Wire electrical discharge machining (WEDM), widely used to fabricate micro and precision parts in manufacturing industry, is a nontraditional machining method using discharge energy which is transformed into thermal energy to efficiently remove materials. A great amount of research has been conducted based on pulse characteristics. However, the spark image-based approach has little research reported. This paper proposes a discharge spark image-based approach. A model is introduced to predict the discharge status using spark image features through a synchronous high-speed image and waveform acquisition system. First, the relationship between the spark image features (e.g., area, energy, energy density, distribution, etc.) and discharge status is explored by a set of experiments). Traditional methods have claimed that pulse waveform of “short” status is related to the status of non-machining while through our research, it is concluded that this is not always true by conducting experiments based on the spark images. Second, a deep learning model based on Convolution neural network (CNN) and Gated recurrent unit (GRU) is proposed to predict the discharge status. A time series of spark image features extracted by CNN form a 3D feature space is used to predict the discharge status through GRU. Moreover, a quantitative labeling method of machining state is proposed to improve the stability of the model. Due the effective features and the quantitative labeling method, the proposed approach achieves better predict result comparing with the single GRU model.

Bearing steel (GCr15) is widely used in key parts of mechanical transmission for its excellent mechanical properties. Electrochemical machining (ECM) is a potential method for machining GCr15, as the machining process is the electrochemical dissolution of GCr15 regardless of its high hardness (>50 HRC). In ECM, NaNO₃ solution is a popular electrolyte, as it has the ability to help in the nonlinear dissolution of many metallic alloy materials, making it useful for precision machining. However, due to high carbon content of GCr15, the electrochemical dissolution of GCr15 is unique, and there is always a black layer with high roughness on the machined surface, reducing the surface quality. In order to improve the electrochemical machining of GCr15 with a high surface quality, the surface characteristics of GCr15 in ECM were investigated. The anodic polarisation curve in the NaNO₃ electrolyte was measured and electrochemical dissolution experiments were conducted with different current densities. SEM, XRD, and XPS were employed to analyse the surface morphology and composition formed on the machined surface at different current densities. The initial results showed that there were two parts (black part and bright part) formed on the machined surface when a short circuit occurred, and the test results suggested that the black part contained a mass of Fe₃O₄ while the bright part was composed of mainly Fe and Fe₃C. Further investigation uncovered that a black flocculent layer (Fe₃O₄) always formed in a low current density (32 A/cm²) with high roughness. With the current density increased, the amount of black flocculent layer was reduced, and Fe₃C particles appeared on the machined surface. When the current density reached 81 A/cm², the entire flocculent oxide layer was removed, only some spherical Fe₃C particles were inserted on the machined surface, and the roughness was reduced from Ra7.743 μm to Ra1.783 μm. In addition, due to exposed Fe₃C particles on the machined surface, the corrosion resistance of the machined surface was significantly improved. Finally, circular arc grooves of high quality were well manufactured with current density of 81 A/cm² in NaNO₃ electrolyte.

The study explored a deep learning image super-resolution approach which is commonly used in face recognition, video perception and other fields. These generative adversarial networks usually have high-frequency texture details. The relevant textures of high-resolution images could be transferred as reference images to low-resolution images. The latest existing methods use transformer ideas to transfer related textures to low-resolution images, but there are still some problems with channel learning and detailed textures. Therefore, the study proposed an enhanced texture transformer network (ETTN) to improve the channel learning ability and details of the texture. It could learn the corresponding structural information of high-resolution texture images and convert it into low-resolution texture images. Through this, finding the feature map can change the exact feature of images and improve the learning ability between channels. We then used multi-scale feature integration (MSFI) to further enhance the effect of fusion and achieved different degrees of texture restoration. The experimental results show that the model has a good resolution enhancement effect on texture transformers. In different datasets, the peak signal to noise ratio (PSNR) and structural similarity (SSIM) were improved by 0.1–0.5 dB and 0.02, respectively.

Multi-channel high-speed wire electrical discharge machining (HSWEDM) has shown great potential in enhancing the cutting rate of metal workpieces. However, the mechanism of multi-channel discharges in this technique remains unclear. In this paper, the equivalent circuit and processing model of the multi-channel HSWEDM were developed to investigate the discharge characteristics. It was found that the equipotential between electrodes is the primary factor causing electrical signal coupling between channels, hindering the achievement of synchronous discharge. To address this issue, a novel power supply with a decoupling circuit was devised. By utilizing the combined effect of electrode wire resistance and current limiting resistance (R_c), a potential difference was induced between electrodes in different channels, enabling electrical signal decoupling and facilitating synchronous discharge. The impact of R_c on synchronous discharge was examined, revealing that a reduction in R_c can increase the gap voltage of non-breakdown channels, thereby enhancing the discharge ratio. Finally, cutting rate experiments were conducted. When the new power supply was used for electrical signal decoupling, the cutting rates of multi-channel WEDM were significantly improved. Compared to single-channel HSWEDM, the cutting rates of two-channel and four-channel HSWEDM are enhanced by 84.06% and 247.83%, respectively.