This thesis presents a new method for Condition Based Ultrasonic Monitoring to be applied in conjunction with a lubrication distribution controller. As part of this thesis, algorithms were developed using ultrasonic sensors to control the application of lubrication to machinery. The controller sensors detect an ultrasonic signal from rolling or sliding machine elements. This signal then alerts the controller to dispense the proper amount of lubrication when needed, as opposed to a time schedule based on average performance or history. The work from this thesis will be used to help reduce equipment downtime and maintenance cost when utilized in an industrial environment.

The goal of the study presented in this thesis was the improvement of estimation and monitoring procedures for condition monitoring of belt tension and misalignment in belt-driven automated material handling systems widely used in modern semiconductor manufacturing systems. In pursuit of this goal, two 3-factor, 3-level experiments were designed to study how belt vibration characteristics depend on changes in belt length, belt tension, belt misalignment, and initial location of the excitation of belt vibration. Dependent variables in each of the experiments were drawn from a denoised frequency spectrum calculated from an Autoregressive model of the belt vibration time-series. A feature vector was developed from the Autoregressive features via variance based sensitivity analysis. Results showed that belt vibration characteristics were sensitive to changes in all of the independent variables examined. These results motivated the design of a device to improve the standardized technique widely used to monitor belt tension in belt-driven material handling systems. Reducing variance in the belt length and the location of the initial excitation of belt vibration yielded a reduction of tension estimate standard deviation an order of magnitude, as compared to a human performing the standardized technique. Thus, the use of this device provided higher belt tension estimate resolution. Future work that could lead to a less intrusive technique is presented. / text

Production of steel powder is done by atomization of a molten steel stream. Atomization is done by feeding high pressure nitrogen gas through nozzles, creating jets of gas which scatter the molten steel stream into powder. The steel powder falls through the atomization tower whilst it cools and solidifies. Finally, the steel powder is transported for further processing. The compressor is used for two main purposes, to compress the nitrogen gas to desired pressure and enable recycling of nitrogen gas. As nitrogen is inert and do not react with its surrounding, the gas can be recycled. Filtering nitrogen gas from the atomization process, one is able to reuse the gas, which is led to the inlet side of a compressor. A closed loop is thus created which is economically important. In 2021 a major compressor failure occurred, which caused large production losses. iba systems is a commercially available product extensively utilized in the Swedish steel industry for data acquisition, production monitoring and generating key performance indicators. Therefore, this thesis investigates what modules and functionality iba systems have to offer. Process and machine signals are studied to assess both their utility in predicting machine failure and relevant iba modules for the predictive maintenance purposes, based on a literature review. This thesis shows the possibility to implement an anomaly detection to detect abnormal behavior, related to historic compressor failure. Estimating when maintenance is needed is possible but requires implementation of new sensors to obtain useful information, mainly vibration data from machinery. Anomaly detection is implemented using ibaAnalyzer. Additional analysis is done in Matlab.

This research serves to evaluate the ability of optical detection techniques to ascertain the lubricity of revenue service track from a moving platform for railroad applications. A literature review is presented that covers the rail vehicle dynamics that drive the need of Top-of-Rail lubrication and directly affect the manner in which the Top-of-Rail Friction Modifiers (TORFM) and flange grease both spread down rail and eventually wear away. This literature review also highlights previous research in the field of rail lubrication and the benefits that rail lubricants, specifically TORFM, provide for the railroads. Finally, the literature review covers the governing optical principals inherent to the synchronous spot radiometer that has been developed for use in the research as a gloss ratio instrument and also addresses the drawbacks and challenges inherent to applying this type of instrument in the railroad industry. The research then overviews previous rail lubricity sensors developed by the Railway Technologies Laboratory (RTL) at Virginia Tech and the lessons learned from their application. The preceding field testing conducting with a modified second generation rail lubricity sensor and a rail push car is briefly summarized with emphasis on the drawbacks and issues that were used to develop the third generation sensor used for this research. The development of the third generation sensor is covered, including the issues that it attempts to solve from its predecessor and the governing optical principals that govern the operation of the sensor. The laboratory evaluations conducting to commission the sensor are also covered in preparation for deploying the new third generation sensor in medium speed, medium distance revenue service testing. This includes a shakedown run on a siding in Riverside, VA prior to conducting mainline in-service testing. Finally, this research thesis covers the in-service testing on revenue track conducted with the new third generation rail lubricity sensor and the accompanying remote-controlled (RC) rail cart. The two components, when combined, create a Lubricity Assessment System which is capable of being operated at speeds upwards of 10 mph remotely from a follow hy-rail truck. The data collected from this field test is analyzed for the lubricity assessments that are able to be drawn from this initial phase of field service testing. The conclusions from this testing affirm the ability of optical methods to determine and evaluate Top-of-Rail (TOR) lubricity from a moving platform. Specifically, the new sensor is able to identify several local phenomena that demonstrate the high potential for errant evaluation of rail lubricity evaluation from spot check based methods that are solved by evaluating the track in a continuous, moving fashion. Based on the continuous moving data collected for this test, several new signal traits such as the spatial frequency (wavenumber) associated with the passing freight cart wheels in the lubricity signal and the phantom applicator effect of transient lubricity conditions at the entrances and exits of curves can be detected and investigated. The success of this research indicates the continued evaluation of lubricity signals from a moving platform is warranted and suggests the potential for introducing one of these systems to various track metrology cars deployed throughout the United States railroads. / Master of Science / The United States railroads have been employing rail lubricants to the rails since the beginning of the industry and have recently, in the past 20 years, introduced another type of lubricant: Top-of-Rail Friction Modifiers (TORFM). TORFM creates a third body layer between the train wheels and the Top-of-Rail surface to minimize asset wear of both the wheel and rail and to increase the train efficiency. As the United States railroads embrace Precision Scheduled Railroading (PSR), increased train efficiency can allow the railroads to run longer trains with fewer locomotives. This increases the efficiency and capability of the rail network and also decreases the fuel burned and the amount of rail and wheel wear. TORFM has been proven to be effective and is widely adopted, but the railroads are still in need of tools to determine the presence and absence of these thin and often nearly invisible layers of lubricant on the rail surface. This research uses lasers as tool to quantify the level of lubrication of the rail surface. The presence of rail lubricants, such as TORFM, on the rail surface change the amount of light that is reflected and scattered off the shiny steel surface. These changes are often small but can be captured by photodetectors housed in the instrument. By plotting the detected sensor values, trends in the lubricity signal can be tracked and evaluated to determine the presence or absence of rail lubricants and assess the overall quality of lubrication on the rail at specific locations down track. The research in this thesis takes existing methods that were used for single spot inspections and adapts them to a moving platform. The moving platform is able to continuously scan the Top-of-Rail surface as the instrument moves along and generates continuous moving evaluations of rail lubricity. This can be especially important when the lubricity is not uniform and allows for trends in the data to be analyzed to provide more consistent and precise evaluations of the lubricity trends down rail. Optical tools like this sensor, which are by nature non-contact sensors, can easily be adapted to existing track measurement railcars and deployed system wide. This solves a strong need for railroad engineers: to be able to identify the presence of rail lubricants and evaluate the effectiveness of their lubrication practices.

In the trends of technical improvements and automatization is it important for companies to keep up with the developments to be competitive on the market. SwePart Transmissions AB is a company that manufacture and develop gearboxes for the currently growing robot arms industry and the main task with this study is to investigate how to apply condition based monitoring on a new gearbox from the company. The work considers vibration analysis and testing new ideas in the oil analysis field. The tests that were performed are based on measuring the difference in impedance or magnetic field due to the increasement of wear. The results of the tests are not clear. This thesis is the beginning of a big project and therefore lies the value of this work in the new ideas and suggestions for further work.

Prognostics and health management (PHM) is a major tool enabling systems to evaluate their reliability in real-time operation. Despite ground-breaking advances in most engineering and scientific disciplines during the past decades, reliability engineering has not seen significant breakthroughs or noticeable advances. Therefore, self-awareness of the embedded system is also often required in the sense that the system should be able to assess its own health state and failure records, and those of its main components, and take action appropriately. This thesis presents a radically new prognostics approach to reliable system design that will revolutionise complex power electronic systems with robust prognostics capability enhanced Insulated Gate Bipolar Transistors (IGBT) in applications where reliability is significantly challenging and critical. The IGBT is considered as one of the components that is mainly damaged in converters and experiences a number of failure mechanisms, such as bond wire lift off, die attached solder crack, loose gate control voltage, etc. The resulting effects mentioned are complex. For instance, solder crack growth results in increasing the IGBT’s thermal junction which becomes a source of heat turns to wire bond lift off. As a result, the indication of this failure can be seen often in increasing on-state resistance relating to the voltage drop between on-state collector-emitter. On the other hand, hot carrier injection is increased due to electrical stress. Additionally, IGBTs are components that mainly work under high stress, temperature and power consumptions due to the higher range of load that these devices need to switch. This accelerates the degradation mechanism in the power switches in discrete fashion till reaches failure state which fail after several hundred cycles. To this end, exploiting failure mechanism knowledge of IGBTs and identifying failure parameter indication are background information of developing failure model and prognostics algorithm to calculate remaining useful life (RUL) along with ±10% confidence bounds. A number of various prognostics models have been developed for forecasting time to failure of IGBTs and the performance of the presented estimation models has been evaluated based on two different evaluation metrics. The results show significant improvement in health monitoring capability for power switches. Furthermore, the reliability of the power switch was calculated and conducted to fully describe health state of the converter and reconfigure the control parameter using adaptive algorithm under degradation and load mission limitation. As a result, the life expectancy of devices has been increased. These all allow condition-monitoring facilities to minimise stress levels and predict future failure which greatly reduces the likelihood of power switch failures in the first place.

Samarbetspartnern som denna studie har gjorts för har som mål att konstruera en detektor för tågtrafiken som bygger på bildigenkänning och artificiell intelligens. Problemet är att de lösningar som finns idag är dyra, och därför är en förutsättning att den ska vara byggd med konsumentprodukter för att få ner kostnaden samt att den ska vara enkel att installera och underhålla. Flera ramverk för objektigenkänning existerar, men dessa bygger på olika metoder och tekniker. Studien har därför utförts som en fallstudie vars syfte har varit att jämföra två välanvända ramverk för objektigenkänning för att identifiera olika för- och nackdelar gällande noggrannhet och tidsåtgång vid träning och test med hjälp av dessa ramverk. Även vilka olika utmaningar som stötts på under tillvägagångssättet har lyfts fram. Studien sammanfattar sedan dessa för att skapa idéer och diskussion för hur dessa skulle kunna implementeras på den nya tågdetektorn. Ramverken som har jämförts är OpenCV och Google TensorFlow. Dessa bygger på olika objektigenkänningstekniker, i huvudsak kaskadklassificering och neurala nät. Ramverken testades med en datamängd på 400 bilder på olika tågfordon där hjulaxlarna användes som parameter för objektigenkänningen. Testerna bedömdes efter kriterier gällande noggrannhet, tidsåtgång för träning samt komplexitet för konfiguration och användning. Resultatet visade att OpenCV hade en snabb träningsprocess, men visade låg precision och en mer komplex konfigurerings- och användningsprocess. TensorFlow hade en långsammare träningsprocess, men visade istället bättre precision och en mindre komplex konfigurering. Slutsatsen av studien är att TensorFlow visade bäst resultat och har mest potential att användas i den nya tågdetektorn. Detta baseras på studiens resultat samt att den bygger på modernare tekniker med neurala nät för objektigenkänning. / The research in this thesis is conducted with the partners aim to construct a new train detection system that uses image recognition and artificial intelligence. Detectors like these that exists today are expensive, so the construction is going to be based around the use of consumer electronics to lower the cost and simplify installation and maintenance. Several frameworks for object detection are available, but they use different approaches and methods. This thesis is therefore carried out as a case study that compares two widely used frameworks for image recognition tasks. The purpose is to identify advantages and disadvantages regarding training and testing when using these frameworks. Also highlighted is different challenges encountered in the process. The summary of the results is used to form ideas and a discussion about how to implement a framework in the new detection system. The frameworks compared in this study are OpenCV and Google TensorFlow. These frameworks use different methods for object detection, mainly cascade classifiers and convolutional neural nets. The frameworks were tested using a dataset of 400 images on different trains where the wheel-axles were used as the object of interest. The results were analyzed based on criteria regarding precision, total training time and also complexity regarding configuration and usage. The results showed that OpenCV had a faster training process but had low precision and more complex configuration. TensorFlow had a much longer training process but had better precision and less complex configuration. The conclusion of the study is that TensorFlow overall showed the best result and has a better potential for implementation in the new detection system. This is based on the results from the study, but also that the framework is developed with a more modern approach using convolutional neural nets for bject detection.