| Ferromagnetic materials composed of steel and its alloys are widely used in petrochemical,aerospace,bridge construction,manufacturing equipment,military equipment,infrastructure,etc.for its good mechanical properties.For pressure-bearing equipment,stress concentrations,defects,fatigue damage,etc.generated during the service of its structural components are the main causes of material failure.Magnetic nondestructive testing technology(MNDTT)can effectively detect stress,defects and fatigue damage in ferromagnetic materials and provide technical support for safety assessment and maintenance decision of in-service components,which is important to avoid safety accidents and ensure safe and stable operation of components.However,due to the insufficient research on the mechanism of MNDTT,there is a large error between the quantitative theoretical results and experimental results,resulting in insufficient accuracy of forward and inverse analysis.The selection of process methods and related parameters involved in the whole testing process depends on long-term working experience and lack of guiding process methods.The quantitative experiments load the stress below 0.5σs,and use the linear coefficient between the detection signal and elastic stress as the identification parameter of stress,resulting in large errors in quantitative results and inability to accurately detect the working stress,which leads to deviations in the evaluation of the stress damage state of the sample,in addition,there is a lack of quantitative research on MNDTT for defects/early fatigue damage.The above problems make it difficult for MNDTT to accurately assess the safety status of components and to inhibit the continuation of the damage development and the occurrence of malignant accidents.Based on this,this paper focuses on the problems during MNDTT of ferromagnetic materials,and carries out systematic research work through theoretical modeling,simulation analysis and quantitative experiments:(1)Research on the physical mechanism involved in the magnetization process and establish a nonlinear magnetization constitutive model of ferromagnetic materials.The magnetization mechanism of ferromagnetic materials is revealed from the perspectives of magnetic domain structure,magnetization characteristic curves and energy changes.The mechanism of tensile and compressive stress on the magnetization state is analyzed,and the coupling mechanism between stress and magnetization is clarified.Based on the laws of thermodynamics,magneto-mechanical coupling mechanism,magnetic domain theory,a new magnetostriction model was established,and the effect of elasto-plastic deformation on magnetization is transformed into an equivalent magnetic field to form a new effective magnetic field equation,which is combined with the Frohlich-Kennelly equation and energy balance equation to establish a new nonlinear magnetization constitutive model.The quantitative analysis of magnetostriction,magnetization and hysteresis shows that the theoretical results of the model proposed in this paper are closer to the experimental results than the existing models.In particular,the accuracy of the prediction of magnetic properties under compressive stress is greatly improved.The differences and physical mechanisms of different models in predicting the magnetization and hysteresis of ferromagnetic materials are discussed.(2)MNDTT simulation analysis of stress/defects in ferromagnetic materials.Based on the electromagnetic field theory and the characteristics of MNDTT,the simulation analysis process of MNDTT were determined.The influence of post-treatment and heat treatment on the magnetic characteristic measurement experiment was studied,the treatment method of the magnetic characteristic measurement specimen was determined.The B-H curve obtained from the magnetic characteristic measurement experiment is fitted with the nonlinear magnetization constitutive model proposed in this paper,then the B-H curve under arbitrary load was predicted.The simulation model of MNDTT was established to study the influence of excitation frequency and excitation voltage amplitude on the detection signal,and the excitation frequency of 10Hz and voltage of 25V were determined for the simulation analysis in this paper.The current peak difference and the characteristic parameters of the peak electric signal curve were proposed as the quantitative evaluation indexes of stress and defects,revealing the nonlinear positive correlation between the peak current difference and stress,and the negative correlation between the characteristic parameters and defect length.(3)Research on quantitative MNDTT experiments of stress/defects in ferromagnetic materials.By means of XRD,EBSD,MFM and VSM,the changes of microstructure,magnetic domain morphology and magnetic characteristic parameters of ferromagnetic materials under different annealing temperatures were studied.The mechanism of stress relief annealing on the magnetic properties of the materials were revealed,and the final selection of stress relief annealing temperature range for Q345R steel was 400～600℃.The quantitative experimental results show that the current difference signal increases nonlinearly with the increase of elastic stress(the relationship between them is hyperbolic tangent),and decreases nonlinearly with the increase of plastic stress.Comparing the results of MNDTT and X-ray inspection of residual stress,the effectiveness of inversion of stress by hyperbolic tangent function is proved.When using the detection signal to predict stress,compared with the BP neural network,the particle swarm optimized BP neural network algorithm reduced the maximum relative error of the predicted value from 20.08%to 14.65%,a reduction of 5.43%.The characteristic parameter gradually decreases from a constant greater than 1 to a constant less than 1 with the increase of defect length,gradually increases with the increase of defect width,and first increases and then decreases with the increase of defect depth,but the characteristic parameters corresponding to the experimental results of defect width and depth are less than 1,indicating that the change of defect width and depth will not change the position of the long and short axes of the signal curve.A double-exponential attenuation model of the detection signal based on the probe lifting off was established,and the coefficient of determination between the theoretical results and experimental results is above 0.996,indicating that the double exponential attenuation model is accurate and effective for quantitative analysis of the probe lifting effect.(4)Research on MNDTT for early fatigue damage of ferromagnetic materials.Based on viscoplastic model,fatigue model and nonlinear magnetization constitutive model proposed in this paper,a method was constructed to evaluate the early fatigue damage of ferromagnetic materials using magnetic signals.The comparison with experimental results showed that the method can accurately analyze the magneto-mechanical behavior of ferromagnetic materials during fatigue process.The change law of magnetic signal of ferromagnetic material with fatigue cycle in constant and alternating magnetic fields were investigated.Based on magnetic signal,a model of early fatigue damage evolution of ferromagnetic materials was proposed,and the approximate linear relationship between early fatigue damage and magnetic signal was revealed,indicating that it is feasible and effective to use magnetic signal to characterize early fatigue damage.Through the systematic research in this paper,a nonlinear magnetization constitutive model of ferromagnetic materials is established,and a quantitative evaluation method of MNDTT for stress/defects/fatigue early damage of ferromagnetic materials is proposed,which helps to further understand the physical mechanism and quantitative law of detection signal in the process of MNDTT of ferromagnetic materials,and improves the research framework of MNDTT,which is of great significance to promote the progress of MNDTT and the development of related advanced manufacturing technology and equipment. |
PDF Full Text Request