No description available.

The power losses occurring under two-dimensional magnetising conditions have received a great deal of attention in recent years. They are very important in the three-phase electrical machines, and have been investigated for a long time in order to understand the physical processes and the mechanism of such losses. However, the accuracy of the measurements of rotational power loss is proved to be not sufficient and the differences between clockwise and anticlockwise power losses have been unexplained for many years. A computerised two-dimensional single sheet tester for the measurements of power losses in electrical sheet steel have been developed. A unique adaptive iterative digital feedback algorithm has been developed it allows magnetisation under controlled flux density and magnetic field patterns in wide range of frequency from 1 Hz to 1000 Hz and in various magnetising yokes. A new type of flux density sensor has been proposed and discussed. The experimentation did not confirm theoretical predictions. Further work needs to be carried out on this subject. It was found that the power losses measured in clockwise and anticlockwise direction of rotation differ significantly and that this difference is caused mainly by the angular displacement of the flux density and magnetic field sensors. The error of averaging from clockwise and anticlockwise losses has been estimated and the technique for minimising the difference has been mathematically proposed and implemented for the first time. Also, several factors contributing to the clockwise-anticlockwise loss difference have been discussed and their influence assessed. It has been found that the peak value of power losses measured under controlled flux density or controlled magnetic field patterns differ, and that this difference depends on the anisotropy of the sample. It has been proved that the peak losses for controlled magnetic field might be as much as six times of the losses measured under standardised alternating magnetisation. The quasi-static hysteresis component of power loss has been investigated for a non-oriented electrical sheet steel. Despite very high flux density of 2.0 T the static loss does not drop to zero. At higher frequencies the eddy current component of the total power loss takes over and causes the peak of the power loss characteristic to vanish. The effect is more visible for thick laminations, as expected.

Direct, real time domain observations in electrical steel during its magnetisation at 50 Hz and 75 Hz was limited due to low the limited sensitivity and frame sampling rate of the camera systems available. In this work, a high efficiency Kerr magneto-optic microscope was coupled to a recent commercially available high speed intensified camera. The superior efficiency of the developed system enabled domain dynamics measurement to be made at magnetisation frequencies up to 100 Hz in the real time mode and up to 1.8 kHz in stroboscopic mode. The application of the system was focussed on the study of grain-oriented electrical steel. Observations in this work revealed that the behaviour of the domain walls under sinusoidal excitation is far from being ideal sinusoidal motion. The deviations from the ideal behaviour can be summarised as non-sinusoidal, non uniform and non-repetitive domain wall motion. The dependence of the non- uniformity of wall motion and cycle-to-cycle non-repeatability on magnetisation conditions studied by direct real-time observation revealed that under certain conditions the non-uniformity and non-repeatability is much higher. Recommendations have been made on how to reduce losses by adjusting the magnetisation conditions and "magnetic history" of the steel core. The observations confirmed previous researchers' suggestions that the perfection of grain-to-grain orientation and grain size uniformity would improve wall motion uniformity, hence reducing power loss in the material. In addition to Kerr-microscopy observations, the application of a magnetic indicator imaging technique was demonstrated. Moving domains under the insulating, opaque coating on commercial grain-oriented electrical steel were observed using this method. The influence of surface roughness and the effect of coating tension on the domain structure and distortion of bar domains under dynamic magnetisation was observed. New experimental techniques and image processing programs have been developed such as the first attempt of automatic domain wall recognition and analysis by computer vision methods. It was demonstrated that the system and software tools were flexible for use with other materials such as steel, NiFe and CoFe alloys, amorphous ribbons and wires, ferrites and magnetic films.

The magnetic Barkhausen effect, which appears as abrupt changes in magnetisation, occurs when a ferromagnetic material is subjected to an external varying magnetic field. The origin of the effect is known to be primarily due to the discontinuous domain wall motion through a material caused by imperfections in the material. The Barkhausen effect is very sensitive to the changes in the microstructure and stress. Owing to this sensitivity, Barkhausen measurements can be used for the non destructive evaluation of a ferromagnetic material. Previous research has been mainly based on low frequency magnetisation to generate Barkhausen noise with a typical excitation frequency less than one Hertz. In this work, a measuring system is presented which is capable of capturing Barkhausen noise signals at a magnetisation frequency up to 100 Hz. Enwrapping coils and the ferrite cored surface sensor methods were used to measure Barkhausen noise, and although the surface sensor was oriented perpendicular to the sample, all the sensors produced very similar trends. Due to the Barkhausen noise signal being of stochastic nature, methods of analysing Barkhausen noise signals was investigated. RMS, total sum of amplitudes, power spectrum and kurtosis all showed repetitive characteristics and were used to analyse Barkhausen signals. A relationship between Barkhausen noise and the hysteresis component of total power loss is presented. The variation of Barkhausen noise signals and the tangential component of surface field over the surface of a grain-oriented 3%SiFe electrical steel was also examined and it was found that there is a strong indication that surface Barkhausen noise activity is influenced by grain-to-grain misorientation. In addition, the results infer that the hysteresis component of loss varies spatially with grain structure. An ancillary study was conducted on cold rolled low-carbon steel and thermally aged Fe lwt%Cu alloy to examine the effects of Barkhausen noise due to rolling reduction.

No description available.

In this research, cobalt ferrite thin films were deposited at 523 K which eliminates the need for annealing at higher temperatures and offers hope for integration of the thin films into micro-electromechanical devices.

A system approach to variable speed drives consisting of two low voltage induction motors (Y and A connection) fed by matrix and two-level DC-link converters is developed (chapter 7). The ratio of iron to copper winding losses is studied based on the induction motor equivalent circuit and constant V/f converter control. Electrical efficiency increased by up to 10% (partial load and low speed) in Y-connected motors fed by the matrix converter. While similar values of efficiency were observed in A-configured motors, up to 8% higher output power is extracted from the machines under two-level DC-link converter supply.

The sub-micron surface region of poly(ethylene terephthalate) (PET) films and bottles has been investigated. Uniaxially drawn and annealed PET films were characterised by bulk and grazing incidence X-ray diffraction (GIXD). Surface-specific diffraction patterns collected below the critical angle (ɑc) for total external reflection reveal a radial shift in diffraction peaks, whereas bulk signals collected above this critical angle occur at the predicted locations. It is calculated that this shift can be described by a ~5 % reduction of the a, b and c unit cell imrameters for PET crystals within the sub-micron surface region. This shift is less apparant in the un-oriented specimens. The crystallite orientation is consistently higher within the surface region of the drawn PET films than the bulk. There is a good agreement between the X-ray diffraction and FTIR orientation measurement over the range of films analysed. Polarised FTIR-ATR spectroscopy was used to map the molecular orientation along the inner and outer surfaces of stretch-blown PET bottles. This work reveals complex orientation trends. For 2-litre bottles, the inner surface shows consistently higher orientation levels than the outer surface. There are also two orientation switches at the shoulder and base regions on the outer surface, but only one switch on the inner wall. In all cases, the maximum orientation occurs on the inner surface at 4-6 cm from the base. This analysis was repeated for various bottles, including examples produced from various polyester co-polymers and a series of bottles stretch-blown after different preform heating times. Heating for 5 seconds longer than the recommended time prior to the stretch seriously reduces the orientation in the finished bottle. Lowering the polymer molecular weight has a similar effect. The use of the popular 1340 cm"' PET IR band for crystallinity evaluation was questioned. It is concluded that the use of this band for crystallinity calculation is erroneous if the specimen possesses molecular orientation. It is shown that the 1340 cm"' band height reveals the orientation trends in highly drawn specimens of low crystallinity. Simple measurements reveal that the 1340 cm ' transition moment for this band lies approximately in the chain direction.

Haynes Alloy 230 is a nickel-based sheet material used for combustor components in a number of small industrial gas turbines manufactured by Siemens. During normal operating service the material is subjected to high temperatures and cyclic mechanical and thermal stresses, which can lead to degradation of the microstructure and mechanical properties of the alloy, and hence limit component design life. This PhD thesis describes studies of the effects of long-term high temperature exposure on hardness and microstructural changes in creep rupture tested and thermally exposed samples of HA230. Advanced FEGSEM was used for documentation of the microstructural evolution in the temperature range of 500-1170°C following exposure up to 30000 hours. For identification of alloy phases, which could potentially act as indicators of the average exposure temperatures experienced for specific service periods, XRD and analytical TEM techniques were used. Primary carbides M12C and secondary carbides, M12C and M23C6, were identified. No TCP phases were observed within the microstructure of the tested samples, but the secondary M23C6 forms large ‘pool-like’ precipitates’ above 990°C which are very hard and brittle and could be detrimental for the material (e.g. reductions in ductility, high temperature strength). A hardness and microstructurally based model was used for the assessment of an ex-service HA230 transition duct and could be an aid for evaluation of the remnant life of HA230 combustor components. The tests were carried out as part of the COST 538 technology programme between 2004 and 2009.

Composite materials are important for their improved strength and wear resistance compared to pure materials. Metal composites have been electrodeposited from aqueous solutions although the instability of colloidal dispersions makes it difficult to obtain reproducible composite compositions. In this thesis metal composites are prepared using Deep Eutectic Solvents, DESs, which are a type of ionic liquid. The electro-deposition of copper and copper composites from DESs based on ethylene glycol and urea were investigated. The mass transport in both electrolytes was found to be diffusion controlled. The mechanism of copper nucleation is studied using chronoamperometry and it is shown that 3D progressive nucleation leads to bright nanostructure deposits. It was found that dispersed colloidal particles are stable over a prolonged period of time in DESs. This work uses an electrochemical quartz crystal microbalance (EQCM) to monitor both the current efficiency and the inclusion of inert particles into the copper co-deposits. The technique showed the majority of second phase was dragged onto the surface instead of sedimented. The effect of the addition of surfactants was also studied. DESs were also used in the electro-deposition of silver composites with alumina and silicon carbide. It was found that improved mechanical properties such as hardness and wear resistance could be obtained. The addition of LiF to the Type 3 deep eutectic solvents was also found to further improve the quality of the mechanical properties and led to smoother surfaces with lower friction coefficients. The final part of this study involved the deposition of nickel composites with the aim of producing coatings with hardness similar to chromium. The addition of organic additives as brighteners (ethylene diamine) leads to smooth and bright finishes. Two types of composites were added to the nickel electrolytes; SiC 1-3μm and Teflon μm. The surface morphology was found to change with particle size and type. The mechanical properties were studied and found to be affected by the type and the concentration of particulate incorporated in the nickel deposits.