Audition is a fundamental tool for biological organisms to detect sound. Sensing the weakest sounds requires metabolically active processes that enhance the nanometre-size oscillations of sound sensors. Mosquitoes are a model system for active audition, incorporating many of the properties found in vertebrate active mechanics. In the mosquito, plumose antennae are used as oscillators that are driven by the flow of particles caused by sound. These antennae are adapted to be resonantly tuned to the female sound, such that they respond maximally to the female wingbeat. As well as their excellent sensitivity, mosquito antennae behave nonlinearly, changing the bandwidth of the oscillator in response to sound intensity. Here they are shown, using Laser Doppler vibrometry, to exhibit hitherto unseen nonlinear features in the antenna, and indeed in audition in general. Amplification and hysteresis of the antennal oscillator occur when presented with a single-frequency stimulus. This nonlinear behaviour is shown to occur when the neuronal sensory ensemble entrains to the sound stimulus. The synchronized neurones can then generate coherent force at the stimulus frequency to modify the parameters of the antennal oscillator, namely natural frequency and dissipation. The mechanism is proposed to be derived from twice-frequency forcing of mechanosensory neurones, evidenced by electrophysiological signalling at twice the stimulus frequency which, when modelled, reproduces remarkably well the experimental response. Such a phenomenon is predicted to harbour the potential for pattern formation in the activity of neurones, a property that would be unparalleled in peripheral auditory organs. This work has therefore revealed the remarkable evolution of both a sensitive and functional sound receiver that relies on the collective behaviour of thousands of neurones to act as one.

For satisfactory production or conversion of continuous material in web form it is generally important to control the tension in the web. The material may be plastics, paper or metal strip; conversion processes include web-fed printing or packaging machines and strip rolling mills. As speeds of production increase it is more likely that relatively high frequency tension perturbations may occur due to interaction between material properties and machine performance characteristics. This could result in frequent breakage of the web or unsatisfactory quality of the product. It is therefore of paramount importance to be able to measure and control tension variations accurately and with an adequate speed of response. The objectives of this investigation are to determine, by direct experiment and analytically, the limitations in static accuracy and dynamic performance of equipment for measurement of web tension.

Energy consumption is one of the primary concerns in a resource constrained visual sensor network (VSN) with wireless transceiving capability. The existing VSN design solutions under particular resource constrained scenarios are application-specific, whereas the degree of sensitivity of the resource constraints varies from one application to another. This limits the implementation of the existing energy efficient solutions within a VSN node, which may be considered to be a part of a heterogeneous network. This thesis aims to resolve the energy consumption issues faced within VSNs because of their resource constrained nature by proposing energy efficient solutions for sensing nodes characterisation. The heterogeneity of image capture and processing within a VSN can be adaptively reflected with a dynamic field-of-view (FoV) realisation. This is expected to allow the implementation of a generalised energy efficient solution that will adapt with the heterogeneity of the network. In this thesis, a FoV characterisation framework is proposed, which can assist design engineers during the pre-deployment phase in developing energy efficient VSNs. The proposed FoV characterisation framework provides efficient solutions for: 1) selecting suitable sensing range; 2) maximising spatial coverage; 3) minimising the number of required nodes; and 4) adaptive task classification. The task classification scheme proposed in this thesis exploits heterogeneity of the network and leads to an optimal distribution of tasks between visual sensing nodes. Soft decision criteria is exploited, and it is observed that for a given detection reliability, the proposed FoV characterisation framework provides energy efficient solutions which can be implemented within heterogeneous networks. In the post-deployment phase, the energy efficiency of a VSN for a given level of reliability can be enhanced by reconfiguring its nodes dynamically to achieve optimal configurations. Considering the dynamic realisation of quality-of-information (QoI), a strategy is devised for selecting suitable configurations of visual sensing nodes to reduce redundant visual content prior to transmission without sacrificing the expected information retrieval reliability. By incorporating QoI awareness using peak signal-to-noise ratio-based representative metric, the distributed nature of the proposed self-reconfiguration scheme accelerates the decision making process. This thesis also proposes a unified framework for node classification and dynamic self-reconfiguration in VSNs. For a given application, the unified framework provides a feasible solution to classify and reconfigure visual sensing nodes based on their FoV by exploiting the heterogeneity of targeted QoI within the sensing region. From the results, it is observed that for the second degree of heterogeneity in targeted QoI, the unified framework outperforms its existing counterparts and results in up to 72% energy savings with as low as 94% reliability. Within the context of resource constrained VSNs, the substantial energy savings achieved by the proposed unified framework can lead to network lifetime enhancement. Moreover, the reliability analysis demonstrates suitability of the unified framework for applications that need a desired level of QoI.

There is current debate in literature as to the main source of photoinduced super-hydrophilic (PSH) nature on Ti02 films with groups segregated into two schools of thought: (i) those attributing the effect to UV induced modifications of the Ti02 surface; and (ii) those attributing the effect to the well-known photo-oxidation properties of Ti02 in removing hydrophobic organic layers from the Ti02 surface. This thesis presents acquisition of experimental goniometric data required to allow theoretical analysis of the super-hydrophilic nature on as prepared mesoporous-Ti02 (m-Ti02), in order to determine the source of PSH. Evidence to advocate that surface reconstruction does not occur instantaneously upon ultra-band gap illumination, the Ti02 surface is inherently wet with a contact angle, 8, being 4°, and the photo-induced super-hydrophilic nature is due to the latter effect, (ii), is also discussed. Experiments demonstrating the change in the Ti02 surface during UV band-gap illumination in the presence of organics using micro-gravimetric techniques, mainly the quartz crystal microbalance (QCM), are presented and exploration and applications of PSH-related phenomena in real world sensing devices are explored.

Broadcasting, coverage, duty cycling, and capacity improvement are some of the important areas of interest in Wireless Networks. We address different problems related with broadcasting, duty cycling, and capacity improvement by sensing different network conditions and dynamically adapting to them. We propose two cross layer broadcasting protocols called CASBA and CMAB which dynamically adapt to network conditions of congestion and mobility. We also propose a broadcasting protocol called DASBA which dynamically adapts to local node density. CASBA, CMAB, and DASBA improve the reachability while minimizing the broadcast cost. Duty cycling is an efficient mechanism to conserve energy in Wireless Sensor Networks (WSNs). Existing duty cycling techniques are unable to handle the contention under dynamic traffic loads. Our proposed protocol called SA-RI-MAC handles traffic contention much more efficiently than RI-MAC without sacrificing the energy efficiency. It improves the delivery ratio with a significant reduction in the latency and energy consumption. Due to limited battery life and fault tolerance issues posed by WSNs, efficient methods which ensure reliable coverage are highly desirable. One solution is to use disjoint set covers to cover the targets. We formulate a problem called MDC which addresses the maximum coverage by using disjoint set covers S1 and S2. We prove that MDC is NP-complete and propose a √n-approximation algorithm for the MDC problem to cover n targets. The use of multi-channel MAC protocols improves the capacity of wireless networks. Efficient multi-channel MAC protocols aim to utilize multiple channels effectively. Our proposed multi-channel MAC protocol called LCV-MMAC effectively utilizes the multiple channels by handling the control channel saturation. LCV-MMAC demonstrates significantly better throughput and fairness compared to DCA, MMAC, and AMCP in different network scenarios.

The intrinsic stability of fibre optic based sensing systems offer a platform that is suited to hazardous waste detection in a wide range of environments. Over the last few years Cranfield University has been working on the development of chemical sensors using optical fibres in combination with a group of chemical recognition molecules called calixarenes. Calixarenes semi-selectively with a range of solvents of interest makes them useful for chemical detection systems. This work has primarily been focused on the use of calixarenes in sensing benzene and other hazardous solvents. However, this approach could potentially be expanded for use in a wide range of chemical and even biological recognition systems. The initial aim of this project was to build on the previous work in fibre optic sensing at Cranfield and explore approaches to improve and extend the performance of the sensor system. The project first focused on improving the techniques used in the Langmuir-Blodgett (LB) deposition of calixarenes. Initial studies in this area highlighted one critical experimental error associated with the use of dry Wilhelmy plates to monitor the surface pressure of the Langmuir film. Dry filter paper plates take up to 2 hours to give stable data, with a drift of up to 10% in the measured surface pressure. It is shown that this problem can be avoided by using pre-soaked plates. To provide an alternative to the Wilhelmy plate surface pressure senor, an optical fibre surface pressure sensor was developed, measuring changes in the meniscus forming properties of a liquid. The sensor consists of a tapered single mode silica fibre, mounted with a small curvature and positioned with the tapered region of the fibre immersed in the water. The performance of the fibre optic sensor is comparable with that of the conventional Wilhelmy plate surface pressure sensor showing linearity of greater than 0.9. Following the analysis of the experimental systems used in the construction of the sensors, the project then focused on the chemistry of the materials and their suitability for LB coating. A variety of these materials were spread as Langmuir monolayers and their behavior upon compression measured. Long chain-substituted resorcinarenes gave more stable monolayers than their short chain analogues. The incorporation of long chain surfactants led to large increases in surface area, demonstrating that both resorcinarenes and surfactants are located at the water surface, except for one system where a bilayer structure is potentially formed. Further work on the behavior of the materials involved the alteration of the dipole-dipole interaction of the monolayer materials with the subphase. The modification of this interaction through the introduction of dipole altering additives, including alcohols and hydrogen peroxide, to the aqueous subphase was investigated. The resulting isotherms of the materials showed a reduction in the surface pressure and area per molecule required in order for the monolayer to reach its point of collapse. This ability to shift the point of collapse has application in the optimisation of Langmuir-Blodgett coating of surfaces. Within this project the sensing properties of a fibre sensor were also modelled extensively in order to determine the theoretical sensing limits of a fibre optic vapour sensor. The model showed that the sensing goals of 1ppm originally envisaged for this project were unobtainable due to the low number of gas molecules interacting with the sensor. However, this led to the proposal of a new application of the system in sensing contaminants in water, where the same limitations would not apply. The results for the sensor system tested in water show how significantly more sensitive the system is to toluene contamination in water than it is to toluene vapour. These results demonstrate the utility of the developed system for many pollutant-sensing applications, include crude oil detection.

This thesis presents an investigation into the fabrication, modelling and optimisation of multi- layered unimorph (bending) EAP dielectric elastomer actuators. Fabrication of dielectric elastomer (DE) EAPs is non-trivial as it requires the production of very thin dielectric films «100 urn) to maximise performance for a given input voltage, and flexible conductive electrodes. A novel dielectric elastomer actuator (DEA) fabrication method based on spray deposition is presented. The characterisation and optimisation of this technique are described and fabrication of very thin DE films « 40 urn) with high accuracy demonstrated. The fabrication method is relatively simple and inexpensive compared to com- peting methods, and is readily adaptable for full automation for mass production of DEAs. Multilayer DEAs are difficult to fabricate due to the need for electrical connection of the DE layers. The fabrication of multilayered unimorph DEAs (MUDEAs) using spray deposition is presented, good repeatability and reliability (- 80% success rate) are demonstrated. The MUDEAs exhibit large tip deflections (- 19 mm) and resonant operation with substantial strain amplification (- 1000%) is achieved. The fabricated actuators are also used to develop proof- of-concept wing flapping designs for MA V applications capable of simultaneous flapping and (minimal) pitching. Models of MUDEAs are important for producing optimal actuator designs. Current model- ling approaches presented in the literature are inadequate for multilayer DEA configurations with non-uniform geometries. A novel finite element (FE) approach for modelling MUDEAs is presented. The model is used to simulate the deflection of two, three and four layer MU- DEAs with inhomogeneous layer geometries. The results are validated against experiments and show good agreement with the data. The Maxwell stress in MUDEAs is also derived and shown to be the same as that for an unconstrained (i.e. free standing, with no external forces or loads applied) DEA. The validated 3D FEA model is used to investigate the effects of varying the number of layers on MUDEA deflection, for several DE layer thicknesses and substrate materials. The results show that there is an optimum number of layers for which MUDEA tip deflection is. a maximum and that it is more beneficial to use actuators with thinner DE layers, rather than thicker layers, for a given combined DE layer thickness.

Magnetic induction tomography (MIT) is a medical imaging technique that uses magnetic fields to image the electrical properties of the human body. In this work, a numerical model has been described and used to simulate MIT systems. A reconstruction algorithm, based on the sensitivity matrix method, has been used to reconstruct images of the internal conductivity distributions of samples, from simulated and experimental measurements. Images of conductivity contrasts of the magnitude encountered in human body have been successfully reconstructed. An initial investigation has made into wave propagation delays in MIT.

The aim of this research is to explore the use of non-invasive impact testing techniques for the on-site interrogation of industrial process vessels to determine their contents. To illustrate the method the work has been targeted at the design, construction and / implementation of a new level gauge. This has been developed in collaboration with Tracerco and sponsored by Johnson Matthey Plc. The research is motivated by the need for better understanding of industrial processes to yield, for instance greater efficiencies, less waste and higher quality products. Ideally instruments are inexpensive, non-invasive, easy to apply and reliable. This thesis presents a novel approach to process monitoring using impact excitation and processing of the resulting acoustic signals. Unlike many applications the impact and detection of signals occurs outside the wall of the vessel and is based on acoustic rather than ultrasonic signals. The acoustic problem is tackled through the analysis of low-frequency signals that are generated using impact excitation. Initial experiments revealed opportunities to extract key features associated with the volume of liquid within the container using this technique. Finite element models based on fluid-structure interaction have corroborated the experimental results. Results suggest the possibility to detect liquid level from the resulting frequency response curves. Based on these results a new acquisition system, including software and hardware specifically designed to detect impact generated waves, has been constructed and tested using a state-of-the-art array of sensors mounted on the outer wall of the vessel. During the course of the project various signal processing techniques have been considered including Fourier analysis and spectrograms. The effectiveness of the resulting equipment and methodology has been demonstrated on a scaled storage tank. Liquid level determination on the vessels studied yielded a maximum error of 2% of the actual level for un stirred conditions in the operative range. Additional experimentation and signal processing have suggested the possibility to implement other process diagnostic techniques based on acoustic transmission-emission. These, including identification of inhomogeneities and stirring and bubble flow characterisation, will be further explored along with the transfer of the developed technology into the collaborating company.

No description available.