Abstract
dc:description.abstractIn this thesis work will be presented which demonstrates the potential benefits and current obstacles to the creation of a geometry morphing procedure for the computational simulation of hot corrosion and related phenomenon in gas turbines using Reynolds-averaged Navier Stokes (RANS) computational fluid dynamics (CFD) simulations. Non-uniform distribution of gaseous species as a result of film coolant injection was observed on a simple single hole test case. RANS simulation showed that the gas composition along the centreline, downstream of a cooling hole injection site, is predominately determined by the coolant gas composition for multiple cooling hole diameters downstream. This observation, of interest when studying the rates of surface chemical reactions, persisted when the same techniques where applied to an nozzle guide vane (NGV) cascade. A simplified model for oxidation and hot corrosion of elemental nickel is suggested. This model was used to drive a geometry morphing technique, based on a level-set geometry representation, of an NGV cascade. The susceptibility of geometry deformation to surface temperature and gas species concentration, being influential variable for hot corrosion, suggests geometry-based feedback in similar scenarios. The potential of highly data-driven hot corrosion modelling was also explored. As well as predicting relative surface loss after a prescribed operational time, estimates of surface roughness were performed. In locations on the model geometry susceptible to hot corrosion, the predicted values for root mean squared roughness, between 30 µm and 48 µm, were comparable with a representative values of 36 µm, obtained by Bons et al. (2001) through the measurement of in-service turbine components suffering from hot corrosion. Peak to valley roughness predictions displayed a similar agreement with published representative values. Work was also performed on predicting material gain or loss in response to the erosion or deposition of impacting particles over a range of particle sizes, temperature and velocities. While particulate effects where not incorporated into the hot corrosion modelling, this is another example of where the geometry morphing procedure used in this work could be of use. The resulting particle impact distributions were qualitatively in line which other published works demonstrated sensitivity to both particle size and softening temperature. However, the particle modelling was not quantitatively accurate and over-predicted particle deposition by orders of magnitude.
Degree
thesis:*- Name dc:type.qualificationname
- Doctor of Philosophy (PhD)
- Level dc:type.qualificationlevel
- Doctoral
- Grantor dc:publisher.institution
- University of Cambridge
- Year dc:date.issued
- 2025
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Lewis, Maximilian
- Advisors dc:contributor.advisor
-
- Dawes, Bill
- Atkins, Nicholas
Subjects
dc:subject × 5Rights
dc:rightsIdentifiers
dc:identifier.*- DOI dc:identifier.doi
- https://doi.org/10.17863/CAM.125031
- OAI identifier oai:identifier
- oai:www.repository.cam.ac.uk:1810/395581