No description available.

No description available.

No description available.

Digitized by Kansas Correctional Industries

Digitized by Kansas Correctional Industries

The behavior of different parameterizations of mixed layer physics when used in an oceanic general circulation model (OGCM) having coarse resolution of the upper ocean is examined. The method of parameterization is expected to have an important effect on the resulting sea surface temperature, and hence affect the model's overall fidelity from the viewpoint of air-sea interaction. Tests of three possible parameterizations differ in the manner in which the mixed layer depth is determined: predetermination, diagnostic determination, or prognostic determination. The sea surface temperature is taken to be equivalent to the top OGCM layer temperature in the first two methods, while it is found prognostically in the third method. Results show that for typical forcing cases such as strong insolation, weak surface cooling or weak winds, mixing is insufficient to cause heat transfer between the top two OGCM layers, which occupy the uppermost 500 m of the model. The predetermined and diagnostically determined mixed layer depth parameterizations reduce to a diffusive mixing parameterization, while the prognostic approach satisfactorily models mixed layer depths for all forcing cases. The prognostic method also agrees most closely with the results of a mixed layer model and with observations. / Graduation date: 1978

We study the electronic band structure of bulk within a six-band bond-orbital model. All interaction parameters of this model involved are directly related to parameters for describing bulk bands near the zone center in the k¡Ep finite difference method. To study the conduction-valence band mixing effect, we calculate the electronic band structure for the InAs-GaSb superlattice, within a six-band bond-orbital model. In the InAs-AlSb-GaSb-AlSb superlattice, we find that energy-gap of this material will change rapidly with different AlSb layer thickness. This indicates that the e-X line observed in far-infrared cyclotron-resonance spectrum is originated from conduction-valence band mixing effect. This result is in good agreement with the experimental results.

Abstract Stirred tanks are typically used in many reactions. The quality of mixing generated by the impellers can be determined using either experimental and simulation methods, or both methods. The experimental techniques have evolved from traditional approaches, such as the application of hot-wire anemometry, to more modern ones like laser Doppler velocimetry (LDV). Similarly, computational fluid dynamics (CFD) simulation techniques have attracted a lot of attention in recent years in the study of the hydrodynamics in stirred tanks, compared to the empirical modelling approach. Studies have shown that the LDV technique can provide very detailed information on the spatio-temporal variations in a tank, but the method is costly. For this reason, CFD simulation techniques may be employed to provide such data at a lower cost. In recent years, both integrated experimental and CFD approaches have been used to determine flow field and to design various systems. Both CFD and LDV data reveal the existence of flow maldistribution caused by system design features, and these in turn show that the configurations that have, over the years, been regarded as standard may not provide the optimal operating conditions with regards to the system homogeneity and power consumption. The current trends in CFD studies point towards an increasing application of more refined grids, such as in large eddy simulation, to capture turbulent structures at microscales. This trend will further improve the quality of the simulation results for processes such as precipitation, in which micromixing and reaction kinetics are important.

No description available.

The role of molecular mixing (as opposed to molecular-collision transport) in the description of turbulent diffusion in continuum framework is examined. This is done by comparing a new virtual fluid parcel treatment with the classical fluid particle treatment of the BMDFE (Basic Macroscopically Describable Fluid Element). It is found that the classical fluid particle treatment conceptually excludes molecular mixing between different BMDFEs, due to its postulated constraint that individual BMDFEs maintain their integrities in motion. The new virtual fluid parcel treatment, on the other hand, conceptually incorporates molecular mixing between different BMDFEs, by relaxing this constraint to permit disintegration of individual BMDFEs. The main improvement made by the new virtual fluid parcel treatment lies in the introduction of a feedback mechanism in the form of physically coupled disintegration and integration of the BMDFEs. This improvement suggests that molecular mixing is a controlling agent of the mixing mechanism in every time-step of turbulent diffusion, whose significance would increase cumulatively. By applying the two treatments to the evolution of the diffusion cloud on the level of single time-step diffusion redistribution, it is shown that molecular mixing persistently and cumulatively influences the evolution of the diffusion cloud by reducing the diffusion distribution variance. This indicates that the exclusion of molecular mixing in the classical fluid particle treatment would lead to a potential mathematical-physical inconsistency in the description of turbulent diffusion by exaggerating the diffusion distribution variance. The results of this analysis are qualitatively supported by experiments of passive scalar diffusion in water flow with moderate turbulence intensity. As a preliminary test, a simplified numerical modeling of scalar diffusion based on the virtual fluid parcel treatment is executed in two wind tunnel models. In this case, measur