In this study, the wear behaviour of a powder metallurgically produced AlCoCrFeNiTi0.5 high-entropy alloy (HEAs) is investigated at elevated temperatures. Spark plasma sintering (SPS) of inert gas atomised feedstock enables the production of dense bulk material. The microstructure evolution and phase formation are analysed. The high cooling rate in the atomisation process results in spherical powder with a microstructure comprising two finely distributed body-centred cubic phases. An additional phase with a complex crystal structure precipitates during SPS processing, while no coarsening of microstructural features occurs. The wear resistance under reciprocating wear conditions increases at elevated temperatures due to the formation of a protective oxide layer under atmospherical conditions. Additionally, the coefficient of friction (COF) slightly decreases with increasing temperature. SPS processing is suitable for the production of HEA bulk material. An increase in the wear resistance at elevated temperature enables high temperature applications of the HEA system AlCoCrFeNiTi0.5.

The standard processing architectures are enough to satisfy a lot of applications by employing already existing stream processing frameworks which are able to manage distributed data processing. In some specific cases, having geographically distributed data sources requires to distribute even more the processing over a large area by employing a geographically distributed architecture.‌ The issue addressed in this work is the reduction of data movement across the network which is continuously flowing in a geo-distributed architecture from streaming sources to the processing location and among processing entities within the same distributed cluster. Reduction of data movement can be critical for decreasing bandwidth costs since accessing links placed in the middle of the network can be costly and can increase as the amount of data exchanges increase. In this work we want to create a different concept to deploy geographically distributed architectures by relying on Apache Spark Structured Streaming and Apache Kafka. The features needed for an algorithm to run on a geo-distributed architecture are provided. The algorithms to be executed on this architecture apply the windowing and the data synopses techniques to produce a summaries of the input data and to address issues of the geographically distributed architecture. The computation of the average and the Misra-Gries algorithm are then implemented to test the designed architecture. This thesis work contributes in providing a new model of building geographically distributed architecture. The experimental results show that, for the algorithms running on top of the geo distributed architecture, the computation time is reduced on average by 70% compared to the distributed setup. Similarly, and the amount of data exchanged across the network is reduced on average by 99%, compared to the distributed setup. / Standardbehandlingsarkitekturer är tillräckligt för uppfylla behoven av många tillämpningar genom användning av befintliga ramverk för flödesbehandling med stöd för distribuerad databehandling. I specifika fall kan geografiskt fördelade datakällor kräva att databehandlingen fördelas över ett stort område med hjälp av en geografiskt distribuerad arkitektur. Problemet som behandlas i detta arbete är minskningen av kontinuerlig dataöverföring i ett nätverk med geo-distribuerad arkitektur. Minskad dataöverföring kan vara avgörande för minskade bandbreddskonstnader då åtkomst av länkar placerade i mitten av ett nätverk kan vara dyrt och öka ytterligare med tilltagande dataöverföring. I det här arbetet vill vi skapa ett nytt koncept för att upprätta geografiskt distribuerade arkitekturer med hjälp av Apache Spark Structured Streaming och Apache Kafka. Funktioner och förutsättningar som behövs för att en algoritm ska kunna köras på en geografisk distribuerad arkitektur tillhandahålls. Algoritmerna som ska köras på denna arkitektur tillämpar “windowing synopsing” och “data synopses”-tekniker för att framställa en sammanfattning av ingående data samt behandla problem beträffande den geografiskt fördelade arkitekturen. Beräkning av medelvärdet och Misra-Gries-algoritmen implementeras för att testa den konstruerade arkitekturen. Denna avhandling bidrar till att förse ny modell för att bygga geografiskt distribuerad arkitektur. Experimentella resultat visar att beräkningstiden reduceras i genomsnitt 70% för de algoritmer som körs ovanför den geo-distribuerade arkitekturen jämfört med den distribuerade konfigurationen. På liknande sätt reduceras mängden data som utväxlas över nätverket med 99% i snitt jämfört med den distribuerade inställningen.

Companies analyse large amounts of sensitive data on clusters of machines, using a framework such as Apache Hadoop to handle inter-process communication, and big data analytic tools such as Apache Spark and Apache Flink to analyse the growing amounts of data. Big data analytic tools are mainly tested on performance and reliability. Security and authentication have not been enough considered and they lack behind. The goal of this research is to improve the authentication and security for data analytic tools.Currently, the aforementioned big data analytic tools are using Kerberos for authentication. Kerberos has difficulties in providing multi factor authentication. Attacks on Kerberos can abuse the authentication. To improve the authentication, an analysis of the authentication in Hadoop and the data analytic tools is performed. The research describes the characteristics to gain an overview of the security of Hadoop and the data analytic tools. One characteristic is that the usage of the transport layer security (TLS) for the security of data transportation. TLS usually establishes connections with certificates. Recently, certificates with a short time to live can be automatically handed out.This thesis develops new authentication mechanism using certificates for data analytic tools on clusters of machines, providing advantages over Kerberos. To evaluate the possibility to replace Kerberos, the mechanism is implemented in Spark. As a result, the new implementation provides several improvements. The certificates used for authentication are made valid with a short time to live and are thus less vulnerable to abuse. Further, the authentication mechanism solves new requirements coming from businesses, such as providing multi-factor authenticationand scalability.In this research a new authentication mechanism is developed, implemented and evaluated, giving better data protection by providing improved authentication.

This thesis aims to provide recommendations for interfacing two memory-safe programming languages, Rust and SPARK. The comparison of Rust and C/C++, as well as SPARK and C/C++, has been thoroughly researched. However, this level of investigation hasn’t been as extensive between memory-safe languages. SPARK is a subset of Ada, a programming language with a long track record in safety-critical applications. Rust is a multi-paradigm language that has gained popularity since its emergence roughly a decade ago. Both languages implement ownership systems and type safety, which are instrumental in developing safe and reliable software. Memory-safety bugs in code written in C/C++ account for about 70 % of major vulnerabilities in the industry, as reported by organizations such as Microsoft, Apple, and Google. While tools to avoid memory errors exist, code cannot be made safe retroactively, which creates a need for languages designed with safety in mind, such as Rust and SPARK. Combining those two languages is a very promising path for low-level and systems programming and safety-critical applications (automotive, avionics, medical etc.). This thesis identifies best practices for safe bindings between Rust and SPARK, focusing on ensuring memory safety, type safety, and ownership, and then analyzes how to maintain or transfer those properties. The thesis provides clear recommendations based on control study programs and a real-world example using BBQueue, a circular buffer library suitable for embedded systems implemented in both languages. The identified best practices can be used for safe bindings and automated tools/code generation. This thesis fills a gap in existing research, primarily focusing on combining safe and unsafe languages. It provides a valuable contribution to software security and safety-critical systems. / Denna avhandling syftar till att ge rekommendationer för att koppla samman två minnessäkra programmeringsspråk, Rust och SPARK. Sådana kombinationer har redan studerats utförligt mellan Rust/SPARK och de osäkra språken C/C++, men det finns relativt lite information och forskning angående minnessäkra språk. SPARK är en delmängd av Ada, ett språk som har använts i decenier i säkerhetskritiska applikationer. Rust är ett multiparadigmatiskt språk som har blivit populärt sedan dess skapande för ungefär ett decennium sedan. Båda språken implementerar ägarskapssystem och typsäkerhet, vilka är avgörande för att utveckla säkra och tillförlitliga programvaror. Minnesfel i C/C++ står för cirka 70 procent av de stora sårbarheterna i branschen, som rapporterats av organisationer såsom Microsoft, Apple och Google. Även om det finns verktyg för att undvika minnesfel kan koden inte göras säker retroaktivt, vilket skapar ett behov av språk som är designade med säkerhet i åtanke. Att kombinera Rust och SPARK är en mycket lovande väg för systemprogrammering, låg nivå och säkerhetskritiska applikationer (bilindustri, flygavionik, medicinskt etc). Denna avhandling identifierar bästa praxis för säkra bindningar mellan Rust och SPARK, med fokus på att säkerställa minnessäkerhet, typsäkerhet och ägarskap, och analyserar sedan hur man kan behålla eller överföra dessa egenskaper. Avhandlingen ger klara rekommendationer baserade på kontrollstudieprogram och ett exempel med hjälp av BBQueue, ett cirkulärt buffertbibliotek som används i inbyggda system. Dessa resulat kan användas för att skapa säkra bindningar, manuelt eller med ett automatiserat verktyg. Denna avhandling fyller en lucka i befintlig forskning, som främst fokuserar på att kombinera säkra och osäkra språk. Den bidrar värdefullt till programvarusäkerhet och säkerhetskritiska system.

Prediction of flow and combustion in IC engines remains a challenging task. Traditional Reynolds Averaged Navier Stokes (RANS) methods and emerging Large Eddy Simulation (LES) techniques are being used as reliable mathematical tools for such predictions. However, RANS models have to be further refined to make them more predictive by eliminating or reducing the requirement for application based fine tuning. LES holds a great potential for more accurate predictions in engine related unsteady combustion and associated cycle-tocycle variations. Accordingly, in the present work, new advanced CFD based flow models were developed and validated for RANS and LES modelling of turbulent premixed combustion in SI engines. In the research undertaken for RANS modelling, theoretical and experimental based modifications have been investigated, such that the Bray-Moss-Libby (BML) model can be applied to wall-bounded combustion modelling, eliminating its inherent wall flame acceleration problem. Estimation of integral length scale of turbulence has been made dynamic providing allowances for spatial inhomogeneity of turbulence. A new dynamic formulation has been proposed to evaluate the mean flame wrinkling scale based on the Kolmogorov Pertovsky Piskunow (KPP) analysis and fractal geometry. In addition, a novel empirical correlation to quantify the quenching rates in the influenced zone of the quenching region near solid boundaries has been derived based on experimentally estimated flame image data. Moreover, to model the spark ignition and early stage of flame kernel formation, an improved version of the Discrete Particle Ignition Kernel (DPIK) model was developed, accounting for local bulk flow convection effects. These models were first verified against published benchmark test cases. Subsequently, full cycle combustion in a Ricardo E6 engine for different operating conditions was simulated. An experimental programme was conducted to obtain engine data and operating conditions of the Ricardo E6 engine and the formulated model was validated using the obtained experimental data. Results show that, the present improvements have been successful in eliminating the wall flame acceleration problem, while accurately predicting the in-cylinder pressure rise and flame propagation characteristics throughout the combustion period. In the LES work carried out in this research, the KIVA-4 RANS code was modified to incorporate the LES capability. Various turbulence models were implemented and validated in engine applications. The flame surface density approach was implemented to model the combustion process. A new ignition and flame kernel formation model was also developed to simulate the early stage of flame propagation in the context of LES. A dynamic procedure was formulated, where all model coefficients were locally evaluated using the resolved and test filtered flow properties during the fully turbulent phase of combustion. A test filtering technique was adopted to use in wall bounded systems. The developed methodology was then applied to simulate the combustion and associated unsteady effects in Ricardo E6 spark ignition engine at different operating conditions. Results show that, present LES model has been able to resolve the evolution of a large number of in-cylinder flow structures, which are more influential for engine performance. Predicted heat release rates, flame propagation characteristics, in-cylinder pressure rise and their cyclic variations are also in good agreement with measurements.

L'alliage Si-Ge est utilisé depuis de nombreuses années dans les modules thermoélectriques dans les sondes spatiales de la NASA. Ils convertissent la chaleur résultant de la désintégration radioactive de matériaux riches en un ou plusieurs radio-isotopes en électricité. Cet alliage est performant à haute température (à partir de 700 °C), c'est pourquoi il trouve également un fort intérêt dans l'industrie automobile. De nombreuses recherches dans ce secteur s'orientent vers la thermoélectricité, notamment vers des modules fonctionnant à haute température pour permettre la réduction de consommation de carburant.La meilleure composition de l'alliage en thermoélectricité est Si0,8Ge0,2. Le facteur de mérite réduit (ZT) de ces matériaux est généralement proche de 0,75 et de 0,45 à 700 °C pour les types n et p respectivement. Le germanium étant très onéreux, la composition retenue dans cette étude est Si0,92Ge0,08. Pour améliorer les performances de la composition choisie et se rapprocher de celles de la meilleure composition, la voie retenue est l'incorporation de nanoinclusions à base de siliciures de molybdène dans le matériau, permettant la diminution de la conductivité thermique.L'alliage Si-Ge est synthétisé par mécanosynthèse, et densifié par SPS. Les dopants utilisés sont le phosphore et le bore pour les types n et p respectivement. Le taux de dopage optimal est de 0,7 %. Ainsi, les ZT obtenus à 700 °C sont égaux à 0,7 et 0,5 pour les types n et p respectivement. La nature des inclusions stables dans la matrice est déterminée par la méthode CalPhad qui permet l'obtention du diagramme ternaire Mo-Si-Ge. La phase MoSi2 apparait alors comme étant la seule phase stable dans la matrice Si0,92Ge0,08. La fraction volumique optimale de molybdène est de 1,3 % lorsque les matériaux sont densifiés à 1280 °C. Le ZT obtenu est supérieur à 1 à 700 °C pour le type n, et proche de 0,8 pour le type p. L'ajout de nanoinclusions a permis d'augmenter les performances de 43 % et de 60 % à 700 °C.

Due to the continuous work in the automobile industry to reduce the environmental impact, reduce fuel consumption and increase efficiency, new technologies need to be developed and implemented in vehicles. For spark ignited engines, one technology that has received more attention in recent years is long route Exhaust Gas Recirculation (EGR), which means that exhaust gases after the turbine are transported back to the volume before the compressor in the air intake system of the engine. In this work, the components of the long route EGR system is modeled with mean value engine models in Simulink, and implemented in a existing Simulink engine model. Then different methods for estimating the mass flow over the long route EGR system are compared, and the transport delays for the recirculated exhaust gases in the engines air intake system are modeled. This work is based on measurements done on an engine rig, on which a long route EGR system was installed. Finally, some ideas on how a long route EGR system on a gasoline engine can be controlled are presented based on the results in this thesis work.

近年來雲端運算技術的發展與電腦設備效能提升，使得以大量電腦主機以水 平擴充的方式組成叢集運算系統，成為一可行的選擇。Apache Hadoop 是Apache 基金會的一個開源軟體框架，它是由Google 公司的MapReduce 與Google 檔案 系統實作成的分布式系統，可以管理數千台以上的電腦群集。Hadoop 利用分散 式檔案系統HDFS 可以提供PB 級以上的資料存放空間，透過MapReduce 框架 可以將應用程式分割成小工作分散到叢集中的運算節點上執行。 此外，企業累積了巨量的資料，如何處理與分析這些結構化或者是非結構化 的資料成了現在熱門研究的議題。因此傳統的資料挖掘方式與演算法必須因應新 的雲端運算技術與分散式框架的概念，進行調整與改良，發展新的方法。 關聯規則是分析資料庫龐大的資料中，項目之間隱含的關聯，常見的應用為 購物籃分析。一般情形下會在特定的維度與特定的顆粒度範圍內挖掘關聯規則， 但這樣的方式無法找出更細微範圍下之規則，例如挖掘一個年度的交易資料無法 發現消費者在聖誕節為了慶祝而購買的商品項目間的規則，但若將時間限縮在 12 月份即可挖掘出這些規則。 Apriori 演算法是挖掘關聯規則的一個著名的演算法，透過產生候選項目集 合與使用自訂的最小支持度進行篩選，產生高頻項目集合，接著以最小信賴度篩 選獲得關聯規則的結果。若有k 種單一項目集合，則候選項目集合最多有2𝑘 − 1 個，計算高頻項目時則需反覆掃描整個資料庫，Apriori 這兩個主要步驟需要耗費 相當大量的運算能力。 因此本研究將資料庫分割成多個資料區塊挖掘關聯規則，再將結果逐步更新 的演算法，解決大範圍挖掘遺失關聯規則的問題，結合spark 分散式運算的架構 實作程式，在電腦群集上平行運算減少關聯規則的挖掘時間。 / Because of improving technique of cloud-computing and increasing capability of computer equipment, it is feasible to use clusters of computers by horizon scalable a lot of computers. Apache Hadoop is an open-source software of Apache. It allows the management of cluster resource, a distributed storage system named Hadoop Distributed File System (HDFS), and a parallel processing technique called MapReduce. Enterprises have accumulated a huge amount of data. It is a hot issue to process and analyze these structured or unstructured data. Traditional methods and algorithms of data mining must make adjustments and improvement to new cloud computing technology and concept of decentralized framework. Association rules is the relations of items from large database. In general, we find association rules in fixed dimensions and granular database. However, it might loss infrequent association rules. Apriori algorithm is one famous algorithm of mining association rule. There are two main steps in this algorithm spend a lot of computing resource. To generate Candidate itemset has quantity 2𝑘 − 1, if there are k different item. Second step is to find frequent, this step must compare all tractions in the database. This approach divides database to segmentations and finds association rules of these segmentations. Then, we combine rules of segmentations. It can solve the problem of missing infrequent itemset. In addition, we implement this method in Spark and reduce the time of computing.

Spark plasma sintering (SPS) is a powder consolidation technique used to rapidly densify a variety of material systems. SPS is capable of precisely controlling material microstructures and achieving non-equilibrium phases due to rapid heating and cooling rates through the simultaneous application of pressure and direct current. Due to these characteristics, SPS is an ideal processing technique for high temperature ceramics which require processing at temperatures greater than 1500°C. Due to the desirable properties obtained on small diameter materials processed by SPS, larger and more complex geometries are desired while maintaining sample microstructures. In order to accurately scale ceramics produced by SPS, a finite element model must be developed that can be used as a predictive tool. My research focuses on developing a finite element model for the spark plasma sintering furnace at the University of Arizona and validating modeled results using experimentally obtained data. Electrical and thermal conductivity as functions of temperature vary widely among different grades of commercially available electrode grade graphite at constant density. Modeled material properties are optimized in order to calibrate modeled results to experimentally obtained data (i.e. measured current, voltage, and temperature distributions). Sensitivity analysis is performed on the model to better understand model physics and predictions. A calibrated model is presented for 20mm ZrB2 and Si3N4 discs. Sample temperature gradients are experimentally confirmed using grain size and β-Si3N4 phase composition. The model is used to investigate scale up from 20mm to 30mm discs and 30mm rings as well as effects of processing conditions on β-Si3N4 content.

A retificação é um dos processos mais utilizados em operações de acabamento, geralmente caracterizada pelo resultado dimensional atingido e baixas rugosidades. Apesar de seu uso intenso na indústria, há aspectos ainda desconhecidos do processo e de sua estabilidade, sendo considerados tempos excessivos de centelhamento para atingir boas condições dimensionais. Este trabalho propõe um novo modelo para determinar o menor tempo de centelhamento necessário para que as deformações do sistema máquina-rebolo-peça sejam absorvidas e a dimensão objetivo da peça alcançada. Esse modelo partiu do equacionamento descrito por S. Malkin para obtenção da rigidez da máquina e avançou na obtenção da constante característica do tempo τ de forma experimental. A comprovação do modelo proposto deu-se em dois experimentos, sendo o primeiro experimento aplicado para determinar o valor de τ em função da velocidade de avanço, segundo o modelo de Malkin. Comprovou-se que o τ varia, principalmente em baixas velocidades de avanço. O segundo experimento teve como objetivo determinar qual o menor tempo de centelhamento necessário para que as deformações do sistema máquina-rebolo-peça sejam plenamente absorvidas e compará-las com o modelo proposto. Os resultados teórico e experimental comprovaram que, para as velocidades de mergulho 3,0 e 2,5 mm/min, com τ = 0,219 e 0,204 s, respectivamente, o tempo mínimo de centelhamento de 1,0 s foi suficiente para absorver as deformações elásticas inseridas no sistema. De forma comparativa, o modelo proposto foi mais eficaz que o de Malkin para o cálculo do tempo de centelhamento e do erro diametral da peça para relações t1/τ menores ou iguais a 20, sendo t1 o tempo em segundos da fase de retificação em desbaste. Conclui-se que o modelo proposto neste trabalho mostrou-se eficiente para predizer o tempo mínimo necessário para centelhamento e a redução diametral da peça, considerando o valor experimental de τ. / Grinding is one of the processes most commonly used in finishing operations, usually characterized by good dimensional results and low roughness values. Despite the extensive use in industry, there are still unknown aspects of the grinding process and its stability, with excessive spark-out times being used to achieve good dimensional requirements. This research work proposed a new model to determine the shortest spark-out time required to absorb the machine-wheel-workpiece system deformations and to achieve the desired workpiece diameter. This model derived from the equation described by S. Malkin to obtain the machine stiffness and consisted in experimentally obtaining the time characteristic constant τ. The verification of the proposed model was done in two experiments. The first one was performed to determine the value of τ according to the radial infeed velocity (vf), as proposed by S. Malkin. It was shown that the τ varies, especially at low feed rates. The second objective was to determine the shortest spark-out time needed, in which all machine-wheel-workpiece system deformations were fully eliminated and to compare it to the one calculated by the proposed model. The theoretical and experimental results showed that for radial infeed velocities of 3.0 and 2.5 mm/min, with τ = 0.219 and 0.204 s, respectively, the spark-out time of 1 s was long enough to ensure that the elastic deformations introduced into the system were fully absorbed. Comparatively, the proposed model was more effective than the Malkin\'s one to predict the spark-out time and the diametric error for the ratio t1/τ lower that 20, being t1 the time in seconds for the rough grinding stage. As a conclusion, the proposed model proved to be effective in predicting the minimum spark-out time and the workpiece diameter reduction, when considering the experimental value of τ.