/library/oar/handle/123456789/62480 2025-11-10T12:37:48Z /library/oar/handle/123456789/67332 Title: The development and optimisation of the B-train system for the ELENA ring Abstract: In synchrotrons, the vertical magnetic field produced by dipole magnets guides charged particles along a closed horizontal orbit, determined by the equilibrium of the centripetal force and the Lorentz force. As a result, the knowledge of the magnetic field value in real-time is essential to control the trajectory of the particles. Subsequently, a direct magnetic field measurement inside a reference magnet, known as a B-train, is used to derive an estimate of the average magnetic field in the ring. As part of a site-wide consolidation project, all the B-train systems at CERN are being replaced with upgraded electronics, software, and sensors. The Extra Low ENergy Antiproton (ELENA) ring is a new decelerator being built at CERN; hence a B-train system is required to be installed and commissioned. The ELENA ring presents challenges to magnetic field measurement systems, such as very low magnetic field and long cycle times. The aim of this thesis is to commission a measurement model for the ELENA B-train, as well as validating and optimising the instrumentation. A measurement model is thus developed in this thesis, and applied to identify the ring and reference magnets, the sensors, as well as the electronic acquisition chain. An uncertainty estimation helps identify the biggest sources of error in the measurement, providing an insight on potential improvements in the measurement system. The ELENA B-train system is validated through a series of tests, characterising the behaviour of the sensors, the stability and the accuracy of the instrument in operating conditions, confirming a relative reproducibility and accuracy better than 2 units (10−4). A potential improvement in the measurement by changing the position of the local field sensors is investigated, and the feasibility of applying a displacement is discussed for the ELENA B-train. A magnetic field model based on the decomposition of the field into different components is finally proposed. The model is tested using the ELENA magnetic cycles, exhibiting a relative accuracy of 1 unit, giving confidence in its prediction ability in low-energy synchrotrons. Description: PH.D. 2020-01-01T00:00:00Z /library/oar/handle/123456789/62594 Title: A remotely configurable delay generator for the HMPID L0 trigger Abstract: This work deals with the design, implementation, and testing of a remotely configurable delay generator which will be used in the High Momentum Particle Identification Detector (HMPID) as part of the LHC-wide upgrade in preparation for Run3 (2021 - 2023). While during the previous two runs (2010 - 2018) HMPID worked with three trigger levels, this will no longer be the case as the triggering mechanism will change. Although HMPID was triggered by the Level-Zero (L0) trigger, which arrives at 1.2µs after an event, this will no longer be the case as the L0 trigger will arrive later. Instead, the Level-Minus (LM) trigger will be used which arrives at 800ns after an event. As such, a delay generator is required such that the trigger signal is delayed by at least 400ns so that the detector is triggered at the peak of the charges on the pads. The work undertaken and described in this dissertation is therefore classified into three parts. The first part relates to the design, implementation, analysis, and testing of a digital delay generator that is based on a shift-register architecture. This delay generator has been implemented on a Xilinx Virtex-5 FPGA and can achieve a delay range of 525ns, with a resolution of 1ns. The second part focuses on the design, analysis, optimisation, and testing of a novel delay generator that is based on a delay locked loop (DLL) architecture that is calibrated in closed loop, and then used in open-loop configuration. While also describing the proposed architecture, this part focuses on the optimisations performed on a symmetric rail-to-rail delay element to maximise linearity. The delay generator was implemented in the X-FAB XH018 technology, and has a range varying from 935ns to 183ns with a resolution of 2ns. Finally, the last part of this work focuses on the firmware upgrades that had to be performed in the detector such that it can handle the new triggering scheme. Through the use of a new firmware architecture, the readout rate of the firmware has been increased from a maximum of 4.57kHz to 14.4kHz, without sacrificing any data integrity. Description: PH.D. 2020-01-01T00:00:00Z /library/oar/handle/123456789/62579 Title: Design and optimisation of high performance resonating micro-scanners through a multiphysics investigation Abstract: Resonating micro-scanners based on MEMS fabrication technologies have been extensively implemented in laser beam scanning systems for micro-display and imaging applications. The optical resolution obtained from laser beam scanning micro-mirrors for micro-display applications is dependent on the mirror size, the scanning frequency and scan angle amplitude. However, other factors need to be considered in the design of such devices, namely, the structural strength, dynamic deformation and the scanning efficiency. Understanding the structural, fluidic and electrostatic characteristics of resonant micro-scanners is crucial in order to maximize the optical performance of such devices. In this dissertation, the multi-physical domains governing the micro-scanner operation will be evaluated using experimentally-validated computational fluid dynamics and finite element models. The numerical simulation models presented will be utilised (i) to investigate the validity of a number of analytical equations found in literature, (ii) to propose analytical equations with improved accuracy and range of validity for microscanner performance predictions and (iii) to develop and optimize novel micro-scanner designs. While micro-scanner designs with angular vertical comb actuation are the focus in this work, a number of research outcomes are also applicable to high frequency resonating micro-scanners employing other actuation methods. A detailed investigation is performed on the two principal limiting factors in micro-scanners for high performance applications: air damping and dynamic deformation. Three-dimensional transient Navier–Stokes simulations are performed to analyse the complex air flow interactions of a high frequency scanning micro-mirror. The damping effects due to device thickness and the depth of the mirror cavity are also evaluated. It is shown that analytical damping models which are applicable to resonant MEMS devices, are not valid within the operating range of high performance micro-scanners. On the other hand good agreement in the overall quality factor is achieved between Navier-Stokes simulation and measurement results. The inertia loading acting on the mirror plate, as the micro-scanner oscillates in out-ofplane rotation, results in the dynamic deformation of the mirror surface. A detailed analysis is presented on the micro-mirror design aspects contributing to dynamic deformation. An improved analytical equation for dynamic deformation predictions is proposed, which takes into consideration the two-dimensional mirror plate twist. A comparison among a number of layout designs was carried out with the aim of improving the dynamic deformation distribution on the mirror surface. A significant improvement in dynamic deformation was achieved with the inclusion of a gimbal-type structure between the mirror plate and the torsional beams. A design optimization scheme based on numerical simulations and meta-modelling methodologies is introduced. The minimization of dynamic deformation is considered as one of the design objectives, which include output parameters related to micro-mirror optical performance, structural reliability and gas damping characteristics. The design optimization scheme was implemented in order to develop three high performance resonating micro-scanner prototypes using the SOIMUMPs MPW process. The simulations and measurements presented in this dissertation demonstrate that the optical performance of resonating micro-scanners fabricated from a single device layer can be maximized by incorporating the indirect-drive configuration and a gimbal-type mirror support structure. Description: PH.D. 2020-01-01T00:00:00Z /library/oar/handle/123456789/62519 Title: Exploiting nonlinearities in MEMS : applications to energy harvesting and RF communication Abstract: Traditionally, in engineering, nonlinear behaviour is avoided, however engineering applications that are intended to create a nonlinear relationship between inputs and outputs also exist. In this thesis, it is shown that exploiting nonlinear phenomena in MEMS design is instrumental in providing counter intuitive solutions to an application involving a vibrational energy harvester and another two designs with applications to communication signal processing. Vibrational energy harvesters at MEMS scale are generally a challenge since at these scales resonant frequencies are in the kHz range and this makes them insensitive to the lower frequencies that are more abundant in the environment. One solution is to include a nonlinear spring such that the harvester becomes sensitive to broadband base excitations. In this work, one such broadband harvester is designed by making use of a ‘quintic’ stiffness, buckling (bistable) spring. The novel aspect in this work can be attributed to the topological arrangement of the two buckling beams and the mass. The arrangement allows only the required beam modes to dominate and together with the designed beam boundary conditions, it is possible to replace the non-linear partial differential equation model (resulting from continuum mechanics) with a simpler nonlinear differential equation. It is demonstrated that this simpler model can still capture the salient characteristics of the complex buckling behaviour; replacing complex finite element analysis simulations with simple numerical solutions of differential equations and hastening the design process. Although the design was constrained geometrically to satisfy this simpler mathematical model, it is demonstrated that these constraints do not impinge negatively on the harvesting capabilities. The harvester has a power destiny of 0.13 mW cm-3 at 3.5g ms-2 at 560 Hz of vibrational excitation. The second design involves a torsionally vibrating plate which is capable of binary phase shift keying demodulation. This plate is driven by electrostatic forces and electrostatics provide signal mixing. The target application is demodulation of signals encoded according to the 802.15.4 standard which describes a low data rate BPSK signalling with a carrier frequency of 868 MHz and a chip rate of 300 kchips/s. It is shown that the torsional plate has a damped resonant frequency of 1.54 MHz and this being greater than the 3rd harmonic of the data rate recovers the baseband signal successfully with 20 V peak of actuation voltages. At normal temperature and pressure, the resulting Q-factor was found to be 60 which narrows the frequency response and as a result the baseband signal recovered is slightly oscillatory. This same torsional plate is investigated under higher actuation voltages and it is shown that when actuation voltages exceed 75 V, nonlinear spring behaviour dominates the response and chaotic trajectories in phase-space appear. At these higher voltages, this device can be used for different purposes, for example, as a hardware random number generation and a chaotic carrier generator. One drawback of using electrostatics for mixing purposes is that apart from the required pure mixing components, spurious products also appear. This is due to the quadratic relationship in the electrostatic interaction and these would need to be filtered out mechanically. However, it is shown that with a differential electrostatic drive using the same torsional plate, these spurious products are attenuated and the resulting plate displacement becomes practically proportional to pure signal mixing. This relaxes the bandwidth-selectivity trade-off in the mechanical filtering and consequently relieves some of the dimensional constraints of the torsional plate. With this possibility, an in-phase/quadrature mixer is designed that is able to demodulate different quadrature amplitude modulated signals with drive voltage levels at 17 Vrms, a footprint of around 40,000 µm2 and giving output voltage levels of 0.18 Vrms for the in-phase and quadrature signals.Two features are considered novel in this design; the width of application and the ability to approximate pure mixing. These features are a result of the adopted torsional topology. In the final design, also related to communication signal processing, a MEMS device is presented that is able to convert a BPSK signal to a simpler amplitude shift keying modulation scheme. Although the structure involves also rotational motion, the topology is very different and much more complex than the designs mentioned previously. A mathematical model was developed and validated against finite element analysis simulation results and this was used to obtain optimised dimensions using a hybrid particle swarm optimisation algorithm. The design, with a footprint of 2.9 mm2, was fabricated and experimentally validated. It was tested with carrier frequencies ranging from 174 kHz to 1 MHz at a binary phase shift keying (BPSK) data rate of 6.6 kbps and with carrier amplitudes of 9.7 V, resulting in an amplitude shift keying (ASK) modulation index of 0.79 at the output sensors and a power consumption of 2.9 𝜇W. The novelty of this device is that it provides a MEMS solution for BPSK to ASK conversion, a function that has always been realised in CMOS as the first stage to BPSK demodulation. The device is capable of meeting current specification requirements (data rates, power consumption and footprint) for implantable medical devices. It is demonstrated that the power consumption is low enough such that it provides an attractive alternative to CMOS realisations. Moreover, being a MEMS, has potential for integration with MEMS sensors and harvesters in wireless sensor network nodes. Description: PH.D. 2020-01-01T00:00:00Z