No description available.

No description available.

Anion intercalation in the graphite positive electrode of a dual-ion battery requires high potential (> 4.3 V vs Li+/Li), which aggravates parasitic reactions involving electrolyte decomposition and Al corrosion, manifesting in poor coulombic efficiency, cycle life, and quick self-discharge. This study aims to investigate the stability of anion-intercalated graphite electrodes in a 4 M solution of lithium bis(fluorosulfonyl)imide (LiFSI) in ethyl methyl carbonate (EMC) using both in-situ and ex-situ Raman spectroscopy. The concentrated electrolyte is essential as it limits parasitic reactions at the cathode-electrolyte interface (CEI) occurring in parallel to anion intercalation. Using electrochemical methods including cyclic voltammetry, and post-mortem electron microscopy it was confirmed that the Al current collector is largely stable at potentials as high as 5.2 V in the electrolyte under consideration; no dissolved Al species were detected using EDX characterization. Results from the cyclic voltammetry study also indicate that parasitic reactions can be mitigated when the cut-off potential is limited to 5.0 V leading to higher coulombic efficiency (CE = 94 %) and more stable discharge capacity (85.17 mAh g-1). However, extending the potential to 5.1 and 5.2 V results in the discharge capacity increasing by almost 20 mAh g-1, though at the expense of the coulombic efficiency, which decreases from 94 to 76 %. Upon raising the cut-off potential to 5.3 V, the CE significantly decreased (20.62 %) as a result of extensive solvent decomposition ultimately leading to much quicker capacity fading.  Based on SEM images taken after 50 cycles, graphite particles did not sustain any structural or morphological change during cycling regardless of the cut-off potentials applied. Further tests were conducted on Li-graphite DIBs using galvanostatic methods in the range from 3 to 5 V, and at different specific currents (20, 50, and 100 mA g-1). Though the cells exhibited good performance in terms of capacity retention, and cycle life at all currents, the coulombic efficiency tended to decrease as the test currents were lowered. This observation confirms the presence of parasitic reactions which are only visible when the experimental timescale is sufficiently long. At 50 and 100 mA g-1, the CE reached > 98 % which further verifies the kinetic aspect of electrolyte decomposition reactions. It is evident that self-discharge sustained in the course of open-circuit potential (OCP) relaxation of the fully charged cell can reveal the stability of the electrolyte and the anion-intercalated graphite. Raman spectroscopy measurements conducted in-situ and ex-situ on graphite electrodes charged and discharged to a series of potential cut-offs reveal the existence of self-discharge leading to extraction of anions from the graphite particles. This was demonstrated through the spectral appearance of E2g2(i) band next to E2g2(b) band at a fully intercalated state, as opposed to the in-situ spectrum, which only showed one intercalated band (E2g2(b)). It can be concluded that concentrated electrolytes (such as 4 M LiFSI in EMC) only provide kinetic stability and are unable to entirely inhibit parasitic reactions at the interface. This further highlights the need for electrolyte additives that can create a more stable interfacial passivation layer on the positive electrode so that more reversible anion intercalation can be attained.

Over time the level of inoculant will decrease due to fading and this needs to be compensated bymore additions of inoculant. When casting CGI400 at Scania a one-step method is used, and the inoculant additions are based on the last ladle from the previous oven. Longer stops in production will result in the previous oven not being representable and more inoculant should be added. A method to establish the inoculation level of the base iron in the oven would make it easier to control the process. This study was about inoculation measurements with the help of thermo- analysis of the melt in the oven to get a better understanding how it fluctuates between ovens and changes over time. The study is also about how to implement this tool into production in a foundry. With the help of thermo- analytic measurements, the inoculation level of the melt was established, and the method was also used to see how the level of inoculant changed over. To establish the inoculation additions needed, the saturation point of inoculant was investigated. The results from the study shows that the starting level of inoculant is too low in the melt it also shows that the inoculation level fluctuates over time. The conclusions that can be drawn from this study is that thermos analysis can be used in production to establish the inoculation level, but moremeasurements needs to be conducted to determine its accuracy. / Över tid så minskar ympnivån i en smälta på grund av fading och detta måste kompenseras med mertillsatser av ymp. Vid gjutning av CGI400 på Scania används en en-stegs metod där man baserar ymptillsatsen i skänken på vad ympnivån var i sista skänken från den tidigare ugnen. Vid längre produktionsstopp kommer inte ympnivån från ugnen innan vara representativ utan mer ymp måste tillsättas. En metod för att bestämma ympnivån av basjärnet i ugnen skulle göra det lättare att styra processen. Denna studie handlar om ympmätning med hjälp av termisk analys i bas smälta för att få en bättre förståelse för hur den fluktuerar mellan ugnar och ändras över tid. Studien handlar också om hur man ska kunna applicera denna mätmetod i produktionen på ett gjuteri. Med hjälp av termo analytiska mätningar skulle grundnivån av ymp i smältan bestämmas och även för att se hur nivån ändrades över tiden. För att bestämma tillsatsen av ymp som borde tillföras undersöktes mättnadspunkten av ymp och hur den kunde tillsättas i skänk. Resultaten från studien visar att grundnivån av ymp är för låg och resultaten visar också att nivån av ymp fluktuerar både upp och ner ökar tid vilket inte var vad man hade förväntat sig och även det finns flera faktorer som kan påverka nivån. Slutsatsen som kan dras är att termisk analys kan användas i produktion för att bestämma ympnivån, men mer mätningar behövs för att bestämmadess noggrannhet.

No description available.

No description available.

No description available.

Tensile strength modelling is usually done to predict the mechanical properties of lamellargraphite iron considering microstructural features. This work attempts to create a simplifiedmodel incorporating cooling rate and solidification time without considering themicrostructural features. This model will save time and cost in industry with the presence of acommercially available software such as Magmasoft which simulates solidification time andcooling rate. A plate model was designed for this purpose as the test geometry to createvariation in solidification time and cooling rate. By altering fraction solid, thermalconductivity, specific heat capacity in Magmasoft, a good fit was created between simulatedcooling curve and experimental cooling curves. The experimental UTS data of samples fromthree moulds were investigated and a regression model was created using statistics toolMinitab. The effect of solidification time and alloying on the graphite length Lmax was studiedfor twelve samples from each mould. Quantification of the effect of cooling rate and alloyingon the pearlitic properties of grey iron like matrix microhardness, pearlite inter-lamellarspacing was also investigated in this work.The developed model has sixty three percent correlation and explains UTS well in terms ofsolidification time and cooling rate. Microhardness measurements show that there is an almostlinear relationship between the cooling rate and microhardness of the matrix structure.Microhardness data also provides an overview of the pearlite fineness/interlamellar spacing.Analysis of the outliers showed that the presence of free ferrite on a fully pearlitic structurereduces the UTS significantly. Comparison of the regression model obtained from this workwith previous work showed that, there is a reduction in the predicted strength with this model.The effort to identify the reason for this reduction was not successful and needs furtherinvestigation. Pearlite inter-lamellar spacing measurement was not accurate. The relationshipbetween pearlite interlamellar spacing and matrix microhardness needs to be investigated inthe future using a better technique for pearlite spacing measurement. This will be useful tounderstand the effect of cooling rate on pearlite spacing and consequently on the UTS of greycast iron. / Draghållfasthetsmodellering görs vanligtvis för att förutsäga de mekaniska egenskaperna av lamellärt grafitjärn. Detta arbete har försökt att skapa en förenklad modell som innehåller kylhastighet och stelningstid utan att överväga mikrostrukturella egenskaper. Modellen kommer att spara tid och kostnad i industrin tillsammans med kommersiellt tillgänglig mjukvara som Magmasoft som simulerar stelningstiden och kylningshastighet. En plattformig modell utformades för detta ändamål som testgeometrin för att skapa variation i stelningstid och kylningshastighet. Genom att ändra fraktion fast fas, termisk konduktivitet och specifik värmekapacitet i Magmasoft skapades en bra anpassning mellan simulerade och experimentella kylkurvor. Experimentella draghållfasthetsdata (UTS) för prover från tre gjutningar undersöktes och en regressionsmodell skapades med hjälp av statistikverktyg Minitab. Effekten av stelningstid och legeringshalt på grafitlängden Lmax studerades för tolv prover från varje form. Effekten av kylhastighet och legering på de perlitiska egenskaperna hos grått järn som matrismikrohårdhet och perlitlamellavstånd undersöktes också i detta arbete.  Den utvecklade modellen har sextiotre procent korrelation och förklarar UTS väl med avseende på stelningstid och kylningshastighet. Mikrohårdhetsmätningar visar att det finns ett nästan linjärt förhållande mellan kylhastigheten och mikrohårdheten hos matrisstrukturen. Mikrohårdhetsdata ger också en översikt över perlitens finhet/interlamellära avstånd. Analys av outliers visade att närvaron av fri ferrit på en fullständigt perlitisk struktur minskar UTS betydligt. Jämförelse av regressionsmodellen erhållen från detta arbete med tidigare arbete visade det att det finns en minskning av den förutsagda styrkan med denna modell. Ansträngningen att identifiera orsaken till denna minskning var inte framgångsrik och behöver ytterligare undersökas. Perlit mellan lamellär avståndsmätning var inte korrekt. Förhållandet mellan perlitens interlamellära avstånd och matrismikrohet måste undersökas i framtiden med hjälp av en bättre teknik för perlit-avståndsmätning. Detta kommer att vara användbart att förstå effekten av kylhastighet på perlitavståndet och följaktligen på UTS av grå gjutjärn

This work investigated a novel vapor chamber for efficient heat spreading and heat removal. A vapor chamber acting as a heat spreader enables for more uniform temperature distribution along the surface of the device being cooled. First, a vapor chamber was studied and compared with the traditional copper heat spreader. The thickness of vapor chamber was kept 1.35 mm which was considered to be ultra-thin vapor chamber. Then, a new geometrical model having graphite foam in vapor space was proposed where the graphite foam material was incorporated in vapor space as square cubes. The effects of incorporating graphite foam in vapor space were compared to the vapor chamber without the embedded graphite foam to investigate the heat transfer performance improvements of vapor chamber by the high thermal conductivity graphite foam. Finally, the effects of various vapor chamber thicknesses were studied through numerical simulations. It was found that thinner vapor chamber (1.35 mm thickness) had better heat transfer performance than thicker vapor chamber (5 mm thickness) because of the extreme high effective thermal conductivities of ultra-thin vapor chamber. Furthermore, the effect of graphite foam on thermal performance improvement was very minor for ultra-thin vapor chamber, but significant for thick vapor chamber. The GF could help reduce the junction temperature by 15-30% in the 5-mm thick vapor chamber. Use of GF embedded vapor chamber could achieve 250-400 Watt per Centimeter square local heat removal for power electronics. The application of this is not only limited to electronic devices but actuator and avionics cooling in aircrafts, thermal management of electronics in directed energy weapon systems, battery thermal management for electric and hybrid vehicles, smart phones cooling, thus covering a wide gamut of heat flux applications.

It is a well known fact that scientific programs spend most of their running time in executing loops operating on arrays. The polyhedral model is a mathematical framework which was designed to optimize this process. Its use in automatic paralyzation of nested loops goes back to the work of Kuck (1978), who showed that the domain of nested loop with affine lower and upper bounds can be described in terms of a polyhedron, and the seminal work of Karp, Miller and Winograd (1967) on scheduling systems of uniform recurrence equations. For a long period of time Graphite (it is part of the GNU Compiler Collection) was relying on CLooG library to produce SSA intermediate representation from the polyhedral model. The Integer Set Library (ISL) recently became available in this project to be used as a back end of CLooG. ISL is nowadays mature enough to replace CLooG with own code generation that sometimes is better. That is why the following goal was chosen: Creation of SSA generator based on Integer Set Library. / Как известно, большая часть времени исполнения программ, связанных с научными расчетами, тратится на выполнение циклов и взаимодействие с массивами. С целью оптимизации этого процесса был разработан математический фреймворк — полиэдральная модель. Первые её применения можно встретить в работах Дэвида Кука и Ричарда Карпа от 1978 и 1967 соответственно. С целью практического использования полиэдральной модели разрабатывались различные фреймоворки и библиотеки, поддерживающие работу с ней. Такими примерами служат библиотеки CLooG и ISL. В отличие от библиотеки CLooG, специализированной для генерации кода из поэлидарльной модели, библиотека ISL позволяет выполнять различные операции с целочисленными множествами и отношениями между ними. Со временем в библиотеке ISL появилась поддержка собственной генерации абстрактного синтаксического дерева. В рамках магистерской работы была поставлена следующая цель: Cоздание генератора промежуточного представления SSA из полиэдральной модели полностью независимого от библиотеки CLooG.