Image of the surface of gas turbine blade as a diagnostic signal

This paper outlines a non-destructive method that is suitable for evaluation of condition demonstrated by gas turbine blades and is based on digital processing of images acquired from the blade surface in visible light. To enable high clearness of these images the particular attention is paid to the problem of how to provide optimum conditions for investigations and mitigate geometrical distortions of images acquired from maintenance operations. The paper demonstrates that there are relationships between operation lifetime of blades and discoloration of their surfaces due to overheating of the blade material. These relationships are revealed by digital analysis of images acquired for the blade surfaces and expressed as statistical parameter of the first and second order. To improve unambiguity of the analysis results a low-pass filter was applied. It was demonstrated that these relationships are suitable for evaluation how much the status of the blade material microstructure is altered