Feature Defection Model And Algorithm Characterization Using Electromagnetic Heat Nondestructive Testing

Non-destruction Test and Evaluation(NDTE) is a key technical to ensure operation safety of engineering equipment and the quality of product manufacturing. Using non-destructive testing evaluation, the size, location and shape of the the workpiece defect is provided. NDTE is a method of using acoustic, the magneto- optical effect, electrical properties, and eddy current to detect damage of the sample, without prejudice to workpiece performance. One recent development has been the combination of eddy current and thermography techniques. The results of this illustrate the advantage of different depth defect characterisation, with the technique known as eddy current pulsed thermography(ECPT) or pulsed eddy current(PEC) thermography for area imaging of defects without scanning. ECPT has been applied to detect and evaluate defects for damage of sample measurement and monitoring, and achieved rich results. Based on the pulsed eddy current thermography technology, the mathematical models are built to analyse the position, the type and etc. of the damage. However, it is difficult to apply a link between the mathematical models and the material physical properties of defect. As there are many types of defects in materials, people need to find a simple model to improve the accuracy of defect dectection using electromagnetic technology. In this paper, using the physical effect of pulsed eddy current thermography, thermo-optical flow is designed for the defect detection of complex component. Combining theory and experiment, the sensitivity, accuracy, generality of thermo-optical-flow model are analysed through impact damage detection and micro defect detection of complex components.The main contents of this thesis are:1) First of all, the principle of pulse eddy current thermography method is introduced. And in order to achieve the optimization device of pulsed eddy current thermograph the influence of thermal imager sampling frequency for defect detection and material texture analysis is investigated.2) Based on the pulsed eddy current thermography, thermo-ptical-flow model is derivated to trace the the heat flow, which is quite sensitivity to the property variation of material due to the vary of both heating propagation and volume of heat. In order to quantitatively analyse these heat flows, thermo-ptical-flow-entropy model and thermo-optical-flow-divergence model are calculated to quantify the differences of heat propagation caused by material changes(the degree of the damage) for the assessment of defect detection.3) The thermo optic flow model is applied for the study of early fatigue on gear and impact damage on composite materials, and thermo-ptical-flow-entropy model and thermo-optical-flow-divergence model is used for quantitatively analysis. Futhermore, the positive and negative values of the divergence and their relationship between composite structures and material state are investigated, and the relationships between thermo-optical flow entropy and the degree of fatigue are established.