Atualmente, a grande concorrência de mercado faz com que as empresas busquem maior confiabilidade e qualidade em seus produtos, aumentando assim a confiança e fidelidade de seus clientes. Além disso a diminuição dos gastos com garantia e a possibilidade de extensão da mesma é um grande diferencial de mercado, e pode ser obtido com a realização de estudos de confiabilidade dos sistemas e equipamentos. Este trabalho propõe um método para esse estudo e apresenta um caso exemplo, aplicando o método sugerido, e, propondo melhorias. O presente trabalho reúne conceitos básicos de confiabilidade, apresentando as ferramentas mais utilizadas e explana sobre análise de dados de retorno em garantia. São propostas também algumas soluções para vencer as dificuldades possíveis de serem encontradas. O método consiste em adequar os dados obtidos, de um banco de dados qualquer, para se realizar uma análise de confiabilidade partindo do conhecimento do produto através de materiais apropriados e da execução da árvore funcional, seguida da Análise de Modos e Efeitos de Falha, e, realizando a análise dos dados de garantia, classificando o tipo de registro, como suspenso, completo ou agrupado, e atribuindo a aproximação dos dados a uma distribuição estatística adequada para se calcular os parâmetros e obter a confiabilidade do equipamento ou sistema. A partir dessa análise dos dados é possível se obter relações de sazonalidade entre falhas, entre falhas e lotes de produção, entre tempos até a falha, além de estimativas de confiabilidade e taxa de falha, entre outras. Além disso, é possível se realizar a validação da análise de modos e efeitos de falha a partir dos dados de retorno em garantia. A aplicação do método é realizada através de um caso exemplo, central telefônica. O estudo segue o modelo proposto e, de posse da estimativa de confiabilidade, é possível se calcular os custos de garantia e realizar melhorias, principalmente no banco de dados de retorno em garantia, para que a análise seja cada vez mais precisa. / Nowadays, the great market competition makes that the companies look for higher reliability and quality in their products, the customers' fidelity trust. Besides the decrease of the expenses with warranty makes possible its extension which is a great market differential this objective can be accomplished with product reliability – based analysis. The objective of this work is to propose a method for that study and to present an example of analysis, proposing improvements and validation the suggested method. The present work gathers basic reliability concepts, presenting the most used tools for data associated with warranty period. The method consists on adapting the failure data, of any database, to develop a reliability analysis, following basic steps such as product analysis based on functional trees, followed by the Failure Mode, Effects and Analysis, to define the critical components and, to accomplish the analysis of the warranty data, the failure data are classified as suspended complete or contained, and an appropriate statistical distribution is adjusted to evaluate the reliability of the equipment or system. Starting from that data analysis it is possible to obtain seasonal relationships among faults, between faults and production lots, among times to failures. Besides, it is possible to validate the Failure Mode, Effects and Analysis based on warranty registers. The application of the method is demonstrated through an example, involving the reliability analysis of an analogic PABX of. The study follows the proposed model and, based on reliability estimative, it is possible to calculate the warranty costs and to propose possible product and data collection improvements, mainly in the warranty database in order to improve the reliability analysis results.

The quality of a product is an important factorto establish and preserve a sustainable market- and business position in theglobal business environment. Measuring techniques and quality management has astrong correlation. To ensure the functionality of a product, there are needsto measure the product towards specific limitations. Mathematical modeling is away of visualizing a process capability, and its parameters. Sandvik Coromantin Gimo are producing different types of face milling cutters. The company hasa hypothesis about one of its carbide inserts, and its geometrical specificationlimits. The hypothesis is that different geometrical deviations affect themilling cuts edges in different ways, when it is applied and installed in themilling cutter body. The purpose is to develop a mathematical model, which canshow what changes affects the most, which will lead to less waste in theproduction process, by predicting the actual position of the cut in the millingcutter body. The method is based on relevant literature in the main areas ofmeasuring methods and its techniques and statistics. This theory, incombination with the method, is a profound base in the development of themathematical model, and analysis of the cuts. The geometrical deviations weremeasured by using three different measuring methods. This had to be done in orderto verify the mathematical model and its theoretical output, and then compareit to the actual measured data. The result of the study was the development ofa mathematical model, and the analysis of 4000 individual measurements. Theconclusion of the analysis is that certain geometrical deviations are affectingthe position of the milling cuts edges more than others. By using themathematical model, the simulation output will show a theoretical value of themilling cuts geometry, and make it possible to predict the milling cutsdimensional deviation. The mathematical model is validated, based on three mainfactors. These are Gauge Repeatability and Reproducibility, statisticalevaluation, and the development of the mathematical model itself. There are manyparameters affecting the milling cuts geometry, and the mathematical modelshould not be used as the only source of verifying the cut, but rather as acomplementary tool that makes a qualified prediction.

Este trabalho tem por objetivo apresentar metodologia e modelo de simulação para comparação de arranjos de subestações de distribuição através de análise de confiabilidade estática de esquemas de manobra. A comparação entre alternativas de arranjos de subestações, sob o ponto de vista de confiabilidade, será feita de duas formas, a saber: - Comparação direta dos índices de confiabilidade obtidos pela modelagem dos arranjos de subestações, com a utilização de um eficiente algoritmo que se baseia na determinação dos cortes mínimos de uma rede genérica; - Avaliação técnico-econômica dos arranjos de subestações, através de análise com enfoque de múltiplos objetivos, considerando-se os objetivos de minimização dos custos (investimento, operação e manutenção) e de maximização da confiabilidade, através do custo da energia não distribuída. Esta análise é feita para um período de estudo, considerando o crescimento da carga a ser atendida. Os arranjos de subestação podem ser fixos durante todo o período de estudo, ou podem evoluir ao longo do tempo, por exemplo, considerando um primeiro estágio, com uma unidade transformadora, e um segundo estágio, com duas unidades. / This dissertation aims at presenting a methodology and simulation model to compare different distribution substation configurations through a static reliability evaluation model applicable to different switching schemes. The comparison of the substation configuration alternatives, regarding reliability analysis, is carried out by the two following ways: - Straight-forward comparison of reliability indices obtained by modeling the substation configuration alternatives through an efficient algorithm based on the determination of minimum cut sets applied to a generic network; - Technical and economical evaluation of the substation configuration alternatives, through a multiple objective framework, by considering the minimization of costs (investment, operation and maintenance) e maximization of a reliability index, which is the costs due to the non-supplied energy. This analysis is carried out for a study period, considering the corresponding load growth. Substation configurations can be either fixed for the entire study period or can evolve along time. For instance, for the first stage a single transformer unit is used and, for the second stage, the configuration comprises two transformer units.

Denna rapport behandlar hur ett införskaffande av en ny rullformningsmaskin skulle kunna påverka produktionen på ArcelorMittal Construction Sverige AB i Karlstad. Detta är ett examensarbete för högskoleingenjör i maskinteknik vid fakulteten för hälsa, natur- och teknikvetenskap på Karlstads universitet under våren år 2017. Planerna i dagsläget är att ersätta tre av de nuvarande maskinerna med en ny effektivare maskin som klarar av att tillverka de valda maskinernas olika profiler. Den nya maskinen kan mycket snabbt byta profil från en till en annan. Det tar ca en minut att genomföra ett verktygsbyte med den nya maskinen jämfört med ca 90 minuter i dagsläget för en av de gamla maskinerna. För att med större säkerhet få reda på nuvarande ställtider genomfördes SMED som är en metod för att reducera ställtider. Arbetet med att reducera ställtiderna visar att med en bättre planering innan stället så kan ställtiden reduceras från 90 minuter till 83 minuter. Det stora problemet som var gemensamt för alla maskinerna i dagsläget var allt kringarbete som måste utföras, speciellt var det den manuella emballeringen som var den riktiga flaskhalsen i produktionen. Emballeringen utförs av samma operatör som kör maskinen och måste därför med jämna mellanrum avbryta körningen för att få tid att emballera produkten så att de kan placeras i färdigvarulagret för utleverans. Med den nya maskinen är det planerat att ha en helautomatisk emballering för att öka den tid som maskinen är produktiv.  De nuvarande maskinerna har tillsammans under 2017 producerat plåt med en total verklig hastighet av 10,84 [m/min]. Den nya maskinen måste därför ha en minsta verkliga hastighet på 10,84 [m/min]. Beräkningar av nuvarande genomsnittliga orderstorlekar visar att den nya maskinen klarar minst av att producera med en verklig hastighet på 15,2 [m/min]. Detta medför att den nya maskinen klarar av att ersätta de nuvarande med en god marginal. Den verkliga beräknade hastigheten för den nya maskinen baserat på att coilbyte och profilbyte utförs mellan varje körning, vilket inte är fallet i dagsläget utan flera ordrar med samma form och råmaterial oftast körs efter varandra. Den nya maskinen medför en förbättring på minst 40 % i verklig hastighet. ArcelorMittal Construction Sverige AB har i dagsläget ett relativt stort råvarulager för att de snabbt ska kunna leverera den produkt som kunden beställer. Produkter går att få i nästan vilken färg som helst och i flera olika tjocklekar vilket medför att många olika produkter finns i råvarulagret. Fokuset i värdeflödesanalysen har därför varit på en av de mer frekventa råvarorna som passerar genom fabriken. För denna produkt fanns för nuvarande ett lager på 900 ton vilket är en buffert på ca 60 dagar. Då råvaruleveranser kommer dagligen så kan man minska råvarulagret till en buffert på istället 20 dagar och därmed minska det bundna kapitalet som är investerat i dessa råvaror.

Este trabalho tem por objetivo apresentar metodologia e modelo de simulação para comparação de arranjos de subestações de distribuição através de análise de confiabilidade estática de esquemas de manobra. A comparação entre alternativas de arranjos de subestações, sob o ponto de vista de confiabilidade, será feita de duas formas, a saber: - Comparação direta dos índices de confiabilidade obtidos pela modelagem dos arranjos de subestações, com a utilização de um eficiente algoritmo que se baseia na determinação dos cortes mínimos de uma rede genérica; - Avaliação técnico-econômica dos arranjos de subestações, através de análise com enfoque de múltiplos objetivos, considerando-se os objetivos de minimização dos custos (investimento, operação e manutenção) e de maximização da confiabilidade, através do custo da energia não distribuída. Esta análise é feita para um período de estudo, considerando o crescimento da carga a ser atendida. Os arranjos de subestação podem ser fixos durante todo o período de estudo, ou podem evoluir ao longo do tempo, por exemplo, considerando um primeiro estágio, com uma unidade transformadora, e um segundo estágio, com duas unidades. / This dissertation aims at presenting a methodology and simulation model to compare different distribution substation configurations through a static reliability evaluation model applicable to different switching schemes. The comparison of the substation configuration alternatives, regarding reliability analysis, is carried out by the two following ways: - Straight-forward comparison of reliability indices obtained by modeling the substation configuration alternatives through an efficient algorithm based on the determination of minimum cut sets applied to a generic network; - Technical and economical evaluation of the substation configuration alternatives, through a multiple objective framework, by considering the minimization of costs (investment, operation and maintenance) e maximization of a reliability index, which is the costs due to the non-supplied energy. This analysis is carried out for a study period, considering the corresponding load growth. Substation configurations can be either fixed for the entire study period or can evolve along time. For instance, for the first stage a single transformer unit is used and, for the second stage, the configuration comprises two transformer units.

An analysis and modeling method of the diagnostic characteristics of a mechanical or electromechanical system is presented. Diagnosability analysis is especially relevant given the complexities and functional interdependencies of modern-day systems, since improvements in diagnosability can lead to a reduction of a system's life-cycle costs. The diagnosis process of a mechanical system, involving an observation phase and a testing phase, is described, as well as how failure types (the way particular system failure modes occur) impact the diagnostic process. Failure and diagnostic analysis leads to system diagnosability modeling with the Failure Modes and Effects Analysis (FMEA) and component-indication relationship analysis. Finally, methods are developed for translating the diagnosability model into mathematical methods for computing metrics such as distinguishabilty, testability, and Mean Time Between Unscheduled Removals (MTBUR). These methods involve the use of matrices to represent the failure and replacement characteristics of the system. Diagnosability metrics are extracted by matrix multiplication. These metrics are useful when comparing the diagnosability of proposed designs or predicting the life-cycle costs of fault isolation. / Graduation date: 2000

Design phase decisions based on diagnosability lead to lower system costs and, in turn, higher quality products by means of reducing maintenance time and increasing system reliability. A case for diagnosability is presented. Functions of diagnosability are expounded upon including life cycle costs, statistical analysis, and design criterion to emphasize the necessity of diagnosability analysis early in the design phase. A diagnosability prediction metric is developed for system modeling of component failure rates and unjustified removals. The metric emphasizes ambiguity of system component indications as well as system structure. The metric is evaluated using historical data from the bleed air control system (BACS) on the Boeing 737-300. Seven design changes are suggested based on improving system diagnosability by changing component functions, modifying indications, and adding or changing sensors. The resulting designs are compared via Boeing's life cycle cost mechanism, DEPCOST model, based on cost improvements. It is shown that system improvements based on this prediction technique will increase the quality of a product since increased diagnosability decreases life cycle costs. / Graduation date: 1996

The median-times-to-failure (t₅₀’s) for straight dual-damascene via-terminated copper interconnect structures, tested under the same conditions, depend on whether the vias connect down to underlaying leads (metal 2, M2, or via-below structures) or connect up to overlaying leads (metal 1, M1, or via-above structures). Experimental results for a variety of line lengths, widths, and numbers of vias show higher t₅₀’s for M2 structures than for analogous M1 structures. It has been shown that despite this asymmetry in lifetimes, the electromigration drift velocity is the same for these two types of structures, suggesting that fatal void volumes are different in these two cases. A numerical simulation tool based on the Korhonen model has been developed and used to simulate the conditions for void growth and correlate fatal void sizes with lifetimes. These simulations suggest that the average fatal void size for M2 structures is more than twice the size of that of M1 structures. This result supports an earlier suggestion that preferential nucleation at the Cu/Si₃N₄ interface in both M1 and M2 structures leads to different fatal void sizes, because larger voids are required to span the line thickness in M2 structures while smaller voids at the base of vias can cause failures in M1 structures. However, it is also found that the fatal void sizes corresponding to the shortest-times-to-failure (STTF’s) are similar for M1 and M2, suggesting that the voids that lead to the shortest lifetimes occur at or in the vias in both cases, where a void need only span the via to cause failure. Correlation of lifetimes and critical void volumes provides a useful tool for distinguishing failure mechanisms. / Singapore-MIT Alliance (SMA)

The effects of interconnect length on the reliability of dual-damascene Cu metallization have been investigated. As in Al-based interconnects, the lifetimes of Cu lines increase with decreasing length. However, unlike Al-based interconnects, no critical length exists, below which all Cu lines are ‘immortal’. Furthermore, we found multi-modal failure statistics for long lines, suggesting multiple failure mechanisms. Some long Cu interconnect segments have very large lifetimes, whereas in Al segments, lifetimes decrease continuously with increasing line length. It is postulated that the large lifetimes observed in long Cu lines result from liner rupture at the bottom of the vias, which allows continuous flow of Cu between the two bond pads. As a consequence, the average lifetimes of short lines and long lines can be higher than those of lines with intermediate lengths. / Singapore-MIT Alliance (SMA)

Thesis (M.S.)--University of Delaware, 2007. / Principal faculty advisor: David R. Legates, Dept. of Geography. Includes bibliographical references.