This research work covers the development of a novel compact model for a Floating Gate nMOSFET that will be added to the QUCS library. QUCS (Quite Universal Circuit Simulator) is a GPL simulation software package that was created in 2006 and is continuing to develop and evolve, and there is already a substantial library of components, devices and circuits. Fundamentally, a floating gate device is an analogue device, even though modern technology uses it mainly as a non-volatile memory element there are numerous uses for it as an analogue device. A study has been carried out with regards to the principles of the physical phenomenon of charge transfer through silicon dioxide to a floating gate. The study has concentrated on the physical properties of the fabricated device and the principles of charge transfer through an oxide layer by Fowler–Nordheim principles. The EKV2.6 MOSFET was used as the fundamental device for the model that has been adapted by the addition of the floating gate. An equivalent circuit of an FGMOSFET was developed and analysed theoretically. This was then formulated into the QUCs environment and created as a compact model. Simulations were carried out and the results analysed to compare with the theoretical expectations and previous research works. It is well documented that the creation of equivalent circuits for floating gate devices is complicated by the fact that the floating gate is isolated as a node and as such cannot be directly analysed by simulators. For the equivalent circuit created, circuit analysis was achieved by the introduction of high value resistances connected in parallel with the capacitive elements that are representative of the incursion of the floating gate that is intrinsic to a floating gate device. The resistance elements were of such value that the time constants were of the order of 10000s and did not interfere with the simulation. The equations from the analysis were formulated and the anticipated responses were shown. The analytical equations developed were then used within the QUCS environment with explicit use of EDDs(Equation Defined Devices) to create a novel model of a FGMOSFET. Simulations of the model created were carried out with a range of voltage pulses being applied to the tunnelling terminal to affect a charge transfer to the floating gate by means of Fowler-Nordheim principles. The changes of the charge stored on the floating gate were clearly shown by the measured anticipated associated shift in the Threshold voltage. Simulated results have been compared with previous research and development work and the new model is considered effective. Also because of the ability of the QUCS software to allow the variation of the multiplicity of the parameters associated with the fabrication process, it is considered to be adaptable to a range of modern floating gate device structures and materials.

This project aims to investigate the influence of UV on the properties under natural and artificial weathering. The real-time outdoor weathering exposures provide the most accurate results but, they are very slow and manufacturers cannot afford to wait in order to see if a new or improved product formulation is really an improvement. So, accelerated testing methods are designed to simulate natural weathering with the combined action of the most weathering damaging factors; UV radiation, oxygen, temperature and water. Whatever the application, there is a concern regarding the durability of the products because if its useful lifetime can be estimated in short time, their maintenance and replacement can be planned in advance. In this study, recycled PET samples were exposed to UV outdoors and to accelerated weathering up to 13,000 hr. The mechanical behaviour (tensile and impact), thermal behaviour (DSC), molecular structure analysis (FTIR), flow characteristics (MFI) and surface properties (colour and gloss) of the samples were investigated. The whole body of the samples was substantially unaffected by degradation for long exposure time and this is confirmed by MFI and DSC results. However, the surface of the samples is affected which is seen in FTIR analysis, colour and gloss change. This explains that the photodegradation is a surface effect. FTIR analysis shows an increase in the gauche ratio and decrease in the trans ratio for both types of weathering indicating a less ordered structure after the UV radiation due to chain scission by Norrish Type I and II mechanism. Carboxylic acid is formed as a result of degradation and increases with the extent of UV exposure in both types of weathering, the formation of carboxylic acid by Norrish II is dominated over Norrish I. Colour and gloss measurement shows that the effect of degradation takes place strongly in the first 2000 hr of exposure in natural weathering and extended up to 5000 hr in accelerating weathering due to the formation of microcracks. Mechanical tests show that the UV effect was not significant on bulk properties such as yield stress and elongation at yield for accelerated weathering samples up to 1000 hours, then dropped by 62% and 57% respectively up to 13000hr of exposure and remained unchanged for the whole period of exposure for outdoors samples. The drop in the failure stress and elongation at failure for accelerated weathering samples right from the beginning indicate the formation of microcracks is from the early stages of exposure. For outdoor samples, failure stress remained unchanged for the first 1000 hr of sunlight exposure, then decreases progressively with increasing exposure in the environment. The elongation to failure is unchanged up to 13000 hours. After 13000 hrs of exposure to sunlight, r-PET samples failed to break while those exposed to UV lamps failed in a brittle manner under impact after 250 hrs and this indicates the transition from ductile to brittle behaviour just after 10 days of accelerated UV exposure because of crack formation. For accelerated weathering samples, the impact strength remained unchanged in the first 1000 hr of exposure, then a decrease up to 5000 hr of exposure when the effects of flaws become significant and dropped sharply by 85% after 5000hr of exposure. The correlation between both types of weathering shows that one year in natural weathering is equivalent to one and a half months in accelerated weathering according to colour measurements. The effect of accelerated weathering is much bigger than the natural weathering and this is due to higher radiation dosage, temperature and humidity during the test which accelerate chain scission rate that lead to faster crack growth.

This work is concerned with an examination of how the Merchant Navy Navigators use the information provided by the different aids. It is divided into three main parts. First are a series of experiments where seafarers studying for their Master’s certificate were presented with cards containing information from a number of different navigational aids, and asked to plot the position lines on a chart and then decide where they would consider the ships' position to be. The second part was the design and construction of a simple non-interactive simulator based on slide displays and video recordings. The aids used were visual bearings, radar, Decca and the echo sounder. The information was taken from instrument readings recorded on board the training ship "Sir John Cass" during a voyage along the East coast from Southwold to Harwich. A number of flashing lights were included in the design to simulate the keeping of a lookout. In the third part a similar group of subjects were asked to navigate a ship on two simulated voyages of half an hour's duration each, while at the same time to log the number of lights they observed. As a measure of their navigational ability they were asked to prepare a course to steer and an E.T.A. for a point about half an hour's steaming ahead of the position at the end of each exercise. A total of 17 dependent variables were identified during the experiment and these were tested in pairs for correlation. From the results of these experiments it was possible to produce an order of the subjects' preference for the different aids, to demonstrate that the subjects preferred to use only two position lines when fixing their position and to examine how the subjects used these aids.

An AOP lumped crystal transverse modulator has been developed, based upon a novel crystal cut, which operates at much lower driving voltages than presently available commercial state of the art modulators. With the latter, the modulation bandwidth is limited by the requirement for driving voltages in the range of 115V to 250V, which are difficult and expensive to produce using existing semiconductor technology. The reduction in driving voltage leads to an increased modulation bandwidth; low voltage wideband drivers are readily available. The modulator's novelty lies in a composite crystal scheme which allows a light beam to be reflected back and forth many times. The voltage requirement is reduced by a factor dependent on the number of times the light beam passes through the crystals. This modulator does not require a half wave plate and can therefore operate at any optical wavelength as is also the case with some of the existing designs. But in addition, unlike the existing designs, this novel modulator allows for multicolour multibeam operation. Using four AOP crystals (15.0mm length, 16.5mm width and 2.5mm thickness) a half wave voltage of 65V has been achieved after seven passes. The extinction ratios obtained were in the range of 50:1 to 100:1. A theory developed and supported by experimental results shows that the extinction ratio depends upon the angle of propagation of the light bean through the crystals. Improvements in this area are suggested. Finally, a technique for extending the operational bandwidth of any modulator is presented. The bandwidth of the multipath modulator was 70MHz.

Considerable interest has recently been shown in the field of marine traffic engineering. Real life data sources made available for maritime studies are often expensive and inconvenient to collect. The marine radar simulator presents the researcher with a relatively inexpensive and readily available source of navlgatlonal data. With the improvement of remote vessel traffic monitoring systems the potential for inexpensive real life data analysis Is enhanced. The work of this study has been to allow the analysis of data archived from the Channel Navigation Information Service Automatic Data Processing system (CNIS AOP) installed at St. Margarets Bay Dover using contemporary digital computer graphical facilities, and to compare mariners' behaviour In a real life and simulator collision avoidance situation. For this comparison certain navigational situations known as encounters have been automatically detected using an extension of the Range to Domain Over Range Rate (RDRR) method (Colley et al 1983), referred to as the RDRR+ technique. A statistical comparison has been completed using non parametric techniques.

Integral equation techniques are used to solve interior and exterior (including cascade) Neumann problems. The equations are in terms of the boundary values of the derivatives of the harmonic function sought. The equations are derived using the complex variables forms of Green’s theorem. They are solved numerically using low order complex polynomial approximations. The solution of problems defined on regions with corners is considered. A complex integral representation which enables a regular function to be generated from its derivatives is derived. Approximations to the representation are given for the case when first order derivatives are known. These approximations are geared to the utilisation of the numerical solutions of the aforementioned integral equations. Beam element solutions of stress and vibration problems require knowledge of such section properties as torsional stiffness, warping stiffness and shear centre co-ordinates. These section properties are usually expressed as integrals taken over a uniform beam’s cross section. It is shown that they can be written as boundary integrals involving the beam’s torsion and warping functions. These functions are determined using the previously mentioned integral equations together with the representation of a regular function. The major practical application of the above techniques is to the analysis of turbomachinery blading at an early stage in the design process. Finally, the solution of plane elastostatic problems using complex potentials is considered. A 1st Kind integral equation for the first derivative of one of the Goursat functions is introduced. The equation is shown to be derivable from a 2nd Kind equation due to Muskhelishvili. The solution of the equation is discussed in terms of a simple numerical method. Where appropriate the numerical methods described in this thesis are illustrated by examples.

The corrosion and degradation of Pb-Sn bearing overlays Is a bilateral process . During service, the lubricant becomes corrosive due to the breakdown of inhibitors with subsequent oxidation occurring at operating temperatures (120-170[degrees]). ...

Some aspects of the corrosion fatigue crack initiation in EN1 type mild steel have been studied. Results of the electrochemical investigations show that the application of both static and compressive stresses enhance anodic dissolution. Application of cyclic stress assists the breakdown of the surface film resulting in localised attack, Metallographic studies indicate that a sulphur enriched band of ferrite exists around the non-metallic inclusions and corrosion occurs in this contaminated band with and without applied stress. Cyclic stress enhances the formation and coalescence of micropits at and around the inclusions leading to the nucléation of cracks. Preferential attack also occurs at the ferrite - pearlite interfaces, at slip band - matrix interfaces, and at grain boundaries. All these modes of attack can lead to crack nucleation. It has also been found that non-metallic inclusions can act as sites for hydrogen blisters; this may assist the nucléation of cracks.

A novel barcode system design to achieve high data storage using more than one layer is introduced theoretically and tested partially in the laboratory. Compared to other existing barcode systems, diffraction gratings are used as core elements in the barcode symbol. As any other barcode system, the novel model requires a source of light, the barcode symbol and photodiode detectors. Theoretical background from optics has been used to design the entire system along with all the positioning of its components. After part-testing the design in laboratory, the barcode system design has been changed to achieve better results. Experiments have showed that the initial proposed Light Emitting Diode (LED) source light cannot deliver 5mm spot light over a range of 50cm and therefore, white Light Amplification by Stimulated Emission of Radiation (LASER) light has been adopted as replacement. The diffractions from the barcode symbol are captured by detectors built with SI photo diodes, which are designed to detect this range of wavelengths. The barcode symbol is composed of small 5mm by 5mm grating modules and the largest possible symbol size defined is 80 modules (5cmx5cm). Experimental works have proved that intensity of the light can be used to uniquely identify each grating rather than the entire spectrum diffracted. A better design is proposed where the detectors are positioned under the barcode symbol and capture the light intensity of the first diffracted order. Theoretical investigations state that diffraction gratings with different lines per mm diffract different sets of wavelengths spectrum. This characteristic allows a set of unique gratings to be used in the barcode symbol which hence allow data to be represented or stored. Character (Char) sets are defined to help encode and decode data in the barcode symbol. High data storage has been achieved through the use of two layers. Multiple layers offer the possibility to increase the number of unique sets of gratings which in turn increase the data representation capacity. Using two layers with 16 unique sets of gratings has proved to be able to store around 100 bytes of data. The system has the potential to use more than two layers and using 4 layers with 16 unique gratings per layer will achieve 200 bytes. The thesis has proved through theoretical and experimental work that diffraction gratings can be used in barcode system to represent data and multiple layers adds the benefit of increasing data storage. Further work is also suggested.

Filters are integral components in all wireless communication systems, and their function is to permit predefined band of frequencies into the system and reject all other signals. The ever-growing demand in the use of the radio frequency (RF) spectrum for new applications has resulted in the need for high performance microwave filters with strict requirements on both inband and out-of-band characteristics. High selectivity, high rejection, low loss and extremely wide spurious-free performance are required for both transmitter and receiver channels. In addition, these devices need to be highly compact, easy to integrate within transceivers and should be amenable to low cost manufacturing. High selectivity is essential to enable the guard band between adjacent channels to be reduced thus improving the efficiency of the RF spectrum and hence increasing the capacity of the system. A low insertion-loss, high return-loss and small group-delay in the passband are necessary to minimize signal degradation. A wide stopband is necessary to suppress spurious passbands outside the filter’s bandwidth that may allow spurious emissions from modulation process (harmonic, parasitic, intermodulation and frequency conversion products) and interfere with other systems. The EMC Directive 89/336/EEC mandates that all electronic equipment must comply with the applicable EN specification for EMI. This thesis presents the research work that has resulted in the development of innovative and compact microstrip bandpass filters that fulfil the above stringent requirements for wireless communication systems. In fact, the proposed highly compact planar microstrip filters provide an alternative solution for existing and next generation of wireless communications systems. In particular, the proposed filters exhibit a low-loss and quasi-elliptic function response that is normally only possible with filter designs using waveguides and high temperature superconductors. The selectivity of the filters has been improved by inserting a pair of transmission zeros between the passband edges, and implementing notched rejection bands in the filter’s frequency response to widen its stopband performance. The filter structures have been analysed theoretically and modelled by using Keysight Technologies’ Advanced Design System (ADS™) and Momentum® software. The dissertation is essentially composed of four main sections. In the first section, several compact and quasi-elliptic function bandpass filter structures are proposed and theoretically analysed. Selectivity and stopband performance of these filters is enhanced by loading the input and output feed-lines with inductive stubs that introduce transmission zeros at specified frequencies in the filter’s frequency response. This technique is shown to provide a sharp 3-dB roll-off and steep selectivity skirt with high out-of-band rejection over a wide frequency span. In addition, the 3-dB fractional bandwidth of the filters is shown to be controllable by manipulating the filter’s geometric parameters. Traditional microwave bandpass filters are designed using quarter-wavelength distributed transmission-line resonators that are either end-coupled or side-coupled. The sharpness of the filter response is determined by the number of resonators employed which degrades the filter’s passband loss performance. This results in a filter with a significantly larger footprint which precludes miniaturization. To circumvent these drawbacks the second section describes the development of a novel and compact wideband bandpass filter with the desired characteristics. The quasi-elliptic function filter comprises open-loop resonators that are coupled to each other using a stub loaded resonator. The proposed filter is shown to achieve a wideband 3-dB fractional bandwidth of 23% with much better loss performance, sharp skirt selectivity and very wide rejection bandwidth. The third section describes the investigation of novel ultra-wideband (UWB) microstrip bandpass filter designs. Parametric study enabled the optimization of the filter’s performance which was verified through practical measurements. The proposed filters meet the stringent characteristics required by modern communications systems, i.e. the filters are highly compact and miniature even when fabricated on a low dielectric constant substrate, possess a sharp quasi-elliptic function bandpass response with low passband insertion-loss, and ultra-wide stopband performance. With the rapid development of multi-band operation in modern and next generation wireless communication systems, there is a great demand for single frequency discriminating devices that can operate over multiple frequency bands to facilitate miniaturization. These multi-band bandpass filters need to be physically small, have low insertion-loss, high return-loss, and excellent selectivity. In the fourth section two miniature microstrip dual-band and triple-band bandpass filter designs are explored. A detailed parametric study was conducted to fully understand how the geometric parameters of the filters affected their performance. The optimized filters were fabricated and measured to validate their performance.