The power electronics market is valued at $23.25 billion in 2019 and is projected to reach $ 36.64 billion by 2027. Power electronic systems (PES) have been extensively used in a wide range of critical applications, including automotive, renewable energy, industrial variable-frequency drive, etc. Thus, the PESs' reliability and robustness are immensely important for the smooth operation of mission-critical applications. Power semiconductor switches are one of the most vulnerable components in the PES. The vulnerability of these switches impacts the reliability and robustness of the PES. Thus, switch-health monitoring and prognosis are critical for avoiding unexpected shutdowns and preventing catastrophic failures. The importance of the prognosis study increases dramatically with the growing popularity of the next-generation power semiconductor switches, wide bandgap switches. These switches show immense promise in the high-power high-frequency operations due to their higher breakdown voltage and lower switch loss. But their wide adaptation is limited by the inadequate reliability study. A thorough prognosis study comprising switch degradation modeling, remaining useful life (RUL) estimation, and degradation-aware controller development, is important to enhance the PESs' robustness, especially with wide bandgap switches. In this dissertation, three studies are conducted to achieve these objectives- 1) Insulated Gate Bipolar Transistor (IGBT) degradation modeling and RUL estimation, 2) cascode Gallium Nitride (GaN) Field-Effect Transistor (FET) degradation modeling and RUL estimation, and 3) Degradation-aware controller design for a PES, solid-state transformer (SST). The first two studies have addressed the significant variation in RUL estimation and proposed degradation identification methods for IGBT and cascode GaN FET. In the third study, a system-level integration of the switch degradation model is implemented in the SST. The insight into the switch's degradation pattern from the first two studies is integrated into developing a degradation-aware controller for the SST. State-of-the-art controllers do not consider the switch degradation that results in premature system failure. The proposed low-complexity degradation-aware and adaptive SST controller ensures optimal degradation-aware power transfer and robust operation over the lifetime.

The degradation of aging energy infrastructure systems has the potential to increase the risk of failure, resulting in power outage and costly unplanned maintenance work. Therefore, the development of scientific and cost-effective life cycle management (LCM) strategies has become increasingly important to maintain energy infrastructure. Since degradation of aging equipment is an uncertain process which depends on many factors, a risk-based approach is required to consider the effect of various uncertainties in LCM. The thesis presents probabilistic models to support risk-based life cycle management of energy infrastructure systems. In addition to uncertainty in degradation process, the inspection data collected by the energy industry is often censored and truncated which make it difficult to estimate the lifetime probability distribution of the equipment. The thesis presents modern statistical techniques in quantifying uncertainties associated with inspection data and to estimate the lifetime distributions in a consistent manner. Age-based and sequential inspection-based replacement models are proposed for maintenance of component in a large-distribution network. A probabilistic lifetime model to consider the effect of imperfect preventive maintenance of a component is developed and its impact to maintenance optimization is illustrated. The thesis presents a stochastic model for the pitting corrosion process in steam generators (SG), which is a serious form of degradation in SG tubing of some nuclear generating stations. The model is applied to estimate the number of tubes requiring plugging and the probability of tube leakage in an operating period. The application and benefits of the model are illustrated in the context of managing the life cycle of a steam generator.

The degradation of aging energy infrastructure systems has the potential to increase the risk of failure, resulting in power outage and costly unplanned maintenance work. Therefore, the development of scientific and cost-effective life cycle management (LCM) strategies has become increasingly important to maintain energy infrastructure. Since degradation of aging equipment is an uncertain process which depends on many factors, a risk-based approach is required to consider the effect of various uncertainties in LCM. The thesis presents probabilistic models to support risk-based life cycle management of energy infrastructure systems. In addition to uncertainty in degradation process, the inspection data collected by the energy industry is often censored and truncated which make it difficult to estimate the lifetime probability distribution of the equipment. The thesis presents modern statistical techniques in quantifying uncertainties associated with inspection data and to estimate the lifetime distributions in a consistent manner. Age-based and sequential inspection-based replacement models are proposed for maintenance of component in a large-distribution network. A probabilistic lifetime model to consider the effect of imperfect preventive maintenance of a component is developed and its impact to maintenance optimization is illustrated. The thesis presents a stochastic model for the pitting corrosion process in steam generators (SG), which is a serious form of degradation in SG tubing of some nuclear generating stations. The model is applied to estimate the number of tubes requiring plugging and the probability of tube leakage in an operating period. The application and benefits of the model are illustrated in the context of managing the life cycle of a steam generator.

No / Greedy and contention-based forwarding schemes were proposed for mobile ad hoc networks (MANETs) to perform data routing hop-by-hop, without prior discovery of the end-to-end route to the destination. Accordingly, the neighboring node that satisfies specific criteria is selected as the next forwarder of the packet. Both schemes require the nodes participating in the selection process to be within the area that confronts the location of the destination. Therefore, the lifetime of links for such schemes is not only dependent on the transmission range, but also on the location parameters (position, speed and direction) of the sending node and the neighboring node as well as the destination. In this paper, we propose a new link lifetime prediction method for greedy and contention-based routing which can also be utilized as a new stability metric. The evaluation of the proposed method is conducted by the use of stability-based greedy routing algorithm, which selects the next hop node having the highest link stability.

The ability to handle and process data for information extraction is getting more and more important. Using extracted data from the business to improve productivity is seen as an important part in developing the business processes. In this thesis, industrial robots and their survival times are analyzed. The work is about predicting the probability that a specific robot will fail during a specified time period. Also, survival analysis is conducted where the median lifetime and conditional median lifetime for industrial robots are estimated. Two approaches are used, logistic regression and survival analysis. A logistic regression model is made to predict the probability for different industrial robots to break during a specified time period. The logistic model achieves an accuracy of 0.694 with even higher accuracy regarding high – and low risk robots. The survival analysis uses a Cox PH model to check validity for proportional hazards and then a parametric model with Weibull distribution is fitted. The parametrical survival model is used to estimate the median lifetime and the remaining median lifetime for the robots. The estimated probabilities and lifetimes can be used as an indication of which robots are in risk of failure.

<p>For complex technical systems in the electricity and pulp and paper industries, maintenance management addresses how to exploit physical assets in the most profitably way. This is a difficult task that requires taking into consideration parameters of totally different natures – e.g. reliability data, operating costs, condition of technical systems, the environment and rules and regulation.</p><p>An incorrect estimate of a residual lifetime can result in a premature renewal with accompanying high capital costs. If, however, renewal is delayed, a breakdown may occur which can cause major damage to technical equipment and a loss of income due to outages. Because of the complexity of many technical systems, it can be hard to select adequate data to use when making decisions about renewal strategies. To cope with this, one approach is to use less detailed models that are operated by skilled analysts.</p><p>This work demonstrates the advantage of such an approach by proposing two methods applied in a joint methodology that has its origins in RCM. The methodology consists of Dynamic Lifetime Model (DLA) and the Condition Based Index (CBI). The DLA method copes with the financial risk associated with the point in time for when a renewal is carried out and the CBI method uses critical parameters to estimate the condition of a technical system. The two methods together create a quantitative connection between reliability, maintenance and financial risk. A case study based validation of the methodology was carried out at SCA Ortvikens paper mill on a refiner system and Forsmark nuclear power plant. Lessons learned from the case study showed that the methodology could be used to identify which components could cause costly breakdown. By using the methodology a manager gets a decision support tool for estimating short-term and long-term consequences of decisions regarding maintenance management in order to maximize utility of the system concerned</p>

For complex technical systems in the electricity and pulp and paper industries, maintenance management addresses how to exploit physical assets in the most profitably way. This is a difficult task that requires taking into consideration parameters of totally different natures – e.g. reliability data, operating costs, condition of technical systems, the environment and rules and regulation. An incorrect estimate of a residual lifetime can result in a premature renewal with accompanying high capital costs. If, however, renewal is delayed, a breakdown may occur which can cause major damage to technical equipment and a loss of income due to outages. Because of the complexity of many technical systems, it can be hard to select adequate data to use when making decisions about renewal strategies. To cope with this, one approach is to use less detailed models that are operated by skilled analysts. This work demonstrates the advantage of such an approach by proposing two methods applied in a joint methodology that has its origins in RCM. The methodology consists of Dynamic Lifetime Model (DLA) and the Condition Based Index (CBI). The DLA method copes with the financial risk associated with the point in time for when a renewal is carried out and the CBI method uses critical parameters to estimate the condition of a technical system. The two methods together create a quantitative connection between reliability, maintenance and financial risk. A case study based validation of the methodology was carried out at SCA Ortvikens paper mill on a refiner system and Forsmark nuclear power plant. Lessons learned from the case study showed that the methodology could be used to identify which components could cause costly breakdown. By using the methodology a manager gets a decision support tool for estimating short-term and long-term consequences of decisions regarding maintenance management in order to maximize utility of the system concerned / QC 20111216

This thesis is dealing with the issue of residual stress present in the wall of a polymer pipe and the influence on its lifetime. Experimental data obtained by the ring slitting method are evaluated and the tangential residual stress in the wall of polypropylene pipe is determined. The evaluation is carried out using a new methodology based on the curved beam theory. The method is verified using numerical simulation. 3D numerical model is used to verify the behavior of the pipe, when there are both tangential and axial residual stresses present in the pipe wall, because the presence of axial stress causes a rise in the magnitude of tangential residual stress. A correction of the tangential stress values corresponding to the pipe length is then proposed. It is shown, that the distribution of the tangential residual stress does not depend significantly on the dimensions or ma-terial of the extruded pipe and a general equation is proposed to describe the distribution. This general distribution is then involved in the calculations of the pipe lifetime that are carried out using a method based on the linear elastic fracture mechanics. A significantly lower lifetimes are obtained when taking the residual stress into account.

The aim of this thesis is to describe behavior of cracks in layered polymer materials. Quasi-brittle fracture (through the initiation and subsequent crack propagation mechanism) under low stresses is the most common mode of failure of polymer materials. In this case plastic deformations are localized in the vinicity of the crack tip and linear elastic fracture mechanics description of the crack behavior can be used. The knowledge of fracture parameters change during the crack propagation in multilayer body is a key point for establishing of the maximum load and consequently for the assessment of the residua lifetime. In contrast to homogeneous bodies the estimation of stress intensity factors for multilayer (composite) structure is numerically more elaborated and the fracture mechanics approach is complicated by the existence of interfaces between single layers, where material parameters are changed by a step. Special attention is paid to the configuration of a crack growing close to the material interface and along the interface. For the crack with tip on the material interface the effective values of stress intensity factor based on the crack stability criteria are estimated. It is shown that under special conditions (depending mainly on the elastic mismatch of materials) the existence of material interface has positive influence on the lifetime of the multilayered structure.

Die Dissertation widmet sich der rechnerischen Lebensdauerabschätzung laufender Drahtseile, die in verschiedensten Applikationen verwendet werden. Konkret wird die Methode Leipzig, als eine der etablierten Vorgehensweisen für die Ermittlung ertragbarer Biegewechselanzahlen für die Betriebs- und Lebensdauer laufender Drahtseile umfassend beschrieben und weiterentwickelt. Die Arbeit stellt somit den Stand der Technik in Bezug auf diese Methode dar. Neben den detaillierten Darstellungen der Herleitung der Methode, werden auch weiterführende Herangehensweisen für eine umfangreiche Analyse von Seiltrieben beschrieben sowie Hinweise und Einschätzungen zur aktuellen normativen Entwicklung hinsichtlich Auslegung und Nachweis von Drahtseilen gegeben. In Bezug auf unterschiedliche Drahtnennfestigkeiten, Verdichtungsverfahren und Grade der inneren Seilschädigung, wurden Versuchsreihen durchgeführt und ausgewertet, um dessen verbesserte Berücksichtigung innerhalb der Lebensdauerabschätzung mit der Methode Leipzig herzuleiten. Im Ergebnis werden Anpassungen für eine veränderte Interpretation der Gesamtbeanspruchung vorgestellt, die nachweislich die Genauigkeit der Lebensdauerabschätzung erhöhen und damit der steigenden Relevanz dieser Analysen Rechnung trägt. / The dissertation is dedicated to the calculative lifetime estimation of running wire ropes, which are used in a wide variety of applications. Specifically, the Leipzig method is comprehensively described and further developed as one of the established procedures for determining endurable numbers of bending cycles for the service life and lifetime of running wire ropes. The thesis thus represents the state of the art with regard to this method. In addition to the detailed presentation of the derivation of the method, further approaches for a detailed analysis of rope drives are described as well as information and assessments on the current normative development with regard to the design and proof of wire ropes. With regard to different nominal wire strengths, compaction methods and degrees of internal rope deterioration, test series were carried out and evaluated in order to derive their improved consideration within the lifetime estimation with the Leipzig method. As a result, adjustments for a modified interpretation of the overall stress are presented, which have been shown to increase the accuracy of the lifetime estimation and thus take into account the increasing relevance of these analyses.