In this paper, a new nonlinear magneto–mechanical coupling constitutive model, based on thermodynamic theory and the hysteresis effect, introduces the 6th order term of magnetization in the whole magnetic field and the global coupling factor related to the magneto–mechanical coupling into the magnetostrictive strain. Compared with the Zheng Xiao-Jing–Liu Xin-En model and Jiles–Atherton model, the nonlinear model precisely describes the hysteresis of the material and the magnetostrictive strain under the compressive stress and the nonlinear change of the magnetostrictive strain with stress and magnetization. The comparison of the prediction results of different models with the existing experimental data shows that the proposed model has a better prediction ability of the mid-field magnetostrictive strain, low-field magnetic flux density, and high-field magnetic flux density than other models. In addition, this paper also investigates that the effects of stress and plastic deformations on the hysteresis curve, coercivity, and remanence. The difference and mechanism of elasto–plastic deformations affecting the magnetization process of materials are obtained. Above all, the results in this paper are of great significance for the development of non-destructive testing inversion technology.

As seen in the Jiles–Atherton (J–A) model and its modified form, the linear relationship between the magnetization coefficient and the stress deviates significantly from the experimental results. It is necessary to introduce many parameters that are difficult to obtain or unknown to describe the effect of elastoplastic deformation on magnetization or hysteresis, such as shape coefficient, pinning coefficient, and molecular field coefficient. In this paper, a new nonlinear magneto-elastoplastic model for ferromagnetic materials is established based on the magneto-mechanical coupling effect, and both the sixth-order term of magnetization and the nonlinear equation of the magnetization coefficient are introduced into the magnetostriction equation. In the models established in this paper, the elastoplastic deformation equivalent magnetic field is introduced into the effective magnetic field, and the Frohlich–Kennelly equation is used to describe the anhysteretic magnetization. After comparing the prediction results of different models with the available experimental results, it is observed that the proposed model in this paper exhibits superior prediction ability for magnetostrictive strain, magnetization, and hysteresis phenomena under different stresses. This paper has also analyzed the mechanism of the effect of elasto-plastic loading and residual stress on the hysteresis in different models as well as the differences between them. The determination coefficient of the proposed model in this paper is closer to 1, which is better than the existing models, indicating that it has a better fitting effect and is of great significance to the development of quantitative non-destructive testing technology.

In recent years, the powerful feature extraction capabilities of deep learning have attracted widespread attention in the field of bearing fault diagnosis. To address the limitations of single-modal and single-channel feature extraction methods, which often result in incomplete information representation and difficulty in obtaining high-quality fault features, this paper proposes a dual-channel parallel multimodal feature fusion model for bearing fault diagnosis. In this method, the one-dimensional vibration signals are first transformed into two-dimensional time-frequency representations using continuous wavelet transform (CWT). Subsequently, both the one-dimensional vibration signals and the two-dimensional time-frequency representations are fed simultaneously into the dual-branch parallel model. Within this architecture, the first branch employs a combination of a one-dimensional convolutional neural network (1DCNN) and a bidirectional gated recurrent unit (BiGRU) to extract temporal features from the one-dimensional vibration signals. The second branch utilizes a dilated convolutional to capture spatial time–frequency information from the CWT-derived two-dimensional time–frequency representations. The features extracted by both branches were are input into the feature fusion layer. Furthermore, to leverage fault features more comprehensively, a channel attention mechanism is embedded after the feature fusion layer. This enables the network to focus more effectively on salient features across channels while suppressing interference from redundant features, thereby enhancing the performance and accuracy of the dual-branch network. Finally, the fused fault features are passed to a softmax classifier for fault classification. Experimental results demonstrate that the proposed method achieved an average accuracy of 99.50% on the Case Western Reserve University (CWRU) bearing dataset and 97.33% on the Southeast University (SEU) bearing dataset. These results confirm that the suggested model effectively improves fault diagnosis accuracy and exhibits strong generalization capability.

This study explores the potential effects of the aerodynamic parameters on the performance of the galloping piezoelectric energy harvester. By considering the geometric configurations, a bluff body cross-sectional shape evolution approach is proposed using Boolean operations on the polygons and forty-eight different cross-sectional shapes with the protruding and depressed features are considered. Computational fluid dynamics is employed to perform a time-varying simulation of the aerodynamic characteristics. The effects of the aerodynamic parameters on performance are investigated computationally using a distributed parameter electromechanical coupling model. The critical wind speed, maximum output power, and the slope of the power versus wind speed curve are introduced as the performance evaluation parameters. The results show that the rear-side protruding feature and the top-side and bottom-side depressed feature have significant potential to enhance the performance. Furthermore, a symmetrical structure of the cross-sectional shape in the downstream direction should be prioritized over asymmetric designs.

A numerical tool for simulating the detection signals of electromagnetic nondestructive testing technology (ENDT) is of great significance for studying detection mechanisms and improving detection efficiency. However, the quantitative analysis methods for ENDT have not yet been sufficiently studied due to the absence of an effective constitutive model. This paper proposed a new magneto–mechanical model that can reflect the dependence of relative permeability on elasto–plastic deformation and proposed a finite element–infinite element coupling method that can replace the traditional finite element truncation boundary. The validity of the finite element–infinite element coupling method is verified by the experimental result of testing electromagnetic analysis methods using TEAM Problem 7. Then, the reliability and accuracy of the proposed model are verified by comparing the simulation results under elasto–plastic deformation with experimental results. This paper also investigates the effect of elasto–plastic deformation on the transient magnetic flux signal, a quantitative hyperbolic tangent model between Bz_pp (peak–peak value of the normal component of magnetic flux signal) and elastic stress, and the exponential function relationship between Bz_pp and plastic deformation is established. In addition, the difference and mechanism of a magnetic flux signal under elasto–plastic deformations are analyzed. The results reveal that the variation of the transient magnetic flux signal is mainly due to domain wall pinning, which is significantly affected by elasto–plastic deformation. The results of this paper are important for improving the accuracy of quantitative ENDT for elasto–plastic deformation.

Defects are a direct cause of failure in ferromagnetic components, which can be evaluated via electromagnetic non-destructive testing (ENDT) methods. However, the existing studies exhibit several limitations (e.g., inaccurate quantification, over-reliance on algorithms, and non-intuitive result presentation, among others) in quantitative defect evaluation. To accurately describe the quantitative relationship between ENDT signals and defect dimensional parameters, the electromagnetic model and electromagnetic induction model are introduced in this paper to elucidate the physical mechanism of ENDT, as both models provide a basis for the selection of the constitutive relationship for simulation analysis. Then, a higher precision three-dimensional nonlinear finite element simulation model is established, and the effects of the excitation parameters and detection positions on signal characteristics are investigated. The simulation results indicate that the excitation frequency influences both the detection depth and sensitivity of ENDT, while the voltage amplitude affects the peak strength of the magnetic signal. Consequently, the excitation parameters are determined to be a 10 Hz frequency with a 25 V amplitude. Based on the characterization of signal peaks at positions of 0°, 90°, 180°, and 270°, the characteristic parameter K_A of the peak electrical signal curve is proposed as a quantitative index for evaluating defects. The quantitative experimental results show that K_A is related to the defect dimension. Specifically, the K_A value monotonically decreases from a constant greater than 1 to a constant less than 1 as the defect length increases, K_A is positively correlated with the defect width, and K_A follows a parabolic trend (first increase and then decrease) as the defect depth increases. Notably, K_A values associated with defect width and depth remain below 1. All the above findings provide a basis for evaluating defect dimensions. The results of this paper provide a novel ENDT method for evaluating defects, which is of great significance for improving the accuracy of ENDT and promoting its engineering applications.