Thesis (M. Phil.)--University of Hong Kong, 2006. / Title proper from title frame. Also available in printed format.

Thesis (M.Phil.)--Hong Kong University of Science and Technology, 2005. / Includes bibliographical references (leaves 84-86). Also available in electronic version.

Th.--Méc. des fluides--Toulouse--I.N.P., 1979. N°: 42.

Turbulence in open channel flows need to be clarified in several aspects. Perhaps one of the most difficult tasks is to understand the structure o f flows having small relative submergence for which standard boundary layer theory is not applicable. Furthermore, rivers and natural channels are often characterized by very porous and rough beds where water flows and interact with the free stream inducing transport mechanisms which play an important role in the flinctioning o f stream ecosystems. In order to design numerical models able to simulate such transport processes, there is a need to understand and predict the turbulence properties of both the free stream and the sub-surface flow. To achieve this objective, it is crucial to understand how the permeability o f the bed influences the velocity statistics in the free stream and in turn how the free stream induces turbulence within the permeable bed. The purpose of this study was to carry out a series of laboratory experiments to gain insight into: 1) the turbulence characteristics of open channel flows having small relative submergence; 2) the influence of permeability on turbulence in open charmel flows; 3) the turbulence properties o f the flow at the interface between the free stream and a permeable bed. The data obtained from these experiments have been analysed and interpreted on the basis of the double-averaged momentum and continuity equations. The results helped on clarifying the definition and meaning of key scaling parameters for flows with small relative submergence such as the friction velocity and the zero plane position. Furthermore, it is shown how permeable beds induce higher resistance to the flow than their impermeable counterparts having the same roughness texture. It is shown that this is mainly due to additional energy dissipation occurring within the porous bed. Measurements of velocities within the bed allowed identifying the dissipation mechanisms and the origin of turbulence in this flow region. Within the bed, two types of turbulence occur: one generated at the pore scale and another one, more energetic, imposed by the large eddies of the free stream.

The present study is aimed at enhancing the effectiveness of the large eddy simulation (LES) approach to the computation of turbulent flows by these two methods: i) developing a superior subgrid scale (SGS) model and ii) improving the economy of LES. First of all, the various existing SGS models are extensively investigated, and their advangages and disadvantages are addressed to highlight the areas requiring improvements. This study leads to the construction of a modified SGS dynamic model. In addition, a detailed derivation of the second-order velocity structure function SGS model is made, correcting an error found in that model. A new multiple mesh method is also designed to accelerate LES. After the above theoretical studies, several low-Reynolds-number channel flow simulations have been performed. Firstly, simulations with varying model constants are carried out, and the results agree with those of Deardorff [14], showing that a model constant of about 0.1 is optimum for channel flows. Secondly, simulations with varying numerical resolutions have been carried out. They reveal that the refinement of the mesh in the direction normal to the wall improves all the turbulence statistics, both higher- and lower-order statistics, over the whole channel, while the refinement of the resolution in the streamwise and spanwise directions improves lower-order statistics over the whole channel, but only improves higher-order turbulence statistics in the central region of the channel. Thirdly, a dissipation-range SGS model (i.e. the Smagorinsky model with low- Reynolds-number modification [67]) is, for the first time, tested and compared with the standard Smagorinsky model. The results obtained show some promise for automatically adjusting the SGS model with Reynolds number. Fourthly, the performance of the modified dynamic SGS model is assessed through a comparison of length scales computed respectively by this modified model, the Germano- Lilly dynamic SGS model and two empirical wall damping functions in conjunction with an optimum model coefficient, which have been successfully used in many simulations of channel flows. Two values of the ratio of filter widths are set for each of the dynamic models. The results have confirmed that the modified dynamic SGS model can be successfully extended to simulate low-Reynolds-number channel flows. Of great promise is that the modified SGS dynamic model gives the correct behaviour of the subgrid eddy viscosity in the region of a plane wall to an accuracy that exceeds the best-tuned wall damping function, and almost collapses with the theoretical behaviour of the length scale near the wall without any tuning and adjustment. In addition, the impact of the choice of the ratio of filter widths on the modified dynamic SGS model is much less than that on the Germano-Lilly model. Finally, simulations using the new and old multiple mesh methods are performed. The instantaneous results just after the interpolation of the coarse mesh velocity field onto the fine mesh show that the fine mesh velocity field created by the new multiple mesh method contains the information of the residual field. In contrast, there is no difference between the fine mesh results obtained by the old method and those from a simulation on the coarse mesh.

At the International Comparison of Turbulence Measuring Instruments, 1970, velocity components and momentum flux measurements were compared using propeller-type Fluxatrons (Hicks, 1970) and sonic anemometers from Kaijo-Denki, Japan and the Institute of Atmospheric Physics, U.S.S.R. There were distinct differences found in the measurements of the vertical velocity from the propeller sensors. The propeller's momentum flux measurements computed from its velocity components were also different. The U¹ propeller was found to be linear for lower frequencies with an associated distance constant of about 7 metre. Measurement of the variance of U¹ for f < 0.16 hz. showed the U¹ propeller in excess of both sonics by 20%. However, with the propeller's high frequency loss beyond f = 0.2 hz. the discrepancy was reduced to only an 8% excess for .00055 hz. < f < 10.8 hz. The W¹ propeller response was non-linear and had an upper cut-off frequency of 1hz. Because of its non-linear response and stalling characteristics at low wind speeds and also its high frequency cut-off the W¹ propeller was observed to measure only about 50% of the total fluctuating W¹ energy available. Analysis of the sonic cospectra of momentum showed that significant contributions to the momentum flux were to be found in the frequency domain 0.001 hz. < f < 5.0 hz. The combined response effects of the propeller were enough to reduce the Fluxatron’s estimate of this momentum flux by 32.5%. / Science, Faculty of / Physics and Astronomy, Department of / Graduate

The Disa type 55A32 X-wire probe has been widely used in turbulence measurements. However, the author was unable to obtain agreement between turbulence measurements made simultaneously with this type of X-wire probe and an ultrasonic anemometer at the same position in the atmospheric boundary layer over the ocean. The nature of the disagreement between the two instruments suggested that there existed an unexpected response of the wires to the cross stream wind component normal to the plane of the X-array. Wind tunnel experiments confirmed this response and attributed most of it to thermal coupling between the two wires of the array via their hot wakes. The prongs and/or probe body were also shown to be contributors to the anomalous responses of the X-wires. Similar experiments carried out with a Thermo-Systems model 1241-20 X-probe (with a sensor length to sensor separation ratio of 5/8 compared with 0.2 or less for the Disa 55A32 X-wire probe) demonstrated that the interference effects were absent (or, at least, insignificant). As a consequence of these findings, the Disa Electronik A/s company of Herlev, Denmark, modified their 55A32, 55A38 and 55A39 lines of X-wire probes to make the length/separation ratio close to unity. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

The objective of this study was to investigate the turbulent transfer mechanisms near the surface. Direct measurements of the turbulent fluxes of momentum, heat, and moisture were made in the atmospheric surface layer: principally, 2 m above a grass surface at Ladner, Canada, and for comparison 8 m above the Atlantic Ocean near Barbados. The spectral correlation coefficients were considered to be a measure of the transfer efficiency as a function of scale size. For momentum transfer the efficiency decreased at all scales as instability increased. It was postulated that this was due to greater amounts of momentum being transferred in bursts of short duration, thus making the spectral correlation coefficient, averaged over sufficient time, smaller. The Ladner results for heat transfer showed that its transfer efficiency increased at all scales when instability increased. The ratios of the transfer efficiency of heat to that of momentum were greater than 1 for most scales, even for near neutral stratifications, and increased to between 2 and 3 for more unstable conditions. The efficiency of moisture transfer, when moisture is a passive scalar, was usually smaller than that for heat transfer and was found to depend on the correlation between moisture fluctuations and those of temperature, which is the active scalar. The results from Barbados pointed out two main differences between the subtropics and mid-latitudes: that the temperature spectrum is much narrower in bandwidth and that the humidity fluctuations make an equally important contribution to buoyancy. These features are reflected in the transfer mechanisms. / Science, Faculty of / Physics and Astronomy, Department of / Graduate

Two systems for remote measurements of the air-sea fluxes of momentum, sensible heat and moisture during moderate to strong winds are described. One employs the dissipation method and the other the Reynolds flux or eddy correlation method. A modified Gill propeller-vane anemometer is the velocity sensor and a method of resolving the vertical velocity component, that accounts for the propeller's non-cosine behavior and avoids its non-linear operating region, is derived. The dynamic responses of the sensors are found from measurements in the actual turbulent conditions of the flux measurements. The results of an experiment on the Bedford tower, a stable platform moored in 59m of water 10 km offshore, are presented. Spectra, cospectra, turbulence statistics and transfer coefficients are calculated from the Reynolds flux velocity and temperature data and found to be comparable to previously reported values. Simultaneous dissipation and Reynolds flux estimates of both the momentum and sensible heat fluxes in up to 20 m/s winds are shown to be in excellent agreement. Also presented are the results of a second experiment where the systems were deployed on the weathership CCGS Quadra. A comparison of ship and tower drag coefficients from the dissipation system, demonstrates that the Bedford tower is essentially an open ocean site. The neutral drag coefficient, CDN, is found, on average, to be nearly constant at 1.14x10⁻³ for winds between 4 and 10 m/s and to increase almost linearly to about 2.18x10⁻³ at 26 m/s. No variation with either fetch (greater than 10 km) or stability is observed. Dissipation estimates of the sensible heat flux from a wide range of conditions are presented. The neutral transfer coefficient, CTN, is found, on average, to vary from about 0.69x10⁻³ in stable stratification to 1.08x10⁻³ in the unstable case. An increase in CTN with increasing wind speed is suggested by only some of the data. Time series of the fluxes are used to investigate additional sources of variation in the transfer coefficients. Their statistical variability about a running mean is seen to be about 10%. Evidence is presented that indicates that persistent departures from average values are related to sea surface conditions. CDN is observed to be significantly smaller, on average, during rising winds than during falling winds or after a change in wind direction. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

The characteristics of several types of heated sensors used for measuring oceanic turbulence have been examined. The author measured the steady heat flux from glass coated ellipsoidal microbead thermistors, computed numerically the unsteady heat flux from an ideal constant temperature flat plate, and analyzed the steady flow calibration data of a paralene-c coated plate thermistor as well as the steady and unsteady calibration data of two conical constant temperature hot film anemometers. In order to understand the behaviour of probes alone, thermal models of these sensors have been developed. The models incorporate a Nusselt number governed steady heat flux from the wetted surface and the thermal effects of a surface coating and a supportive substrate. Derived functional relationships between the steady heat flux and the flow rate agree favourably with the available calibration data. The quasi-steady sensitivity of these probes when used as anemometers or thermometers as well as their signal contamination by temperature or velocity are calculated using the functional heat flux relationships. The substrate and the coating reduce the sensitivity to temperature and to speed as well as the ratio of speed-to-temperature sensitivity. The response of sensors is not governed by the Nusselt number when the boundary layer is unsteady. The unsteady response of a sensor to velocity oscillations is governed by its unsteady viscous boundary layer and may increase with increasing frequency over some frequency ranges. The response bandwidth is wider for velocity than for temperature. The ratio of unsteady temperature-to-velocity sensitivity is highest at zero frequency. Frequency response calibrations methods must realistically simulate the sensor's unsteady viscous boundary layer. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate