Дисертації з теми "Transient heat flux"

Doutoramento em Engenharia Mecânica
The development of a new instrument for the measurement of convective and radiative is proposed, based on the transient operation of a transpiration radiometer. Current transpiration radiometers rely on steady state temperature measurements in a porous element crossed by a know gas mass flow. As a consequence of the porous sensing element’s intrinsically high thermal inertia, the instrument’s time constant is in the order of several seconds. The proposed instrument preserves established advantages of transpiration radiometers while incorporating additional features that broaden its applicability range. The most important developments are a significant reduction of the instrument’s response time and the possibility of separating and measuring the convective and radiative components of the heat flux. These objectives are achieved through the analysis of the instrument’s transient response, a pulsed gas flow being used to induce the transient behavior.
Propõe-se o desenvolvimento de um novo instrumento para medição de fluxos de calor convectivos e radiativos, baseado na operação de um radiómetro de transpiração em regime transitório. Os radiómetros de transpiração atuais baseiam-se em medições de temperatura em regime estacionário num elemento poroso atravessado por um caudal mássico gasoso conhecido. Como consequência da inércia térmica intrinsecamente elevada do elemento sensível poroso, a constante de tempo do instrumento é da ordem dos segundos. O instrumento proposto preservará as vantagens estabelecidas dos radiómetros de transpiração incorporando características adicionais que alargarão a gama de aplicabilidade. As novas características mais importantes serão uma redução significativa do tempo de resposta do instrumento e a possibilidade de medir separadamente as componentes radiativa e convectiva do fluxo de calor. Estes objetivos serão conseguidos através da análise da resposta transitória do instrumento, utilizando-se um caudal pulsado de gás para induzir o comportamento transitório.

This project investigated if ignition could be mathematically predicted when a material is subjected to a transient heat flux. Six timbers commonly used in New Zealand for construction and indoor furnishing timbers were tested in a cone calorimeter at the University of Canterbury. The experiments were run at 50, 35, 20 and 15 kW/m2 incident heat flux. The sample surface temperature and heat release data was collected for each test. From the ignition time data a value for thermal inertia was calculated and using specific heat data from the literature the thermal properties of each material was inserted into a One Dimensional Heat Transfer Model. A second series of tests were conducted on each of the materials tested at constant flux. These new tests involved subjecting the sample to a transient heat flux based on t² fire growth curves. Again surface temperature and heat release data was obtained from the tests. The one dimensional heat transfer model was used to attempt to predict the surface temperature profile and the ignition time when the test conditions were entered into it. It was found that the predicted surface temperature profile generally matched the shape of the measured temperature profile. However the model was unsuccessful in accurately predicting the ignition time in either the constant or transient flux conditions. It is considered that accurate values for the thermal conductivity and the specific heat would be required before the ignition time and temperature profile could be accurately modelled.

A method of performing local transient heat flux measurements in an uncooled axisymmetric hybrid rocket nozzle is presented. Surface temperatures are collected at various axial locations during short duration tests and post processed using finite difference techniques to determine local transient heat fluxes and film coefficients. Comparisons are made between the collected data and the complete Bartz model. Although strong agreement is observed in certain sections of the nozzle, ideal steady state conditions are not observed to entirely validate the Bartz model for hybrid rocket nozzles. An experimental error analysis indicates the experimental heat fluxes are accurate within ±5.2% and supports the accuracy of the results.

The effects of using an insert Heat Flux Microsensor (HFM) versus an HFM deposited directly on a turbine blade to measure heat flux in a transonic cascade are investigated. The HFM is a thin-film sensor, 6.35 mm (0.250") in diameter (for an insert gage, including the housing) which measures heat flux and surface temperature. The thermal time response of both gages was modeled using a 1-D, finite difference technique and a 2-D, finite element solver. The transient response of the directly deposited gage was also tested against insert gages using an unsteady shock wave in a bench test setup and using a laser of known output. The effects of physical gage offset from the blade surface were also investigated. The physical offset of an insert HFM near the stagnation point on the suction side of a turbine blade was intentionally varied and the average heat transfer coefficient measured. Turbulence grids were used to study how offset affects the heat transfer coefficient with freestream turbulence added to the flow. The time constant of the directly deposited gage was measured to be 856 ms compared to less than 30 ms for the insert gages. Model results predict less than 20 ms for both gages and rule out the anodization layer (used for electrical isolation of the directly deposited gage from the blade) as the cause for the directly deposited gage's much slower time response. Offsets of ± 0.254 mm (0.010") at the gage location with an estimated boundary layer thickness of 0.10 mm (0.004") produced a higher average heat transfer coefficient than the 0.000" offset case. Using an insert HFM resulted in a higher average heat transfer coefficient than using the directly deposited gage and reduced the effects of freestream turbulence. To accurately measure heat transfer coefficients and the effects of freestream turbulence, the disruption of the flow caused by a gage must be minimized. Depositing a gage directly on the blade minimizes the effects of offset, but the cause of the slow time response must first be resolved if high speed data is to be taken.
Master of Science

Thesis (M.S.) -- University of Maryland, College Park, 2006.
Thesis research directed by: Dept. of Fire Protection Engineering. Title from t.p. of PDF. Includes bibliographical references. Published by UMI Dissertation Services, Ann Arbor, Mich. Also available in paper.

A multi-region, transient concrete ablation and decomposition model is developed. The model consists of four regions of concrete containing a thermally affected region, a dry (evaporated and chemically dehydrated) region, and a gas-free (decarboxylated) region with ablated concrete at the melt/concrete interface. Each region has an interface where the latent heat of local decomposition reactions is taken into account as heat sinks due to endothermic characteristics of the reactions. The time dependent temperature profiles, and depth and growth rate of the regions are evaluated by use of the heat balance integral method. Solutions are obtained for surface heat fluxes in forms of constant, e ⁻(λ)ᵗ, t⁻(λ) and -At to analyze various melt cooldown schemes. The erosion front progresses with a constant rate proportional to the surface heat flux in case of constant heat flux, and terminates at a finite erosion depth that is logarithmically proportional to the cooldown rate for surface heat flux in forms of ⁻(λ)ᵗ and t⁻(λ). Sensitivity analyses are performed to investigate the effects of important thermophysical parameters. Larger erosion depth and rate is observed for higher thermal conductivity. Decomposition temperatures are found to be significant in ablation. Model results were compared with previous experiments and models, and determined to be valid and accurate for different types of melt/concrete interaction. The model presented in this study is simple yet very detailed and accurate in simulating the actual molten core/concrete interaction (MCCI) phenomena, and in investigating the concrete reaction to the molten core. It not only can be embodied into the MCCI codes currently being developed, but also can be used to determine the containment integrity, and fission products released into the environment and to the public as a stand alone code.

Effective gages to measure skin friction and heat transfer have been established over decades. One of the most important criteria in designing such a gage is the physical size of the gage to minimise the interference of the flow, as well as the mass of these devices. The combined measurement of skin friction and heat flux using one single gage on the other hand, present unique opportunities and with it, unique technical problems.

The objective of this study is therefore to develop a cost-effective single gage that can be used to measure both skin friction and heat flux. The method proposed in this study is to install a coaxial thermocouple into an existing skin friction gage to measure the unsteady temperature on the surface of the gage. By using the temperature history and a computer program the heat flux through the surface can be obtained through an iterative guessing method. To ensure that the heat flux through the gage is similar to the heat flux through the rest of the surface, the gage is manufactured of a material very similar to the rest of the surface.

Walker developed a computer program capable of predicting the heat flux through a surface from the measured surface temperature history. The program is based on an inverse approach to calculate the heat flux through the surface. The biggest advantages of this method are its stability and the small amount of noise induced into the system. The drawback of the method is that it is limited to semi-infinite objects. For surfaces with a finite thickness, a second thermocouple was installed into the system some distance below the first thermocouple. By modifying the computer program these two unsteady temperatures can be used to predict the heat flux through a surface of finite thickness.

As part of this study, the effect of noise induced by the Cook-Felderman technique, found in the literature were investigated in detail and it was concluded that the method proposed in this study is superior to this Cook-Felderman method. Heat flux measurements compared well with measurements recorded with heat flux gages. In all cases evaluated the difference was less than 20%. It can therefore be concluded that heat flux gages on their own can measure surface heat flux very accurately. These gages are however too large to install in a skin-friction gage. The method introduced in this study is noisier than the heat flux gages on their own, but the size which is very important, is magnitudes smaller when using a coaxial thermocouple, to measure the surface temperature history.
Master of Science

This thesis consists of two different research problems. In the first one, the heat transfer characteristic of wavy fin assembly with dehumidification is carried out. In general, fin tube heat exchangers are employed in a wide variety of engineering applications, such as cooling coils for air conditioning, air pre-heaters in power plants and for heat dissipation from engine coolants in automobile radiators. In these heat exchangers, a heat transfer fluid such as water, oil, or refrigerant, flows through a parallel tube bank, while a second heat transfer fluid, such as air, is directed across the tubes. Since the principal resistance is much greater on the air side than on the tube side, enhanced surfaces in the form of wavy fins are used in air-cooled heat exchangers to improve the overall heat transfer performance. In heating, ventilation, and air conditioning systems (HVAC), the air stream is cooled and dehumidified as it passes through the cooling coils, circulating the refrigerant. Heat and mass transfer take place when the coil surface temperature in most cooling coils is below the dew point temperature of the air being cooled. This thesis presents a simplified analysis of combined heat and mass transfer in wavy-finned cooling coils by considering condensing water film resistance for a fully wet fin in dehumidifier coil operation during air condition. The effects of variation of the cold fluid temperature (-5˚C - 5˚C), air side temperature (25˚C - 35˚C), and relative humidity (50% - 70%) on the dimensionless temperature distribution and the augmentation factor are investigated and compared with those under dry conditions. In addition, comparison of the wavy fin with straight radial or rectangular fin under the same conditions were investigated and the results show that the wavy fin has better heat dissipation because of the greater area. The results demonstrate that the overall fin efficiency is dependent on the relative humidity of the surrounding air and the total surface area of the fin. In addition, the findings of the present work are in good agreement with experimental data. The second problem investigated is the heat transfer analysis of confined liquid jet impingement on various surfaces. The objective of this computational study is to characterize the convective heat transfer of a confined liquid jet impinging on a curved surface of a solid body, while the body is being supplied with a uniform heat flux at its opposite flat surface. Both convex and concave configurations of the curved surface are investigated. The confinement plate has the same shape as the curved surface. Calculations were done for various solid materials, namely copper, aluminum, Constantan, and silicon; at two-dimensional jet. For this research, Reynolds numbers ranging from 750 to 2000 for various nozzle widths channel spacing, radii of curvature, and base thicknesses of the solid body, were used. Results are presented in terms of dimensionless solid-fluid interface temperature, heat transfer coefficient, and local and average Nusselt numbers. The increments of Reynolds numbers increase local Nusselt numbers over the entire solid-fluid interface. Decreasing the nozzle width, channel spacing, plate thickness or curved surface radius of curvature all enhanced the local Nusselt number. Results show that a convex surface is more effective compared to a flat or concave surface. Numerical simulation results are validated by comparing them with experimental data for flat and concave surfaces.

Lors d’une insertion accidentelle de réactivité dans un réacteur nucléaire expérimental, la puissance du cœur peut augmenter de manière exponentielle, avec un temps caractéristique allant de quelques millisecondes à quelques centaines de millisecondes. À cause des effets neutroniques et thermohydrauliques, le système peut atteindre les conditions de crise d’ébullition à même d’engendrer une réaction explosive. Bien que la crise d’ébullition ait été largement étudiée en conditions de chauffage stationnaires, ce n’est pas le cas pour les transitoires notamment de type excursions de puissance. Le but de ce travail est donc de comprendre et de prédire la crise d'ébullition sous l’effet d’un chauffage transitoire rapide de l'eau sous fortes sous-saturations à pression modérée. Des campagnes expérimentales ont été réalisées pour étudier la crise d’ébullition dans de telles conditions au moyen de vidéos et de thermographie IR hautement résolues en temps et en espace. L’analyse de ces données a permis de déterminer la dépendance du flux critique en transitoire rapide en fonction des différents paramètres d’intérêt (temps caractéristique d’excursion de puissance, vitesse d’écoulement, sous-saturation, pression, largeur du canal, longueur de chauffe). De plus, une analyse approfondie de ces données a permis de mettre en évidence les mécanismes sous-jacents à la crise d’ébullition dans ces conditions. En convection forcée et avec de fortes sous-saturations, les bulles générées en paroi présentent un comportement pulsant. Ce phénomène assure un transfert de chaleur efficace depuis la paroi vers le fluide environnant. Le déclenchement de la crise d’ébullition se produit lorsqu’une fine couche de fluide adjacente à la paroi atteint les conditions de saturation. Un modèle développé à partir de ces observations met en évidence deux paramètres adimensionnés utiles pour décrire la nature transitoire du processus ainsi que pour identifier le mode de refroidissement dominant. Grâce à la connaissance du flux critique en régime permanent, le modèle permet d’estimer de manière conservative le flux critique en fonction de la période d’excursion de puissance et du sous-refroidissement. Ce modèle est maintenant prêt à être implémenté dans des codes de simulation pour l’étude des transitoires accidentels
In case of a reactivity insertion accident in an experimental nuclear reactor, heat generation in the core can grow exponentially in time, with a power escalation period ranging from a few to a few hundreds of milliseconds. Due to neutronic and thermohydraulic effects, boiling crisis may arise, possibly leading to an explosive reaction. If the boiling Crisis has been widely investigated in steady-state conditions, this has not been the case for transient heat inputs. The aim of the present work is to understand and to predict the transient flow boiling crisis in the conditions of moderate pressure and high subcooling. To this end, an experimental campaign has been realized making use of space and time highly resolved videos and IR thermography covering a wide range of experimental parameters. The analysis of the massive amount of data produced by these experiments gives a better insight on the dependency of the transient Critical Heat Flux to the different parameters of interest (power escalation period, flow velocity, subcooling, pressure, channel width, heating length). Moreover, their fine analysis enables to highlight the underlying mechanisms. For conditions of forced flow and high subcooling, the bubbles generated at the wall present a pulsating behavior. This specific process leads to an efficient heat transfer from the wall to the neighboring fluid. Boiling crisis is stated to occur when a thin layer of liquid contacting the wall reaches the saturation temperature. Starting from these observations, a model is developed which brings to light two non-dimensional parameters useful to describe the transient nature of the process and the dominant cooling processes. With the knowledge of the steady-state CHF, the model permits to conservatively estimate the value of the Critical Heat Flux for any power escalation period and subcooling. This model is now ready for implementation into simulation codes to investigate nuclear accidental transients

On présente les résultats d’une simulation numérique de la convection mixte conjuguée transitoire dans un tube vertical soumis à un flux DED chaleur constant et uniforme. Le fluide pénètre au haut du tube pour se diriger vers le bas ; par conséquent on est en présence d’un écoulement de convection mixte contrariée. Les équations de conservation dans le fluide et dans la paroi sont résolues numériquement en utilisant la méthode des volumes finis. Le couplage pression-vitesse est assuré en utilisant l’algorithme simple. On étudie l’influence des propriétés physiques et géométriques sur l’évolution transitoire des grandeurs thermiques (flux de chaleur à l’interface paroi-fluide et distribution radiale de température) et sur les grandeurs dynamiques (coefficient de frottement et champs vecteurs vitesses)
The proposed survey in this thesis appears in the setting of the conjugated laminar and transient mixed convection in a thick vertical conduct submitted to a constant and uniform heat flux. The fluid penetrates to the top of the conduct to head downwards, therefore one is in presence of opposed mixed convection flow. The governing transport equations were solved using the finite volume formulation and the simple algorithm is adopted. We study the effect of physical and geometrical properties of the physical system on the transient evolution of the thermal magnitudes (interfacial heat flux and radial distribution of the temperature) and the hydrodynamic magnitudes (friction coefficient and vector velocities)

Дисертація присвячена дослідженню теплопередаючих характеристик пульсаційних капілярних теплових труб (ПТТ) в залежності від режимних та експлуатаційних параметрів. Дослідження проводились зі скляною та мідною ПТТ з внутрішнім діаметром, відповідно, 3,8мм та 1мм; кількість петель 4 та 7. Теплоносієм слугувала вода з коефіцієнтом заповнення приблизно 50% від внутрішнього об’єму. Охолодження скляної ПТТ відбувалося за рахунок вільної конвекції повітря, мідної – за рахунок вимушеної конвекції рідини з різними значеннями температури та витрати. Кут нахилу мідної ПТТ до горизонту змінювався від -90° до +90° з кроком 45°. Робота ПТТ умовно розділена на два режими передачі тепла: конвективно-кондуктивний, що відповідає малим значенням підведеної теплової потужності, та пульсаційний, що відповідає середнім та високим значенням підведеної теплової потужності та початку кипіння теплоносія. Величину теплового потоку, за якої відбувається перехід від одного режиму передачі тепла до другого, названо перехідним QПЕРЕХ. В результаті досліджень виявлено вплив режимних (підведений тепловий потік, витрата та температура охолоджувальної рідини) і експлуатаційних (орієнтації в просторі, зовнішні механічні вібрації) на температурний режим, термічний опір та коефіцієнти тепловіддачі ПТТ. Отримана напівемпірична залежність для приблизного розрахунку QПЕРЕХ. Отримані формули для обчислення кількості петель замкнутої та розімкнутої ПТТ в залежності від геометрії капілярної трубки, довжин зон нагріву та конденсації. На базі пульсаційного механізму теплопередачі розроблені новітні пристрої. Порівняння роботи ПТТ з іншими радіаторами показало, що пульсаційні теплові труби найбільш ефективні при необхідності відведення високих теплових потоків (більш ніж 6 Вт/см2).
The dissertation is dedicated to the heat transfer characteristics of pulsating capillary heat pipes (PHP) depending on the regime and operational parameters. The experiments were conducted with glass and copper PHP with the internal diameter, respectively, 3,8mm and 1mm; number of turns 4 and 7. The water was used as a heat carrier; the filling ratio was approximately 50% of the internal volume. Cooling of the glass PHP was carried out by free air convection, and cooling of the copper one was carried out by forced convection of the liquid with different values of temperature and flow rate. The inclination angle of the copper PHP varied from -90° to + 90° in increments of 45 °. The PHP operation can be conditionally divided into two modes of heat transfer that are: convection-conductive mode that corresponds to small values of input heat power and pulsation mode that corresponds to middle and high of input heat power and to the heat carrier boiling. The heat flux called transient takes place at the transition from one mode of heat transfer to another. As a result of experimental studies the temperature of the PHP heating, transport, and condensation areas as well as thermal resistance and heat transfer coefficients are presented depending on the input heat flux and parameters of the cooling fluid. The dependence of the PHP heat transfer characteristics on external mechanical vibrations and PHP orientation in space was researched. The simplified semi-empirical formula for transient heat flux calculating is obtained. Given dissertation also presents a constructional calculation of the PHP number of loops when manufactured depending on the geometry of the capillary tube, and the lengths of the heater and the condenser. On the basis of the pulsation heat transfer mechanism some new heat transfer devices were designed, such as pulsating thermosyphon radiator with PHP. Comparing of the PHP with other cooling systems has shown that it is most effective for rejection of the heat fluxes over 6 W/cm2.
Диссертация посвящена исследованию теплопередающих характеристик пульсационных капиллярных тепловых труб (ПТТ) в зависимости от режимных и эксплуатационных параметров. Исследования проводились со стеклянной и медной ПТТ с внутренним диаметром, соответственно, 3,8мм и 1мм; количество петель 4 и 7. Теплоносителем служила вода с коэффициентом заполнения примерно 50% от внутреннего объема. Охлаждение стеклянной ПТТ осуществлялось за счет свободной конвекции воздуха, медной – за счет принудительной конвекции жидкости с разными значениями температуры и расхода. Угол наклона медной ПТТ к горизонту изменялся от -90° до +90° с шагом 45°. Работа ПТТ условно разделена на два режима передачи тепла: конвективно-кондуктивный, соответствующий малым значениям подведенной тепловой мощности, и пульсационный, соответствующий средним и высоким значениям подведенной тепловой мощности и началу кипения теплоносителя. Величина теплового по- тока, при котором происходит переход от одного режима передачи тепла к другому, называется переходным QПЕРЕХ. В результате экспериментальных исследований представлены зависимости температур в зонах нагрева (ЗН), транспорта (ЗТ) и конденсации (ЗК) ПТТ от времени и подведенного теплового потока. Показано влияние параметров охлаждающей жидкости – расхода и температуры – на величину QПЕРЕХ. Для медной ПТТ стабильный пульсационный режим теплопередачи устанавливается при 30-50 Вт в зависимости от параметров эксперимента. Величина термического сопротивления ПТТ различается только в области конвективно-кондуктивного режима теплопередачи и достигает значений 4-5 °С/Вт, после начала кипения эта цифра снижается на порядок и составляет примерно 0,3-0,6 °С/Вт. Влияние режима теплопередачи сказывается и на величину средних коэффициентов теплоотдачи в ЗН и ЗК ПТТ. Если для конвективно-кондуктивного режима теплопередачи средние коэффициенты теплоотдачи для ЗН составляют 400-450 Вт/(м2·К), а для ЗК – 200-250 Вт/(м2·К), то для пульсационного режима передачи тепла в ПТТ средние коэффициенты теплоотдачи в ЗН достигают 3,5-4 кВт/(м2·К), а в ЗК – 1,8 кВт/(м2·К), т.е. почти в 9 раз больше. Впервые исследована зависимость теплопередающих характеристик ПТТ от внешних механических колебаний. Эксперименты показали, что вибрации практически не оказывают влияния на величину термического сопротивления, однако способствуют тому, что QПЕРЕХ наступает при меньших значениях подведенной мощности. Например, если без вибраций QПЕРЕХ = 45-50 Вт, то для частоты 10 Гц это значение снижается до 40 Вт, а для частоты порядка 40 Гц – до 20-25 Вт. Приведена физическая модель процессов, возникающих в ЗН в момент начала кипения теплоносителя. На основе теплового баланса построена математическая модель, учитывающая зарождение, рост и дальнейший отрыв парового пузырька в ЗН. В результате решения математической модели получена упрощенная полуэмпирическая формула для расчета QПЕРЕХ. Расчетные значения величины QПЕРЕХ превышают экспериментальные данные в среднем на 21%, что не уменьшает работоспособности формулы. В работе представлен конструктивный расчет количества петель ПТТ при её изготовлении в зависимости от геометрии капиллярной трубки, а также длин ЗН и ЗК. Приведена методика инженерного расчета ПТТ. Зная максимальную температуру и геометрические параметры теплонагруженного элемента, а также отводимую мощность и условия охлаждения, можно рассчитать среднюю температуру и термическое сопротивление ПТТ. На основе пульсационного механизма передачи тепла разработаны новые конструкции теплопередающих устройств: пульсационный термосифон и радиатор с ПТТ.

L’étude de l’ébullition transitoire est un enjeu important pour la sureté nucléaire. Un tel phénomène peut se produire lors d’un accident de type RIA (Reactivity Initiated Accident)dans un réacteur nucléaire où le pic de puissance au niveau d’un crayon de combustible peut déclencher une ébullition transitoire conduisant à une forte augmentation de la température de la gaine et à un risque de rupture. Plusieurs études en conditions réacteurs ont permis d’obtenir des courbes d’ébullition transitoires mais la modélisation qui en découle manque encore de fiabilité. Dans le cadre d’une collaboration avec l’Institut de Radioprotection et de Sûreté Nucléaire (IRSN), une expérience modèle a été construite à l’Institut de Mécanique des Fluides de Toulouse (IMFT). Elle génère un écoulement de réfrigérant HFE7000 dans un canal de section semi-annulaire, simulant l’écoulement autour d’un crayon de combustible, dont la partie intérieure, composée d’une feuille de métal, est chauffée rapidement par effet Joule, simulant l’échauffement de la gaine du crayon. La thermographie infra-rouge permet de mesurer la température de la paroi externe du métal. L’application d’une peinture noire sur le métal augmente son émissivité mais aussi la résistance thermique de la paroi. La précision de la mesure de la température d’intérêt a été optimisée en fonction de l’épaisseur de peinture et une correction sur le bilan d’énergie prend en compte ce paramètre. Ces mesures sont couplées avec une caméra rapide qui permet de visualiser les régimes d’ébullition et d’obtenir des tailles de bulles à l’aide de la mise en place d’algorithmes de traitement d’image. On représente sur un diagramme flux-température les transferts thermiques lors des différents régimes en stationnaire et en transitoire. Chaque régime d’ébullition, en conditions stationnaire ou transitoire, est alors passé en revue : la convection, le déclenchement de l’ébullition, l’ébullition nucléée, la crise d’ébullition, l’ébullition en film et le remouillage. Les régimes stationnaires sont correctement modélisés par des corrélations usuelles. La convection transitoire est caractérisée sur toute la paroi et son évolution se rapproche de la solution quasistationnaire. Il est montré que les transferts thermiques lors du passage vers l’ébullition nucléée sont dépendants de la formation d’une importante poche de vapeur qui se propage sur la paroi. Une étude locale de cette propagation est alors nécessaire. Afin de simuler des transitoires de température durant l’ébullition nucléée, un système d’asservissement de type P.I.D. permet d’imposer des créneaux ou des rampes de températures (de 5 à 500 K.s 1 ). Les résultats en ébullition nucléée sont conformes avec ceux de la littérature, tant en conditions stationnaire que transitoire. L’expérience permet d’étudier le transfert de chaleur lorsqu’un film de vapeur se forme et isole la paroi. Ce régime d’ébullition en film, pendant la chauffe ou le refroidissement de la paroi peut ainsi être stabilisée pendant plusieurs secondes avec ce système. On caractérise ainsi les conditions de déclenchement de l’ébullition en film, la dynamique de sa propagation et les transferts une fois établi. Enfin, l’implémentation des caractéristiques physiques de notre expérience dans le code SCANAIR de l’IRSN, permet de commencer à calculer et comparer nos résultats expérimentaux avec les simulations numériques. Des calculs de conduction instationnaire sont notamment considérés en imposant la température mesurée pour analyser nos résultats lors du régime de convection et après le déclenchement de l’ébullition
The study of rapid transient boiling is an important issue in the nuclear safety. Such a phenomenon may occur in the case of a RIA (Reactivity Initiated Accident) in the core of a nuclear reactor powerplant, where a power excursion can trigger the formation of a vapour film around the fuel rod, leading to an important rise of the rod temperature and a risk of failure. Some studies in reactor conditions provided transient boiling curves but the modeling lacks of reliability. In collaboration with the IRSN (Institut de Radioprotection et de Sûreté Nucléaire), an experiment model was built at the Institute of Fluid Mechanics of Toulouse. It generates the flow of a refrigerant, HFE7000, in a semi-annular section channel, whose inner wall is made of a metal foil rapidly heated by Joule effect, simulating the heating of a fuel rod. Infrared thermography is used to measure the temperature of the metal foil, painted with a black paint to increase its emissivity, causing also an increase of the wall thermal resistance. The measurement accuracy of the interest temperature has been optimized according to the paint thickness and a correction on the energy balance takes account this parameter. These measurements are coupled with a high-speed camera that allows visualizing the boiling regimes and get bubble sizes using image processing algorithms. On a flux-temperature diagram, the heat transfers are represented both for steady and transient regimes. Each boiling regime is then reviewed : convection, onset of nucleate boiling, nucleate boiling, boiling crisis, film boiling and rewetting. Steady regimes are correctly modeled by usual correlations. Transient convection is characterized over the whole wall and its evolution is closed to the quasi-steady solution. It is shown that heat transfer during the transition to nucleate boiling are strongly related to the formation of a large vapor phase that spreads on the wall. A local study of this propagation is then necessary. In order to simulate and control transient temperature during nucleate boiling, a P.I.D. is implemented to impose a steady or ramps temperature (from 5 to 500 K.s 1 ). The results in nucleate boiling make it possible to recover the results of the literature in both steady and transient conditions. The experiment allows to study the heat transfer when a vapor film is formed and insulates the wall. The film boiling regime during heating or the cooling of the wall can thus be stabilized for several seconds with this system. The conditions for triggering of film boiling are thus characterized, as its spread dynamic and its transfers once established. Finally, the implementation of the physical characteristics of our experience in IRSN’s SCANAIR code allows us to begin to calculate and compare our experimental results with numerical simulations. Unsteady conduction calculations are applied to the measured temperature to analyze our results during the convection regime and after the onset of boiling

碩士
逢甲大學
應用數學系
88
Discussing the heat flux of the transient state suddenly transfer to the root of two-dimensional rectangular plate fin and cylinder fin. The upside、underside and tip of the rectangular plate fin all have different heat convection effect. According to different Biot Number, there have different temperature distribution and heat transfer state. (1)The analysis of the heat transfer of the two dimensional rectangular via using eigenfunction expansion method to fit、top and bottom boundary condition and to compact variable. And to utilize coordinate transform making the homogeneous of right and left boundary condition of the two dimensional rectangular plate. Similarly, to utilize eigenfunction expansion method to satisfy the right and left boundary condition of two dimensional rectangular plate and to compact variable. The resolute answer of the two dimensional rectangular plate would be achieved via fundamental the differential equation method of the single variable. (2)When we research the analysis of the heat transfer of the two dimensional cylinder fin, as result of the symmetry of the cylinder fin, we only consider one chamfer. The research method primarily uses zero Bessel function to transfer the equation of the cylinder fin to dimensionless heat transfer equation. satisfy that the center temperature is finite , so we could resolve by using function to be eigenfunction expansion. It would be satisfied the right and left boundary condition of the two dimensional cylinder fin via the eigenfunction function expansion. The analytical answer of the two dimensional cylinder would be solved by using a general method of single variable differential equation. To solve the analysis and numerical solution using the method of numerical analysis and the program of Matlab. Compare the equivalent of analysis and numerical solution and non-dimensional heat transfer rate. The center temperature distributions of the rectangular plate have apparent differentia with different the Biot Number and thickness. The center temperature will gradually descend with raising the Biot Number. When the Biot Number of the end arise , then the temperature distribution of the end will show constringency status. When the Biot Number is very small , the slope of he center temperature distribution of the different thick fin is still identical. The diversity of temperature would be more with more small thickness. Heat quantity would be faster accumulated and lost. The temporal heat transfer rate will be difficult to reach steady state with arising thickness. The center temperature distributions of the cylinder fin have apparent differentia with different the Biot Number and thickness. The center temperature will gradually descend with raising the Biot Number. The heat quantity would be more accumulated with more large thickness. When the Biot Number of the end arise , then the temperature distribution of the end will show constringency status. When the Biot Number is very small , the slope of he center temperature distribution of the different thick fin is still identical. The temporal heat transfer rate will be difficult to reach steady state with arising thickness.

碩士
國立交通大學
機械工程系所
96
An experiment is carried out in this study to investigate the transient flow boiling heat transfer performance and associated bubble characteristics of FC-72 flow over a small circular plate subject to a time varying heat flux with the plate flush mounted on the bottom of a horizontal rectangular channel. The imposed heat flux oscillates with time in the form of rectangular waves. The effects of the FC-72 inlet liquid subcooling, mass flux, and imposed heat flux oscillation on the FC-72 transient flow boiling characteristics are explored. In the experiment the inlet liquid subcooling △Tsub ranges from 0 to 10 ℃, the coolant mass flux G varies from 300 to 400 kg/m2s, mean imposed heat flux ranges from 0 to 10 W/cm2 and the amplitude of the heat flux oscillation is fixed at 10%, 30% and 50% of with the period of the heat flux oscillation varied from 10sec. to 30sec., covering the saturated and subcooled flow boiling. The transient flow boiling heat transfer characteristics are illustrated by the measured time variations of the heated plate temperature and boiling heat transfer coefficient. The experimental results show that the time-average FC-72 transient saturated and subcooled flow boiling heat transfer characteristics are not affected to a significant degree by the amplitude and period of the imposed heat oscillation. In fact, they resemble that for the stable flow boiling. However, the imposed heat flux oscillation causes significant temporal oscillations in the boiling heat transfer coefficient, bubble departure diameter and frequency, and active nucleation site density. These physical quantities oscillate at the same frequency as the heat flux oscillation and at a higher , a larger △q/ , and a longer tp they exhibit stronger oscillations except for tp=30 sec. Besides, a slight time lag in Tw oscillation is seen. Moreover, the size of departing bubbles, active nucleation size density and bubble departure frequency decrease as the heat flux is reduced to the low level of -△q. The opposite processes take place for the heat flux raised to the high level of +△q. Furthermore, at the mean imposed heat flux close to that for the ONB in the stable boiling we observe intermittent boiling in the transient saturated flow boiling and subcooled flow boiling. We also find that an increase in the inlet liquid subcooling results in stronger oscillations of these quantities.