The thermal design constraint of power electronic converters under the specific power loss and heat sink is mainly determined by the maximum permissible junction temperature of the power devices. As the power density and switching frequency increase, transient electro-thermal models become more and more important for the thermal management system design of the power electronic converters. In traction inverters, the junction temperature has huge variation because the fundamental frequency and phase currents vary significantly during the load cycles. Thus, the junction temperature estimation becomes extremely important for the reliability of traction inverters. In this thesis, the transient electro-thermal analysis of a traction inverter considering the inter-dependency of the power losses and junction temperature in an iterative process is implemented. Considering the impact of circuit stray parameters on the switching loss, the temperature dependent power loss model is built based on the datasheet values and the measured switching losses. A state-of-the-art thermal model of the entire inverter including the power modules and the heat sink is developed considering the thermal coupling effects of multiple power devices. By using transient thermal simulation, the linearity of the heat transfer process of the entire traction inverter is verified. The impact of the material thermal properties on the thermal impedance is also presented. In addition, the accuracy of the combination of the thermal subsystem models is verified with simulation. The developed transient electro-thermal model is then used to simulate the junction temperature profiles of the inverter under different operating conditions. Finally, the developed model is experimentally verified. By considering the thermal impedance of the thermal grease layer, the simulation results match with the experimental results very well. The proposed electro-thermal model can provide important information for the thermal management system design, package optimization, long-term reliability analysis, and maximum rating characterization of the traction inverters. / Thesis / Master of Applied Science (MASc)

In this thesis work, temperature estimation algorithms based on physical and radiometric approaches are developed. Developed algorithms, firstly, tested on artificial images for different test cases. Following this, algorithms are tried out on real infrared images in order to verify that they are working properly. Finally, temperature estimations are done by including emissivity. Obtained results are compared to the temperature estimation results of a reference infrared camera. All the results and errors obtained during this study are presented and discussed.

Induction motors are largely used in the industry in our days and the temperature monitoring in these motors is extremely important for their operation. Through this monitoring one can estimate when and how often the maintenance will be performed, adjust the proper protection and identify abnormal operation conditions. In some cases the installation of thermal sensors is not possible due to physical or economical restrictions and the utilization of virtual instrumentation is an option. The following work presents a study for a specific condition where the stator temperature for a generic induction motor in a steady state operation is estimated by using an Artificial Neural Network. The purpose of this study is the evaluation of the applicability of this tool for this specific case and it is the staring point for further studies where the feasibility of a general system can be tested. After simulating the proposed Artificial Neural Network using the backpropagation algorithm it was concluded that this is an applicable method to solve this problem and can be extended to more complex systems. / Os motores de indução são largamente usados em vários setores da sociedade. Entre os setores que mais os utilizam está o setor industrial. Existem vários aspectos a serem considerados na sua operação, entre eles a monitoração da sua temperatura de trabalho. Através desse monitoramento pode-se estimar quando e com que freqüência sua manutenção se faz necessária, condições anormais de funcionamento e também o ajuste de proteções. Em alguns casos a instalação de sensores térmicos não é possível devido a limitações físicas ou econômicas e uma opção é a utilização da instrumentação virtual. O trabalho apresentado faz um estudo de uma dada condição de operação de um motor de indução trifásico cuja temperatura será estimada através de Redes Neurais Artificiais. O objetivo desse estudo é avaliar a aplicabilidade dessa ferramenta para esse caso específico como ponto de partida para estudos futuros, onde a viabilidade da aplicação para sistemas mais generalizados possa ser testada. Após a simulação da Rede Neural Artificial proposta utilizando o algoritmo backpropagation, concluiu-se que é possível aplicar essa metodologia para esse caso e isso indica a possibilidade de aplicá-la em sistemas mais complexos.

Due to the limitations of silicon (Si), there is a paradigm shift in research focusing on wide-bandgap-based (WBG) materials. SiC power semiconductors exhibit superiority in terms of switching speed, higher breakdown electric field, and high working temperature, slowly becoming a global solution in harsh medium-voltage (MV) high-power environments. However, to utilize the SiC MOSFET device to achieve those next-generation, high-density, high-efficiency power electronics converters, one must solve a plethora of challenges. For the MV SiC MOSFET device, a high-performance gate-driver (GD) is a key component required to maximize the beneficial SiC MOSFET characteristics. GD units must overcome associated challenges of electro-magnetic interference (EMI) with regards to common-mode (CM) currents and cross-talk, low driving loop inductance required for fast switching, and device short-circuit (SC) protection. Developed GDs (for 1.2 kV, and 10 kV devices) are able to sustain dv/dt higher than 100 V/ns, have less than 5 nH gate loop inductance, and SC protection, turning off the device within 1.5 us. Even with the introduction of SiC MOSFETs, power devices remain the most reliability-critical component in the converter, due to large junction temperature (Tj) fluctuations causing accelerated wear-out. Real-time (online) measurement of the Tj can help improve long-term reliability by enabling active thermal control, monitoring, and prognostics. An online Tj estimation is accomplished by generating integrated intelligence on the GD level. The developed Tj sensor exhibits a maximum error less than 5 degrees Celsius, having excellent repeatability of 1.2 degrees Celsius. Additionally, degradation monitoring and an aging compensation scheme are discussed, in order to maintain the accuracy of the sensor throughout the device's lifetime. Since ultra high-voltage SiC MOSFET devices (20 kV) are impractical, the modular multilevel converter (MMC) emerged as a prospective topology to achieve MV power conversion. If the kernal part of the power-cell (main constitutive part of the MMC converter) is an SiC MOSFET, the design is able to achieve very high-density and high-efficiency. To ensure a successful operation of the power-cell, a systematic design and assessment methodology (DAM) is explored, based on the 10 kV SiC MOSFET power-cell. It simultaneously addresses challenges of high-voltage insulation, high dv/dt and EMI, component and system protections, as well as thermal management. The developed power-cell achieved high-power density of 11.9 kW/l, with measured peak efficiency of n=99.3 %@10 kHz. It successfully operated at Vdc=6 kV, I=84 A, fsw>5 kHz, Tj<150 degrees Celsius and had high switching speeds over 100 V/ns. Lastly, to achieve high-power density and high-efficiency on the MV converter level, challenges of high-voltage insulation, high-bandwidth control, EMI, and thermal management must be solved. Novel switching cycle control (SCC) and integrated capacitor blocked-transistor (ICBT) control methodologies were developed, overcoming the drawbacks of conventional MMC control. These novel types of control enable extreme reduction in passive component size, increase the efficiency, and can operate in dc/dc, dc/ac, mode, potentially opening the modular converter to applications in which it was not previously used. In order to explore the aforementioned benefits, a modular, scalable, 2-cell per arm, prototype MV converter based on the developed power-cell is constructed. The converter successfully operated at Vdc=12 kV, I=28 A, fsw=10 kHz, with high switching speeds, exhibiting high transient immunity in both SCC and ICBT. / Doctor of Philosophy / In medium-voltage applications, such as an electric grid interface in highly populated areas, a ship dc system, a motor drive, renewable energy, etc., land and space can be very limited and expensive. This requires the attributes of high-density, high-efficiency, and reliable distribution by a power electronics converter, whose central piece is the semiconductor device. With the recent breakthrough of SiC devices, these characteristics are obtainable, due to SiC inherent superiority over conventional Si devices. However, to achieve them, several challenges must be overcome and are tackled by this dissertation. Firstly, as a key component required to maximize the beneficial SiC MOSFET characteristics, it is of utmost importance that the high-performance gate-driver be immune to interference issues caused by fast switching and be able to protect the device against a short-circuit, thus increasing the reliability of the system. Secondly, to prevent accelerated degradation of the semiconductor devices due to high-temperature fluctuations, real-time (online) measurement of the Tj is developed on the gate-driver to help improve long-term reliability. Thirdly, to achieve medium-voltage high-power density, high-efficiency modular power conversion, a converter block (power-cell) is developed that simultaneously addresses the challenges of high-voltage insulation, high interference, component and system protections, and thermal management. Lastly, a full-scale medium-voltage modular converter is developed, exploiting the advantages of the fast commutation speed and high switching frequency offered by SiC, meanwhile exhibiting exceptional power density and efficiency.

Gravity on Earth limits the study of the properties of pure fluids near critical point because they become stratified under their own weight. Near the critical point, all thermodynamic properties either diverge or converge and the heating and cooling cause instabilities of the convective flow as a consequence of the expansibility divergence. In order to study boiling, fluctuation and phase separation processes near the critical point of pure fluids without the influence of the Earth's gravity, a number of experiments were performed in the weightlessness of Mir space station. The experimental setup called ALICE II instrument was designed to suppress sedimentation and buoyancy-driven flow. Another set of experiments were carried out on Earth using a carefully density matched system of deuterated methanolcycloxexane to observe critical fluctuations directly. The set of experiments performed on board of Mir space station studied boiling and wetting film dynamics during evaporation near the critical point of two pure fluids (sulfur hexafluoride and carbon dioxide) using a defocused grid method. The specially designed cell containing the pure fluid was heated and, as a result, a low contrast line appeared on the wetting film that corresponded to a sharp change in the thickness of the film. A large mechanical response was observed in response to the cell heating and we present quantitative results about the receding contact lines. It is found that the vapor recoil force is responsible for the receding contact line. Local density fluctuations were observed by illuminating a cylindrical cell filled with the pure fluid near its liquid- gas critical point and recorded using a microscope and a video recorder. Microscopic fluctuations were analyzed both in sulfur hexafluoride and in a binary mixture of methanol cyclohexane. Using image processing techniques, we were able to estimate the properties of the fluid from the recorded images showing fluctuations of the transmitted and scattered light. We found that the histogram of an image can be fitted to a Gaussian relationship and by determining its width we were able to estimate the position of the critical point. The characteristic length of the fluctuations corresponding to the maximum of the radial average of the power spectrum was also estimated. The power law growth for the early stage of the phase separation was determined for two different temperature quenches in pure fluid and these results are in agreement with other experimental results and computational simulations.

Safe driving is a topic of multiple factors where the road surface condition is one. Knowledge about the road status can for instance indicate whether it is risk for low friction and thereby help increase the safety in traffic. The ambient temperature is an important factor when determining the road surface condition and is therefore in focus. This work evaluates different methods of data fusion to estimate the ambient temperature at road segments. Data from vehicles are used during the temperature estimation process while measurements from weather stations are used for evaluation. Both temporal and spatial dependencies are examined through different models to predict how the temperature will evolve over time. The proposed Kalman filters are able to both interpolate in road segments where many observations are available and to extrapolate to road segments with no or only a few observations. The results show that interpolation leads to an average error of 0.5 degrees during winter when the temperature varies around five degrees day to night. Furthermore, the average error increases to two degrees during springtime when the temperature instead varies about fifteen degrees per day. It is shown that the risk of large estimation error is high when there are no observations from vehicles. As a separate result, it has been noted that the weather stations have a bias compared to the measurements from the cars.

With the advent of technology scaling lifetime reliability is an emerging threat in high-performance and deadline-critical systems. High on-chip thermal gradients accelerates localised thermal elevations (hotspots) which increases the aging rate of the semiconductor devices. As a result, reliable operation of the processors has become a challenging task. Therefore, cost effective schemes for estimating temperature and reliability are crucial. In this work we present a reliability estimation scheme that is based on a light-weight temperature estimation technique that monitors hardware events. Unlike previously pro- posed hardware counter-based approaches, our approach involves a linear-temporal-feedback estimator, taking into account the effects of thermal inertia. The proposed approach shows an average absolute error of We then present a counter-based technique to estimate the thermal accelerated aging factor (TAAF), which is an indicator of lifetime reliability. Results demonstrate that the estimation error is within [−3, +5].

Fordonsindustrin är inne i ett skifte, där förbränningsmotorer (ICE) bör-jar fasas ut och det satsas mer på eldrift med hjälp av batterier (BEV).  I detta arbete, som efterfrågades av Volvo Cars, utvecklades en matema-tisk modell som uppskattar den lägsta temperaturen i en elbils högspän-ningsbatteri. Modellen delades initialt upp i två delar, där den ena delen utgick från hur omgivningstemperaturen samt fordonets hastighet på-verkar eventuell värmeövergång mellan batteri och omgivning. Den andra delen utgick från hur batteriets kylsystem påverkar batteriets temperatur, där kylsystemets temperatur och flöde var de styrande fak-torerna. De styrande variablerna filtrerades med lågpassfilter där olika tidskonstanter användes. Den del som för varje tillfälle visade lägst tem-peratur var gällande. Modellen togs fram med hjälp av mätdata från testkörningar i kallt klimat där testbilens batteri var preparerad med ett antal externa temperatur sensorer, och de preparerade sensorerna an-sågs visa batteriets verkliga temperatur. Tidiga resultat visade att mo-dellen för omgivningstemperatur och hastighet visade relativt bra vär-den men saknade inflytande från kylsystemets dynamik medan mo-dellen för kylsystemet fungerade mindre bra i de flesta fall. Modellen modifierades därför till att ha hastighet och omgivningstemperatur som basfunktion och kylsystemets påverkan adderades. Resultatet visade att den senare modellen hanterade systemets dynamik bättre än tidigare modell. I framtida studier bör batteriets termiska massa samt dess egen-uppvärmning tas i beaktning. / The automotive industry is shifting from internal combustion engines (ICE) cars to battery electric vehicles (BEV). In this work, requested by Volvo Cars, a mathematical model was developed that estimates the lowest temperature of an electric car's high-voltage battery. The model was initially divided into two parts, where one part was based on how the ambient temperature and the vehicle's speed affect any heat transfer between battery and surroundings. While the second part was based on how the battery's cooling system affects the battery temperature, where the cooling system temperature and flow were the controlling factors. The controlling variables were filtered with low-pass filters where dif-ferent time constants were used. The part that showed the lowest tem-perature for each occasion was valid. The model was developed using measurement data from test runs in cold climates where the test car's battery was prepared with more temperature sensors than are in a pro-duction battery and the prepared sensors were considered to show the battery's actual temperature. Early results showed that the ambient tem-perature and speed model showed relatively good values but lacked in-fluence from the cooling system, while the model for the cooling system worked less well in most cases. The model was therefore modified to have speed and ambient temperature as the base function and the cool-ing system's impact was added. The results showed that the latter model handled the dynamics of the system better than the previous model. In future studies, the thermal mass of the battery and its self-heating should be considered.

No description available.

The objective of the dissertation is to develop simplified analytical models for typical linear-motion and rotary-motion energy-conversion systems under active DC excitation without tedious numerical-simulation effort, and provide practical implementation of the models in optimal-design and thermal-protection aspects. The model of a vacuum automatic circuit recloser (a typical linear-motion system under DC excitation) is first developed in the form of a non-linear discontinuous eighth-order dynamic system. The model is then used to simulate the transient mechanical and electromagnetic performance during the opening and closing movements of the recloser. Such a model is not found in the literature. Although the model is based on certain simplifying assumptions, the result is validated by high-speed-camera measurements. In addition, the impact of key design variables is explored, based on which an improved recloser design is proposed, and helps to optimize capital and production costs without degrading performance. Further analytical investigation is carried out in modeling an inverter-fed induction motor (IM) (a typical rotary-motion system) with active DC injection. The IM is closed-loop controlled via two popular motor-control algorithms, namely, the direct-torque-control (DTC) algorithm and field-oriented-control (FOC) algorithm. Quantitative relationships between the changes of various machine variables during the active DC excitation are provided in the theoretical analysis. The developed DC-injection model is further simplified for practical implementation. The developed IM model under DC injection results in practical ways to excite a proper amount of DC current directly or indirectly into IM stator windings via different closed-loop motor-control algorithms. In a DTC motor-drive system, the modeling work makes it possible to excite the DC current indirectly inside the motor by superimposing a stator-flux-linkage-bias command in the flux-control loop or a torque-ripple command in the torque-control loop. The proposed flux-linkage-injection and torque-injection methods are the first novel efforts to implement the DC-signal-injection method in a DTC motor-drive system. In addition, the analysis carried out in a standard FOC drive system brings about an improved DC-current-injection approach: the torque ripple in this method is significantly mitigated compared to all existing DC-injection methods in FOC systems. The proposed DC-injection methods, either in a DTC or an FOC system, lead to a simple, low-cost, accurate, and non-invasive thermal-monitoring scheme for closed-loop-controlled IMs, where the stator temperature is indirectly estimated from stator resistance. Furthermore, considering inverter non-idealities, there is a challenge for a typical inverter drive to accurately estimate the DC component of motor terminal voltages. The existing methods are extended to provide a complete study of the real-time signal-processing technique for both DTC and FOC algorithms, and are finally implemented in a custom-built programmable motor-drive system. The experimental results demonstrate that the proposed technique gives accurate and robust stator-temperature estimation, regardless of the operating conditions and cooling modes. The analytical modeling method for the linear-motion and rotary-motion energy-conversion systems can be further extended to other power devices with similar mechanisms, and implemented in optimal design, control, and thermal-protection areas.