No description available.

No description available.

No description available.

No description available.

Computational fluid dynamics (CFD) has been used to analyze the flow fields generated by FDG in the quasi-stagnant configuration to allow that technology to give simultaneous measurements of deposit thickness and strength. Stress field predictions were generated by solving the governing equations using the numerical solver <i>Fastflo™</i>, and validated by comparison with experimental hydrostatic pressure measurements. Moreover, an analytical approximation for the shear stresses imposed during the gauging experiments has been shown to yield reasonably good agreement with the CFD predictions. Particular predictions of note were that the surface shear stress is largest in the area directly underneath the rim of the gauging nozzle, and that shear stresses characteristic of industrial cleaning-in-place systems could be generated in these quasi-stagnant systems by suitable selection of the hydrostatic suction head and the clearance between the gauging nozzle and the surface. This enhanced FDG tool was employed to study the removal characteristics of tomato paste soils and weak calcium sulphate scale deposits. The critical shearing yield stress for the tomato paste was found to be strongly dependent on the extent of baking (ageing), approaching an asymptote as the material was transformed from a soft, malleable paste into a hard, brittle semi-solid. It was also found that after extended deposition and ageing, calcium sulphate layers could not be removed by these gauging flows, implying that liquid velocities customarily used in industry would not be able to dislodge such deposits. Polystyrene co-polymer fouling layers were fabricated in the laboratory. The kinetics and mechanisms of cleaning them from stainless steel surfaces using NaOH, MEK and TPU solvents were successfully investigated using FDG. NaOH was found to be a non-solvent across the range of experimental conditions, in the sense that the polymer films simply swelled to constant thickness after an induction period, without dissolving. In contrast, the polymer films were removed completely in MEK and TPU, more rapidly in the former case, in terms of swelling and dissolution rates, and overall cleaning time. The removal characteristics of different polymer films in MEK and TPU were shown to differ fundamentally, and an improved 'campaign' strategy for polymerization reactors is thereby identified. Experiments on pilot plant samples immersed in MEK and TPU indicated that the results from the laboratory samples are applicable to pilot plant scale.

No description available.

It is known theoretically that infinitely long fluid loaded plates in mean flow exhibit a range of unusual phenomena in the 'long time' limit. These include convective instability, absolute instability and negative energy waves which are destabilized by dissipation. However, structures are necessarily of finite length and may have discontinuities. We have undertaken an analytical and computational study to investigate the response of finite plates, and of plates with local inhomogeneities, to ascertain if these unusual effects might be realized in practice. Analytically, we adopt Crighton & Oswell’s (1991) structural acoustics approach and take a "wave scattering" --as opposed to a "modal superposition"-- view of the fluttering plate problem. First, we derive the energy balance relations for the extended plate (i.e., plate with two sided flow, spring foundation and plate pretension) and define a generalized wave impedance valid for both positive energy waves (PEW) and negative energy waves (NEW). Next, we solve for the scattering coefficients of localized plate discontinuities using a multipole source approach. Our solutions are exact and include the nearfields due to fluid-loading effects. We introduce the concept of power normalized scattering coefficients, and show that overall power is conserved during the scattering process if the sign of the wave energy is preserved. We argue that energy conservation, combined with the presence of NEWs on the plate, are responsible for the phenomenon of over-scattering, or of amplified reflection/transmission. These are scattering processes that draw energy from the mean flow into the plate. Next, we use the Wiener-Hopf technique to solve for the scattering coefficients of a variety of plate leading and trailing edge conditions -- including the flag like configuration of a free trailing edge with wake. We find that the edges are over-reflective in the frequency range where NEWs are present. The exception is a free trailing edge with wake where, remarkably, the wake is found to absorb almost all of the incident wave energy. We use combinations of these upstream and downstream edge reflection matrices to solve for the complex resonance frequencies of long, finite plates immersed in mean flow. Finally, we construct the response of a finite plate by a superposition of infinite plate propagating waves continuously scattering off the plate ends. We solve for the unstable resonance frequencies and temporal growth rates for long plates. We derive upper and lower bounds on the unstable growth rates of finite plates with given edge conditions. We find that a flag-like configuration of a clamped leading edge and a free trailing edge with wake is destabilized for sub-critical flow speeds only for very long plate lengths and only in the presence of convectively unstable waves. We present a comparison between direct computational results and the infinite plate theory. In particular, the resonance response of a moderately sized plate is shown to be in excellent agreement with the long plate analytical predictions.

Prediction of regenerative machine tool chatter requires a knowledge of the cutting forces existing under vibration conditions. Previous direct experimental work has been very limited in its range and has given results which contradicted those derived from observations on regenerative chatter. A description of suitable experimental rigs and the results obtained from these rigs are given. The results show that there are large phase changes of cutting force relative to chip thickness when vibrating. A simplified criterion is given for assessing machining stability and it is shown that the observed phase changes will greatly affect machining stability; in some cases machining will be stable regardless of machine damping. The requirements for machining stability contradict those for long cutting tool life so a compromise must be reached in practical cases. The cutting process is not, in general, amenable to quantitative analysis but theories have been put forward to explain the observed force phase and amplitude changes. There is a reasonable qualitative correlation between experiment and theory for most of the experimental range. Temperature estimates involved in phase change estimates show that there are large tool temperature variations under chatter or milling conditions. These temperature variations and the consequent thermal stresses are likely to cause tool failure with brittle materials such as carbides. The results obtained under steady cutting conditions such as lathe turning have been extended to intermittent cuts as encountered in milling. The tests show that there is no fundamental difference between steady and intermittent cuts. Suggestions are put forward for the large amount of further work required to clarify the mechanisms of the outing process and to extend cutting force knowledge for chatter prevention.

This dissertation describes the experimental investigation of rotating stall in compressors of high hub-tip ratio. Measurements were obtained in builds of one to four stages, covering a wide range of design flow rates and also a change in design reaction. The purpose of the experimental work has been to gain a better understanding of the details of the flow in the stall cell, so as to make possible the prediction of the in-stall performance of axial flow compressors. A new, on-line, phase-lock sampling technique has been developed, which, for the first time, allows the quasi-continuous recording of instantaneous velocity, pressure and flow direction measurements from within the stall cell. The technique provides results of superior quality to any hitherto available, and has been used to study the details of the flow under various test conditions. In particular, the axial and radial profiles of the stall cells in multi-stage compressors of different design flow rates were considered. The effects of changes in design reaction, blade row spacing and the influence of the number of stages in the compressor were also investigated. Finally, measurements were obtained to try to determine the differences between full-span and part-span stall. From a study of the results obtained, a fundamentally new picture of the flow in the stall cell has been built up, which shows features that are at variance with conventional ideas about stall cell structure. It has been shown that within the compressor itself, the edges of the cell do not follow streamlines in the unstalled flow, as might be expected of a dead wake, but rather assume an attitude implying tangential flow across the cell. The presence of this tangential flow is easily detectable from the measurements obtained within the cell, and is supported by the observation of extremely high whirl velocities just upstream of the rotor blades. Comparatively little flow is shown to pass axially through the stalled region, but, because of the opposing accelerating and decelerating influences of the rotor and stator blades, the axial movement of this small amount of fluid has been shown to lead to the dissipation of large amounts of mechanical energy. This results in the overheating of the compressor during stall. Centrifugal effects in the swirling flow ahead of the rotor blades are shown to contribute to strong radial pressure gradients in the stall cell, and are thought to be responsible for the fact that the pressure rise across a stage in the stalled region occurs ahead of the rotor blades. This observation is included in an overall model of compressor performance which is then used to explain the observed relationships between the various time-averaged compressor characteristics. This performance model is also used to provide some explanation of the observation (first made by McKenzie) that the total-to-static pressure rise across a stalled compressor appears to be independent of the blading used, but increases by a fixed amount for each stage in the compressor. Time-averaged measurements were obtained from all the compressor builds tested, and were used in conjunction with the overall flow model to formulate a new correlation for predicting the performance of a stalled compressor. The correlation makes use of the above idea that the pressure rise during stall is independent of the blading used, for both part-span and full-span stall, and relies on the concept of a critical level of cell blockage to distinguish between operation in either of these two modes. The work reveals the effect of the number of stages and the design flow coefficient on the behaviour of the compressor, and demonstrates the unexpected influence of system parameters (such as throttle line slope) on the stalling performance. The correlation also makes it possible to predict the occurrence of either full-span or part-span stall at stall inception from a knowledge of the unstalled characteristic, and allows estimates to be made of the size of the stall/unstall hysteresis loop. Data obtained from the literature are successfully correlated by this approach, and a new basis is therefore established from which to view the stalling behaviour of all axial flow compressors.

No description available.