Research On The Key Technologies Of Natural Fatigue Cracks Based On Eddy Current Pulsed Thermography

Safety technology and equipment development are one of the goals of China's “13th Five-Year Plan”.NDT&E is a key technology to ensure the quality and safety of major engineering equipment and product manufacturing,and is an effective means to ensure material quality and achieve quality control.It has been widely used in aerospace,railway,nuclear industry,power system and machinery manufacturing,and has brought significant economic and social benefits.Non-destructive testing is hailed as the “Quality Guardian” of the industry.Then,the crack is “industrial cancer”.Fatigue cracks have no obvious precursor information before fracture,so it is easy to cause catastrophic consequences,causing huge economic losses and casualties.The traditional magnetic particle non-destructive testing method widely used in the industry is affected by the coating layer on the surface of the workpiece,which greatly reduces the detection sensitivity.It is difficult to implement effective fatigue cracks,especially in the micro crack stage,in key parts of in-service large equipment and components.In recent years,Eddy Current Pulsed Thermograohy(ECPT)is an emerging multiphysics coupled induction thermography method.It has been wide attentioned by detection area,failure analysis,and structure health monitoring due to its large detection area,fast and non-contact detection,high spatial resolution and visual presentation.For the detection and evaluation of fatigue microcracks,there are three major challenges in electromagnetic thermal imaging: first,lateral thermal diffusion will have a fuzzy effect on the thermal image,and the electromagnetic and thermal disturbance generated by the microcrack will be submerged;secondly,many factors will lead to the uneven heating of the workpiece,such as the selection of the excitation coil and the complex geometry of the detection object.It will reduce the spatial resolution of the detection.Finally,for the in-service workpiece,the surface of the test piece is poorly tested and may coexist different types of defects which make it difficult to extract and separate target defects.Aiming at the three challenges of ECPT for natural microcrack detection,this paper,firstly,constructs a mathematical-physical time segmentation model from the electromagnetic-thermal multiphysics mechanism,and establishes the theoretical basis of electromagnetic thermal imaging nondestructive detection and evaluation.The detection time window of the crack reduces the influence of the blurring of the lateral thermal diffusion.Subsequently,how the excitation source parameters improve the detection efficiency of the system and the influence of the sensing architecture design on the electromagnetic-thermal field uniformity are discussed.The effect of uniformity of electromagnetic-thermal field is analyzed for the detection;finally,based on the theoretical model,the optimization of the excitation source is guaranteed,a Thermal Pattern Contrast(TPC)algorithm is proposed for fatigue microcracks using ECPT is studied.The thermal pattern extraction,separation and evaluation of natural microcracks in the in-service impact damage workpieces with complex geometry and harsh surface environment were researched.The main research contents and innovations of this paper are as follows:1)Establish a mathematical-physical time segmentation model by studying the multiphysics mechanism of electromagnetic and thermal fields.The theoretical derivation of each physical stage of electromagnetic thermal imaging is carried out from electromagnetic energy and time scale,and the characteristic time of each stage is studied.The process of ECPT is divided into four different physical phases,and each phase is explained about the physical phenomenon.Starting from the thermal diffusion equation,the multiphysics electromagnetic and thermal parameters are separated.Through simulation and experiment,the artificial crack and natural crack are verified by the mathematical-physical time-segmentation model.The effects of different physical stages in the detection are explained from the two aspects of electromagnetic field and thermal field.Then the conclusion is drawn: It has a good effect on the quantification of cracks on metal surfaces in first stage.Finally,the theoretical model results were verified by using natural edge cracks of turbine blades.The staged problem of ECPT and the influence of physical field on the detection results are solved.The detection time window of microcracks is clarified,and the fuzzification effect of lateral thermal diffusion is reduced,which lays a foundation theoretical basis for fatigue microcracks detection by ECPT.2)The theoretical optimization of excitation parameters of ECPT based on time segmentation model is studied.The control variable method is used to discuss the effects of single excitation parameters: different induction heating times and cooling times,and different excitation currents on the detection results.Theoretically,a clear directional choice scheme is provided for the selection of excitation parameters.Finally,the optimization of the excitation parameters is realized,and the detection efficiency of the system is improved,which provides a prerequisite for the non-destructive testing and evaluation of micro-crack electromagnetic thermal imaging.3)In order to study the influence of electromagnetic-thermal uniformity for the detection results,a uniform electromagnetic thermal excitation coil was designed.Through the design of induction coils with different structures and different numbers of turns,the uniformity of the electromagnetic thermal multiphysics of these coil models is theoretically calculated and compared,and the uniformity optimization design of electromagnetic thermal excitation coils is realized.On this basis,for the complex geometry of the edge,the finite element simulation is used to study how the uniform electromagnetic field excited by the coil is destroyed in the detection and the influence on the uniformity of the thermal field.Finally,from the optimization of excitation parameters and the design of excitation coil,the effect of electromagnetic-thermal field uniformity on the detection of microcracks was studied for the natural microcracks of pressure vessel welds.4)Based on the theoretical model,optimize the excitation source to ensure the thermal model comparison algorithm,and study the non-destructive detection and evaluation of fatigue microcrack by ECPT.By capturing the difference between the thermal diffusion modes between the defect and the non-defect,the detection problem of the weak electromagnetic-thermal disturbance signal is solved.Finally,the feature extraction and separation of the micro-crack weak signal of ECPT is realized.The TPC algorithm mainly has two processes: thermal diffusion motion feature extraction and defect separation.Firstly,the mapping relationship between optical flow algorithm and thermal diffusion theory is established theoretically.The thermal diffusion is characterized by optical flow vector,and the thermal diffusion motion characteristics of defects and non-defects are extracted.Secondly,the optimal fixed frame interval is selected by calculating the relationship between the microcrack and the strong external interference,and then the entire thermal image sequence is subjected to optical flow algorithm which every two frames of thermal images under the fixed interval frame number to form a new data set.This data set globally and dynamically associates the temperature information of individual pixel points in the original thermal image data set,extracts the transient motion and the time-space domain information in the original data,and forms a thermal diffusion data set based on the optical flow field.Through the motion information of thermal diffusion,the difference between the defect and the non-defect thermal pattern is maximized contrast.Finally,the new data set after optical flow processing is taken as input,and the micro-crack thermal pattern separation is performed by principal component analysis to realize the feature extraction and separation of the micro-crack weak signal under ECPT system.5)Finally,we test the impact damage samples which are in the in-service aircraft brake system.There existed many strong interference factors,such as impact of complex geometric structure,harsh impact surface environment,a variety of defects coexist.Based on the platform of eddy current pulsed thermography,the natural fatigue microcracks have been verified by thermal pattern contrast algorithm.Firstly,compared with the standard magnetic particle test results,the effectiveness of the thermal pattern contrast algorithm is determined.Secondly,the robustness of the thermal pattern contrast algorithm is verified by multi sets of different in-service test pieces,from F-score and signal-to-noise ratio(SNR).The results of this algorithm are compared with the results of traditional thermal imaging signal processing methods.It is concluded that the thermal contrast algorithm has an essential improvement on the feature extraction and recognition of the micro-crack weak signal of the impact damage test piece which is in the in-service aircraft brake system.It is realized the detection and evaluation of the fatigue micro-crack under ECPT system.The above research will contribute to the research of natural fatigue microcracks in the field of induction thermography.It will help to predict and control the development and evolution of fatigue microcracks.It provides a theoretical basis for on-line and portable detection system of induction thermography for the key parts of in service largescale equipment and components,which provides a possibility to replace traditional magnetic particle detection(MT).