Abstract
Actual contact between two real mating surfaces happens only at discrete points and so the real contact area is significantly less than the apparent contact area. This leads to a constriction of heat flow through the larger length-scales (out-of-form and waviness), and then a further constriction through the smaller length-scale (surface roughness). Constriction of heat flow leads to a temperature drop across the interface known as the Thermal Contact Conductance (TCC). It has been shown that the existing models and experimental work are limited only to some specific applications. Prior to this study, very limited work has been carried out to understand the effect of TCC on gas turbine applications. The objective of this work is to carry out experimental work and numerical modelling that will add to the volume of work in this field and also will help to increase the understanding about contact conductance in gas turbine technology. Prior to this study no work has been published on the characterisation of the topography of mechanically contacting gas turbine components. In this study the surface topography of a turbine casing from a current Rolls-Royce engine was examined. Manufacturing techniques that produce surface patterns that typify the surface topography applicable to turbomachinery were used to prepare the samples. Performing Fourier analysis on the engine component revealed that the two dominant length-scales applicable to such surfaces are the medium frequency (waviness) and the large frequency (surface roughness) length-scales. A split-tube technique experimental methodology with in-line washers forming multiple interfaces was used to perform various TCC test cases. A parametric study of TCC is performed to verify the effects of the different length-scales of surface topography. The study also aims to understand the effect on TCC of other parameters such as contact pressure, material and operational history that are relevant in gas turbine technology. It was shown that the large scale, repeatable surface deviations played a primary role in the interface contact heat transfer. For accurate prediction of TCC it is vital to accurately simulate the surface topographies. Previous studies fail to accurately characterise machined surfaces. In this study, a Fast Fourier Transformation (FFT) and Inverse Fast Fourier Transformation (IFFT) algorithm were used to perform Fourier analysis on machined surface. The methodology developed has shown that any machined surface could be accurately modelled as long as the manufacturing parameters such as the feed rate and cutting tool geometries are known. As a result the surface topography of any machined gas turbine component could be accurately modelled based on this methodology. Using this idealised surface model the thermo-mechanical contact problem for machined interfaces were investigated using finite element techniques. It was shown that gap conductance plays a crucial role in accurate prediction of TCC for gas turbine applications. It was shown that with the combined effect of the numerical surface modelling along with the finite element modelling technique, it was possible to estimate the TCC of various mating gas turbine components.