The uncertainty in the determination of the momentum and scalar fluxes remains one of the main obstacles to accurate numerical forecasts in low to moderate wind conditions. For example, latent heat fluxes computed from data using direct covariance and bulk aerodynamic methods show that there is good agreement in unstable conditions when the latent heat flux values are generally positive. However, the agreement is relatively poor in stable conditions, particularly when the moisture flux is directed downward. If the direct covariance measurements are indeed accurate, then they clearly indicate that the bulk aerodynamic formula overestimate the downward moisture flux in stable conditions. As a result, comparisons of the Dalton number for unstable and stable conditions indicate a marked difference in value between the two stability regimes. Investigations done for this thesis used data taken primarily at the Air-Sea Interaction Tower (ASIT) during the Coupled Boundary Layers and Air-Sea Transfer (CBLAST) Experiment 2003 from the 20-27 August 2003. Other data from the shore based Martha's Vineyard Coastal Observatory (MVCO) and moored buoys in the vicinity of the ASIT were also incorporated. During this eight day period, the boundary layer was often characterized by light winds, a stably stratified surface layer and a swell dominated wave field. Additionally, the advection of warm moist air over cooler water resulted in fog formation and a downward flux of moisture on at least three occasions. Therefore, a primary objective of this thesis is to present a case study to investigate the cause of this shortcoming in the bulk formula under these conditions by examining the physical processes that are unique to these boundary layers. Particular attention will be paid to the behavior of the Dalton number in a stable marine atmospheric boundary layer under foggy conditions using insights derived from the study of fog formation and current flux parameterization methods.

本論文將詳述對湍流熱對流系統改變邊界條件的三組實驗。 / 第一組實驗中，在寬高比等於1 的圓柱形對流槽中，通過平滑板和由金字塔狀規則排列的粗糙表面的板的組合，我們對五個對流槽進行測量。对實驗測量得到Nu 和Ra 的关系式，进行两个指数函数关系式的叠加--分别是由GL 模型提出的主体主导的（1/2）指数和边界层主导的(1/4)指数关系式。两个关系式的系数之比表明，粗糙板可以提高主体的贡献，而且将系统从边界层主导转变到主体主导。而對各個板上的測量顯示，改變平板的條件，比如恒定熱量或者恒定溫度條件，以及同一個對流槽中另外一塊板是否粗糙板，該平板的傳熱沒有發生變化。而對於粗糙版，讓它處於恒定熱量或者恒定溫度條件，得到不同的結果；同一個槽中另外一塊板是平板或者粗糙板，該粗糙板的傳熱也會受影響。 / 第二組實驗中，在平滑板组成的对流槽，和两块粗糙板组成的对流槽中，我们添加高分子聚合物，进一步研究添加的效果。对于平滑板的对流槽，研究观察到热传递效率Nusselt 数在添加聚合物之后降低。而且，添加的聚合物越多，Nu 降低越多，在聚合物浓度大约120ppm 的时候，Nu 降低了接近12%，而且保持不变。而对于粗糙板组成的对流槽，当聚合物浓度大于120ppm 的时候，Nu 提高了大约4%。 / 第三組實驗，在两块平滑板组成的圆柱形对流槽中，我们也对底板边界层内的速度场进行了实验测量。结果表明，在倾斜角度小于1°的时候，速度边界层的厚度δ[subscript v]与雷诺数的指数关系，与理论给的平流边界层的结果接近。当倾斜角度比较大的时候，边界层厚度与雷诺数的指数随着角度减小。 在不同瑞利数下，同一个倾斜角度时，测量得到的水平速度剖面通过标准化后，得到的形状是一致不变的。而对于不同的角度，标准化的形状之间，是有区别的。 / In this thesis three sets of experiments of turbulent Rayleigh-Bénard convection with modified boundary conditions are presented. / The first set of measurements were made in cylindrical cells with aspect ratio one and with various combinations of smooth and rough plates in the form of regularly-arrayed pyramids. The experimental results suggest that the Nu - Ra relationship can be represented by the combination of two power laws, corresponding to the bulk dominant regime (exponent=1/2) and boundary layer dominant one (exponent=1/4) of the Grossmann-Lohse model. The behaviors of the coefficients of the two power laws suggest that the roughness of the plate can enhance the contribution of bulk and push the system to change from the boundary dominant state to bulk dominant state. A further examination of the individual plates reveal that the heat transport properties of smooth plates are insensitive to the surface and boundary conditions of the other plate of the same cell, whether smooth or rough, or whether under constant flux or constant temperature. The heat transport properties of the rough plates, on the other hand, appear to depend on surface and boundary conditions of the other plate of the same cell. / In the second set of experiments we study the effect of polymer additives in two Rayleigh-Bńard convection cells, one with smooth top and bottom plates and the other with rough top and bottom plates. For the cell with smooth plates, a reduction of the measured Nusselt number (Nu) was observed. Furthermore, the amount of Nu reduction increases with increasing polymer concentration (c), reaching ~12% for c = 120 ppm and an apparent level-off thereafter. For the cell with rough plates, however, an enhancement (~4%) of Nu was observed when the polymer concentration is greater than 120 ppm. / The third set of experiments investigates the properties of the velocity boundary layer in turbulent Rayleigh-Bénard convection in a cylindrical cell when it is tilted with respect to gravity. It is found that at small tilt angles (θ ≤ 1°), the measured viscous boundary layer thickness δ[subscript v] scales with the Reynolds number Re with an exponent close to that for a Prandtl-Blasius laminar boundary layer. For larger tilt angles, the scaling exponent of δ[subscript v] with Re decreases with θ. The normalized mean horizontal velocity profiles measured at the same tilt angle but with different Ra are found to have an invariant shape. But for different tilt angles, the shape of the normalized profiles is different. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Wei, Ping = 改變邊界條件下的熱湍流對流系統 : 傳熱性質及邊界層的實驗研究 / 韋萍. / "November 2012." / Thesis (Ph.D.)--Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 148-158). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts also in Chinese. / Wei, Ping = Gai bian bian jie tiao jian xia de re tuan liu dui liu xi tong : chuan re xing zhi ji bian jie ceng de shi yan yan jiu / Wei Ping. / Table of Contents --- p.vii / List of Figures --- p.xxii / List of Tables --- p.xxiv / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Background of Turbulence --- p.1 / Chapter 1.2 --- Rayleigh-Bénard convection system --- p.4 / Chapter 1.2.1 --- Physical picture --- p.4 / Chapter 1.2.2 --- Structure of Rayleigh-Bénard convection --- p.8 / Chapter 1.2.3 --- Previous investigation about Rayleigh-Bérnard Convection --- p.11 / Chapter 1.3 --- Motivation of the thesis --- p.15 / Chapter 1.4 --- Organization of this thesis --- p.17 / Chapter 2 --- Experimental apparatus and method --- p.18 / Chapter 2.1 --- Experimental apparatus --- p.18 / Chapter 2.1.1 --- The convection cell with smooth conducting plate --- p.18 / Chapter 2.1.2 --- The rough conducting plate --- p.20 / Chapter 2.2 --- Experimental techniques --- p.21 / Chapter 2.2.1 --- Temperature measurements --- p.21 / Chapter 2.2.2 --- Particle Image Velocimetry (PIV) --- p.23 / Chapter 2.3 --- Correction of heat flux measurment --- p.25 / Chapter 3 --- Convective Heat Transfer in Bulk- and Boundary-Dominated Regimes in Turbulent Thermal Convection --- p.28 / Chapter 3.1 --- Introduction --- p.29 / Chapter 3.2 --- Apparatus and experimental method --- p.31 / Chapter 3.3 --- Results and discussion --- p.34 / Chapter 3.3.1 --- Homogenity of temperature measurement over plate --- p.34 / Chapter 3.3.2 --- The heat transport measurement Nu --- p.40 / Chapter 3.3.3 --- Different heat transport behaviors over smooth and rough plates --- p.41 / Chapter 3.3.4 --- Fluctuations of local temperature --- p.47 / Chapter 3.3.5 --- Velocity and Reynolds number measurement --- p.48 / Chapter 3.3.6 --- Comparison with theoretical model and understanding . --- p.50 / Chapter 3.4 --- Summary --- p.52 / Chapter 4 --- Enhanced and Reduced Heat Transport in Turbulent Thermal Convection with Polymer Additives --- p.54 / Chapter 4.1 --- Introduction --- p.55 / Chapter 4.2 --- Experimental setup and methods --- p.58 / Chapter 4.2.1 --- The convection cell and experimental parameters --- p.58 / Chapter 4.2.2 --- Characterization of Polymer --- p.59 / Chapter 4.2.3 --- Measurement of the polymer solution viscosity --- p.62 / Chapter 4.3 --- Results --- p.65 / Chapter 4.3.1 --- The Nusselt number measurement --- p.65 / Chapter 4.3.2 --- Fluctuations of Nu and the local temperature --- p.72 / Chapter 4.3.3 --- Velocity and Reynolds number behavior --- p.83 / Chapter 4.4 --- Discussions --- p.90 / Chapter 4.5 --- Summary --- p.94 / Chapter 5 --- Viscous boundary layer properties in turbulent thermal convection in a cylindrical cell: the effect of cell tilting --- p.97 / Chapter 5.1 --- Introduction --- p.98 / Chapter 5.1.1 --- Boundary layer measurements in turbulent thermal convection --- p.98 / Chapter 5.1.2 --- Organization of the chapter --- p.102 / Chapter 5.2 --- Experimental apparatus --- p.103 / Chapter 5.2.1 --- Convection cell --- p.103 / Chapter 5.2.2 --- PIV measurement --- p.104 / Chapter 5.3 --- Results and discussion --- p.104 / Chapter 5.3.1 --- Temperature profile and fluid properties --- p.106 / Chapter 5.3.2 --- Velocity profiles and the Reynolds number scaling --- p.106 / Chapter 5.3.3 --- The viscous boundary layer and its scaling with Ra and Re --- p.114 / Chapter 5.3.4 --- Fluctuations and statistical properties of the velocity field in the boundary layer --- p.121 / Chapter 5.3.5 --- Properties of shear stresses and near-wall quantities --- p.125 / Chapter 5.3.6 --- Dynamical scaling and the shape of velocity profiles in the boundary layer --- p.130 / Chapter 5.4 --- Summary --- p.138 / Chapter 6 --- Conclusion --- p.141 / Chapter 6.1 --- Conclusion of our work --- p.141 / Chapter 6.1.1 --- Comparison of the thermal convection with modified plates --- p.142 / Chapter 6.1.2 --- The effect of polymer additives in smooth and rough cells --- p.143 / Chapter 6.1.3 --- Velocity field with tilting angles --- p.144 / Chapter 6.2 --- Perspectives for further investigation --- p.146 / Bibliography --- p.148

by Xin-liang Qiu. / Thesis (Ph.D.)--Chinese University of Hong Kong, 1998. / Includes bibliographical references (p. 98-103). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Mode of access: World Wide Web. / Abstracts in English and Chinese. / by Xin-liang Qiu.

長期以來，人們知道壁面受限湍流中的聚合物添加劑將顯著降低摩擦阻力，但是對聚合物在熱對流熱傳輸的影響還沒有太多研究。作為第一步，一項最近的工作[1]研究了在穩態邊界層流中熱量傳輸是怎樣被聚合物添加劑所影響的。在這項工作中[1]，聚合物是用Oldroyd-B模型來描述，這個模型允許聚合物無限伸展而沒有限制。 / 在這篇論文中，我們用一個更加真實的聚合物模型來研究聚合物在穩態邊界層流中對熱量傳輸的影響。我們採用FENE-P（有限擴展非線性彈性Peterlin）模型，在這個模型中，聚合物僅可以被伸展到一個最大的長度。聚合物的有限伸展性由參數L來衡量，它是聚合物最大長度與平衡長度的比例。基於該模型，我們發現，相對於與聚合物溶劑在底板處粘度相同的牛頓流體，熱量傳輸可以被提高或者被降低，這取決於聚合物不同的L值。而在不同的L值下，流場中底板的阻力始終加強。在早期的工作中，可以用一個隨位置改變的有效粘度來理解聚合物的效果。我們探討了聚合物的有效粘度和流速場是怎樣被聚合物改變的，以理解這個問題。我們也對熱量傳輸與不同參數的依賴關係進行了研究，這些參數包括威森博格數，普朗特數和聚合物對零剪切下溶劑粘度作出的貢獻的比例。 / It has long been known that friction drag will be reduced signicantly due to polymer additives in turbulent wall-bounded flows, but the effect of polymers on heat transport in thermal convection has not been studied much. As a rst step, a recent work [1] has studied how heat transport in a steady-state boundary layer flow might be influenced by the addition of polymers. In this work [1], polymers are modeled by the Oldroyd-B model, in which they can be extended innitely without a limit. / In this thesis, we study the effect of polymers on the heat transport in steady-state boundary layer flow using a more realistic model of polymers. We apply the FENE-P (nite extensible nonlinear elastic-Peterlin) model, in which the polymers can only be extended up to a maximum length. The nite extensibility of the polymers is measured by the parameter L, which is the ratio of the maximum length to the equilibrium one. Based on the model, we nd that compared to a Newtonian flow with the same viscosity as that of the polymer solution at the plate, heat transport can be enhanced or reduced depending on L. The fraction drag is always enhanced by the polymers for all different L. In the earlier work, the effect of the polymers has been understood to produce an effective viscosity that is position-dependent. We have explored the effective viscosity of the polymers and how the velocity eld is modied by the polymers to understand our results. We have also studied how the results depend on the different parameters, including Weissenberg number, Prandtl number and the ratio of polymer contribution to the total zero-shear viscosity. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Wang, Yiqu = 有限伸展的聚合物對邊界層流中熱量傳輸的改變 / 王異曲. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 68-69). / Abstracts also in Chinese. / Wang, Yiqu = You xian shen zhan de ju he wu dui bian jie ceng liu zhong re liang chuan shu de gai bian / Wang Yiqu. / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Prandtl-Blasius boundary layer flow --- p.7 / Chapter 3 --- Earlier work with Oldroyd-B polymers --- p.13 / Chapter 4 --- Theoretical formulation of the problem with polymers of finite extensibility --- p.20 / Chapter 4.1 --- Equations of motion --- p.20 / Chapter 4.2 --- Quantities of interest --- p.30 / Chapter 5 --- Checking validity of fixed angle approximation --- p.34 / Chapter 6 --- Results and Discussion --- p.42 / Chapter 6.1 --- Calculations --- p.42 / Chapter 6.2 --- The effect on heat transport --- p.45 / Chapter 6.3 --- The effect on drag --- p.48 / Chapter 6.4 --- The velocity field due to polymers --- p.49 / Chapter 6.5 --- Effective viscosity --- p.55 / Chapter 6.6 --- Dependence on Weissenberg number --- p.58 / Chapter 6.7 --- Dependence on Prandtl number --- p.61 / Chapter 6.8 --- Dependence on the ratio of polymer contribution to the total zero-shear viscosity --- p.64 / Chapter 7 --- Conclusion --- p.66

Eight years (1980-87) of twice-daily high-resolution radiosonde data from a tropical mid-Pacific island station are analyzed to generate time series of the planetary boundary layer (PBL) top pressure, an ageostrophic parameter (related to the divergence), and a parameter of mean PBL specific humidity. Sample spectra of these time series are calculated along with a spectra for nighttime outgoing longwave radiation (OLR) data (centered at a gridpoint near the island station) and compared in order to investigate the relationship between the ageostrophy and dynamical phenomena of the tropical Pacific. Methods of boundary layer top diagnosis are compared. Data derived methods include temperature inversion detection, humidity jump detection, and wind shear detection. The model based diagnosis method uses a Richardson number scheme. Comparisons between data derived methods and model based methods show good agreement. Correlation coefficients for comparisons are all above 0.91. In each sample spectra, the dominant peak corresponds to the frequency of the annual cycle. With an estimated annual cycle removed from each of the time series the nighttime (12 GMT) sample spectra for PBL top pressure shows a peak at a frequency that corresponds to the 30-60 day Intraseasonal Oscillation (ISO). Removal of an estimated annual cycle was unsuccessful for the daytime (0 GMT) sample spectra for PBL top pressure, PBL mean specific humidity, and the ageostrophic parameter. Sample spectra of the ageostrophic parameter show peaks in the ISO frequency range for both 12 GMT and 0 GMT even before removal of the annual cycle was attempted. The ISO range peaks remained in the 12 GMT ageostrophic parameter spectrum when the annual cycle was successfully removed. The nighttime OLR spectrum is dominated by the annual cycle peak and its harmonics. The OLR spectrum also shows a significant peak in the ISO range of periods. Removal of an estimated annual cycle intensifies the peak in the ISO range. The ageostrophic parameter time series is compared to the results of an EOF analysis performed on the same 8 years of radiosonde humidity measurements. The first three EOF modes of specific humidity give sample spectra which are dominated by the annual cycle. With the annual cycle removed the sample spectra of the first and third EOF modes of specific humidity both show significant peaks in the ISO frequency range, while the second EOF mode sample spectrum is dominated by two peaks at frequencies corresponding to periods of approximately 4 and 5 months. The presence of common frequencies in the ISO range of the power spectra of the ageostrophic parameter (a measure of divergence), the nighttime OLR (a measure of deep convection), the nighttime planetary boundary layer top, and the first EOF mode of specific humidity, lends support to theories that relate the ISO to the interaction of dynamics and moist processes. / Graduation date: 1995

Thesis (Ph. D.)--Massachusetts Institute of Technology and Woods Hole Oceanographic Institution, 1984. / Vita. Bibliography: p. 200-206.

The object ofthe present work is to experimentally study thecase ofa turbulent boundary layer subjected to an AdversePressure Gradient (APG) with separation and reattachment. Thisconstitutes a good test case for advanced turbulence modeling.The work consists ofde sign of a wind-tunnel setup, developmentofP article Image Velocimetry (PIV) measurements and evaluationtechniques for boundary layer flows, investigations ofs calingofb oundary layers with APG and separation and studies oftheturbulence structure ofthe separating boundary layer withcontrol by means ofs treamwise vortices. The accuracy ofP IV isinvestigated in the near-wall region ofa zero pressure-gradientturbulent boundary layer at high Reynolds number. It is shownthat, by careful design oft he experiment and correctly appliedvalidation criteria, PIV is a serious alternative toconventional techniques for well-resolved accurate turbulencemeasurements. The results from peak-locking simulationsconstitute useful guide-lines for the effect on the turbulencestatistics. Its symptoms are identified and criteria for whenthis needs to be considered are presented. Different velocityscalings are tested against the new data base on a separatingAPG boundary layer. It is shown that a velocity scale relatedto the local pressure gradient gives similarity not only forthe mean velocity but also to some extent for the Reynoldsshear-stress. Another velocity scale, which is claimed to berelated to the maximum Reynolds shear-stress, gives the samedegree of similarity which connects the two scalings. However,profile similarity achieved within an experiment is notuniversal and this flow is obviously governed by parameterswhich are still not accounted for. Turbulent boundary layerseparation control by means ofs treamwise vortices isinvestigated. The instantaneous interaction between thevortices and the boundary layer and the change in the boundarylayer and turbulence structure is presented. The vortices aregrowing with the boundary layer and the maximum vorticity isdecreased as the circulation is conserved. The vortices arenon-stationary and subjected to vortex stretching. Themovements contribute to large levels ofthe Reynolds stresses.Initially non-equidistant vortices become and remainequidistant and are con- fined to the boundary layer. Theamount ofi nitial streamwise circulationwas found to be acrucial parameter for successful separation control whereas thevortex generator position and size is ofseco ndary importance.At symmetry planes the turbulence is relaxed to a nearisotropic state and the turbulence kinetic energy is decreasedcompared to the case without vortices. <b>Keywords:</b>Turbulence, Boundary layer, Separation,Adverse Pressure Gradient (APG), PIV, control, streamwisevortices, velocity scaling.

The mixing layer (ML) heights inferred from radiosondes, wind profilers, airborne lidar, airborne microwave temperature profiler (MTP), and in-situ aircraft data were compared during the Texas 2000 Air Quality Study in the Houston area. The comparisons and resulting good agreement between the separate instruments allowed for the spatial and temporal evolution of the ML height distribution to be determined across the Houston area on September 1, 2000. A benchmark method was created for determining ML heights from radiosonde data. The ML heights determined using this method were compared to ML heights determined using wind profiler data. The airborne lidar and MTP heights were also compared to the wind profiler heights. This was the first time the MTP was used for estimating ML heights. Because of this, the MTP heights were also compared to the ML heights determined by in-situ aircraft data. There was good agreement between the ML estimates when the instruments were co-located. The comparisons between the benchmark method and the wind profilers were independent of the quality of the profiler heights. The statistics for lidar and the wind profilers were better for the inland profiler comparisons. Even so, the results for coastal profilers were similar to the other comparisons. The results between the MTP and the wind profilers were comparable with the results found between the other instruments, and better, in that the statistics were similar for the both the inland and coastal profilers. The results between the MTP and in-situ aircraft data provided additional support for the use of MTP for determining ML heights. The combination of the inland and coastal wind profilers with the airborne instruments provided adequate information for the spatial and temporal evolution of the ML height to be determined across the Houston area on September 1, 2000. By analyzing the ML height distribution, major features were evident. These features included the shallow ML heights associated with the marine air from Galveston Bay and the Gulf of Mexico, and the sharp gradient of increasing ML heights north of Houston associated with the variation in the inversion depth found on this day.

<p>The object ofthe present work is to experimentally study thecase ofa turbulent boundary layer subjected to an AdversePressure Gradient (APG) with separation and reattachment. Thisconstitutes a good test case for advanced turbulence modeling.The work consists ofde sign of a wind-tunnel setup, developmentofP article Image Velocimetry (PIV) measurements and evaluationtechniques for boundary layer flows, investigations ofs calingofb oundary layers with APG and separation and studies oftheturbulence structure ofthe separating boundary layer withcontrol by means ofs treamwise vortices. The accuracy ofP IV isinvestigated in the near-wall region ofa zero pressure-gradientturbulent boundary layer at high Reynolds number. It is shownthat, by careful design oft he experiment and correctly appliedvalidation criteria, PIV is a serious alternative toconventional techniques for well-resolved accurate turbulencemeasurements. The results from peak-locking simulationsconstitute useful guide-lines for the effect on the turbulencestatistics. Its symptoms are identified and criteria for whenthis needs to be considered are presented. Different velocityscalings are tested against the new data base on a separatingAPG boundary layer. It is shown that a velocity scale relatedto the local pressure gradient gives similarity not only forthe mean velocity but also to some extent for the Reynoldsshear-stress. Another velocity scale, which is claimed to berelated to the maximum Reynolds shear-stress, gives the samedegree of similarity which connects the two scalings. However,profile similarity achieved within an experiment is notuniversal and this flow is obviously governed by parameterswhich are still not accounted for. Turbulent boundary layerseparation control by means ofs treamwise vortices isinvestigated. The instantaneous interaction between thevortices and the boundary layer and the change in the boundarylayer and turbulence structure is presented. The vortices aregrowing with the boundary layer and the maximum vorticity isdecreased as the circulation is conserved. The vortices arenon-stationary and subjected to vortex stretching. Themovements contribute to large levels ofthe Reynolds stresses.Initially non-equidistant vortices become and remainequidistant and are con- fined to the boundary layer. Theamount ofi nitial streamwise circulationwas found to be acrucial parameter for successful separation control whereas thevortex generator position and size is ofseco ndary importance.At symmetry planes the turbulence is relaxed to a nearisotropic state and the turbulence kinetic energy is decreasedcompared to the case without vortices.</p><p><b>Keywords:</b>Turbulence, Boundary layer, Separation,Adverse Pressure Gradient (APG), PIV, control, streamwisevortices, velocity scaling.</p>

Thesis (M.S.)--Joint Program in Oceanography/ Applied Ocean Science and Engineering, Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution, 2003. / Includes bibliographical references (p. 70-72).