M. Tech. Electrical Engineering. / This study proposes a distributive joint power and rate adaptation algorithm (JRPA) in ad hoc networks based on coupled interference minimisation. In the proposed method, the influence of coupled interference was controlled by dynamically adjusting network users' transmit power choices. The users are therefore aware of the current link status while determining their data rates. In addition, every maximize utility of other users as it maximizes its utility due to the inevitable cooperation, hence, improving a collective network performance. Solving this network utility maximization problem results in a supermodular game equivalence where users cooperate to maximise both local and global utility, hence the supermodular game theory concept was used to analyse the optimality and convergence of the proposed solution.

Laminated composites are materials with complex architecture made of continuous fibers embedded within a polymeric resin. The properties of the raw materials can vary from one point to another due to different local processing conditions or complex geometrical features for example. A first step towards the identification of these spatially varying material parameters is to image with precision the displacement fields in this complex microstructure when subjected to mechanical loading. This thesis is aimed to accurately measure the displacement and strain fields at the fiber-matrix scale in a cross-ply composite. First, the theories of both local subset-based digital image correlation (DIC) and global finite-element based DIC are outlined. Second, in-situ secondary electron tensile images obtained by scanning electron microscopy (SEM) are post-processed by both DIC techniques. Finally, it is shown that when global DIC is applied with a conformal mesh, it can capture more accurately sharp local variations in the strain fields as it takes into account the underlying microstructure. In comparison to subset-based local DIC, finite-element based global DIC is better suited for capturing gradients across the fiber-matrix interfaces.

Tube bending is a widely used manufacturing process in the aerospace, automotive, and other industries. During tube bending, considerable in-plane distortion and thickness variation occurs. The thickness increases at the intrados (surface of tube in contact with the die) and it reduces at the extrados (outer surface of the tube). In some cases, when the bend die radius is small, wrinkling occurs at the intrados. In industry a mandrel is used to eliminate wrinkling and reduce distortion. However, in the case of a close bend die radius, use of a mandrel should be avoided as bending with the mandrel increases the thinning of the wall at the extrados, which is undesirable in the manufacturing operation. The present research focuses on additional loadings such as axial force and internal pressure which can be used to achieve better shape control and thickness distribution of the tube. Based on plasticity theories, an analytical model is developed to predict cross section distortion and thickness change of tubes under various loading conditions. Results from both the FEA and analytical model indicated that at the intrados the increase in thickness for bending with internal pressure and bending with combined axial pull and internal pressure was nearly the same. But in the case of bending with the combination of axial pull and internal pressure there was a significant reduction of thickness at the extrados. A parametric study was conducted for the case of bending with combined internal pressure and axial pull and it was seen that with proper selection of the pressure and axial pull wrinkling can be eliminated, thickness distribution around the tube can be optimized, and cross section distortion of the tube can be reduced. Predictions of the analytical model are in good agreement with finite element simulations and published experimental results. The model can be used to evaluate tooling and process design in tube bending.

No description available.

The advent of medical journalism was initially felt to be an answer to the problem of communicating health and medical information to the public. However, currently, there is a concern among scientists with the way the media, newspapers in particular, communicate health and medical information. The concern of the medical community in particular and of the scientific community in general is that newspapers ‘distort’ health and medical information. In order to deal with this ‘perceived’ problem, scientists adopt a mechanical view and propose to solve it by issuing guidelines for journalists to follow when writing health and medical news. Close investigation of journalistic practice shows that many of the proposed guidelines are already present in journalistic practice, and yet, the concern for ‘distortion’ remains. The overall aim of the thesis is to contribute to this issue. Adopting an Applied Linguistics perspective, more specifically, using the discourse analytic methodology of Genre Analysis, the thesis demonstrates that Health and Medical News Reports are first and foremost news stories and that the proposed guidelines fail to achieve the envisaged changes precisely because they seem to be ignorant of this essential reality. In order to reach this conclusion, Genre Analysis is applied to different types of texts with a view to comparing their structures. Some of the text types used have already been described in the literature, but others are analysed for the first time in this thesis. Thus, comparison is made between Health and Medical Research Articles and Health and Medical News Reports, between Popularised Health and Medical Texts and Health and Medical News Reports, between News Texts and Health and Medical News Reports and between Health and Medical Press Releases and Health and Medical News Reports. Genre Analysis shows that Health and Medical News Reports are first and foremost news stories and, therefore, that the discourse of ‘distortion’ is somewhat ‘misguided’. However, because of its nature as a structural analysis, Genre Analysis leaves one important question unanswered, namely the ‘why’ of the discourse of distortion. Although it is beyond the scope of this thesis to investigate this question, in the thesis, it is indicated that a more context-sensitive analysis, using Critical Discourse Analysis (CDA) for example, could fruitfully be pursued. This thesis draws on four types of data. The main data set consists of Health and Medical News Reports published in The Herald and The Guardian between April and May 2007, where possible, corresponding press releases were collected. Email interviews were conducted with authors whose research was reported in the two newspapers. Finally, ethnographic observation of newsrooms and face-to-face semi-structured interviews were conducted with journalists who wrote the reports over a period of one week.

This thesis describes a novel harmonic domain approach for assessing the steady state performance of Flexible AC Transmission System (FACTS) devices. Existing harmonic analysis techniques are reviewed and used as the basis for a novel iterative harmonic domain model for PWM FACTS devices. The unified Newton formulation adopted uses a combination of positive frequency real valued harmonic and three-phase fundamental frequency power-flow mismatches to characterise a PWM converter system. A dc side mismatch formulation is employed in order to reduce the solution size, something only possible because of the hard switched nature of PWM converters. This computationally efficient formulation permits the study of generalised systems containing multiple FACTS devices. This modular PWM converter block is applied to series, shunt and multi-converter FACTS topologies, with a variety of basic control schemes. Using a three-phase power-flow initialisation and a fixed harmonic Jacobian provides robust convergence to a solution consistent with time domain simulation. By including the power-flow variables in the full harmonic solution the model avoids unnecessary assumptions regarding a fixed (or linearised) operating point, fully modelling system imbalance and the associated non-characteristic harmonics. The capability of the proposed technique is illustrated by considering a range of harmonic interaction mechanisms, both within and between FACTS devices. In particular, the impact of transmission network modelling and operating point variation is investigated with reference to ac and dc side harmonic interaction. The minor role harmonic distortion and over-modulation play in the PWM switching process is finally considered with reference to the associated reduction in system linearity.

In this thesis the in-plane pattern distortion resulting from the use of Evanescent Near Field Optical Lithography (ENFOL) masks was investigated. ENFOL is a high resolution low-cost technique of lithography that is able to pattern features beyond the diffraction limit of light. Due to its use of the evanescent near field, ENFOL requires the use of conformable masks for intimate contact. Such masks can stretch and skew as they come into contact with silicon substrates and therefore distort the high resolution features patterned on them. It was desired to measure this distortion to ascertain the patterning performance of ENFOL masks and possibly correct for any uniform distortion found. To this end a sophisticated measuring process was successfully demonstrated. This involved the use of a Raith 150 Electron Beam Lithography (EBL) system with precision laser interferometer stage and metrology software module for automated measurements. Custom software was written for the Raith to enable it to take additional measurements to compensate for electron beam drift. Processing algorithms were then employed to using the measurements to compensate for beam drift and correcting for shift and rotation systematic errors. The performance of the in-plane distortion measuring process was found to have a precision of 60nm. With the ability to measure distortion, ENFOL masks were used to pattern substrates and distortion was found to be large, on the order of 1µm. This is much larger than desired for sub 100nm patterning as is expected of ENFOL. The distortions were non-uniform patterns of localised displacements. This, the observation of Newton's rings beneath a test mask and the observation of a single particle distortion across measurements of the same mask across different loadings in the EBL pointed to particulate contamination causing the distortion. In order to prove beyond doubt that particulate contamination was the cause of the spurious distortions, mechanical modelling using the Finite Element Method (FEM) of analysis was employed. The results from this matched the distortions observed experimentally, particles 20-40µm modelling the observed distortion.

The welding of austenite stainless steel often results in large amount of welding distortion due to its high thermal expansion coefficient and low thermal conductivity. This has created great difficulty in the dimensional control of the welded stainless steel structure, ending up with high manufacturing cost. Researches on the welding distortion of stainless steels were very limited, especially for large weld structures with complex component shapes. The studies of this thesis were initiated with focus on the stainless steel nozzle-to-shell-can weld structures, a very typical structural configuration for pressure vessels used in petrochemical and nuclear power generation industries. Both the experimental and the FEA (finite element analysis), i.e. computational simulation, approaches were taken in the studies which addressed the influences of the welding fixture, the welding sequence, and the welding process on the distortion caused by stainless steel nozzle-to-shell welding. The investigations employed single and multi-nozzle weld test models (called mockups in the thesis) or FEA models. Manual GTAW (gas tungsten arc welding) and SMAW (shielded metal arc welding) processes were selected to represent the most common practice for stainless steel nozzle welding. The FEA simulations were conducted with ABAQUS program using sequentially coupled transient analysis method with lumped weld passes to achieve high computing efficiency. The investigations on the effect of the welding fixture concluded that the contour fixtures introduced in the thesis be effective for reducing the welding distortion for both the single and the multi-nozzle welding. The contour fixtures tend to localize the welding distortion, hence yield less impact on the global distortion of the whole weld structure. The rib-bar fixture, a more common fixture type for multi-nozzle welding, was found resulting in a big jump in the shell plate distortion when the fixture was removed. The studies on the influence of the welding sequence revealed that a progressive approach was more favorable for distortion control under the given nozzle-to-shell weld structure configurations. The best sequence suggested is to start welding at one nozzle, firstly on shell OD (outside diameter) side then on ID (inside diameter) side, then proceed to next neighboring nozzle. The effect of the welding direction of each weld pass was found affecting only the nozzle angular distortion. The experimental data showed that the manual GTAW process developed much higher shell plate distortion than the SMAW process. The reason would be that a higher percentage of the welding heat had been consumed on the base metal. The influence of the weld bead size didn???t appear to be significant. In the FEA study on the effect of the size of the lumped weld pass, the increase in weld bead size even resulted in a decrease in weld distortion. From the FEA simulation point of view, using large lumped pass would be a highly efficient choice without compromising too much in the precision of the distortion prediction. The FEA study confirmed that a decrease in cooling time after welding would result in more welding distortion. The large scale multi-nozzle mockup with rib-bar fixture demonstrated a maximum out-of-plane shell distortion of 16.4mm after the welding of 10 nozzles with GTAW+SMAW process, which suggests that additional measures should be developed to further control the welding distortion.

Loudspeakers were invented over 150 years ago, but the loudspeakers used todayare still based on the same ideas. Traditionally, good sound quality has been obtainedby using expensive materials in the loudspeakers and by allowing themto be big. However, nowadays loudspeakers are wanted in applications such asmobile phones and tablets where size and weight are very limited and there is aconstant desire to decrease production costs. Special small loudspeakers, knownas micro loudspeakers, have been developed for this purpose but due to the severerestrictions in size and manufacturing costs, the sound quality in the microloudspeakers is relatively poor. One problem is that the nonlinearities of thesystem, present in any loudspeaker, become more evident in the case of microloudspeakers and cause noticeable distortion of the sound.This master’s thesis has been performed in cooperation with Opalum (formerlyActiwave), a company specializing in using digital signal processing to improvethe sound in loudspeakers with poor acoustic properties. The objective of thethesis is to investigate ways to increase the sound quality in micro loudspeakersby using nonlinear control. Focus has been on frequencies below the resonancefrequency since the distortion is more noticeable at low frequencies. First, a nonlinearmodel of the micro loudspeaker has been obtained using system identificationstrategies. The model describes the relationship between the voltage overthe voice-coil and the diaphragm displacement. Subsequently, input-output linearisationhas been used to design a controller for the system and the effect onthe distortion has been investigated through experiments. Two different modelstructures have been tested, a physical model based on the Thiele-Small modeland a black-box model with a Hammerstein-Wiener structure. In both cases, thenonlinearities were modelled as polynomials. The controller was then extendedwith an updating algorithm, making it adaptive.The efficiency of the controllers has been proved by experiments, where distortionwas decreased by up to 60 % compared to the case without control. The effectwas largest for low frequencies, around one third of the resonance frequency,but improvements were noted up to about two thirds of the resonance frequency,depending on the loudspeaker unit. The approach using a physical model andthat using a black-box model have shown similar results. / Högtalaren uppfanns för över 150 år sedan men de högtalare som används idagbygger till stora delar på samma teknik. Högkvalitativt ljud har traditionellt uppnåttsgenom att ge högtalaren goda akustiska egenskaper genom att tillåta den attvara stor och tillverkad av dyra material. Utmaningen idag ligger i att högtalarefinns inbyggda i exempelvis mobiltelefoner, vilket innebär att de behöver görassmå, lätta och billiga att producera. För att möta dessa krav har kompromisserkrävts vilket gör att dessa små högtalare, kallade mikrohögtalare, har sämre ljudkvalitet.Ett problem är att de olinjäriteter som finns i alla högtalare blir extraframträdande i små högtalare vilket leder till distorsion och övertoner i ljudsignalen.Detta examensarbete är gjort i samarbete med Opalum (tidigare Actiwave), vilketär ett företag som specialiserar sig på att med hjälp av digital signalbehandlingförbättra ljudkvaliteten för högtalare med akustiskt dåliga egenskaper. Syftetmed examensarbetet har varit att minska distorsionen i en mikrohögtalaremed hjälp av olinjär reglering. Fokus har legat på den lägre delen av frekvensbandet,under resonansfrekvensen, eftersom det är där distorsionen är mest märkbar.Först har en olinjär modell av högtalaren tagits fram genom systemidentifiering.Modellen förklarar sambandet mellan spänningen över högtalarens talspole ochmembranets utslag. I ett nästa steg har en regulator designats utifrån modellenoch regulatorns effekt på distorsionen har utvärderats genom experiment. Två olikamodellstrukturer har undersökts, dels en fysikalisk modell baserad på Thiele-Smallmodellen och dels en svartlådemodell med Hammerstein-Wienerstruktur.I båda fallen har olinjäriteterna modellerats som polynom. Regulatorn har sedanutökats med en uppdateringsalgoritm som gör den adaptiv.Experiment har visat att regleringen bidrog till att minska distorsionen med upptill 60 % jämfört med då systemet kördes utan reglering. Effekten har varit störstför låga frekvenser, kring en tredjedel av resonsnsfrekvensen, men förbättringarhar kunnat ses upp till frekvenser kring två tredjedelar av resonansfrekvensen.Både metoden med en fysikalisk modellstruktur och med en svartlådestrukturhar visat likartade resultat.

Orthogonal frequency division multiplexing (OFDM) has become a popular modulation method in high-speed wireless communications. By partitioning a wideband fading channel into flat narrowband channels, OFDM is able to mitigate the detrimental effects of multipath fading using a simple one-tap equalizer. However, in the time domain OFDM signals suffer from large envelope variations, which are often characterized by the peak-to-average ratio (PAR). High PAR signals, like OFDM, require that transmission amplifiers operate at very low power efficiencies to avoid clipping. In this dissertation, we explore the problems associated with transmitted OFDM signals through peak limited channels. A large part of this work deals with analyzing different distortion metrics and determining which metrics are most useful. We find that the signal-to-noise-plus-distortion ratio (SNDR) is one of the most important metrics in assessing distortion in nonlinear channels. As part of this analysis, we compare sample-based SNDR and symbol-based SNDR and find that using the more comprehensive symbol-based metric as the objective in SNDR maximization algorithms leads to only marginal SNDR improvements. The SNDR perspective is also applied to existing PAR-reduction techniques to compare existing schemes and proposed new schemes. Part of this work involves deriving a SNDR maximizing adaptation of the popular PAR-reduction scheme, selected mapping (SLM). We also compare another popular PAR-reduction method, partial transmit sequence (PTS), to SLM through a variety of metrics including SNDR and found that for any given amount of complexity or side information SLM provided better performance. The next major piece of work in this dissertation addresses synchronization and channel estimation in peak-limited channels for OFDM. We build off of existing work that shows that embedded synchronization energy is a more bandwidth efficient means of synchronization than preamble-base methods. With this, we demonstrate a method for generating embedded sequences that have low PAR, and thus minimize the PAR of the combination OFDM symbol/embedded sequence among all embedded sequences. Next, we extend this work to sequences called joint synchronization-pilot sequences (JSPSs) by deriving the symbol-estimate mean squared error (MSE) pilot placements for the JSPSs and by showing how the JSPSs can be used with SLM for blind detection. Finally, the dissertation concludes with a derivation of the SNDR-optimal transmitter/receiver pairs. Using functional analysis, we show that the SNDR-optimal receivers for peak-limited transmitters are not linear. Instead they follow non-linear functions that depend on the noise and signal distributions.