This thesis describes the research with drag reducing additives which was carried out on a part time basis between 1965 and the present date. As is inevitable with such a rapidly expanding field of study, between 85 and 90% of the published work of others in this field has occurred since the commencement of the present research. When the work was started surprisingly little was known about the drag reduction phenomenon in detail. At an early stage it was decided to attack the problem on a broad front in order to provide as much data as possible for likely practical applications and to throw light on the mechanism causing the effect. The initial approach had of necessity to be largely empirical. The work is concerned with internal flows through rough and smooth pipes, external flow over bluff bodies and other anomalous effects involving heat transfer cavitation and submerged jets. The thesis is conveniently presented in sections which are linked together with a general discussion. Many of the results herein have been previously published by the author in a number of scientific journals and symposia transactions. In some instances similar findings have been reported by other workers almost simultaneously, which is not surprising with such a rapidly developing topic. It is therefore very difficult to designate with certainty to any worker the priority for several of the discoveries. A collection of the publications by the author are presented in the flap at the end of this thesis, and the dates of publication will provide some indication of the time scale of the subject development during the course of this study.

As an, aid to jet pump design and performance analysis, a theoretical investigation on turbulent confined jet mixing in a non-uniform axisymmetric duct typically used in jet pumps and ejectors has been undertaken. A so-called Prandtl-Kolmogorov two-equation turbulence model, with turbulent kinetic energy k and turbulent energy dissipation rate E as the two parameters, is incorporated into the time-mean Navier-Stokes equations to form a complete set of partial differential equations which describes the turbulent flow mathematically. The equations are solved numerically via a primitive pressure-velocity finitedifference procedure using a digital computer. The timemean static pressure, velocities, turbulent kinetic energy and dissipation rate are predicted directly throughout the whole flow field. To validate the computer model, predicted time-mean static pressure and velocity as well as turbulent shear stress for flow in a uniform bore mixing tube are compared with the published results. The method is then extended to predict flows in conical diffusers and typical jet pumps. The predictions are also compared with the available experimental data. A laser Doppler anemometer is used to measure the mean and fluctuating velocities of water jet mixing in a uniform perspex mixing tube with a centrally located nozzle. The measured data which enable turbulent kinetic energy to be evaluated, are compared with the computer predictions to further consolidate the theoretical model. Finally, the computer model is used to predict the performance of a proposed jet pump and to investigate the influence of various geometrical parameters on jet pump performance. The capability of the computer model as a useful design tool is also demonstrated via an optimization procedure to give the optimum geometry for a given design specification.

No description available.

Since O. Reynolds' experiment revealed the transition from laminar state to turbulence in pipe flow (1883), numerous efforts had been exerted in order to understand the mechanism. Linear hydrodynamic stability theory was established in the first half ofthe twentieth century. This theory predicts the growth of small disturbances during the initial stage of transition. As an important achievement, the Tollmien-Schlichting (T-S) instability, which represents the initial evolution of disturbances in two dimensional shear flows, has been described and verified in boundary layer experiments. Once the enonnous benefits from the delay of flow transition (and then turbulence) were acknowledged, diversities of flow control techniques have been motivated. In fact, some relevant techniques have already been demonstrated, such as the drag reduction oflarge aeroplane through laminar flow control. At the same time, the development of control theory makes it possible to implement deliberately designed systems to control transitional flows. After decades of advances with passive control, the concept of active control became widely accepted in fluid mechanics in the early 1980s. . The study in this thesis attempts to find an efficient closed-loop active control scheme to cancel the unstable Tollmien-Schlichting waves in a transitional boundary layer flow. The T-S instabilities was simulated by numerical solutions of the linearised Navier-Stokes equation system. Controllers were designed according to the frequency characteristics ofthe disturbed flowfield and implemented on the basis of the classical control theory in frequency domain. The perfonnance of controllers in the disturbed boundary layer flow along a flat plate were evaluated through numerical simulations and then validated by experiments.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.