No description available.

This thesis aimed to develop a new version of an adsorption calorimeter originally proposed by Ahamat and Tierney. A more rigorous approach was taken in terms of mechanical components, vacuum equipment, calibration of thermoelectric modules, and length of tests. Novel contributions include the new temperature control algorithm, new adsorbent configuration, new adsorbent, and consideration of the effect of residual air on the kinetics of adsorption. Improved design of the hardware reduced the air ingress from 1mbar/hour to <O.05mbar/hour (important when considering the kinetics of adsorption). It also allowed faster assembly and disassembly of the calorimeter, enabling gravimetric cross checks of the equilibrium adsorption uptake. The calorimeter uses thermoelectric modules (TEMs) to simultaneously control the temperature of the base plate and measure the amount of heat required to do so. Calibration of the thermoelectric modules revealed a possible temperature dependence of the TEM parameters, which was accounted for in heat flow calculations. An offset between expected and calculated heat flow could not be eliminated, and became large and inconsistent when the TEM deteriorated through use. Measured electrical resistance was the best indicator of TEM deterioration, and hence was used to monitor TEM wear. Calculated heat flow during condensation tests showed a different steady state offset between the start and end of each experiment. An appropriately modified fitting equation gave reasonable consistency between the calculated and measured mass of condensed water. Tests on loose beads of Type A silica gel (3mm and 1mm diameters) were carried out to identify the effect of the bonding agent previously used. Heat of adsorption was between 2438kJ/kg and 2763kJ/kg for all calorimetric and gravimetric measurements. Gravimetric measurements confirmed the validity of calorimeter measurements. Equilibrium uptake was higher than for Ahamat and Tierney's bonded beads, but within the range of other published data, suggesting previous contamination of some adsorption sites by the bonding agent. Kinetics of adsorption for 3mm diameter beads were similar to those found for bonded beads, but faster for 1mm diameter beads, indicating a slowing of adsorption caused by the bonding agent. Tests on loose grains of SWS-lL (mesoporous silica gel impregnated with calcium chloride) were carried out. High levels of scatter gave a discrepancy between gravimetric and calorimetric heats of adsorption (1972kJ/kg and 2703kJ/kg), but both values were within 15% of published data. Gravimetrically determined equilibrium uptake was generally O.04g/g lower than the calorimetrically determined uptake. Overall, equilibrium uptakes were O.lg/g lower than published data, probably due to non-ideal storage conditions for the sample. Diffusivities of water vapour in SWS-1L were extracted from the Fickian diffusion model, giving values of 2.55xlO-6 m2/s. This was an order of magnitude higher than published experimental data, but fitted well with published modelling data. The high level of scatter seen in most kinetic results was partly explained by modelling the Stefan flow of residual air through the bulk vapour. The model showed that a partial air pressure of 1mbar (the lowest pressure reported by the pressure transducer) could halve the apparent rate of adsorption.

Reliable, fast and sensitive hydrogen detection is central to safety in the emerging hydrogen economy. Extensive research has highlighted nanostructured materials as an approach to meet these demands. In this work, arrays of palladium nanotubes and nanorods have been characterised as optical hydrogen sensors. The fabrication of nanorod and nanotube arrays using a bottom-up template method is presented. Electrodeposition into alumina templates is shown to produce ordered arrays of structures across glass substrates, confirmed through microscopy techniques. Utilising the purpose built characterisation facility, the sensitivity and response of palladium nanotubes have been demonstrated to improve upon nanorods, which in turn offer performance enhancements over traditional thin film sensing elements.

This is the first time that small fluorescent sensors performing complicated computational functions, edge detection, which can be regarded as a new area of research in form of image processing. With a combination of florescent sensor, photo acid generator and sodium carbonate spreading on a filter paper, clear and visualized edges of 1-2 mm width were delivered, based on the light dose-driven ‘off-on-off fluorescent behaviour. The fluorescent intensity of sensor molecules in the irradiated region gradually enhanced upon irradiation of 254 nm UV light, due to generation of protons by photo-decomposition of the photo acid generator in that area. Then, it was slowly quenched by a strong quencher produced along with protons; whereas the fluorescence at the edge of irradiated and unirradiated regions was switched ‘on’ by the diffusion of protons neutralizing the pH there. A number of parameters, such as the pKa value of the fluorescent sensors, the light intensity from the UV lamp at two wavelengths, the number of molecules involved in producing the edge of the object and the values of Ksv between sensors and quenchers were determined. The proof of light dose-driven ‘off-on-off fluorescent behaviour of the molecules in solution and the development of a minimal model were built.

The measurement of rapidly changing temperature fluctuations is a challenging problem due to the inherent limited bandwidth of temperature sensors. This results in a measured signal that is a lagged and attenuated version of the true input signal. Compensation can be performed provided an accurate, parameterised model of the sensor is available. However, the sensor characteristics are strongly dependent on the measurement environment, which is often time-varying and cannot be determined a priori. To account for the changing characteristics, the sensor model must be estimated in-situ, thereby resulting in a blind identification problem. In this study, methods for performing blind characterisation of a two-sensor probe for fast temperature measurement are investigated and developed, with a particular focus on extending existing methods to handle second-order models.

No description available.

No description available.

No description available.

This project looks at the operation of wireless sensor nodes in rapeseed using antennas located close to the ground, through both measurements of the propagation loss with test equipment and through building and operating an experimental wireless sensor network. Rapeseed is a technically challenging crop in which to operate a wireless sensor network, due to its dense foliage from ground level up to the canopy. To facilitate these activities two measurement systems have been devised and constructed. One measurement system was designed to allow the measurement of propagation loss in and around the crop. The other was designed to allow the power profile of sensor nodes to be measured. Both of these systems provided information to guide the design of the experimental wireless sensor network. The measurements with the propagation measurement system along with the received signal strength and message loss results from the deployment provide information on the distances over which communication is likely to be feasible with antennas located close to the ground. Sensor nodes and a base station to collect the data from them were designed based on Microchip's PIC18F45J20 microcontroller, MRF24J40 transceiver chip and MiWi software stack with customisations for power management and to record the received signal strength of packets arriving at the base station. Initial propagation loss measurements in a rapeseed crop were made at Rothamsted Experimental Station in 2011. Further propagation loss measurements and a deployment of the sensor nodes in a rapeseed crop were made at Tatton Dale Farm in 2012.

No description available.