No description available.

The demand for more power and performance from electronic equipment has constantly been growing resulting in an increased amount of heat dissipation from these devices. Thermal management of high-density power control units for hybrid electric vehicles is one such application. Over the last few years, the performance of this power control unit has been improved and size has been reduced to attain higher efficiency and performance causing the heat dissipation as well as heat density to increase significantly. However, the overall cooling system has remained unchanged and only the heat exchanger corresponding to the power control unit (PCU) has been improved. This has allowed the manufacturing costs to go down. Efforts are constantly being made to reduce the PCU size even further and also to reduce manufacturing costs. As a consequence, heat density will go up (~ 200 – 250 W/cm2) and thus, a better high performance cooler/heat exchanger is required that can operate under the existing cooling system design and at the same time, maintain active devices temperature within optimum range (<120 – 125 °C) for higher reliability. The aim of this dissertation was to study the various cooling options based on jet impingement, mini-channel, ribbed mini-channel, phase change material and double sided cooling configurations for application in hybrid electric vehicle and other similar consumer products and perform parametric and optimization study on selected designs. Detailed experimental and computational analysis was performed on different cooling designs to evaluate overall performance. Severe constraints such as choice of coolant, coolant flow-rate, pressure drop, minimum geometrical size and operating temperature were required for the overall design. High performance jet impingement based cooler design with incorporated fin-like structures induced swirl and provided enhanced local heat transfer compared to traditional cooling designs. However, the cooling scheme could manage only 97.4% of the target effectiveness. Tapered/nozzle-shaped jets based designs showed promising results (~40% reduction in overall pressure drop) but were not sufficient to meet the overall operating temperature requirement. Various schemes of mini-channel arrangement, which were based on utilizing conduction and convection heat transfer in a conjugate mode, demonstrated improved performance over that of impingement cooling schemes. Impingement and mini-channel based designs were combined to show high heat transfer rates but at the expense of higher pressure drops (~5 times). As an alternate, mini-channel based coolers with ~1.5 mm size channels having trip strips or ribs were studied to accommodate the design constraints and to enhance local as well as overall heat transfer rates and achieve the target operating temperature. A step by step approach to the development of the heat exchanger is provided with an emphasis on system level design. The computational based optimization methodology is confirmed by a fabricated test bed to evaluate overall performance and compare the predicted results with actual performance. Additionally, one of the impingement based configuration (Swirl-Impingement-Fin) developed during the course of this work was applied to the internal cooling of a turbine blade trailing edge and was shown to enhance the thermal performance by at least a factor of 2 in comparison to the existing pin-fin technology for the conditions studied in this work. / Ph. D.

<p dir="ltr">Current industrial processes are energy and carbon emission intensive. Amidst the growing demand for decarbonization, it is critical to utilize alternate sources of energy and innovative technologies that could improve efficiency and reduce power consumption. In this context, twin-screw compressors are used extensively in commercial and industrial applications. Profile optimization and capacity modulation solutions (e.g., slide valves, variable-speed, etc.) are continuously investigated to improve the performance and operation of the compressors. This study focuses on an exploratory investigation of an additively manufactured twin-screw compressor with internal cooling channels to achieve a near isothermal compression process by evaluating both the potential compressor performance improvement and the structural integrity by means of rotordynamics and fatigue analyses.</p><p dir="ltr">To predict the compressor performance, complex coupling between compression process and heat transfer during the operation of the compressor must be investigated. The interactions between solid (i.e., rotors) and fluid phases (i.e., air and coolant) were modeled using a transient 3D CFD model with conjugated heat transfer (CHT). The CFD model predicted compressor performance parameters such as isentropic efficiency, heat transfer rate, work input and compression forces on the rotors. The performance of the twin-screw compressor with internal cooling channels has been compared with a conventional twin-screw compressor for which experimental data was available. Further investigations have been conducted at different operating conditions, including various pressure ratios, rotational speeds, and mass flow rates to improve the compressor efficiency. The results of the CFD model were used to quantify compression loads, assess the characteristics of the heat transfer processes, and optimize the internal flow through the cooling channels. As the rotors can be affected by stress accumulation and deformations due to their hollowness and reduced wall thickness over time, this study also established a detailed rotordynamic simulation model and a fatigue model using the actual compression forces obtained from previous CFD studies. Both hollow and solid rotors have been analyzed and compared. The bearing loads have been verified against Campbell diagrams whereas the fatigue results have been compared with experimental testing. With the validated model, the hollow rotor compressor durability was analyzed and compared with the conventional rotors. Lastly, a general mechanistic model to better understand bearing loads and frictional losses in a twin-screw compressor is also established and studied.</p><p dir="ltr">The CHT study concluded that the hollow rotor with single-phase internal cooling yielded to an increase in isentropic efficiency of 1% for the higher pressure ratio and 2% for lower pressure ratio at 19,000 RPM. More importantly, the hollow rotors also showed a decrease of 40 K and 20 K in discharge temperatures for the two operating conditions respectively, thereby arriving closer to isothermal conditions and reducing the thermal stresses on the rotors. The rotordynamic study revealed that the male rotor would endure highest amount of von Misses stress reaching up to 338 MPa for the pressure ratio of 3.29 bar and 19,000 RPM. Because of this, a maximum fatigue factor of safety of 5 occurs on the male rotor. From the analyses, the rotors were deemed to be safe and optimized for the designed operating conditions and proof of concept rotors were additively manufacturers with an Inconel alloy through Direct Metal Laser Sintering.</p>

<p dir="ltr">The thermal efficiency of gas turbines improves with higher turbine inlet temperatures (TIT) or compressor outlet pressure. Nowadays, gas turbines achieve TITs up to 1600 °C for power generation and 2000 °C for aircraft. These temperatures far exceed the limits where structural integrity can be maintained. For Ni-based superalloys with thermal barrier coatings, that limit is about 1200 °C. Gas turbines can operate at these high temperatures because all parts of the turbine component that contact the hot gases are cooled so that material temperatures never exceed those limits. </p><p dir="ltr">Gas-turbine vanes and blades are cooled by internal and film cooling with the cooling air extracted from the compressor. Since the extracted air could be used to generate power or thrust, the amount of cooling air used must be minimized. Thus, numerous researchers have investigated fluid flow and heat transfer in internal and film cooling to enable effective cooling with less cooling flow. For internal cooling, significant knowledge gaps persist, notably in ducts with varying cross sections. Reviews of existing literature indicate a lack of studies on flow and heat transfer in cooling ducts that account for the taper in the blade geometry from root to tip for both power-generation and aircraft gas turbines.</p><p dir="ltr">This study investigates the flow and heat transfer in ribbed and smooth tapered U-ducts, under conditions relevant to turbine cooling by using computational fluid dynamics (CFD) and a reduced-order model (ROM) developed in this study. The CFD analysis was based on steady Reynolds-Averaged Navier-Stokes (RANS) equations with the Shear Stress Transport (SST) turbulence model. The CFD analysis examined the effects of rotation number (Ro = 0, 0.0219, 0.0336, 0.0731), Reynolds number (Re = 46,000, 100,000, 154,000), and taper angle (α = 0°, 1.41°) under conditions that are relevant to electric-power-generation gas turbines. CFD results obtained showed increasing the taper angle significantly increases both the friction coefficient and the Nusselt number, regardless of rotation. With rotation at Ro = 0.0336 and Re = 100,000, the maximum increase in the average friction coefficient and Nusselt number due to taper was found to be 41.7% and 36.6% respectively. Without rotation at Re = 46,000, those increases were 11.5% and 14.7% respectively. </p><p dir="ltr">The ROM was derived from the integral continuity, momentum, and energy equations for a thermally and calorically perfect gas to provide rapid assessments of radially outward flow in tapered ducts subjected to constant heat flux. The ROM was used to study the effects of taper angle (α = 0°, 1.5°, 3.0°), ratio of mean radius to hydraulic diameter (Rm/Dh = 45, 150), rotation number (Ro = 0, 0.025, 0.25), Reynolds number (Re = 37,000, 154,000), and thermal loadings (q" = 5×104, 105 W/m2) on the mean density, velocity, temperature, and pressure along the duct. The parameters studied are relevant to both electric-power-generation and aircraft gas turbines. Results obtained show density and pressure variations to be most affected by the rotation number, while velocity along the duct is most affected by the duct’s taper angle. Additionally, it was found that if the taper angle is sufficiently large (α = 3°), then the temperature could reduce along the duct despite being heated because the thermal energy is converted to mechanical energy. When compared to a duct without taper, the mass flow rate of the cooling air could be reduced by up to 44% to achieve the same temperature distribution of the cooling flow along the duct.</p><p dir="ltr">The ROM developed was assessed by comparing against grid-converged CFD results for both ribbed and smooth sections of the duct. The validation study showed the maximum relative errors for density, velocity, temperature, and pressure distributions to be 0.6%, 3.3%, 0.4%, and 0.3% for smooth sections, and 3.2%, 5.6%, 0.9%, and 3.0% for ribbed sections, respectively. Thus, the ROM developed has accuracy comparable to CFD based on steady RANS but is order of magnitude more efficient computationally, making it a valuable tool for preliminary design. </p><p><br></p>

Ce travail de thèse, réalisé dans le cadre d’une convention CIFRE entre TURBOMECA et le CERFACS et en partenariat avec l’IVK, se place dans un contexte d’amélioration des performances des turbines axiales équipant les turboréacteurs d’hélicoptère. Un des points critiques du dimensionnement de tels moteurs est la maitrise de la durée de vie des pales de la turbine haute pression qui font face à de très hautes températures provenant de la chambre de combustion. Les prédictions numériques de l’environnement aérothermique des pales (écoulements dans la veine et système de refroidissement) sont réalisées aujourd’hui dans le milieu industriel à l’aide de la modélisation Reynolds Averaged Navier-Stokes (RANS). Grâce à des capacités de calculs grandissantes, l’approche Simulation aux Grandes Echelles (SGE) offre désormais un nouveau potentiel de prédictions d’écoulements. Les travaux de cette thèse s’intéressent ainsi à la capacité de la SGE à prédire l’écoulement du circuit de refroidissement interne d’une pale de turbine. Pour simplifier l’analyse de ce problème ou plusieurs phénomènes physiques sont en jeu, une progression en trois parties est proposée. La première s’intéresse à l’étude aérothermique de géométries simplifiées de canaux de refroidissement (coude à 180° et canal avec promoteurs de turbulence) en configuration statique. Aux régimes d’écoulement considérés, une approche résolue en paroi avec maillage non-structuré hybride est proposée et validée en vue d’une application industrielle facilitée. La seconde partie étend l’analyse de l’écoulement à un cas de canal avec promoteurs de turbulence en rotation utilisant une méthode de résolution numérique dans un repère absolu. Les investigations des résultats de la SGE fournissent des prédictions moyennes et instationnaires en bon accord avec les expériences disponibles et les travaux précédents aussi bien pour la dynamique de l’écoulement que les transferts de chaleur. Enfin, une troisième partie présente une application de la méthode sur un cas de pale réelle avec couplage thermique entre le circuit de refroidissement et le solide de la pale. Cette dernière partie classée confidentielle n’est pas présente dans le manuscrit disponible publiquement. Les résultats de l’approche résolue en paroi et de la rotation dans le repère absolu comparés aux résultats RANS disponibles pour le cas applicatif montrent d’importante différences locales et ainsi le potentiel de la méthode proposée. / This PhD dissertation, conducted as part of a CIFRE research project between TURBOMECA and CERFACS in partnership with the VKI, deals with improving performance of axial turbines from helicopter engines. One of the most critical design points of such engines is the control of the high pressure turbine blade lifetime which face the high temperatures from the combustor. Today, industrial numerical aerothermal predictions of the flows around the blade (in the vein and in its cooling system) are performed with the Reynolds Averaged Navier-Stokes (RANS). Thanks to the increasing computational power, Large Eddy Simulation (LES) becomes affordable to offer further flow predictions. Therefore, this thesis focuses on the capabilities of the LES to estimate the flow in turbine blade internal cooling channels. To simplify this analysis where several physical phenomenon are present, the problem is described in three parts with increasing complexity. The first part addresses simplified typical geometries of cooling channel (U-bend and ribbed channel) in a static configuration. Considering the flow regime, a wall-resolved approach using a hybrid unstructured mesh is proposed in view of the application on an industrial case. The second part extends the study of the ribbed channel in rotation using an inertial reference frame. LES provides mean and unsteady results in good agreement with the available experimental data and previous works, for the flow dynamic and the heat transfer. Finally, the third part presents the application of the method to an industrial case with conjugate heat transfer between a complex cooling channel and the blade. This last section is not present in the public manuscrit for confidential reasons. Results of the use of the wall-resolved approach in rotation in an inertial frame of reference are compared to RANS predictions and show the potential of the method with high local differences.

Tato práce je zaměřena na systém vodního vstřikování a způsoby zajištění dostatečného množství potřebné kapaliny pro správný chod tohoto systému. Jedním zajímavým a nezávislým řešením je využití odpadních produktů motoru, a to vodní páry produkované spalováním paliva. Pro zajištění kondenzace je nutné zchlazení výfukových plynů pod rosný bod páry, což lze zajistit opatřením výfukového potrubí chladiči. Vzniklý kondenzát je pak potřeba efektivně oddělit od proudu plynů a zachytit. Návrh takovéhoto systému vychází z reálně naměřených dat a jeho následným ověřením pomocí CFD simulace.