Gas turbines are used to convert thermal energy into mechanical energy. The thermal efficiency of the gas turbine is directly related to the turbine inlet temperature. The combustion and turbine technology has improved to such an extent that the operating temperature in the turbine inlet is higher than the melting temperature of the turbine material. Different techniques are used to cope with this problem. One of the most commonly used methods is internal cooling of the turbine blades. Conventionally air from the compressor is used for this purpose but due to higher heat capacity, steam can be used as coolant. This opens up the possibility to increase the gas temperature. In the case of a combined cycle power plant, its availability provides a good opportunity to be used as a coolant. The trailing edge of the gas turbine blades is an important region as it affects the aerodynamics of the flow. The aerodynamics demands a sharp and thin trailing edge to reduce profile losses. The conventional method is the release of a lot of cooling air though a slot along the airfoil trailing edge. However in the case of internal only cooling designs, the coolant is not allowed to leave the channel except from the root section to avoid mixing of the gas in the main flow path with the coolant and loss of cooling medium. The challenge is to design an inner cooling channel, with the cooling medium entering and leaving the blade at the root section, which reduces the metal temperatures to the required values without an increase of the profile losses and at acceptable cooling flow rate and pressure drop. This thesis presents Computational Fluid Dynamic (CFD) based numerical work concentrated firstly on the flow and heat transfer in two-pass rectangular channels with and without turbulator ribs. The aspect ratio of the inlet pass was reduced to accommodate more channels in the blade profile in chord-wise direction. Additionally, the divider-to-tip wall distance was varied for these channels. Their effect on heat transfer and pressure drop was studied for smooth as well as ribbed channels.  It was followed by a numerical heat transfer study in the trapezoidal channel. Different RANS based turbulence models were used to compare the numerical results with the experimental results. Further, new designs to enhance heat transfer in the channel’s side walls (named as trailing edge wall) were studied. These include the provision of ribs at the trailing edge wall only, inline arrangement of ribs at the bottom as well as at the trailing edge wall and a staggered arrangement of these ribs. The final study was a conjugate heat transfer problem with an aim to propose the best internal cooling channel design to reduce the metal temperature of the trailing edge surface for the given thermal and flow conditions. A number of different options were studied and changes were made to get the best possible channel design. The results show that for a two-pass rectangular channel (both smooth and ribbed), the reduction in inlet channel aspect ratio reduces the pressure drop. For a smooth channel the reduction in the width of the inlet pass does not affect the heat transfer enhancement at the inlet pass and outlet pass regions. In case of ribbed channels, heat transfer decreases at the tip and bend bottom with decrease in the width of the inlet pass. Among different turbulence models used to validate numerical results against experimental results for case of trapezoidal channel, the low-Re k-epsilon model is found to be the most appropriate. Using the turbulence model that yields results that are closest to the experimental data, the staggered arrangement of ribs at the trailing edge wall is found to have maximum thermal performance. The results from the conjugate heat transfer problem suggest using steam as coolant if it is available as it requires less mass flow rate to get similar wall temperature values as compared to air at similar thermal and flow conditions. It is also found that staggered arrangement of ribs is the best option compared to others to enhance heat transfer in trailing edge of the gas turbine blade with the pressure drop in the cooling duct in the acceptable range. / Gasturbiner används för att omvandla värmeenergi till mekanisk energi. Den termiska verkningsgraden för en gasturbin är direkt relaterad till turbinen inloppstemperatur. Förbrännings- och turbintekniken har förbättrats så mycket att gastemperaturen i turbininloppet är högre än smälttemperaturen för turbinmaterialet. Olika tekniker används för att hantera detta problem. En av de vanligaste metoderna är intern kylningen av turbinbladen. Konventionellt luft från kompressorn används för detta ändamål, men på grund av högre värmekapacitet kan ånga användas som kylmedel. Detta öppnar för möjligheten att höja gasens temperatur. Vid ett kombikraftverk, ger dess tillgänglighet ett bra tillfälle att användas som kylmedel.   Den bakre delen av turbinbladen är ett viktigt område eftersom geometrin påverkar strömningen. Aerodynamiken kräver en skarp och tunn bakkant för att minska profilförlusterna. Den konventionella metoden för kylning av denna är att släppa ut en stor mängd kylluft genom en spalt längs bakkanten. Men i fallet med enbart inre kylning får kylmediet inte lämna skovelprofilen i strömningskanalen utan endast genom rotsektionen för att undvika blandning av förbränningsluften i turbinens strömningskanal med kylmediet och förlust av kylmedium.   Utmaningen är att utforma en inre kylkanal, i vilken kylmediet kommer in och lämnar bladet i rotsnittet som är tillräckligt bra för att hålla metalltemperaturen på normala värden utan att öka profilförlusterna och med acceptabla kylluftflöden och tryckfall.   Denna avhandling består av ett Computational Fluid Dynamics (CFD) baserat numeriskt arbetet koncentrerat på strömning och värmeöverföring först i två-pass rektangulära kanaler med och utan turbulensalstrande ribbor. Geometrin för inloppspassagen reducerades för att ge utrymme för fler kylkanaler inom bladets profil i kordans riktning. Dessutom varierades mellanväggens avstånd till toppväggen. Effekten på värmeöverföring och tryckfall studerades för båda kanalerna. Därefter följde en numerisk studie av värmeöverföringen i liknande men trapetsformade kanaler. Olika RANS baserade turbulensmodeller användes för att jämföra numeriska och experimentella resultat. Vidare har nya konstruktioner för att förbättra värmeöverföringen i kanalens sidoväggar och bakkant studeras. Dessa inkluderar turbulensribbor på enbart bakkantsväggen samt ribbor på såväl sidoväggar som på bakkantsväggen i linje med och förskjutna mot varandra. Den slutliga studien var ett sammansatt värmeöverföringsproblem bakkantens yta för ett visst angivet tillstånd i form av värmebelastning, tryck, temperatur och flöden. Ett antal olika alternativ har studerats och modifierats för att bästa möjliga kanalutformningen.   Resultaten visar att för en två-pass rektangulär kanal (både släta och ribbade), minskar tryckfallet när inloppskanalens geometri reducerades. För en slät kanal påverkar inte den minskade bredden på inloppskanalen värmeöverförning i inlopps- och utloppskanalerna. Vid ribbade kanaler minskar värmeöverföring vid toppen och på toppväggen med minskad bredd på inloppskanalen. Av de olika turbulensmodeller som används för att validera numeriska resultat mot experimentella för fallet med trapetsformad kanal visade sig låg-Re k-epsilon modellen den mest lämpliga. Genom att använda den turbulensmodell som är närmast experimentella data visar det att geometrin med förskjutna ribbor på bakkantsväggen har maximal termiska prestanda. Resultaten från det sammansatta värmeöverföringsproblemet framhåller användning av ånga som kylmedium om den finns tillgänglig eftersom den kräver mindre massflöde för att få samma värden på väggtemperaturerna jämfört med luft vid samma termiska tillstånd. Det kunde också visas att förskjutna turbulensribbor är det bästa alternativet jämfört med andra för att öka värmeöverföringen i bakkanten av ett gasturbinblad med acceptabelt tryckfall i kylkanalen. / QC 20111108

Dissertation (Ph.D.)--Auburn University, 2005. / Abstract. Vita. Includes bibliographic references.

Thesis (M.S.)--Rochester Institute of Technology, 1983. / Typescript. Includes bibliographical references (leaf 56).

Thesis (M.S.)--California Polytechnic State University, 2009. / Mode of access: Internet. Title from PDF title page; viewed on September 28, 2009. Major professor: Dr. Joseph Mello. "Presented to the faculty of California Polytechnic State University, San Luis Obispo." "In partial fulfillment of the requirements for the degree [of] Master of Science in Mechanical Engineering." "June 2009." Includes bibliographical references (p. 69).

Thesis (M.S.)--California Polytechnic State University, 2009. / Title from PDF title page; viewed on May 26, 2010. Major professor: Patrick Lemieux, Ph.D. "Presented to the faculty of California Polytechnic State University, San Luis Obispo." "In partial fulfillment of the requirements for the degree [of] Master of Science in Mechanical Engineering." "September 2009." Includes bibliographical references (p. 198-201).

- / -

This work is an extension to the spray model of Watkins and Jones (2010). In their model, the spray is characterized by evaluating the three moments Q_2, Q_3 and Q_4 of general gamma number size distribution from their transport equations. The sub-models of drop drag, drop break-up and drop collisions were simulated in terms of gamma distributions. The model is considered as non-vaporising and compared with cases which have low ambient gas temperature and also is strict to a particular set of sub-models for drop drag and break up which they are applicable to produce integrable functions. In this work the model is adjusted to allow a variety of sub-models to be implemented. Three models (TAB, ETAB, DDB) are considered for drop breakup which have been basically introduced to be used with the Droplet Discrete Method (DDM) approach. So in order to implement these models with the model of Watkins and Jones the source terms of the breakup are calculated by grouping the droplets in each cell into parcels which contain a certain number of droplets with similar physical properties (size, velocity, temperature ...). The source terms of each parcel are calculated and multiplied by the number of droplets in these parcels and a numerical integration is then used to obtain the resultant effect of the drop breakup in each cell. The number of drops in each cell is determined from the gamma size distribution. Also three hybrid breakup models (KH-RT, Turb-KH-RT, Turb-TAB) which include two distinct steps: primary and secondary break up model are implemented. The Kelvin- Helmholtz (KH) and the turbulence induced breakup (Turb) models were used to predict the primary break up of the intact liquid core of a liquid jet while the secondary break up is modelled using the TAB model and competition between the KH and the RT models. Both models are allowed to work simultaneously. However it is assumed that if the disintegration occurs due to the RT the KH break up does not occur. In case of drag sub-model, a dynamic drag model is introduced which accounts for the effects of drop distortion and oscillation due to the effects of high relative velocity between the liquid and the surrounding gas. In this model the drag coefficient is empirically related to the magnitude of the drop deformation. The magnitude of drop deformation was calculated by using the TAB model. In this work, the effects of mass and heat transfer on the spray are modelled. An additional equation for the energy of the liquid is solved. The mass transfer rate is evaluated using the model of Godsave (1953) and Spalding (1953) while the Faeth correlation (1983) is used to model heat transfer between the two phases. For all equations of heat and mass transfer between phases, the drop Nusselt and Sherwood number are calculated by using the correlation of Ranz and Marshall. In this model also the liquid surface-average temperature T_l2 which is calculated by Watkins (2007) is used to determine the heat and mass transfer between phases instead of liquid volume-average temperature. It was derived by assuming a parabolic temperature profile within individual drops. All the equations are treated in Eulerian framework using the finite volume method. The model has been applied to a wide range of sprays and compared to a number of experiments with different operating conditions including high liquid injection pressure and high ambient gas density and temperature. A reasonable agreement is found by the ETAB model with most of the data while the TAB and the DDB models continually underestimate the penetration and drop sizes of the spray. The hybrid breakup models perform well and show better agreement with the available experimental data than the single breakup models. In term of high temperature cases, the model correctly captures the effect of evaporation on the different spray properties especially with hybrid break up model.

Thesis (M.Sc.Eng.) PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at open-help@bu.edu. Thank you. / Boiling heat transfer is a critical process in large-scale industrial applications such as steam engines and heat exchangers in power plants, and in microscopic heat transfer devices such as heat pipes and microchannels for cooling electronic chips. Enhancing boiling heat transfer thus has great significance on lots of energy transportation and utilization systems. Recent studies has suggested that micro/nano structured surfaces can produce considerably different boiling heat transfer curves than normal plain surfaces, resulting in different values of the critical heat flux (CHF) and heat transfer coefficient (HTC). In this thesis, pool boiling on several new micro/nano structured surfaces was experimentally investigated to further understand the mechanism of boiling heat transfer and increase boiling performance. We first evaluated enhanced boiling heat transfer on three kinds of micro/nano structured super-hydrophilic surfaces: 1) nanowire coated super-hydrophilic surfaces, 2) hybrid microscale cavity and nanowire structured surfaces and 3) hybrid microscale pillar and nanowire structured surfaces. All three surfaces showed significant enhancement of CHF and HTC compared to plain silicon surfaces. Combined micro/nano structured surfaces presented better performance than nanowire coated surfaces suggesting that both active nucleation density and surface roughness significantly affect boiling heating transfer. Experimental investigations indicate an optimum design both in size (~ 20μ𝑚) and density (between 0 and 10000=cm^2) of cavities for microscale cavity/nanowire structured surfaces. The highest CHF and peak HTC values were obtained on microscale pillar/nanowire structured surfaces. Among the test surfaces, the largest enhancements of CHF and peak HTC were 228% and 298%, respectively, compared to plain silicon surfaces. For a better understanding of the boiling phenomena, pool boiling on super-hydrophobic surfaces was also studied. We found that, for super-hydrophobic surfaces, the major heat transfer mechanism at the initial boiling regime is natural convection of liquid water. In conclusion, micro/nano structured surfaces can greatly influence nucleate boiling heat transfer. The various physical attributes employed with the structured surfaces further revealed the profound influence of surface topography on enhancing boiling heat transfer. / 2031-01-01

This thesis reports the results of an experimental study into the flow and heat transfers associated with both inclined and orthogonally impinging axisymmetric air jets. The majority of previously reported studies have been mainly confined to orthogonally impinging jets in stagnant surroundings. In this investigation, free jets as well as the effects of crossflows are considered. This investigation is primarily concerned with local heat transfer variations. The experimental tests were conducted with a single 12.7 mm diameter jet impinging on a flat surface, and heat transfers were evaluated using a heat-mass transfer analogy (the Chilton-Colburn analogy). The sublimation of naphthalene was employed as the mass transfer technique. The flowfield associated with impinging jets has a significant influence on their heat transfer characteristics. In view of the present limited level of understanding of this 'complex' flowfield, extensive flow visualisation techniques were employed in this present investigation. Those were primarily intended to aid interpretation of the experimental heat transfer results, and also to provide further physical understanding of the flowfields resulting from the interactions between impinging jets and crossflowing streams. The flow and heat transfer tests conducted in the programme of work reported in this thesis covered typical ranges of flow parameters of interest in many practical applications of jet impingement systems. Jet inclinations of 45°, 60°, and 90°, nozzle to target spacings of 2, 4, and 8 nozzle diameters were studied. The Reynolds numbers were 30,200, 32,700 and 55,100 and mass velocity ratios in the range 4.0 to 8.8 were studied. The effects of these parameters on the flow and heat transfers associated with impinging jets are reported. Comparisons were drawn between the heat transfer results and those of previously reported studies where appropriate.

Made available in DSpace on 2014-10-09T12:26:06Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T14:01:49Z (GMT). No. of bitstreams: 1 01064.pdf: 4252441 bytes, checksum: 2fbcdbf2781761be69e44b5c664fd572 (MD5) / Dissertacao (Mestrado) / IEA/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP