allows users to develop advanced software on high level graphical platforms like Simulink® and LabView®, while " hiding" the underlying layers of complex code. This approach allows the advanced hardware, traditionally reserved for software engineers, to be accessed by a much wider audience and is an ideal educational tool. This thesis presents the complete development of the Rapid Application Development Environment (RADE). The RADE system customises the Mathworks Real Time Workshop (RTW) revision 11 for application on both standalone and networked DS? cards. The functionality of the RTW is incorporated into the RADE system. This affords the user seamless code generation, downloading, on-line parameter tuning and on-line data visualisation with storage capability. An added advantage of the RADE system is its easy portability to multiple target platforms, which is demonstrated by its implementation on two different DSP cards. Finally the functionality of the RADE system is demonstrated as an educational tool, with the demonstration of a DC motor speed and position controller. / Thesis (M.Sc.Eng.-University of Natal, Durban, 2001.

Aero engines are made up of a large number of blades which are subject to wear and damage. They are expensive and must be repaired wherever possible. Engines also have small components which are required in small numbers that need to be developed rapidly. The first part of this research work focuses on developing a robust automated blade repair method using robotic welding. Optimal weld parameters were developed for build-up of edges having different thicknesses. Samples with varying Current and varying travel speed were produced and their micro hardness values were compared. Blade profiles were welded upon. The second part involves a methodology for producing small components using rapid prototyping (RP) techniques. This part involves use of 3D robotic welding for layered manufacturing. Tensile samples produced using the metal RP method were tested and results were found to be well above the minimum cast specifications for the given material.

No description available.

Sheet metal forming has found increasing applications in modern industries. To eliminate use of expensive tools during product development, thermal forming, a rapid prototyping process that is flexible enough to decrease costs has been developed. Thermal forming processes use a heat source to perform the required deformation mainly by creating a thermal difference along the thickness of the sheet. Gas flames, lasers and plasma heat sources have been used for sheet metal bending by thermal forming. An alternative to laser and gas flames, plasma-jet forming has been developed that uses a non-transferred plasma arc as a heat source. The plasma-jet forming system uses a highly controllable non-transferred plasma torch as a heat source to create the necessary thermal gradient in the sheet metal that causes the required plastic deformation. Various experiments to produce simple linear bends and other complex shapes have been conducted by using different scanning options and coupling techniques. A computer simulated model using finite element method is being developed to study key parameters affecting this process and also to measure the thermal transient temperature distribution during the process. A predictive model to relate the deformation to the temperature gradient for various materials is being developed. Simulation results that are in accordance to experimental observations will further improve this material forming process to be highly controllable and more accurate