Ionic conductivity of metal oxides is critical for the function of a broad range of different components, such as electrolytes in solid oxide fuel cells and alloys designed for high temperature applications. In both cases the ionic conductivity can be studied by in situ impedance spectroscopy, which is also able to reveal information on the dielectric properties of the metal oxides, and in some cases the influence of their microstructure. The focus of this thesis is on impedance spectroscopy measurements of α-alumina in the temperature range 400-1000 °C. This metal oxide has found extensive use as the protective scale on heat resistant alloys. Some unpublished work on oxygen ion conductivity of yttria-stabilized zirconia is also included. The low electrical conductivity of α-alumina can be a source for errors and misinterpretations during impedance spectroscopy measurements. A major disturbance originates from leakage currents that appear in the experimental setup. These leakage currents are due to conduction through the gas phase around the sample, conduction on the sample surface, or poor insulation in the sample holder. It was shown that below 700 °C, conduction on the sample surface could severely distort the measurement. The magnitude of the distortions appeared to be sensitive to the type of electrodes used. The use of a so-called guard electrode was shown to effectively block the surface conduction in the measurements. Conductivity of metal oxides is known to be dependent on their microstructure. Generally it is believed that ionic conductivity is favoured along grain boundaries and dislocations. The influence of microstructure on conductivity was studied for α-alumina in the temperature range 400-1000 °C. The conductivity of a series of highly pure and dense samples with narrow grain size distributions was measured by impedance spectroscopy. It appeared that the activation energy for conduction increased with decreasing grain size. Results based purely on impendence spectroscopy have some inherently weaknesses. For instance no information on the nature of the charge carrier can be found. Therefore the charge transport in single crystalline α-alumina was simulated by the molecular dynamics method. The results from the simulation were then compared to results from impedance measurements on single crystalline α-alumina. From the simulation it turned out that diffusion of aluminium ions had lower activation energy than diffusion of oxygen. The activation energy of oxygen was close to the measured activation energy, and the mobility of oxygen was higher than for aluminium. Therefore the dominating charge carrier was suggested to be oxygen ions.

We study fundamental graph problems under parallel computing models. In particular, we consider two parallel computing models: Parallel Random Access Machine (PRAM) and Massively Parallel Computation (MPC). The PRAM model is a classic model of parallel computation. The efficiency of a PRAM algorithm is measured by its parallel time and the number of processors needed to achieve the parallel time. The MPC model is an abstraction of modern massive parallel computing systems such as MapReduce, Hadoop and Spark. The MPC model captures well coarse-grained computation on large data --- data is distributed to processors, each of which has a sublinear (in the input data) amount of local memory and we alternate between rounds of computation and rounds of communication, where each machine can communicate an amount of data as large as the size of its memory. We usually desire fully scalable MPC algorithms, i.e., algorithms that can work for any local memory size. The efficiency of a fully scalable MPC algorithm is measured by its parallel time and the total space usage (the local memory size times the number of machines). Consider an 𝑛-vertex 𝑚-edge undirected graph 𝐺 (either weighted or unweighted) with diameter 𝐷 (the largest diameter of its connected components). Let 𝑁=𝑚+𝑛 denote the size of 𝐺. We present a series of efficient (randomized) parallel graph algorithms with theoretical guarantees. Several results are listed as follows: 1) Fully scalable MPC algorithms for graph connectivity and spanning forest using 𝑂(𝑁) total space and 𝑂(log 𝐷loglog_{𝑁/𝑛} 𝑛) parallel time. 2) Fully scalable MPC algorithms for 2-edge and 2-vertex connectivity using 𝑂(𝑁) total space where 2-edge connectivity algorithm needs 𝑂(log 𝐷loglog_{𝑁/𝑛} 𝑛) parallel time, and 2-vertex connectivity algorithm needs 𝑂(log 𝐷⸱log²log_{𝑁/𝑛} n+\log D'⸱loglog_{𝑁/𝑛} 𝑛) parallel time. Here 𝐷' denotes the bi-diameter of 𝐺. 3) PRAM algorithms for graph connectivity and spanning forest using 𝑂(𝑁) processors and 𝑂(log 𝐷loglog_{𝑁/𝑛} 𝑛) parallel time. 4) PRAM algorithms for (1 + 𝜖)-approximate shortest path and (1 + 𝜖)-approximate uncapacitated minimum cost flow using 𝑂(𝑁) processors and poly(log 𝑛) parallel time. These algorithms are built on a series of new graph algorithmic primitives which may be of independent interests.

No description available.

Oxid uraničitý (UO2) je v současnosti nejčastěji používaným materiálem jakožto palivo v komerčních jaderných reaktorech. Největší nevýhodou UO2 je jeho velmi nízká tepelná vodivost, a protože se při štěpení UO2 v jaderném reaktoru vytváří velké množství tepla, vzniká v UO2 peletě velký teplotní gradient. Tento teplotní gradient způsobuje vznik velkého tepelného napětí uvnitř pelety, což následně vede k tvorbě trhlin. Tyto trhliny napomáhají k šíření štěpných plynů při vysoké míře vyhoření paliva. Tvorba trhlin a zvýšený vývin štěpného plynu posléze vede ke značnému snížení odolnosti jaderného paliva. Tato práce se zabývá problematikou zvyšování tepelné vodivosti jaderného paliva na modelu materiálu (CeO2). V této práci jsou studovány podobnosti chování CeO2 a UO2 při konvenčním slinováním a při „spark plasma sintering.“ Způsob jak zvýšit tepelnou vodivost použitý v této práci je včlenění vysoce tepelně vodivého materiálu, karbidu křemíku (SiC), do struktury CeO2 pelet. Od karbidu křemíku je očekáváno, že zvýší tok tepla z jádra pelety, a tím zvýší tepelnou vodivost CeO2. V této práci je také porovnávána podobnost chování SiC v CeO2 matrici s chováním SiC v UO2, které bylo popsáno v literatuře.

No description available.

No description available.

This licentiate thesis concerns one dimensional flow simulation of turbocharged spark ignited engines. The objective has been to contribute to the improvement of turbocharged SI engines’ performance as well as 1 D simulation capabilities. Turbocharged engines suffer from poor gas exchange due to the high exhaust pressure created by the turbine. This results in power loss as well as high levels of residual gas, which makes the engine more prone to knock. This thesis presents an alternative gas exchange concept, with the aim of removing the high exhaust pressure during the critical periods. This is done by splitting the two exhaust ports into two separate exhaust manifolds. The alternative gas exchange study was performed by measurements as well as 1-D simulations. The link between measurements and simulations is very strong, and will be discussed in this thesis. As mentioned, turbocharged engines are prone to knock. Hence, finding a method to model knock in 1-D engine simulations would improve the simulation capabilities. In this thesis a 0-D knock model, coupled to the 1-D engine model, is presented / QC 20101112

A pulsed Nd:YAG laser is used to study laser spark ignition of methane counter-flow diffusion flames with the use of helium and argon as diluents to achieve a wide range of variations in transport properties. The global strain rate and Damkohler number on successful ignition were investigated for the effects of Lewis number and transport properties, which are dependent on the diluent type and dilution level. A high-speed camera is used to record the ignition events and a software is used for pre-ignition flow field and mixing calculations. It is found that the role of effective Lewis number on the critical global strain rate, beyond which ignition is not possible, is qualitatively similar that on the extinction strain rate. With the same level of dilution, the inert diluent with smaller Lewis number yields larger critical global strain rate. The critical Damkohler number below which no ignition is possible is found to be within approximately 20% for all the fuel-inert gas mixtures studied. When successful ignition takes place, the ignition time increases as the level of dilution of argon is increased. The ignition time decreases with increasing level of helium dilution due to decreases in thermal diffusion time, which causes rapid cooling of the flammable layer during the ignition process. However, the critical strain for ignition with helium dilution rapidly decreases as the dilution level is increased. The experimental results show that with the increase of strain rate the time to steady flame decreases, and that with the increase of dilution level time for the flame to become steady increases. For the same level of dilution, the time for steady flame is observed to be longer for He-diluted flames than for Ar-diluted flames due to its thermal diffusivity being larger than that of Ar.

<p>A novel two-loop estimation and control strategy is proposed to reduce the natural gas (NG) spark-ignition (SI) engine tail pipe emissions, with focus on the outer loop development. In the outer loop, an fractional oxidation state (FOS) estimator consisting of a three-way catalyst (TWC) model and an extended Kalman-filter is used to estimate the real-time TWC's FOS, and a robust controller is used to control the first-half TWC's FOS by manipulating the desired engine lambda (i.e., air–fuel equivalence ratio; lambda=1 at stoichiometry). The outer loop estimator and controller are combined with an industry-production baseline inner loop controller, which controls the engine $\lambda$ based on the desired lambda value. This novel two-loop control strategy reduces more CH4 and NOx emissions over no-outer-loop control strategy and the conventional two-loop control strategies through simulation. </p> <p><br></p> <p>Engine with and without fuel cut-off are both investigated. Although fuel cut-off brings better fuel economy, it also over-oxidizes the TWC during fuel cut events, which makes the FOS-based controller's competence in NOx reduction over non-FOS-based controllers less significant. By comparing simulation results with and without fuel cut-off, it shows huge potential for much better emission result if fuel cut-off's side effect can be alleviated. Considering that fuel cut-off generally being cutting engine fueling during zero load periods and introducing unreacted oxygen into the after-treatment system, the best way of dealing with the issue is to cut off or reduce the oxygen input to the TWC during those events. Several advanced engine technologies such as cylinder deactivation and exhaust gas re-circulation are good candidates to approach this issue. </p> <p><br></p> <p>An industry-production Cummins B6.7N natural gas SI engine was installed in the Ray W. Herrick Laboratories for study of variable valve actuation (VVA) technology, for the purpose of evaluating/improving SI engine's fuel efficiency, emission reduction, and engine knock resistance. A one-dimensional, physics-based natural gas SI engine model was investigated and calibrated in GT-Power software. To calculate the burn rates in the cylinder, three different pressure analysis methods were investigated and implemented. It is observed that all six cylinders' pressure curves are different, which in turn render different burn rates cylinder-to-cylinder. Cylinder with a higher peak cylinder pressure has a faster burn rate. Each operating condition has its unique pressure curve, and their burn rates are different under different operating conditions. Considering that the burn rate profile can vary cylinder-to-cylinder and operation-to-operation, to make the GT combustion model work for a larger range of loads, a fixed burn rate model may help in the preliminary research phase, but a predictive combustion model is more preferable.</p> <p><br></p> <p>The GT-Power model's VVA capability is investigated, where intake valve closure (IVC) modulation and cylinder de-activation (CDA) are built and analyzed. To mitigate TWC's over-oxidation issue during engine's fuel cut-off events, the CDA is implemented and simulated to demonstrate its benefit on further emission and fuel consumption reductions.</p>

There is a pressing need for creative new data analysis methods whichcan sift through scientific simulation data and produce meaningfulresults. The types of analyses and the amount of data handled by currentmethods are still quite restricted, and new methods could providescientists with a large productivity boost. New methods could be simpleto develop in big data processing systems such as Apache Spark, which isdesigned to process many input files in parallel while treating themlogically as one large dataset. This distributed model, combined withthe large number of analysis libraries created for the platform, makesSpark ideal for processing simulation output.Unfortunately, the filesystem becomes a major bottleneck in any workflowthat uses Spark in such a fashion. Faster transports are notintrinsically supported by Spark, and its interface almost denies thepossibility of maintainable third-party extensions. By leveraging thesemantics of Scala and Spark's recent scheduler upgrades, we forceco-location of Spark executors with simulation processes and enable fastlocal inter-process communication through shared memory. This provides apath for bulk data transfer into the Java Virtual Machine, removing thecurrent Spark ingestion bottleneck.Besides showing that our system makes this transfer feasible, we alsodemonstrate a proof-of-concept system integrating traditional HPC codeswith bleeding-edge analytics libraries. This provides scientists withguidance on how to apply our libraries to gain a new and powerful toolfor developing new analysis techniques in large scientific simulationpipelines.