Manufacturing Variability in High Pressure Turbine Rotor Blades
Repository URI
Repository DOI
Change log
Authors
Abstract
This thesis is a compilation of three studies that give a detailed understanding of the impact of manufacturing variability on high pressure turbine rotor blade aerodynamic and proposes several means by which the effects of manufacturing variability can be mitigated.
A prerequisite for the aerodynamic studies is the ability to process the measurements of real blades. Structured-light scanning was used measure the manufactured blades and geometry processing scripts were developed to extract the manufacturing variability once the blades were aligned to the design-intent coordinate system. The extracted manufacturing variability were then mapped to computational fluid dynamics (CFD) models and assessed aerodynamically.
For the first study, blade sections were analysed using MIT’s Multiple Blade Interacting Streamtube Euler solver (MISES). The distribution of blade surface curvature was found to influence the blade surface velocity distributions, which consequently affect the development of the boundary layers, ultimately contributing to changes in momentum deficit and aerodynamic blockage. The recommendation here was to measure the blades with sufficient resolution to resolve curvature in order for a meaningful performance assessment to be made.
The second study analysed the measured blades using Reynolds-Averaged Navier-Stokes (RANS) CFD. From this study, an exit flow angle-based blade reorientation procedures and a functional sentencing methodology were proposed. For a statistically non-representative sample, these changes to the manufacturing process reduces the spread in stage capacity from ~±0.5% to ~±0.15% while increasing the manufacturing yield rate from ~68% to ~97%.
The third and final study was performed using RANS CFD to understand the multi-passage behaviour of assembled HPT rotor blade rows. This study included the shroud and the flow structures around the tip of the blade were found to be the result of the interaction between the horse-shoe vortex and the pressure fields of the blades. In a multi-passage environment where periodicity is relaxed, the flow structures were also dependent on how the blades interacted with one another. However, despite the change in flow structures, averaging results obtained from single-passage CFD models gave good approximations of multi-passage stage capacity and mass-averaged exit flow angle.
Description
All but one of the figures have been granted clearance by the copyright holders.
One figure has been redacted for confidentiality reasons.
Date
Advisors
Coull, John D