AbstractThe need for metrological systems with new capacities to perform measurements that go beyond the limits of existing measurement systems is being driven by recent technological advancements and innovation. Structures of machines distort due to thermal and finite stiffness effects which affect their performance hence the need for structural monitoring. At the design stage of machines and instruments, metrology must be embedded and accessed as part of the machines' and instruments' functioning. Structures of machines distort due to thermal and finite stiffness effects which affect their performance hence the need for structural monitoring. This need is not met by existing measurement systems and is the drive for innovation of new metrological systems. The result of this may take the form of metrological systems embedded in machines
and instruments during the design stage to add structural monitoring functionality to the devices.
The performance specification for embedded sensing techniques is very strict regarding dimensional stability. The materials currently used in precision measurement systems have some limitations that preclude their use for most applications. Different methods and materials are being investigated for future precision measurement systems. The main properties considered are good dimensional stability, light weight, and low cost. Although strain measurement of a structural part is ubiquitous in many safety-critical systems, it is rarely employed in precision machinery to allow for error correction. This could be owing to the difficulty of installing multiple sensors to characterise complex bending of structural parts and the associated cost of embedding precision sensors. Every sensor has shortcomings, and knowing these limitations is important for accurately assessing
a system's performance. Many well-defined parameters can provide a detailed characterisation of a sensor's output, rather than just specifying the quality of a sensor in terms of accuracy. Using special configurations and techniques, a designer may also minimise the impact of a sensor's inherent limitations.
In this thesis, the design of a new structural deformation monitoring system based on a low-cost displacement sensing methodology using off-the-shelf slotted photomicrosensors and connecting rods is investigated. The uncertainty contributors are investigated and a technique for in-situ calibration of the sensor is developed. This
is especially important for embedded applications where it is impractical or costly to remove the sensors.
The sensor design incorporates a differential voltage technique that reduces the influence of variation in input voltage, ambient light variations and differential thermally induced expansion between the sensor circuit and shutter. Traditional rigid link mechanisms move due to rolling or sliding but this is undesirable as they can induce hysteresis, wear, and backlash. To overcome this limitation and provide precise displacements at the application of a force, a flexure mechanism was designed. The output of the sensor showed 90% correlation at zero lag to the output of a laser interferometer.
Misalignment of the sensors can occur due to vibration for example, which can affect the calibration accuracy. In an embedded application, it is impossible to ascertain which sensors have been misaligned during installation without removing the sensors. This is also true for long term variability such as drift or possible misalignment or damage during maintenance of the machinery on which the sensor is installed. To facilitate in-situ calibration, a self-excitation technique was developed utilising controlled voltage excitation to detect and correct for variation during the lifetime of the sensor while it is embedded on a machine. The voltage calibration technique achieved 60% reduction of errors due to shutter movements.
A limitation of the flexure mechanism is the introduction of additional stiffness. To reduce the stiffness introduced by the flexure mechanism and reduce the cost of machining intricate parts of the sensor by traditional means, an improved 3D printed design of the sensor is created. The sensor was configured together with suitable connecting rods to form a framework for measuring structural deformation. Simulation results show the reduced flexure stiffness resulting in 30% reduction in compression/tension losses of the connecting rods. The results also show that the improved design measures over 50% more strain than the traditionally manufactured sensor.
Finally, the characterisation results are used to develop an uncertainty budget for a system that utilises the newly developed sensor for structural deformation monitoring. A total uncertainty of 4 µm was achieved for a system that uses the sensor for structural monitoring.
|Date of Award||2023|
|Supervisor||Simon Fletcher (Main Supervisor), Andrew Longstaff (Co-Supervisor) & Naeem Mian (Co-Supervisor)|