No description available.

No description available.

No description available.

No description available.

A falling body viscometer with self-centring sinkers has been developed and used to measure the viscosities of benzene, carbon tetrachloride and eight halogenated hydrocarbons at pressures up to 500 MN m⁻² in the temperature range 25˚C to 100˚C. Two isotherms of water have also been measured at pressures up to 1000 MN m⁻² . Details of the viscometer and its pressurising equipment are given together with an analysis of the performance of the system. The results are estimated to be accurate to within ±2 per cent, and show good agreement with other measured data where available. The change of viscosity with pressure for the halogenated hydrocarbons is generally similar to that of other simple liquids. The results show that liquids having molecules of similar shape also have a similar change in relative viscosity with pressure. Theories of liquid viscosity are reviewed and constants required by Eyring's significant structure theory are obtained for more than sixty liquids using literature data. Methods are derived for calculating these constants from correlations with readily available critical properties or chemical structure. The correlations are tested using the new measurements and literature data. These tests show that the methods derived work well if one or two values of viscosity at atmospheric pressure can be used but are less reliable if only structural information is available. Since at least one measured viscosity isavailable for most liquids it is concluded that the method will be useful for predicting viscosities at other temperatures and pressures.

Low-speed tests have been carried out on a branched annular diffuser system having a geometry similar to that employed in some gas turbine engine combustion systems. The system comprised a straight walled pre-diffuser followed by a sudden area expansion in which the flow was divided between two concentric annuli separated by a bluff body simulating a combustion chamber. The overall geometric area ratio was maintained at 2.0 and all tests were carried out with fully developed flow at inlet. The design flow split between the outer and inner annuli was 2.15:1. The system was tested with five different pre-diffuser geometries to show the effect of increasing the area ratio, increasing the included angle and canting the pre-diffuser. For each pre-diffuser geometry the influence of varying the flow split and the axial distance between pre-diffuser outlet and combustion chamber head (dump gap) were investigated. In addition to determining the overall performance characteristics, the pressure losses for the inner and outer flow fields were calculated and the losses further sub-divided in order to identify regions of high loss.

Stepping motors have been in existence for over thirty years. The first stepping motors were electrical but were limited in their power output because of the relatively high inertia of the electric motors. In order to overcome this difficulty hydraulic torque amplification has been used. The resulting hydraulic position control systems then have the advantage over conventional positioning systems of not requiring the usual electrical feedback links. A further increase in the power output of these systems can be obtained by controlling a standard hydraulic motor to perform in a stepping mode. A stepping system has been developed to give accurate positioning for a range of load conditions, but under certain conditions where the load had a very slight inertia a steady final position was not obtained and additional system damping was required to give a satisfactory response, Various means of optimising the stepping performance characteristics were investigated experimentally with various degrees of success. Each method generally improved one aspect of motor performance, such as rise time or initial overshoot, at the expense of another. The method to be used in practice would depend on the load and the requirements surrounding its use. A detailed analysis has been made of the effect of changes in the main operating parameters during the stepping phase. This analysis has led to a better understanding of the operation of the system. The behaviour of the stepping motor was simulated using several different models on both analogue and digital computers. These simulations enabled the effects of modifications to the stepping motor design to be assessed. Fundamental design changes were considered which would be difficult and expensive to alter in practice. The means of sequential control of the stepping system was also considered. In particular pure hydraulic systems were investigated as these eliminate the need for interfaces. This led to a proposal for a sequential control system based on miniature moving part hydraulic elements. The thesis considers the various applications of stepping devices. It also analyses a possible application for a stepping motor system as a governor for a diesel engine. This was tested by means of a computer simulation and the results showed that this type of system could give accurate yet relatively simple control with some advantages over standard engine governors.

The suitability of a prototype flowmeter for the measurement of transient flows in high pressure hydraulic systems is examined. The flowmeter uses hot film anemometry and is capable of measuring both steady-state and dynamic flow velocities. The flowmeter has been subjected to an extensive series of calibration trials under steady-state flow conditions over a wide range of hydraulic fluid temperature and pressures. Dynamic calibration has also been obtained for the meter response when subjected to sinusoidal and step changes in flow. An investigation into the generation of pressure and flow transients in a simple pump-pipeline-termination hydraulic system is also undertaken using the prototype flowmeter. Employment of suitable analytical techniques enables the harmonic content of the complex pressure and flow waveforms to be determined over ten harmonics of pump piston frequency. A digital computer model of the hydraulic system has been developed based on the Impedance concepts, and the results compared with those obtained experimentally. The accuracy of the simulation for pressure and flow ratios at positions on the pipeline is considered to be good throughout the frequency range. Methods have been developed to determine the source impedance spectra of the pump from experimental data. Inclusion of the predicted source impedance values in the model gave good agreement between theoretical and experimental pressure and flow waveforms. Having obtained the source impedance enabled the pump flow spectra and instantaneous flow to be established using the theoretical model and the experimental flow harmonics measured at a point on the pipeline. This latter part of the work gives a method of remote measurement of pump discharges which would be extremely valuable for test and development work. Further slight modification would allow it to be used for test and development of valve characteristics.

The aim of this investigation was to develop a computer-aided design procedure for fluid couplings and torque converters. Three different classes of fluid transmission devices are considered i) The coupling in which input and output torques are equal throughout the working range, with efficiency rising from zero to a maximum at the minimum available slip condition; ii) The torque converter which is designed to give an increased torque with maximum efficiency at a turbine rotar speed which is less than that of the pump impeller; iii) The converter coupling in which the torque ratio varies from a maximum with the turbine runner stalled to approximately unity at coupling speed. The analyses are carried out by assuming that the fluid circulating within the closed circuit has a distributed velocity profile which increases linearly from zero at the mean radius, the radius about which the fluid circulates. That is the circulating motion is a forced vortex. The analyses have been developed by applying the fundamental one-dimensional flow relationships for continuity, moment of momentum and energy to a single streamtube and integrating over the limits of the device considered. This basic procedure is applied to all different types of hydrodynamic devices considered- Each device, e.g. baffled coupling, partially filled coupling, torque converter, etc., have their own specified problems which are considered and described separately. In all cases the losses occurring in the flow path have been considered to be composed of three components (i) incidence or shock loss, (ii) skin-friction loss and (iii) secondary flow circulation or bend loss. Empirical procedures are developed to describe these losses and their relative magnitudes compared. The theoretically predicted performance of each hydrodynamic device is compared with simpler theoretical procedures, available in the literature, and with experimental results available from industry. In general good agreement between theory and experiment is obtained. It is shown that the theoretical approaches developed here are an improvement on earlier ones where these were available, while in certain other cases, particularly those of the baffled and partially filled couplings, the approach may be claimed to be novel.

This thesis describes work concerned with the theory and development of a method of analysis of the flow in regenerative machines and its use to calculate the flow details and predict the overall performance of specified pumps. A flow model is assumed based on the view that these machines operate as genuine rotodynamic devices. The equations derived reveal that a) the familiar Euler's equation has to be supplemented by an extra term accounting for the tangential pressure gradient so that it can apply to rotors of regenerative machines, and b) to compute the head rise, the tangential displacements of flow in the rotor have to be determined. The meridional velocity field is represented by a streamfunction given by an empirical expression which is determined iteratively by solving the flow along streamlines and applying the condition that the net head rise per flow cycle should exactly match the product of the circumferential displacement and pressure gradient. An attempt is made to write both the momentum and energy equations such that the effect of friction is incorporated with reasonable consistency. Conventional models are used to estimate skin friction losses in the blade passages and in the channel, losses at entry to impeller blading and the effect of slip. However, towards compeltion of the work it was realised that the tangential pressure gradient should enhance the slip and discussion is presented of this aspect. In modelling the flow in the port and stripper regions, leakage and carry-over losses are included. However, a convincing model of the ports losses which accounts for their-design remains to be established. Computer programs written to implement the method on a digital computer are described and samples of the results of their use to investigate the flow in pumps of various configurations are discussed. It is found that the method is generally stable and convergent and it seems to enable the magnitude and interdependence of various flow parameters to be quantified. The computed overall performance characteristics led to the realisation that the slip effect is a function of the tangential pressure gradient. They can also be interpreted as suggesting that the length of the effective pumping passage varies with the pressure gradient. Suggestions are made for the improvement of the accuracy of prediction and refinement of the procedure. This work provides some novel findings regarding both the theory of regenerative machines and the procedures for the solution of their internal flows. 40-word summary suitable for retention in an automatic data processing system: 'Regenerative (side-channel) machine theory is investigated. Euler's equation requires an extra term, computable only from the flow tangential displacements in the rotor. The circumferential pressure gradient increases the relative eddy (slip). Overall performance predictions made employ detailed internal flow calculations'.