TU Darmstadt / ULB / TUbiblio

Experimental Study of Water Jet Impingement Cooling of Hot Steel Plates

Karwa, Nitin (2012)
Experimental Study of Water Jet Impingement Cooling of Hot Steel Plates.
Technische Universität Darmstadt
Dissertation, Erstveröffentlichung

Kurzbeschreibung (Abstract)

Liquid jet impingement cooling is critical in many industrial applications. Principle applications include extracting large heat flux from metal parts, such as hot fuel bundle post-loss-of-coolant-accident in nuclear reactors, heat treatment of steel plates post-hot-processing, etc. The ability of liquid jets to extract high heat flux at controlled rates from metal parts, with temperatures as high as 800-1000 ºC, at moderate flow rates has made them indispensable in these applications. Due to the complexity of the process, the mechanism of flow boiling heat transfer during jet impingement cooling is not well understood. Resultantly, the presently used design approaches are based more on experience and rule of thumb than science. The principle challenge in the study of jet impingement cooling for these high temperature applications has been the lack of reliable instrumentation for measuring the cooling rates. To add to this, the conjugate nature of boiling heat transfer, especially on low conductivity metal like steel, makes this problem very complicated to understand. Thus, much of the state of art on this subject has been limited to experiments where either the conjugate problem has not been addressed or the tests have been performed at temperatures that are much lower than in the above mentioned applications. The basic objective of the present work is to contribute to the understanding of the thermo-hydrodynamic phenomenon occurring during the cooling of a hot steel plate with an impinging water jet. This work also complements a parallel study being conducted at the Institute of Fluid Mechanics and Aerodynamics (Technische Universität Darmstadt), in which the complex transport processes are being treated theoretically and validated against the experimental results of this work. To achieve the objective, transient cooling experiments have been performed on an instrumented stainless steel AISI-type 314 cylinder. To measure the temperature variation within the stainless steel cylinder during the transient cooling, fast-response thermocouples have been embedded within holes that are precisely drilled though its bottom flat face. The cylinder is induction heated to a homogeneous initial temperature of 900 ºC and is subsequently cooled by means of an axisymmetric subcooled free-surface water jet that impinges on its top flat face (impingement surface). During the cooling, each thermocouple output has been recorded at the rate of 100 samples per second. A two-dimensional axisymmetric inverse heat conduction analysis using these measured temperature data has been performed to estimate the temporospatial variation of temperature and heat flux on the impingement face. Both low and high speed images have been recorded to visualize the two-phase flow. These images and the estimated heat transfer distribution are used to understand the boiling mechanism. The effect of jet parameters, namely subcooling and impingement velocity, on the heat transfer process has been studied. Additionally, the effect of spent liquid accumulation over the impingement surface has been studied in few exploratory plunging jet experiments. This study presents a systematic methodology for the measurement and estimation of the temporospatial variation of heat transfer on the impingement surface of a hot steel plate. Three distinct regions, with difference in the extent of liquid-wall contact, have been identified on the impingement surface from the recorded images. i) A wetted region surrounds the jet stagnation region. Nucleate boiling is the principle heat transfer mode in this region. The outer periphery of this region is called the wetting front. No boiling activity has been observed in the high speed images, most likely because the bubbles were small and were unable to reach the liquid free-surface. The maximum heat flux position is determined to be within this region. As the wetted region grows in size with time, the maximum heat flux position also moves radially outwards. The wetting front and maximum heat flux position velocity reduce with increasing radial distance from the impingement point because the liquid velocity and subcooling reduce at the wetting front. Likewise, the wetting front velocity increases with jet velocity and subcooling. ii) The liquid gets deflected at the wetting front due to the efflux of large vapor bubbles beyond the maximum heat flux position. A term ``wetting front region' has been coined in this thesis to describe this region. The width of this region could not be determined from the high speed images. Transition boiling within a thin superheated liquid film that is continuously replenished by the bulk flow is proposed to be the probable reason for the high heat flux in this region. Further, the radial heat conduction to the wetted region is also significant here. iii) The impingement surface outside the wetting front region is dry. The dry surface slowly cools down due to film boiling and radial heat conduction to the wetting front region. The film boiling rate is very low in the impingement region. After deflecting away from the impingement surface in the wetting front region, the liquid film breaks into droplets over this region. Looking from the side, droplet deflection angle is observed to be small; still these droplets do not come into direct contact with the impingement surface, as has been confirmed by looking down from the top. The velocity of the splashed droplets has been determined by analyzing the high speed images. It has been found that the drop velocity is much lower than the liquid film velocity calculated at the wetting front position using single-phase flow relations suggested by Watson. It has been hypothesized that the liquid film in the wetted region is decelerated by the bubbles growing on the impingement surface. Further, measurements reveal that the drop velocity increases with decreasing subcooling, which means that the film and the droplet are accelerated in the radial outward direction by the vapor released in the wetting front region. It has been shown that the rewetting temperature (analogous to the Leidenfrost temperature for a sessile droplet), which refers here to the temperature below which the direct liquid-wall contact is re-established and the heat flux increases, in both the impingement and radial flow regions is significantly higher than that reported in the literature for pool boiling. Removal of bubbles by the flowing liquid in the early stages of their growth and then their rapid condensation within the subcooled liquid avoids the buildup of vapor near the hot wall, which is the likely reason for the enhancement of the rewetting temperature. This observation confirms that high heat fluxes can be removed at large wall superheats by impinging liquid jets, as practiced in the industry. The boiling curve shifts to higher heat flux and superheat with the increase in the jet velocity and subcooling. The maximum heat flux and surface temperature at maximum heat flux increase with both the jet velocity and subcooling. Area-weighted average boiling curves have been determined, which clearly show the enhancement in the heat transfer with jet velocity over the average surface superheat of 100 to 800 K. The enhancement in jet subcooling is, however, noticeable only in the wall superheat range of 300 to 700 K. The maximum heat flux and surface temperature at maximum heat flux decrease with radial distance from the stagnation point before reaching a constant value. The radial distribution of maximum heat flux condition has been classified into three regions based on the relative size of the hydrodynamic/thermal boundary layer and the liquid film. In the plunging jet impingement studies, it has been found that the wetting front growth slightly slows down due to accumulation of the spent liquid over the impingement surface. Area-weighted average boiling curves show that the heat transfer reduces due to accumulation.

Typ des Eintrags: Dissertation
Erschienen: 2012
Autor(en): Karwa, Nitin
Art des Eintrags: Erstveröffentlichung
Titel: Experimental Study of Water Jet Impingement Cooling of Hot Steel Plates
Sprache: Englisch
Referenten: Stephan, Dr.-Ing. Peter ; Tropea, Dr.-Ing. Cameron ; Tacke, Dr.-Ing. Karl-Hermann
Publikationsjahr: 10 Juli 2012
Ort: Darmstadt, Germany
Verlag: tuprints
Datum der mündlichen Prüfung: 30 Mai 2012
URL / URN: urn:nbn:de:tuda-tuprints-30419
Kurzbeschreibung (Abstract):

Liquid jet impingement cooling is critical in many industrial applications. Principle applications include extracting large heat flux from metal parts, such as hot fuel bundle post-loss-of-coolant-accident in nuclear reactors, heat treatment of steel plates post-hot-processing, etc. The ability of liquid jets to extract high heat flux at controlled rates from metal parts, with temperatures as high as 800-1000 ºC, at moderate flow rates has made them indispensable in these applications. Due to the complexity of the process, the mechanism of flow boiling heat transfer during jet impingement cooling is not well understood. Resultantly, the presently used design approaches are based more on experience and rule of thumb than science. The principle challenge in the study of jet impingement cooling for these high temperature applications has been the lack of reliable instrumentation for measuring the cooling rates. To add to this, the conjugate nature of boiling heat transfer, especially on low conductivity metal like steel, makes this problem very complicated to understand. Thus, much of the state of art on this subject has been limited to experiments where either the conjugate problem has not been addressed or the tests have been performed at temperatures that are much lower than in the above mentioned applications. The basic objective of the present work is to contribute to the understanding of the thermo-hydrodynamic phenomenon occurring during the cooling of a hot steel plate with an impinging water jet. This work also complements a parallel study being conducted at the Institute of Fluid Mechanics and Aerodynamics (Technische Universität Darmstadt), in which the complex transport processes are being treated theoretically and validated against the experimental results of this work. To achieve the objective, transient cooling experiments have been performed on an instrumented stainless steel AISI-type 314 cylinder. To measure the temperature variation within the stainless steel cylinder during the transient cooling, fast-response thermocouples have been embedded within holes that are precisely drilled though its bottom flat face. The cylinder is induction heated to a homogeneous initial temperature of 900 ºC and is subsequently cooled by means of an axisymmetric subcooled free-surface water jet that impinges on its top flat face (impingement surface). During the cooling, each thermocouple output has been recorded at the rate of 100 samples per second. A two-dimensional axisymmetric inverse heat conduction analysis using these measured temperature data has been performed to estimate the temporospatial variation of temperature and heat flux on the impingement face. Both low and high speed images have been recorded to visualize the two-phase flow. These images and the estimated heat transfer distribution are used to understand the boiling mechanism. The effect of jet parameters, namely subcooling and impingement velocity, on the heat transfer process has been studied. Additionally, the effect of spent liquid accumulation over the impingement surface has been studied in few exploratory plunging jet experiments. This study presents a systematic methodology for the measurement and estimation of the temporospatial variation of heat transfer on the impingement surface of a hot steel plate. Three distinct regions, with difference in the extent of liquid-wall contact, have been identified on the impingement surface from the recorded images. i) A wetted region surrounds the jet stagnation region. Nucleate boiling is the principle heat transfer mode in this region. The outer periphery of this region is called the wetting front. No boiling activity has been observed in the high speed images, most likely because the bubbles were small and were unable to reach the liquid free-surface. The maximum heat flux position is determined to be within this region. As the wetted region grows in size with time, the maximum heat flux position also moves radially outwards. The wetting front and maximum heat flux position velocity reduce with increasing radial distance from the impingement point because the liquid velocity and subcooling reduce at the wetting front. Likewise, the wetting front velocity increases with jet velocity and subcooling. ii) The liquid gets deflected at the wetting front due to the efflux of large vapor bubbles beyond the maximum heat flux position. A term ``wetting front region' has been coined in this thesis to describe this region. The width of this region could not be determined from the high speed images. Transition boiling within a thin superheated liquid film that is continuously replenished by the bulk flow is proposed to be the probable reason for the high heat flux in this region. Further, the radial heat conduction to the wetted region is also significant here. iii) The impingement surface outside the wetting front region is dry. The dry surface slowly cools down due to film boiling and radial heat conduction to the wetting front region. The film boiling rate is very low in the impingement region. After deflecting away from the impingement surface in the wetting front region, the liquid film breaks into droplets over this region. Looking from the side, droplet deflection angle is observed to be small; still these droplets do not come into direct contact with the impingement surface, as has been confirmed by looking down from the top. The velocity of the splashed droplets has been determined by analyzing the high speed images. It has been found that the drop velocity is much lower than the liquid film velocity calculated at the wetting front position using single-phase flow relations suggested by Watson. It has been hypothesized that the liquid film in the wetted region is decelerated by the bubbles growing on the impingement surface. Further, measurements reveal that the drop velocity increases with decreasing subcooling, which means that the film and the droplet are accelerated in the radial outward direction by the vapor released in the wetting front region. It has been shown that the rewetting temperature (analogous to the Leidenfrost temperature for a sessile droplet), which refers here to the temperature below which the direct liquid-wall contact is re-established and the heat flux increases, in both the impingement and radial flow regions is significantly higher than that reported in the literature for pool boiling. Removal of bubbles by the flowing liquid in the early stages of their growth and then their rapid condensation within the subcooled liquid avoids the buildup of vapor near the hot wall, which is the likely reason for the enhancement of the rewetting temperature. This observation confirms that high heat fluxes can be removed at large wall superheats by impinging liquid jets, as practiced in the industry. The boiling curve shifts to higher heat flux and superheat with the increase in the jet velocity and subcooling. The maximum heat flux and surface temperature at maximum heat flux increase with both the jet velocity and subcooling. Area-weighted average boiling curves have been determined, which clearly show the enhancement in the heat transfer with jet velocity over the average surface superheat of 100 to 800 K. The enhancement in jet subcooling is, however, noticeable only in the wall superheat range of 300 to 700 K. The maximum heat flux and surface temperature at maximum heat flux decrease with radial distance from the stagnation point before reaching a constant value. The radial distribution of maximum heat flux condition has been classified into three regions based on the relative size of the hydrodynamic/thermal boundary layer and the liquid film. In the plunging jet impingement studies, it has been found that the wetting front growth slightly slows down due to accumulation of the spent liquid over the impingement surface. Area-weighted average boiling curves show that the heat transfer reduces due to accumulation.

Alternatives oder übersetztes Abstract:
Alternatives AbstractSprache

In vielen industriellen Anwendungen hat die Kühlung mit Hilfe von Flüssigkeitsstrahlen eine große Bedeutung. Diese Kühltechnik wird hauptsächlich dann verwendet, wenn hohe Wärmestromdichten abgeführt werden müssen wie z.B. zur Notkühlung von Brennstabbündeln bei Kühlmittelverluststörfällen in Kernreaktoren oder Abschreckung bzw. Wärmebehandlung von Metallbauteilen. Die Flüssigkeitsstrahlkühlung ist unerlässlich, da mit moderaten Volumenströmen hohe Wärmestromdichten von heißen Metallbauteilen, welche Temperaturen weit über 800 bis 1000 °C haben, abgeführt werden können. Aufgrund der hohen Komplexität der Wärmeübertragung in einem solchen Kühlprozess, sind die physikalischen Mechanismen nicht vollständig verstanden. Die gegenwärtigen Auslegungen solcher Kühlsysteme basieren daher mehr auf Erfahrungen und Faustformeln als auf Wissenschaft. Die grundsätzliche Herausforderung im Bezug auf die Untersuchung der Wärmeübertragung beim Aufprall von Flüssigkeitsstrahlen in Hochtemperaturanwendungen, ist der Mangel an betriebssicheren und zuverlässigen Messgeräten zur Bestimmung der Abkühlraten. Darüber hinaus wird dieses Problem durch den Charakter der gekoppelten Wärmeübertragung während des Siedevorgangs auf schwach wärmeleitenden Metallen wie Stahl sehr kompliziert. Daher ist die Forschung auf diesem Gebiet zum großen Teil auf Experimente begrenzt, in denen das gekoppelte Wärmeübertragungsproblem nicht adressiert wird oder die Versuche bei geringeren Temperaturen durchgeführt werden als die, die in den oben genannten Anwendungen vorliegen. Das grundlegende Ziel dieser wissenschaftlichen Arbeit ist es, einen Beitrag zum besseren Verständnis über das thermo-hydrodynamische Phänomen, welches während des Abschreckens einer heißen Edelstahlplatte mit einem Wasserstrahl eintritt, zu leisten. Um dieses Ziel zu verwirklichen, sind transiente Abschreckexperimente an einem auf die Anfangstemperatur von 900 °C homogen aufgeheizten, instrumentierten Edelstahlzylinders (AISI 314) durchgeführt worden, auf dessen Oberfläche ein unterkühlter, zylindrischer Wasserstrahl achsensymmetrisch aufprallt. Die Abkühlraten wurden während der Abschreckvorgänge von Thermoelementen erfasst. An Hand der gemessenen Temperaturdaten wurde eine zwei-dimensionale, achsensymmetrische, inverse Analyse der Wärmeleitung vorgenommen, um die zeitliche und örtliche Änderung der Oberflächentemperatur und der Wärmestromdichte an der Aufprallfläche zu bestimmen. Um den Siedemechanismus zu verstehen, wurde zusätzlich die Zweiphasenströmung mit niedriger und hoher Kamerageschwindigkeit visualisiert und mit der Verteilung der Wärmeübertragung verglichen. Der Einfluss der Unterkühlung und der Aufprallgeschwindigkeit des Wasserstrahls auf den Wärmetransportprozess wurde untersucht. Drei verschiedene Bereiche mit unterschiedlichen Ausmaßen des Flüssigkeit-Wand-Kontaktes wurden auf der Aufprallfläche identifiziert. (i) Ein zentraler, benetzter Bereich, in dem einphasige Konvektion und Blasensieden die grundlegenden Wärmetransportmechanismen sind. In den Hochgeschwindigkeitsaufnahmen wurden jedoch keine Siedeaktivitäten beobachtet, da die Bläschen sehr klein sind und die freie Oberfläche der Flüssigkeit nicht erreichen. Die Position der maximalen Wärmestromdichte befindet sich innerhalb dieses Bereiches. Wenn der benetzte Bereich mit der Zeit wächst, so wandert auch die Position der maximalen Wärmestromdichte radial nach außen. Die Wachstumsrate des benetzten Bereichs steigt mit zunehmender Geschwindigkeit des unterkühlten Wasserstrahls. (ii) Die Flüssigkeit wird am äußeren Umfang des benetzten Bereichs aufgrund des Abflusses von verhältnismäßig großen Dampfbläschen nahe der Position der maximalen Wärmestromdichte verformt. Übergangssieden innerhalb eines dünnen Flüssigkeitsfilmes, der kontinuierlich mit Flüssigkeit nachgespeist wird, wird als der Wärmetransportmechanismus im benetzten Frontbereich angenommen. (iii) Die Aufpralloberfläche außerhalb des zentralen, benetzten Bereichs ist trocken und der~Wärmeübergang ist hier sehr gering. Nach der Ablösung, weg von der Aufpralloberflächem, zerfällt der Flüssigkeitsfilm in Tröpfchen. Die Geschwindigkeit fort spritzender Tröpfchen wurde an Hand der Hochgeschwindigkeitsaufnahmen analysiert. Dabei wurde eine viel geringere Tropfengeschwindigkeit ermittelt als die Gesch\-windigkeit des Flüssigkeitsfilms, die mit der von Watson vorgeschlagenen Korrelation für einphasige Strömungen errechnet wurde. Es wird vermutet, dass Dampfbläschen, die auf der Aufpralloberfläche innerhalb des benetzten Bereiches wachsen, den Flüssigkeitsfilm abbremsen. Es wird gezeigt, dass die Wiederbenetzungstemperatur in den Bereichen des Aufpralls und der radialen Strömung deutlich höher ist als die, die in der Literatur für das Behältersieden berichtet wird. Wahrscheinlich erfolgt durch die strömende Flüssigkeit eine Ablösung mit anschließend schneller Kondensation der ausfliegenden Dampfbläschen in ihrer frühen Wachstumsphase innerhalb des unterkühlten Flüssigkeitsfilmes. Dies verhindert die Entstehung von Dampf nahe der heißen Wand und führt schließlich zu einer Erhöhung der Wiederbenetzungstemperatur. Diese Beobachtung bestätigt, dass, wie in der Industrie angewendet, hohe Wärmestromdichten bei sehr großen Wandüberhitzungen durch den Aufprall von Flüssigkeitsstrahlen abgeführt werden können. Der maximale Wert der Wärmestromdichte und die zugehörige Temperatur nehmen mit wachsendem Abstand vom Staupunkt ab bevor sie einen konstanten Wert erreichen. Im Gegensatz dazu, steigen sie mit zunehmender Wasserstrahlgeschwindigkeit und -unterkühlung. Mit Erhöhung der Unterkühlung und der Geschwindigkeit des Wasserstrahls verschiebt sich die Siedekurve zu höheren Wärmestromdichten und Wandüberhitzungen.

Deutsch
Freie Schlagworte: Free-surface jet impingement, plunging jet impingement, quenching, boiling, maximum heat flux condition, rewetting temperature
Sachgruppe der Dewey Dezimalklassifikatin (DDC): 500 Naturwissenschaften und Mathematik > 530 Physik
600 Technik, Medizin, angewandte Wissenschaften > 620 Ingenieurwissenschaften und Maschinenbau
Fachbereich(e)/-gebiet(e): 16 Fachbereich Maschinenbau > Fachgebiet für Technische Thermodynamik (TTD)
16 Fachbereich Maschinenbau
Hinterlegungsdatum: 13 Jul 2012 15:52
Letzte Änderung: 05 Mär 2013 10:01
PPN:
Referenten: Stephan, Dr.-Ing. Peter ; Tropea, Dr.-Ing. Cameron ; Tacke, Dr.-Ing. Karl-Hermann
Datum der mündlichen Prüfung / Verteidigung / mdl. Prüfung: 30 Mai 2012
Export:
Suche nach Titel in: TUfind oder in Google
Frage zum Eintrag Frage zum Eintrag

Optionen (nur für Redakteure)
Redaktionelle Details anzeigen Redaktionelle Details anzeigen