Self-repairing method for micro-defects on surface of CuCr contact

Abstract

The application discloses a self-repairing method for CuCr contact surface microdefects, and the method comprises the following steps: pretreating a CuCr contact; performing laser remelting on the pretreated CuCr contact; and performing post-treatment optimization on the laser-remelted CuCr contact to obtain a CuCr contact after self-repairing of surface defects. The application can improve the compactness, grain refinement degree and service performance of the CuCr contact surface, and can also prevent the generation of secondary defects.

Description

Technical Field

[0001] This application belongs to the field of electrical contact material technology, specifically relating to a self-healing method for micro-defects on the surface of CuCr contacts. Background Technology

[0002] CuCr alloy has become the preferred material for electrical contacts due to its excellent electrical and thermal conductivity, high hardness, good resistance to arc erosion and weldability.

[0003] However, CuCr contacts are prone to various surface micro-defects (such as micropores and microcracks) during manufacturing and operation, which can seriously affect their performance and lifespan. For example, in the manufacturing process of CuCr contacts, factors such as uneven powder particle size, fluctuations in pressing density, and improper control of sintering temperature and time may lead to defects such as micropores, inclusions, and porosity on the contact surface. In the plastic processing and machining processes such as forging and rolling, if the deformation is not properly distributed or the surface finish of the die is insufficient, microcracks or scratches may also occur on the contact surface.

[0004] Currently, the main methods for repairing micro-defects on the surface of CuCr contacts include mechanical grinding and polishing, vacuum brazing, plasma spraying, and traditional laser remelting. However, these methods all have significant limitations. For example, mechanical grinding and polishing can only remove very shallow defects on the contact surface, and its effect on repairing deep micro-defects is limited. It can also easily lead to a decrease in dimensional accuracy and stress concentration. Vacuum brazing requires filler metal, which can easily reduce the electrical and thermal conductivity of the repaired area, and high temperatures may cause grain growth in the substrate. Plasma spraying coatings have low bonding strength with the substrate and are prone to peeling off. Traditional laser remelting technology has core defects: First, the parameters are not optimized for the compositional characteristics of CuCr alloy and the specific micro-defect size, resulting in incomplete defect closure and poor grain refinement. Second, the protection measures are rudimentary, and surface oxidation is prone to occur during the repair process, forming secondary defects. Third, the subsequent processing is rough, and residual microburrs can easily cause tip discharge under high-voltage conditions. Fourth, some solutions require additional steps such as laser cladding, which requires the addition of cladding materials to change the original composition of the contact, making the process cumbersome and costly. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this application is to provide a self-repair method for micro-defects on the surface of CuCr contacts. This application aims to achieve efficient and precise repair of micro-defects on the surface of CuCr contacts through a closed-loop process.

[0006] To achieve the above objectives, this application provides the following technical solution: A self-healing method for micro-defects on the surface of a CuCr contact is disclosed. The method includes: pre-treating the CuCr contact; laser remelting the pre-treated CuCr contact; and post-processing optimization of the laser-remelted CuCr contact to obtain a CuCr contact with self-healed surface defects.

[0007] Optionally, the preprocessing of the CuCr contact includes: scanning the CuCr contact to obtain a multimodal image of the CuCr contact surface; classifying defects and assessing risks on the CuCr contact surface based on the multimodal image; adaptively generating laser parameters based on the risk assessment results; and performing digital twin simulation of the laser repair process.

[0008] Optionally, the defect classification and risk assessment of the CuCr contact surface based on multimodal images includes: performing defect classification and risk assessment on the multimodal images of the CuCr contact surface using an adaptive defect identification and assessment network.

[0009] Optionally, the adaptive defect identification and assessment network includes: a multi-scale feature extraction module, an attention-enhanced fusion module, a defect classification and regression module, and a risk assessment decision module. The multi-scale feature extraction module extracts local detail features and global contextual information of surface defects on the contact, achieving efficient capture and fusion of features across scales from micrometers to macrometers. The attention-enhanced fusion module adaptively focuses on key areas of surface defects on the contact and suppresses background interference. The defect classification and regression module simultaneously performs surface defect type identification and geometric parameter regression, outputting multi-dimensional quantitative information including defect category, location, contour, and size / depth. The risk assessment decision module combines defect geometric features and process rules to perform differentiable reasoning, classifying the risk levels of various defects.

[0010] Optionally, the adaptive generation of laser parameters based on risk assessment results includes: adaptively generating laser parameters by constructing a laser parameter decision model, which includes: a feature encoding layer for receiving the three-dimensional geometric features of defects and risk rating results obtained from the preprocessing stage and encoding them into high-dimensional structured feature vectors; a rule embedding and reasoning layer for converting prior process knowledge into learnable logical mappings to achieve knowledge-guided parameter adjustment; and a parameter generation and optimization layer for generating initial laser parameters based on encoded features and rule constraints through neural network mapping, and coupling a physics-based molten pool simulation proxy model for iterative optimization to output a combination of laser parameters that meets the repair quality threshold.

[0011] Optionally, the laser remelting of the pretreated CuCr contact includes: constructing a laser remelting environment; setting laser remelting parameters; and performing self-repair on the CuCr contact surface in the constructed laser remelting environment based on the set parameters.

[0012] Optionally, the construction of the laser remelting environment includes: placing the CuCr contact in a fully enclosed laser remelting process chamber; introducing a mixed atmosphere of ultra-high purity argon and trace amounts of reducing gas into the chamber; and employing a dual-layer airflow field design inside and outside the chamber.

[0013] Optionally, the self-repairing of the CuCr contact surface in the constructed laser remelting environment includes: using a high-frequency pulse-continuous composite laser mode to self-repair the contact surface.

[0014] Optionally, the post-processing optimization of the laser-remelted CuCr contact to obtain a CuCr contact with self-healed surface defects includes: setting magnetic polishing parameters; and polishing the contact surface based on the set magnetic polishing parameters.

[0015] Optionally, the setting of magnetic grinding parameters includes: adaptive setting based on the three-dimensional morphology scanning results of the CuCr contact surface, wherein the magnetic field strength adopts a gradient control mode; the grinding speed is combined with the eddy current damping effect for frequency conversion control; and the grinding time is determined by self-termination based on real-time roughness feedback.

[0016] Compared with the prior art, the beneficial effects of this application are as follows: This application achieves efficient and precise repair of micro-defects on the surface of CuCr contacts through closed-loop control of the entire process, and proactive optimization of surface structure. It can not only completely eliminate original defects such as micropores and microcracks on the contact surface, but also effectively suppress secondary defects such as oxidation, overheating, and re-solidification burrs during the repair process. Finally, a high-quality repair layer with smooth surface, refined grains, dense structure and uniform performance is obtained, thereby improving the electrical contact reliability, mechanical stability and service life of CuCr contacts. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a self-healing method for micro-defects on the surface of a CuCr contact provided in one embodiment of this application; Figure 2 This is a macroscopic schematic diagram of the surface of a CuCr contact with surface defects. Figure 3 This is a microscopic schematic diagram of the surface of a CuCr contact with surface defects. Figure 4 These are macroscopic and microscopic schematic diagrams of the CuCr contact surface after single grinding and sand oxidation; Figure 5 This is a macroscopic schematic diagram of the CuCr contact surface after repair according to this application; Figure 6 This is a microscopic schematic diagram of the CuCr contact surface after being repaired according to this application. Detailed Implementation

[0018] Specific embodiments of this application will now be described in detail with reference to the accompanying drawings. While specific embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art.

[0019] It should be noted that certain terms are used in the specification and claims to refer to specific components. Those skilled in the art will understand that different terms may be used to refer to the same component. This specification and claims do not distinguish components based on differences in terminology, but rather on differences in function. The terms "comprising" or "including" used throughout the specification and claims are open-ended and should be interpreted as "comprising but not limited to." The following descriptions in the specification are preferred embodiments for carrying out this application; however, these descriptions are for the purpose of understanding the general principles of the specification and are not intended to limit the scope of this application. The scope of protection of this application shall be determined by the appended claims.

[0020] To facilitate understanding of the embodiments of this application, the following will provide further explanation and description with reference to the accompanying drawings and specific embodiments, and the accompanying drawings do not constitute a limitation on the embodiments of this application.

[0021] Figure 1 This is a flowchart illustrating a self-healing method for micro-defects on the surface of a CuCr contact, provided in one embodiment of this application. Figure 1 As shown, the repair method includes the following steps: S100: Pretreatment of CuCr contacts; S200: Laser remelting of pretreated CuCr contacts; S300: Post-processing optimization is performed on the CuCr contacts after laser remelting to obtain CuCr contacts with self-healing surface defects.

[0022] In this embodiment, the self-repair method for micro-defects on the surface of CuCr contacts proposed in this application achieves a closed-loop process from intelligent defect identification and precise repair to complete surface optimization through a three-step synergy of "pretreatment - laser remelting - post-treatment optimization". Ultimately, it can form a repair layer with a smooth surface, uniform grains, and dense structure, thereby improving the electrical contact performance and service stability of CuCr contacts.

[0023] In another exemplary embodiment, step S100, the pretreatment of the CuCr contact includes the following steps: S101: Scan the CuCr contact to obtain a multimodal image of the CuCr contact surface; In this step, this embodiment employs a composite surface imaging system constructed using a white light interferometer and a confocal laser scanning microscope to acquire high-precision, multi-scale, multi-modal three-dimensional morphology of the CuCr contact surface. The system uses the white light interferometer to obtain surface contour information with a large field of view and high vertical resolution, and combines this with the high lateral resolution and optical tomography capabilities of the confocal laser scanning microscope to achieve depth-sensitive detection of micron / submicron level defects on the CuCr contact surface. Furthermore, based on the multimodal images acquired by this system, a three-dimensional point cloud model containing complete topological information and material characteristics is constructed after registration and fusion processing. On this basis, this embodiment introduces digital image correlation (DIC) technology to perform full-field strain analysis and deformation tracking of the three-dimensional point cloud data. This not only enables quantitative characterization of the geometric parameters (size, depth, distribution density, and gradient distribution) of defects such as pores, microcracks, and inclusions on the CuCr contact surface, but also allows for automatic identification and classification of defect types through feature matching and pattern recognition algorithms. This provides holographic and structured surface state information input for adaptive parameter decision-making in subsequent repair processes.

[0024] S102: Classify defects on the contact surface and conduct risk assessment based on multimodal images; In this step, this embodiment introduces an adaptive defect recognition and evaluation network based on attention mechanism and multi-scale feature fusion to perform real-time analysis and evaluation of the acquired multimodal images of the contact surface. This network includes a multi-scale feature extraction module, an attention-enhanced fusion module, a defect classification and regression module, and a risk assessment decision module. The multi-scale feature extraction module extracts local detail features and global contextual information of defects on the contact surface, achieving efficient capture and fusion of features across micrometer to macroscopic scales. This module includes an input layer, a parallel dual-branch structure, and an output layer. The input layer receives registered and fused white light interferometry and confocal laser scanning images, unifying the size to 512×512×3 (RGB analog channels). The dual-branch structure includes a local detail branch and a global context branch. The local detail branch uses a 3-layer stacked lightweight convolution group (Depthwise Separable Convolution + BatchNorm + ReLU) to progressively extract high-frequency features such as edges and textures of micrometer-level defects from the image. The global context branch uses dilated convolution combined with a lightweight self-attention mechanism to capture the overall pattern and regional correlation of defect distribution. The output of the dual-branch structure undergoes feature concatenation and dimensionality reduction in the fusion layer, and is then output by the output layer to obtain a multi-scale fused feature map. .

[0025] The attention-enhanced fusion module adaptively focuses on key areas of defects on the contact surface and suppresses background interference. This module includes a channel attention submodule and a spatial attention submodule to fuse feature maps at multiple scales. Enhancement is achieved through several methods. The channel attention submodule employs a squeeze-and-excitation mechanism to adaptively weight feature channels, thereby highlighting the expression of defect-related features. The spatial attention submodule dynamically focuses on multi-scale fused feature maps using spatial pyramid pooling and deformable convolution. The defect spatial region is identified and background interference is suppressed. Furthermore, the local and global features output from the two channels are adaptively concatenated to generate an enhanced feature map. .

[0026] The defect classification and regression module is used to simultaneously perform contact surface defect type identification and geometric parameter regression, outputting multi-dimensional quantitative information including defect category, location, contour, and size / depth. This module enhances feature maps. As input, classification and regression tasks are executed in parallel, specifically including a classification head and a regression head. The classification head uses a lightweight fully connected layer and Softmax to output the probability of the defect category (porosity / microcrack / inclusion / no defect). The regression head, through coordinate regression and mask prediction branches, simultaneously outputs the bounding box position of the defect, the pixel-level contour mask, and calculates its geometric parameters (depth, shape factor, edge sharpness, distribution density, etc.).

[0027] The risk assessment and decision-making module combines defect geometric features with process rules to perform differentiable reasoning and classify various defects into risk levels. This module includes a feature encoding layer, a rule embedding network, and a risk assessment layer. The feature encoding layer encodes the geometric parameters output by the regression head and the confidence scores output by the classification head into high-dimensional structured feature vectors. The rule embedding network introduces a differentiable rule reasoning layer to transform prior process knowledge (e.g., "depth > 8μm → high risk") into learnable weight mappings for knowledge-guided decision-making. The risk assessment layer consists of two stacked fully connected layers and a sigmoid activation function. It outputs a risk score (0-1) for each defect based on the structured feature vector, classifying them into three levels: low risk (e.g., 0-0.3), medium risk (e.g., 0.3-0.7), and high risk (e.g., 0.7-1). It also dynamically calculates the key impact weights of each defect on subsequent laser repair processes, providing a quantitative basis for intelligent planning of laser parameters.

[0028] S103: Adaptively generate laser parameters based on risk assessment results; In this step, this embodiment constructs a laser parameter decision model based on the three-dimensional features of the defect and the risk assessment results. The decision model adopts a three-level linkage mechanism of feature encoding, rule embedding and parameter generation. Specifically, it includes: a feature encoding layer, which is used to receive the three-dimensional geometric features of the defect (size, depth, distribution density, shape factor, etc.) and risk rating results obtained from the preprocessing stage, and encodes them into high-dimensional structured feature vectors through a fully connected network and an embedding layer. The rule embedding and reasoning layer has a built-in differentiable rule reasoning module that transforms prior process knowledge (e.g., "power should be increased if depth > 8μm" and "spot spacing should be reduced in dense areas") into learnable logical mappings to achieve knowledge-guided parameter adjustment. The parameter generation and optimization layer, based on encoded features and rule constraints, generates initial laser parameters (power, speed, spot shape, path, etc.) by mapping through neural networks (e.g., including an input layer, hidden layers (composed of 3 to 4 fully connected layers), an attention control submodule (introducing a lightweight self-attention mechanism to dynamically weight key process dimensions in the input features), and multi-task output heads (e.g., including laser power output head, scanning speed output head, spot shape parameter output head, and path planning parameter output head)). Iterative optimization is performed by coupling a physics-based molten pool simulation proxy model, and finally outputs a combination of laser parameters that meets the repair quality thresholds (e.g., molten pool density, upper limit of thermal stress).

[0029] Furthermore, the model dynamically generates laser power, scanning path, and energy density distribution based on power dynamic mapping rules and adaptive planning of the beam path. The power dynamic mapping rules are expressed as follows: Based on defect depth (Unit: μm) and laser power A nonlinear mapping relationship is established between (unit: W) as shown below:

[0030] The adaptive planning of the light spot path is represented as follows: Based on the defect distribution density and shape, a space-filling curve (such as a Hilbert curve) is used to optimize the laser scanning path, reducing idle travel and improving repair efficiency and uniformity. For densely populated defect areas, the spot spacing is automatically reduced to 0.03~0.05mm to achieve complete coverage.

[0031] S104: Perform digital twin simulation of the laser repair process.

[0032] In this step, before performing laser repair, this embodiment uses a high-fidelity finite element thermal-fluid coupling model to perform a full-process digital twin simulation of the laser parameters to be used. This model takes the pre-processed three-dimensional surface defect geometry as input, coupling the laser heat source, material thermal properties, and dynamic boundary conditions. By solving the energy, momentum, and phase transition control equations, it accurately simulates the morphological evolution of the molten pool, temperature field distribution, melt flow behavior, and solidification structure evolution under laser irradiation, and predicts potential secondary defects such as porosity and hot cracks. The simulation results are compared in real time with preset repair quality thresholds (such as molten pool density, thermal stress limit, defect tolerance, etc.). Based on sensitivity analysis or a surrogate model, the laser power, scanning path, and spot parameters are iteratively optimized until all predicted indicators meet the process requirements. Finally, a closed-loop optimized set of laser parameters is output, achieving predictable control of the repair process and ensuring one-time forming quality.

[0033] In another exemplary embodiment, step S200, the laser remelting of the pretreated CuCr contact, includes the following steps: S201: Constructing a laser remelting environment; In this step, the CuCr contact to be repaired is placed in a fully enclosed laser remelting process chamber. A dynamic atmosphere control system is activated, introducing a mixed atmosphere of ultra-high purity argon (purity ≥99.999%) and trace amounts of reducing gases (such as hydrogen with a volume fraction of 0.5%–1.0%) into the chamber to suppress the thermodynamic stability of oxides on the CuCr contact surface. Furthermore, the chamber employs a dual-layer airflow design: the outer layer maintains a positive pressure of 0.12–0.15 MPa argon gas, while the inner layer is a laminar flow purification zone. Ultrasonic atomization of nano-level passivating agents (such as calcium fluoride suspension) forms a temporary protective film on the substrate surface. Simultaneously, an oxygen content detector and mass spectrometer are used to monitor multiple impurities such as oxygen, water, and nitrogen in real time. When the oxygen content falls below 200 ppm, a plasma-assisted purification module is activated, further adsorbing residual oxygen-containing impurities through glow discharge. Ultimately, the oxygen content of the process atmosphere is stably controlled below 10 ppm, and the water content is reduced to below 5 ppm, thereby achieving an ultra-clean remelting environment.

[0034] The above laser remelting process is carried out in a fully enclosed chamber. By using a mixed atmosphere of high-purity argon and trace amounts of reducing gas, and by combining double-layer airflow and plasma-assisted purification technologies, the oxygen and water content is stably controlled below 10 ppm and 5 ppm, respectively. This can fundamentally suppress the formation of secondary defects such as oxidation and gas absorption, thereby avoiding the generation of secondary defects on the CuCr contact surface.

[0035] S202: Set the laser remelting parameters and perform self-repair on the CuCr contact surface in the constructed laser remelting environment based on the set parameters.

[0036] In this step, based on the three-dimensional features and risk assessment results of CuCr contact surface defects obtained in the preprocessing stage, this embodiment calls a trained adaptive laser process model to generate a multivariate synergistic combination of laser parameters. Furthermore, this embodiment employs a high-frequency pulse-continuous composite laser mode for self-repair of the contact surface. The pulse component operates with nanosecond-level pulse widths and kilohertz-level repetition frequencies to achieve precise transient heating and fine-tuning of thermal stress in micron-level defect areas, suppressing heat-affected zone expansion and abnormal grain growth. The continuous component provides stable energy input during pulse intervals, maintaining dynamic equilibrium of the molten pool, promoting uniform diffusion and convection mixing of components within the melt, and preventing elemental segregation and uneven microstructure. The laser power is gradient-adjusted according to the depth distribution, type, and risk level of the defects, with a range of 200–480W. Power density spatial modulation technology is introduced to achieve localized focusing and enhanced energy distribution in areas with dense defects or greater depth, improving defect filling efficiency and metallurgical bonding quality. Furthermore, the laser scanning speed and spot morphology are dynamically matched and adaptively adjusted based on defect characteristics. For example, in microcrack regions, a low-speed scan (e.g., 50–100 mm / s) combined with an elliptical spot (e.g., 0.2 mm major axis, 0.1 mm minor axis) is used. The anisotropy of the spot shape enhances the heat input to the crack sidewalls, promoting the wetting and metallurgical bonding of the molten metal to the crack walls. In pore regions, a high-speed scan (e.g., 150–300 mm / s) combined with a circular spot, along with pulse interval control, is used to induce controlled oscillations in the molten pool, accelerating gas escape from the pores and achieving complete closure and dense filling of the pores. In addition, the spot scanning path is planned based on an improved space-filling algorithm (e.g., adaptive Hilbert curve) to ensure that the overlap rate between adjacent scanning trajectories is controlled between 30% and 50%, avoiding the generation of unfused and overheated zones.

[0037] Under laser irradiation, controlled melting occurs on the surface of the CuCr contact within a depth of approximately 50–100 μm. Through the synergistic effect of Marangoni convection and gas / liquid interfacial tension, the molten metal drives the complete filling of micropores and the metallurgical closure of microcracks. This also promotes the uniform dispersion of the chromium-rich phase within the molten pool through forced convection. After melting, relying on the high thermal conductivity of the CuCr alloy itself, and with the directional heat dissipation provided by the bottom active temperature control substrate (set temperature 80–120°C), the molten pool achieves rapid directional solidification. This ultimately forms a dense remelted layer on the CuCr contact surface with uniformly refined grain sizes of 3–8 μm and free of oxide inclusions.

[0038] It should be noted that the entire laser remelting process is carried out under a fully enclosed ultra-low oxygen dynamic atmosphere. The oxygen content in the process chamber is monitored in real time and stably controlled below 10 ppm, and the water content is below 5 ppm, in order to prevent secondary oxidation of the contact surface and the inhalation of ambient gases during the repair process, and to ensure the purity of the chemical composition of the repair layer and the quality of the interface metallurgy.

[0039] In another exemplary embodiment, step S300, which involves post-processing optimization of the laser-remelted CuCr contact to obtain a CuCr contact with self-healing surface defects, includes the following steps: S301: Set magnetic abrasion parameters; In this step, the laser-remelted CuCr contact is transferred to a closed-loop controlled six-axis linkage magnetic polishing machine. This system integrates an in-situ surface morphology monitoring and dynamic parameter feedback module. In the polishing machine, the polishing media adopts a composite functional design, with a multi-level gradient structure of stainless steel-based magnetically responsive needle-shaped abrasive as the core. The surface is coated with a composite layer of nano-diamond and cerium oxide through chemical plating, forming a "hard-soft" synergistic polishing layer. The polishing media with this structure, through the synergistic effect of "hard-soft," can achieve efficient and precise polishing of the CuCr contact surface. Specifically, the multi-level gradient structure of the stainless steel-based magnetically responsive needle-shaped abrasive can achieve multi-dimensional composite motion under a time-varying magnetic field, adapting to surface undulations to maintain uniform contact pressure, and precisely cutting micron / submicron-level protrusions through its sharp tip. The surface chemically plated nanodiamond and cerium oxide composite coating further enhances the synergistic polishing effect: nanodiamonds, with their ultra-high hardness, achieve mechanical micro-cutting, efficiently removing microburrs and resolidified layers remaining in the remelted layer; cerium oxide particles, through chemical mechanical polishing (CMP), selectively soften and dissolve the surface material under frictional heat and surface activity, thereby achieving atomic-level material removal under low pressure. This "hard-soft" synergistic grinding layer not only significantly improves polishing efficiency but also effectively inhibits surface scratches, microcrack propagation, and subsurface damage during the polishing process, ultimately obtaining a high-quality surface with high smoothness, intact structure, and no residual stress concentration.

[0040] In addition, the polishing slurry uses a nanofluid with pH buffering and passivation functions. It is made by in-situ mixing of deionized water, ionic liquid (such as 1-ethyl-3-methylimidazolium tetrafluoroborate), and α-alumina microparticles of different sizes (0.1μm, 0.3μm, 0.5μm) in a smart ratio. This fluid can release corrosion-inhibiting ions in real time during the polishing process, thereby inhibiting the surface activity of the contact.

[0041] Furthermore, the grinding parameters are adaptively set based on the three-dimensional morphology scanning results of the CuCr contact surface. For example, the magnetic field strength adopts a gradient control mode (dynamically adjustable from 1200 to 1800 Gs), the grinding speed is combined with the eddy current damping effect for frequency conversion control (dynamically optimized in the range of 200 to 400 r / min), and the grinding time is determined by self-termination based on real-time roughness feedback. The overall process window is pre-verified through a digital twin model to ensure optimal matching between material removal rate and surface integrity.

[0042] S302: Polish the contact surface based on the set magnetic abrasive parameters.

[0043] In this step, after the magnetic polishing system is activated, functionalized needle-shaped abrasives, driven by a three-dimensional time-varying magnetic field, perform multi-scale synergistic polishing of the contact remelted layer surface through multi-mode composite motion (including precession, nutation, and axial vibration). The system integrates a white light interferometry online monitoring module to acquire real-time data on surface roughness (Sa), profile peak-valley height, and subsurface damage layer depth. It also dynamically adjusts the magnetic field distribution and abrasive trajectory using machine learning algorithms to achieve "shape-preserving" polishing. During polishing, nanofluids form a molecular-level lubricating film under the influence of frictional heat and an electric field, while simultaneously inhibiting micro-area corrosion through an electrochemical passivation mechanism. The nanodiamond and cerium oxide components on the abrasive surface achieve synergistic effects of mechanical micro-cutting and chemical mechanical polishing (CMP), efficiently removing micron / submicron-level burrs and resolidified layers remaining after remelting. When the system detects that the surface roughness is stable at Sa≤0.1μm and the subsurface grain structure is undamaged (confirmed by online laser ultrasonic testing), the polishing process is automatically terminated, ultimately obtaining a high-performance CuCr contact surface with a smooth surface, complete grain structure, and no residual stress concentration, meeting the service requirements under high voltage and high current conditions.

[0044] Figure 2 The macroscopic morphology of the CuCr contact surface with surface defects is presented, consisting of... Figure 2 As can be seen, no obvious macroscopic damage or large-area breakage was observed on the contact surface. However, several scattered micro-pits were uniformly distributed on the flat substrate, presenting a localized shallow pit morphology. In addition, slight, thin, linear marks were also visible on the contact surface. These marks were mostly micro-scratches or shallow textures, possibly related to slight mechanical actions during material preparation or processing. Although the defect size was small and the distribution was relatively sparse, such micro-geometric discontinuities could still affect the electric field distribution, contact performance, and long-term service stability of the contact surface.

[0045] Figure 3 This study demonstrates the microscopic structural features of CuCr contacts with surface defects. From... Figure 3Numerous tiny pores, approximately 1–3 μm in diameter, are clearly visible on the surface. These pores are circular or elliptical in shape, and some are interconnected. Oxidation impurities adhere to the pore walls, indicating that surface oxidation may have occurred during processing or storage. Simultaneously, microcracks of 20–50 μm in length and 3–8 μm in depth are present on the contact surface. The crack edges exhibit a rough morphology and are accompanied by minute inclusions, further exacerbating localized stress concentration and structural discontinuities. Furthermore, chromium-rich regions are unevenly distributed in an "island-like" pattern within the matrix, with minute interfacial gaps between them and the surrounding Cu matrix. This results in a loose and uneven overall surface microstructure. These microscopic defects collectively affect the material's density, mechanical properties, and electrical contact reliability.

[0046] Figure 4 This study showcases the macroscopic and microscopic structural features of a CuCr contact surface after a single mechanical grinding and sand oxidation treatment. Macroscopic observation reveals that previously visible fine scratches and linear marks have been largely eliminated, resulting in a relatively smooth surface. However, a small number of incompletely removed micropores remain identifiable. Further microscopic analysis reveals that these pores are diffusely distributed on the surface, with pore sizes remaining at the micrometer level. Oxidation impurities still adhere to the pore walls, indicating that simple grinding and sand oxidation failed to completely fill and purify the defects. Furthermore, microscopic gaps remain at the interface between the chromium-rich phase and the matrix, with a loose local microstructure exhibiting a degree of compositional and structural inhomogeneity. This suggests that while this treatment method improves surface morphology, its effectiveness in repairing deep microscopic defects is limited.

[0047] Figure 5 The macroscopic morphology of the CuCr contact surface repaired using the method described in this application is shown. From Figure 5 As can be seen, the micro-dimples and fine linear marks present before repair have been completely eliminated, and no new defects have been generated on the contact surface, exhibiting an overall flat and smooth morphology. After repair, the contact surface is uniform and consistent, without obvious undulations or local abnormalities, while maintaining the same dimensional accuracy as before repair. This indicates that the method described in this application can effectively eliminate micro-defects on the surface while precisely controlling the material removal and surface forming process, achieving good preservation of the geometric integrity of the workpiece.

[0048] Figure 6 The microstructure characteristics of the CuCr contact surface repaired using the method described in this application are shown. From Figure 6As can be seen, the original fine pores with a diameter of 1–3 μm have been completely and densely filled, and the microcracks with a length of 20–50 μm and a depth of 3–8 μm have also been completely closed. There are no obvious pores or cracks remaining on the surface and near the surface. The chromium-rich region is uniformly and diffusely distributed in the matrix, forming a tightly fused interface with the copper matrix, without visible gaps or inclusions. The repaired surface layer forms a significantly refined equiaxed grain structure, with the grain size uniformly refined to 3–8 μm. The overall structure is dense and uniform, without oxide impurities or loose areas, indicating that the method described in this application can effectively promote the homogenization of the structure and the refinement of the grains while achieving defect self-repair, thereby improving the integrity of the material surface and its service performance.

[0049] The above embodiments are only for illustrating the technical concept and features of this application, and are intended to enable those skilled in the art to understand the content of this application and implement it accordingly. They should not be construed as limiting the scope of protection of this application. All equivalent changes or modifications made in accordance with the spirit and essence of this application should be included within the scope of protection of this application.

Claims

1. A self-healing method for micro-defects on the surface of a CuCr contact, characterized in that, The method includes: Pretreatment of CuCr contacts; The pretreated CuCr contacts are then laser remelted. Post-processing optimization was performed on the CuCr contacts after laser remelting to obtain CuCr contacts with self-healing surface defects.

2. The method according to claim 1, characterized in that, The pretreatment of CuCr contacts includes: The CuCr contact was scanned to obtain multimodal images of the CuCr contact surface; Defect classification and risk assessment of CuCr contact surface based on multimodal images; Laser parameters are adaptively generated based on risk assessment results; Digital twin simulation of the laser repair process.

3. The method according to claim 2, characterized in that, The defect classification and risk assessment of CuCr contact surfaces based on multimodal images includes: An adaptive defect identification and evaluation network is used to classify defects and assess risks in multimodal images of CuCr contact surfaces.

4. The method according to claim 3, characterized in that, The adaptive defect identification and evaluation network includes: The module comprises a multi-scale feature extraction module, an attention-enhanced fusion module, a defect classification and regression module, and a risk assessment and decision-making module. The multi-scale feature extraction module is used to extract local detail features and global contextual information of defects on the contact surface, achieving efficient capture and fusion of features across scales from micrometers to macrometers; The attention-enhanced fusion module is used to adaptively focus on critical areas of contact surface defects and suppress background interference; The defect classification and regression module is used to simultaneously perform contact surface defect type identification and geometric parameter regression, and output multi-dimensional quantitative information including defect category, location, contour, and size and depth. The risk assessment and decision-making module is used to perform differentiable reasoning by combining the geometric characteristics of defects with process rules, and to classify the risk levels of various defects.

5. The method according to claim 2, characterized in that, The adaptive generation of laser parameters based on risk assessment results includes: Laser parameters are adaptively generated by constructing a laser parameter decision model, which includes: The feature encoding layer receives the three-dimensional geometric features of defects and risk rating results obtained from the preprocessing stage and encodes them into high-dimensional structured feature vectors. The rule embedding and reasoning layer is used to transform prior knowledge of the process into a learnable logical mapping, enabling knowledge-guided parameter adjustment. The parameter generation and optimization layer generates initial laser parameters through neural network mapping based on encoded features and rule constraints. It then couples a physics-based molten pool simulation proxy model for iterative optimization, outputting a combination of laser parameters that meets the repair quality threshold.

6. The method according to claim 1, characterized in that, The laser remelting of the pretreated CuCr contacts includes: Constructing a laser remelting environment; Set the laser remelting parameters and perform self-repair on the CuCr contact surface in the constructed laser remelting environment based on the set parameters.

7. The method according to claim 6, characterized in that, The construction of the laser remelting environment includes: The CuCr contact is placed in a fully enclosed laser remelting process chamber; A mixed atmosphere of ultra-high purity argon gas and trace amounts of reducing gas is introduced into the chamber; A dual-layer airflow field design is adopted inside and outside the cavity.

8. The method according to claim 6, characterized in that, The self-healing of the CuCr contact surface in the constructed laser remelting environment includes: A high-frequency pulse-continuous composite laser mode is used to perform self-repair on the contact surface.

9. The method according to claim 1, characterized in that, The post-processing optimization of the laser-remelted CuCr contact to obtain a self-healing CuCr contact with surface defects includes: Set the magnetic abrasive parameters; The contact surface is polished based on the set magnetic abrasive parameters.

10. The method according to claim 9, characterized in that, The setting of magnetic abrasive parameters includes: Adaptive settings were implemented based on the three-dimensional morphology scanning results of the CuCr contact surface, wherein... The magnetic field strength is controlled using a gradient modulation mode; The grinding speed is controlled by frequency conversion using the eddy current damping effect. The grinding time is automatically terminated based on real-time roughness feedback.

Images