Abstract
The discovery of X-rays revolutionised medical imaging, never before had a technique offered a method of non-invasively imaging the human body. X-ray computed tomography (XCT) built on this, allowing not only the imaging but the reconstruction of a scanned object into a 3D volume to be realised. For both fields the primary use in the formative years was medical imaging, for which the technologies were uniquely suited. Both however were also adopted as means of non-destructive evaluation (NDE) for engineering applications. This largely occurred for the same reasons the technologies thrived in the medical field. X-rays can be used to image the internal structure of objects without line of sight, due to their penetrative capabilities.Both X-ray imaging and XCT are however not without drawback. Most obviously X-rays are a form of ionising radiation, this was unknown in the early years of using X-rays as a medical imaging tool and led to significant numbers of illness and fatality amongst the early pioneers. Another issue highlighted by early research into X-rays and XCT was the low resolution and long processing times. Whilst between the first X-ray to the first use of XCT these had both improved, the issues remain prevelant.
Today XCT has been adopted into a wide range of industrial engineering disciplines, including porosity analysis and metrology. Current research in these fields works to develop specialist methods and parameters to further advance XCT’s usefulness as a tool for the non-destructive analysis. The use of XCT in fields such as metrology, has led to the need to establish traceability and an understanding of error sources in XCT. Traceability is the ability to trace measurements taken back to a pre-existing standard. For most metrology applications this standard is the meter. Current research has highlighted that extracted areal surface texture parameters show a level of variation not seen in other measurement methods.
This thesis investigates how instability in several key processes of the generation of X-rays may propagate through to extracted areal surface texture parameters taken using XCT. The filament is a key component in the generation of X-rays and a consumable that requires changing on a semi-regular basis to ensure continued function of the XCT. Responsible for the emission of electrons that are used to generate X-rays it is superheated to allow for thermionic emission to occur, this causes its degradation over time. Theoretically as the cross-sectional area of a part emitting electrons is altered so with the electron beam emitted. This is investigated with several studies assessing if a pattern could be established between the filament’s age and extracted parameter results. It is shown that no pattern could be established though variation in extracted surface parameters is seen throughout the filament’s life, with larger jumps present if the filament is changed.
The changing of a filament is a procedure after which the machine requires refocusing, this process is investigated, and it is shown that by refocusing the machine without changing the filament a similar variation to that noted after a filament change is achieved. To quantify this a method of measuring the focal spot of a XCT is required, several methods were considered and one selected as a base from which a novel method was developed. The novel method was developed to allow for the measurement of the focal spot alongside a surface artefact.
The method proved viable returning data in line with literature and existing methods. The thesis presents work showing the development of this method and its application in measuring both the machine’s focal spot and a surface artefact simultaneously. Measured focal spot diameter is shown to shift and links to the variation noted in areal surface texture parameters are presented.
The work shown in this thesis was carried out to investigate how variables inherent to XCT propagate into extracted areal surface texture parameters, the methods developed are applicable to any cone beam XCT with little alteration. Results presented show that users of cone beam XCT should take into consideration the effects of variation in the XCT process when performing non-destructive evaluation. The work also highlights the need for further work in developing XCT for surface metrology.
| Date of Award | 18 Sep 2023 |
|---|---|
| Original language | English |
| Supervisor | Liam Blunt (Main Supervisor) & Paul Bills (Co-Supervisor) |