Study Of Infrared Inspection Technique For Aircraft Turbine Engine Blades

Turbine blades are the crucial heated end components of gas turbine engine, which have directly impact on the thrust-to-weight ratio and safety of engine. Because of the speciality of turbine blade materials, the complexity of manufacture processing and the high cost, several nondestructive techniques have been proposed and used for blades inspection tasks. For conventional in nondestructive testing methods, such as X-ray radiography and ultrasonics, there are some drawbacks for blade inspection. The application of X-ray technology is restricted by the radiation leakage problem and high expenditure on maintaining X-ray system.In this dissertation, in order to meet safety requirements and improve efficiency and accuracy, thermal wave nondestructive testing technique has been used to test defects of turbine blades. Thermal wave nondestructive testing is different from traditional infrared nondestructive testing techniques. The surface and subsurface structure information of specimen may be obtained through controlling the stimulation media actively and monitoring the temperature distribution of specimen. Both metal and non-metal material defects can be inspected using the technique. This method has superiorities, including high testing efficiency, large visual field, quantitative analysis based on digital image processing and no radiation.In this thesis, the hardware circuits and the software system of thermal wave nondestructive testing system are illustrated and several key issues in testing experiments are mainly discussed. The main research content and innovation are as follows:(1) The thermal wave nondestructive testing system for turbine blades of aircraft engine has been well designed. The installation and debugging tasks of the system have been accomplished. The controlling circuit, the most important part of the system, has been designed and tested. The system is semi-automatically controlled, and automotive controlling the image acquisition of infrared image sensor, auto-focusing of infrared image sensor, moving of 3-D platform, rotating of 360°accurate electric driven table and electric valves have been realized through data communication between PC and single-chip microcomputer.(2) Different heat medias and stimulation methods have been used and vast defects testing experiments have been done, based on in-depth analysis of defects characters of thermal barrier coatings and cooling ducts of turbine blades.(3) Pulse coupled neural network, which is a model derived from a neural mammal model, has been established to enhance image. The experiment results show that the algorithm can enhance contrast and brightness of image effectively and remove noise in images efficiently, especially for the infrared image with low contrast and brightness.(4) The influences of four factors (surface emissivity, noises, stimulation methods and acquisition time of images) on the results of experiments are amply analyzed. Methods of solving these problems are presented.