Metal sheet blanking surface quality control method

Abstract

The invention relates to a metal sheet blanking surface quality control method which comprises the following steps: applying micro-plastic excitation to a metal sheet before blanking to obtain an instantaneous recovery parameter of a material under low-energy deformation so as to judge a shear response state of the metal sheet; judging the deformation stability of the material in the current stress field according to the shear response state, and classifying the characteristics as shear offset or strain concentration characteristics; aiming at a shearing offset type material, a sliding shear path of low-speed feeding and gap reduction is adopted; aiming at a strain concentration type material, adopting a splitting path of high-speed rushing-in and strengthening front angle fracture; monitoring derivative changes of blanking force and a displacement curve in the blanking process, identifying a fracture critical stage, dynamically adjusting the feeding speed of a punch, and regulating and controlling the fracture morphology through acceleration or deceleration; and the edge springback behavior of the punch in the return stroke stage is detected at the blanking tail end, if a warping trend exists, the tail end calendaring correction action is activated, and secondary shaping of the edge area is achieved by combining an obtuse-angle pressing head or a conical structure of the punch with material springback kinetic energy.

Description

Technical Field

[0001] This invention relates to the field of metal plastic processing and intelligent manufacturing technology, specifically to a method for controlling the surface quality of stamped metal sheets. Background Technology

[0002] Existing technologies, such as the Chinese patent document CN111318605A (Precision Blanking Press and Method for Operating a Precision Blanking Press), primarily improve the stability of the blanking force by implementing closed-loop force control between the two drive units. The force control unit adjusts the reaction force based on sensor feedback to improve blanking accuracy and repeatability. However, while this structure improves the overall motion coordination of the press to some extent, it still has significant shortcomings and limitations in controlling the surface quality of the blanked metal sheet. First, this patent focuses on mechanical feedback control at the mechanical system level, with its core objective being to balance the driving force between the upper and lower press units, rather than directly addressing the microscopic shearing behavior or cross-sectional morphology changes in the blanking area. Its force feedback signal mainly originates from macroscopic load changes and cannot identify the strain response characteristics of the material in the microplastic stage. Therefore, it cannot predict the shear stability and fracture sensitivity of the material in the early stages of blanking, nor can it achieve adaptive path adjustment for different microstructures (such as bainite, pearlite, and coarse-grained regions).

[0003] Secondly, although this scheme achieves closed-loop force control, its control response cycle depends on the sampling rate and computational delay between the mechanical sensor and the controller, generally in the millisecond range. This is still considered lagging control for the microsecond-level stress concentration and crack initiation stages during the punching process. Especially in the punching process of high-strength steel or multilayer composite materials, the fracture behavior exhibits transient nonlinear evolution. This control method cannot achieve real-time adjustment of speed or gap within the critical fracture window, resulting in unstable proportions of bright bands on the cross-section and difficulty in suppressing tear edges. Thirdly, this patent does not consider the influence of differences in the microstructure and anisotropy of the metal sheet before punching on surface quality. Although the force control unit can monitor overall load changes through sensors, it cannot obtain the microscopic response parameters of the material, such as the elastic recovery time constant, displacement hysteresis, and microcrack signs. Therefore, a uniform force control strategy is still used under different materials or different rolling directions, lacking a targeted variable parameter adjustment mechanism. This easily leads to uneven stress distribution in the shear zone, rough fracture surfaces, thickened burrs, asymmetrical collapse angles, and other quality problems. Fourth, this device relies solely on the kinematic system of the press for process control, lacking multi-channel signal fusion and intelligent recognition capabilities. In actual blanking processes, factors affecting cross-sectional quality include multi-dimensional parameters such as punch feed speed, die clearance, lubrication status, and material microstructure. This solution, however, uses only a single mechanical closed-loop control, failing to establish a dynamic coupling judgment logic of force-displacement-fracture morphology, and thus cannot actively correct edge springback or fracture warping at the blanking end. Compared to modern intelligent control approaches, this technology lacks a microscopic level of dynamic sensing and does not incorporate a material property feedback model.

[0004] Fifth, the sensor feedback accuracy of the force control unit depends on the mechanical stiffness and force transmission path. When the punching load fluctuates at high frequency, the sensor signal suffers from mechanical hysteresis and noise interference, causing the controller's compensation signal to deviate from the actual requirements, resulting in secondary force oscillations. These oscillations of the reaction force can cause stress disturbances within the shear band, leading to microcrack propagation and roughness extension on the fracture surface. Furthermore, this scheme does not process or filter the derivative characteristics of the force and displacement signals during punching, making it unable to identify the force abrupt change inflection point at the critical fracture stage. Therefore, it often lacks accurate judgment and dynamic speed adjustment mechanisms during the fracture transition stage. Sixth, this method only considers the equipment motion perspective and does not propose any pre-judgment or excitation testing of the material before punching, nor does it establish corresponding process windows for different material properties. In actual production, the strain rate sensitivity and thermal response of materials such as steel, aluminum alloys, and magnesium alloys differ significantly. Using a unified closed-loop force control algorithm can easily lead to overcompensation or undercompensation, resulting in tearing or incomplete separation of the fracture layer. Furthermore, the system does not analyze or optimize the formation mechanism of surface defects such as burrs, collapsed corners, and microcracks after punching, and lacks a self-learning mechanism to feed back the punching results to the control parameters.

[0005] In summary, while existing technologies have improved the coordinated control performance of presses at the mechanical level, they remain at the level of macroscopic load feedback and drive compensation. They lack a closed-loop control logic for the blanking surface quality based on the microscopic response characteristics of materials, and lack real-time dynamic adjustment mechanisms and adaptive organizational capabilities. Therefore, in practical applications, it remains difficult to achieve systematic control over the proportion of bright bands, burr height, fracture smoothness, and edge warping behavior of the blanking cross-section, limiting the promotion and performance optimization of blanking processes in precision forming fields such as high-strength steel and heterogeneous composite materials. Summary of the Invention

[0006] The purpose of this invention is to provide a method for controlling the surface quality of stamped metal sheets, thereby addressing some of the drawbacks and shortcomings pointed out in the background art.

[0007] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: a method for controlling the surface quality of metal sheet punching, the main structure of which includes: applying micro-plastic excitation to the metal sheet before punching, including small-amplitude knocking or pre-stretching disturbance, to obtain the instantaneous recovery parameters of the material under low-energy deformation, the parameters including deformation recovery rate, displacement hysteresis or microcrack signs, to determine the shear response state of the metal sheet.

[0008] The deformation stability of the material under the current stress field is determined based on the shear response state, and it is classified as shear offset type or strain concentration type. In the punching control step, for shear offset type materials, a low-speed feed and a sliding shear path with reduced gap are adopted; for strain concentration materials, a high-speed punching and a splitting path with enhanced front angle fracture are adopted.

[0009] During the punching process, the derivative changes of the punching force and displacement curves are monitored to identify the critical fracture stage and dynamically adjust the punch feed speed. The fracture morphology is controlled by accelerating or decelerating. At the end of the punching process, the edge springback behavior during the punch return stage is detected. If there is a warping trend, the end rolling correction action is activated. The obtuse angle punch head or conical structure is combined with the material springback kinetic energy to achieve secondary shaping of the edge area.

[0010] Furthermore, the pre-stretching perturbation includes intermittent stretching along multiple directions to identify differences in shear response due to material anisotropy; the acquisition of the instantaneous recovery parameters includes synchronously recording the elastic recovery path before and after the perturbation using a laser displacement sensor; the judgment of the shear response state adopts a multi-parameter weighted decision criterion, including the combined ratio of deformation rate and microcrack indications.

[0011] Furthermore, the intermittent stretching in multiple directions includes tensile excitation along at least two principal axis directions, including the material rolling direction and the perpendicular direction, to identify directional shear-sensitive areas; wherein the time interval and amplitude of the tensile disturbance increase in a stepwise manner to observe the anisotropic response variation law of the material under different stress levels; the laser displacement sensor is deployed on the upper and lower surfaces of the material to achieve synchronous monitoring of in-plane and out-of-plane elastic recovery paths.

[0012] Furthermore, the elastic recovery path analysis includes evaluating the maximum displacement hysteresis, recovery time constant, and stress unloading slope change; the evaluation of microcrack indications in the multi-parameter weighted decision criterion is performed by a material surface micro-strain imaging system, combined with grayscale perturbation to identify crack initiation locations; the shear response capability is quantified using an integral nonlinear function, with the function form as follows:

[0013]

[0014] in, It is a shear response capability index used to characterize the stress release and crack propagation characteristics of materials under perturbation conditions; These are the start and end times of the disturbance loading and unloading processes, respectively; It is a displacement hysteresis function per unit time, reflecting the degree of hysteresis in the excitation response; The micro-strain perturbation intensity coefficient is obtained from the micro-strain imaging system; Elastic recovery time constant, which describes the stress recovery rate; This is the change in stress unloading slope, used to characterize the elastic release rate during the unloading phase; The function representing the change in microcrack density per unit area indicates the crack initiation rate. It is the surface roughness factor of the material, used to balance the effect of microcrack propagation on surface energy; It is an inverse hyperbolic cosine function, used to amplify and correct the nonlinear sensitivity of crack propagation.

[0015] When the calculated shear response capacity index Greater than the set threshold When the material is determined to be shear offset type, it is adapted to the sliding shear path control mode; when Less than the threshold When the strain is concentrated, it is adapted to the splitting path control mode, thereby realizing the adaptive selection of the punching path and the optimized control of surface quality.

[0016] Furthermore, the shear offset material includes steel types with bainitic phase structure or coarse grain characteristics; the reduction in die gap in the sliding shear path is dynamically adjusted during the punching process to match the change in shear band width, wherein the high-speed punching in the splitting path adopts nonlinear speed segmented control to form a stress concentration induction band; wherein the monitoring of the derivative of the punching force and displacement curve uses a dual-channel dynamic sampling system for filtering; the identification of the fracture critical stage is based on the combined judgment that the shear force change rate exceeds a first threshold and the punching speed fluctuation exceeds a second threshold.

[0017] Furthermore, the identification of the bainitic phase structure is determined by the response behavior of thermoelectric properties or hysteresis curves; the determination of coarse grain characteristics is based on micron-level scanning electron microscopy and in-plane structural homogeneity analysis; the nonlinear velocity segmented control of the splitting path includes three stages: a pre-acceleration stage, a stress-locking stage, and a fracture-progression stage.

[0018] Furthermore, the determination of the thermoelectric properties is based on applying pulsed thermal disturbance to the material before punching and simultaneously measuring the voltage response change rate to distinguish the structural differences between bainite and pearlite; the determination of the response behavior of the hysteresis curve is obtained by a Hall sensor array under low-frequency magnetic field scanning to detect the hysteresis loop area and coercivity characteristics of the material.

[0019] Furthermore, the thermoelectric response signal obtained during the bainitic phase identification process is combined with the shape of the hysteresis curve for feature matching; while the field resolution of the scanning electron microscopy imaging in the grain coarseness feature determination is not less than 0.3~0.5µm, and the structural homogeneity is analyzed in combination with the root mean square error of the image grayscale distribution.

[0020] Furthermore, the in-plane structural homogeneity analysis includes extracting the grain boundary direction distribution density index based on two-dimensional Fourier transform to quantify the uniformity of coarse grain arrangement; after identifying coarse grains, the die gap control strategy is modified to compensate for the risk of irregular fracture zones; the control speed of the pre-acceleration section is 20% to 40% of the punching target speed, which is used to induce the initial plastic stress of the material.

[0021] Furthermore, the control strategy of the stress-locking section is to maintain a constant speed, while the system monitors the shear force fluctuation in real time to determine whether it has entered the critical fracture range; and the acceleration of the punch in the fracture advance section is not less than 10~15m / s², so that the shear zone passes through the fracture stage.

[0022] The beneficial effects of this invention are as follows: The method for controlling the surface quality of stamped metal sheets provided by this invention achieves accurate identification of the anisotropic shear response of the metal sheet by applying micro-plastic excitation to the material before stamping and combining it with multi-directional intermittent tensile perturbation. A laser displacement sensor is used to synchronously monitor the elastic recovery path before and after the perturbation. Combined with characteristic parameters such as maximum displacement hysteresis, recovery time constant, and stress unloading slope, the deformation coordination and crack initiation trend under the material's microstructure can be effectively perceived. Furthermore, the detection accuracy of microcrack signs is further improved through a micro-strain imaging system and grayscale perturbation feature recognition technology. A shear response capability index is constructed using the combined ratio of deformation rate and microcrack density change rate, enabling adaptive control of slip shear or splitting along the stamping path.

[0023] This method introduces a high-speed / low-speed dynamic path switching mechanism during the blanking process and combines it with a nonlinear speed segmented control strategy to adjust the splitting path in stages. It can adjust the die clearance and punch feed strategy in real time based on the bainitic structure or coarse grain characteristics of the material. A dual-channel filtering monitoring system identifies the critical fracture stage, and then dynamically adjusts the punch speed to optimize the fracture morphology. Simultaneously, a springback correction action is introduced at the end of the blanking process to actively shape any potential edge warping. In summary, this method significantly improves the proportion of bright bands in the blanked section, reduces burr height and edge deformation, and possesses excellent adaptability and engineering applicability. Attached Figure Description

[0024] Figure 1 This is a flowchart of the metal sheet punching surface quality control method of the present invention.

[0025] Figure 2 This is a diagram showing the relationship between the multi-parameter response determination and path adaptive control function before punching in this invention.

[0026] Figure 3 This is a flowchart of the metal sheet material parting and path adaptive punching control of the present invention.

[0027] Figure 4 This is a flowchart of multi-directional disturbance and path adaptive control before punching in Embodiment 1 of the present invention.

[0028] Figure 5 This is a flowchart of the multi-type steel sheet punching and parting identification and adaptive path control in Embodiment 2 of the present invention. Detailed Implementation

[0029] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0030] Combined with appendix Figure 1The present invention discloses a method for controlling the surface quality of metal sheet punching. Before the punching begins, a micro-plastic excitation is applied to the metal sheet to be processed. The micro-plastic excitation includes a small-amplitude tapping operation controlled by energy input or a controlled pre-stretching disturbance. The tapping can be applied to the material surface by a low-energy mechanical pulse, while the pre-stretching is achieved by applying a short-term reversible tensile force to the metal sheet at a preset displacement rate using an axial loading system. This low-energy disturbance method activates the microstructure response of the material surface. Subsequently, immediately after the excitation is completed, the data acquisition module is activated to obtain the instantaneous recovery parameters of the material under low-energy deformation. These parameters cover, but are not limited to, three key behavioral characteristics: first, deformation recovery rate, which characterizes the material's rebound speed and deformation repair capability after the removal of external force; second, displacement hysteresis, which describes the degree of path asymmetry during loading and unloading and indirectly reflects internal friction characteristics; and third, microcrack indications, which are detected by a high-precision strain or surface disturbance imaging device to determine whether microcrack sources or stress concentration areas appear on the surface of the metal sheet. Real-time measurement and analysis of these three behavioral parameters yield the shear response state of the material under the current stress environment. The response state is divided into different deformation stability levels and serves as the basis for punching path planning and parameter setting, providing physical support and input for path decisions in the subsequent punching process.

[0031] Based on the shear response data obtained before punching, the deformation stability of the metal sheet material under the current stress field is further determined. This determination is based on a multi-dimensional parameter comprehensive evaluation, including multiple physical quantities such as deformation recovery capacity, microcrack initiation tendency, displacement hysteresis amplitude, and strain concentration performance. The system performs weighted analysis on the above parameters to determine whether the material has strong structural stability or is prone to deformation drift during shearing. If the analysis results show that the material has a directional shift trend, unstable displacement in the shear zone, or a tendency for early crack propagation during shearing, it is classified as a shear shift type material; if the material exhibits obvious strain concentration, a clear fracture initiation location, and strong energy concentration capability, it is determined to be a strain concentration characteristic material.

[0032] In the blanking control process, differentiated path control strategies are adopted based on the aforementioned classification results to adapt to the changes in cross-sectional quality caused by different shearing behaviors. For shear offset materials, the system selects the sliding shear path mode to control the blanking process. This path uses a low-speed feed method and dynamically reduces the die gap during blanking. By controlling the stable punch speed, smooth blade entry, and uniform stress in the shearing zone, the material forms a continuous and smooth shear cross-section at a relatively slow deformation rate, effectively avoiding tearing, corner collapse, and burr propagation. For strain-concentrated materials, a splitting path mode is used for control. This mode quickly establishes a stress concentration area through a high-speed punching strategy, combined with a reinforced front corner structure designed at the front of the punch. The front corner area forms a strong fracture induction zone during high-speed impact, which helps the fracture to advance rapidly along the predetermined path, achieving the control target of a short and clear shear cross-section, a high proportion of bright surface, and a regular fracture interface.

[0033] A real-time monitoring and dynamic control mechanism is introduced during the punching process. By continuously tracking the derivative changes of the punching force and displacement curves, the critical state of the material transitioning from the elastic deformation stage to the fracture stage is identified. This monitoring process is completed by a dual-channel dynamic sampling system. One channel collects the punching force signal, and the other channel simultaneously collects the punch displacement information. The sampling results are processed by derivative calculation and filtering to generate a force-displacement rate curve. The system determines that the material has entered the critical fracture stage based on the sudden drop in force, gradient reversal point, or abrupt change in curve slope. After identifying this stage, the control module immediately adjusts the punch feed rate. When the material exhibits a delayed fracture propagation or a bright band contraction trend, the punch speed is appropriately increased to promote stress concentration and rapid formation of the fracture interface. When excessively rapid fracture or local tearing tendency is detected, the feed rate is reduced to slow down the fracture process, thereby making the energy release in the shear zone more uniform and obtaining a smooth and flat cross-sectional structure.

[0034] At the end of the punching process, the system further monitors the return stroke of the punch, focusing on detecting the elastic recovery characteristics of the material edge at the moment the punch leaves the workpiece. When warping, rebound, or stress concentration tendencies are detected in the edge area, the end-of-pipe rolling correction action is immediately activated. This action is implemented through an obtuse-angled indenter or a rolling structure with a gradually tapered profile at the bottom of the punch. A slight back pressure is applied to the edge area at the initial stage of the punch return stroke. The material's own springback kinetic energy and the indenter's profile create a counteracting effect, allowing the edge area to undergo secondary plastic shaping during the springback process.

[0035] Combined with appendix Figure 2In the blanking preparation stage, a pre-stretching disturbance is applied to the metal sheet to be processed. The disturbance is an intermittent tensile loading method, in which small tensile displacements are applied sequentially along multiple principal axes, such as the rolling direction and the perpendicular direction, so that the metal sheet produces slight reversible deformation under different stress orientations. The intermittent loading is performed by alternating short-cycle force application and unloading, and the stretching amplitude is set according to the material thickness and yield strength, so that the differences in the orientation of the micro-grains inside the material can be revealed.

[0036] Immediately after the tensile perturbation is completed, a high-precision laser displacement sensor is used to obtain the elastic recovery path of the material before and after the perturbation. The sensor is deployed on the upper and lower surfaces of the metal sheet to simultaneously capture micro-displacement changes in the in-plane and out-of-plane directions, recording the displacement curves of the entire loading, unloading, and recovery stages. The system obtains the instantaneous recovery parameters of the material through data comparison, including the deformation recovery rate, hysteresis displacement, and springback characteristics during the stress unloading stage.

[0037] Based on the instantaneous recovery parameters, the system comprehensively judges the shear response state of the material. The judgment algorithm adopts a multi-parameter weighted decision criterion, using deformation rate and microcrack indications as the main judgment indicators. The coupling relationship between the two is expressed by calculating a joint ratio, where deformation rate reflects the plasticity and ductility of the material, and microcrack indications reflect the degree of brittle initiation of the structure. A higher joint ratio indicates a significant difference between the crack initiation rate and the plastic response under perturbation, indicating a shear shift tendency; a lower ratio indicates that the strain energy release is balanced during loading and recovery, indicating strain concentration. Through this weighted decision criterion, the system can dynamically classify and model the response of materials in different batches and with different microstructures, thereby selecting the optimal path strategy in subsequent punching control.

[0038] This method employs multi-principal intermittent tensile excitation to identify the directional shear characteristics of materials and achieve accurate stress response analysis. In the pre-punching stage, a dual-principal tensile excitation unit is set up along the rolling direction and perpendicular to the material to be processed, applying alternating loading to the material. The tension in each principal axis direction is applied in a low-amplitude, graded, incremental manner, with the loading duration and interval increasing sequentially according to a stepwise pattern. This allows the material to undergo micro-plastic deformation and elastic recovery processes under different stress levels. This loading method effectively induces differences in grain orientation, enabling the gradual identification of directional shear-sensitive regions at different stress stages, thereby obtaining the stability variation law of their anisotropic response.

[0039] Throughout the tensile excitation process, high-precision monitoring is achieved using laser displacement sensors. These sensors are deployed on both the upper and lower surfaces of the material to simultaneously acquire elastic recovery paths in both in-plane and out-of-plane directions. The system compares the displacement curves acquired by the sensors over time to obtain the micro-displacement characteristics of the material during loading and unloading, and then analyzes the consistency of strain recovery under stress in different directions. This method comprehensively reflects the stability distribution of the material's internal structure under multi-directional stress coupling, enabling quantitative identification of differences in directional shear response, and providing fundamental data support for intelligent matching of subsequent punching path parameters and cross-sectional quality control.

[0040] By performing multidimensional coupled analysis of the dynamic response parameters of the material's perturbation process, quantitative calculation of shear response capability and adaptive control of the punching path were achieved. During the micro-plastic excitation stage before punching, the system records the displacement and strain response curves of the material in real time during loading and unloading. Data is simultaneously acquired through a high-precision displacement sensor and a micro-strain imaging system to comprehensively analyze the elastic recovery path. The analysis includes three core indicators: maximum displacement hysteresis, recovery time constant, and stress unloading slope change. Maximum displacement hysteresis reflects the difference in deformation recovery after the removal of external force, reflecting its inherent damping characteristics; the recovery time constant characterizes the time required for the material to return from the perturbed state to a steady state, reflecting the stress relaxation rate; and the stress unloading slope change describes the stress release rate and energy dissipation characteristics of the material during the unloading stage.

[0041] To achieve synergistic analysis of material microcrack behavior and deformation response, the system extracts grayscale perturbation information of the crack initiation region based on a surface microstrain imaging device. By performing time-series analysis of the grayscale field distribution changes, the location and propagation trend of microcracks can be determined. The obtained data serves as input to a multi-parameter weighted decision criterion, forming a comprehensive evaluation system for the material's shear response together with the deformation rate parameter. To overcome the limitations of single-parameter linear models on complex behaviors, this invention proposes an integral nonlinear function for quantifying the shear response capability of materials, the expression of which is:

[0042]

[0043] in, It is a shear response capability index used to characterize the stress release and crack propagation characteristics of materials under perturbation. and These represent the start and end times of the disturbance loading and unloading processes, respectively. It is a displacement hysteresis function per unit time, used to reflect the degree of hysteresis in the excitation response; The micro-strain perturbation intensity coefficient is obtained by a micro-strain imaging system. is the elastic recovery time constant, describing the stress recovery rate of the material after unloading; This is the change in stress unloading slope, used to characterize the elastic release rate during the unloading phase; This is a function representing the change in microcrack density per unit area, used to indicate the crack initiation rate; It is the surface roughness factor of the material, used to balance the influence of surface energy during the microcrack propagation process; It is an inverse hyperbolic cosine function, used to amplify and correct the nonlinear sensitivity of crack propagation, enabling the model to capture the stress concentration effect before the fracture abrupt change.

[0044] The system performs real-time calculations on the function integral results, and if the shear response capability index... Greater than the set threshold This indicates that the material exhibits strong deformation drift and crack activity under disturbance conditions, classifying it as a shear-shifting material. In this case, the blanking control system selects the sliding shear path mode, using low-speed feed and reduced die clearance to ensure a continuous and smooth cross-section. Less than the threshold If the material exhibits concentrated strain and stable fracture behavior, it is identified as a strain-concentrated material. The system then switches to the splitting path mode, using high-speed penetration and enhanced front angle fracture to form a high-quality fracture interface.

[0045] function The derivation process includes: the elastic deformation hysteresis behavior of the metal sheet under slight perturbation, and the displacement response hysteresis. This parameter describes the response delay between the perturbation input and the structural feedback; however, this parameter itself is difficult to couple directly with the microcrack behavior, therefore a coefficient characterizing the energy density of the micro-strain response is introduced. , representing the local tensile response strength of a material under unit disturbance.

[0046] To address the comparability of elastic recovery rates between different materials, the denominator of the formula is designed as follows: The elastic recovery time constant is about to be Change in stress unloading slope A linear combination, as a comprehensive resistance to the overall stress release rate of the material, has a physical meaning equivalent to the recovery difficulty.

[0047] To further characterize the potential abrupt change behavior of crack initiation in materials, we introduce... This represents the microcrack density variation function per unit area, i.e., the crack propagation rate over time. Considering that crack propagation typically exhibits a nonlinear or even abrupt change, this term is expressed as... The transformation process utilizes the monotonic convexity of the inverse hyperbolic cosine function to enhance its influence in high-crack-density regions, while simultaneously introducing a surface roughness suppression factor. This is to reflect the influence of surface morphology on energy dissipation during microcrack propagation.

[0048] Finally, the above behavioral mechanism is expanded into a continuous function and integrated using integral notation. For the entire disturbance period to The evolutionary behavior is fully covered to obtain the response index. Its overall form possesses a complete physical explanation, mathematical solvability, and adaptability to the microscopic behavior of materials.

[0049] Combined with appendix Figure 3 Metal sheet materials are classified and managed according to their shear response. Shear-offset materials include steels with bainitic phase structures and steels with obvious coarse grain characteristics. For these materials, after identification through a microstructure assessment process before punching, sliding shear path control is prioritized during the punching stage. The sliding shear path maintains stable shear band generation through low-speed feed, and dynamically reduces the die gap using an online gap fine-tuning mechanism. The gap adjustment is updated in a closed loop based on real-time measured changes in the shear band width, matching the cutting edge contact state with the local deformation zone of the material, thereby suppressing corner collapse and tearing and increasing the proportion of bright bands. Gap adjustment is completed in micron-level steps, preferentially implemented in the early stage of shear band expansion and the critical transition zone, ensuring continuous and uniform energy input during the cross-section formation process.

[0050] For materials with strain concentration characteristics, the system switches to splitting path control. The splitting path employs a high-speed inrush strategy and implements nonlinear velocity segmentation control to form a stable stress concentration induction zone. The velocity segmentation includes at least a pre-acceleration segment, a stress-locking segment, and a fracture advance segment. Pre-acceleration is used to quickly establish contact stiffness, stress locking is used to stabilize the crack initiation position, and fracture advance is used to briefly traverse the fracture zone to reduce burrs and secondary tearing. The segment boundaries are updated based on real-time mechanical feedback to ensure that the stress concentration position is consistent with the crack propagation direction, thereby obtaining a regular fracture surface and a high proportion of bright surface.

[0051] To support the aforementioned path control strategy, this invention employs a dual-channel dynamic sampling system for process monitoring. One channel acquires the punching force signal, while the other channel simultaneously acquires the punch displacement signal. Both signals undergo denoising and anti-spurious spike processing. The force channel preferentially uses a median and low-pass fusion filter, while the displacement channel preferentially uses a differential prior and moving average fusion filter. Subsequently, the derivatives of the force and displacement curves are calculated to obtain the rate of change and slope information. The system uses a combined threshold strategy to identify the critical fracture stage. When the shear force change rate exceeds the first threshold and the punching speed fluctuation exceeds the second threshold, it is determined that the system has entered the critical fracture zone. Upon entering the critical zone, the controller immediately adjusts the feed rate according to a predetermined priority, slightly decelerating the sliding shear path to extend the bright band and briefly increasing the speed of the splitting path to avoid filamentary tearing.

[0052] Before punching, the microstructure of the metal sheet is determined. The identification of the bainitic phase structure is accomplished through the response behavior of thermoelectric properties or hysteresis curves. In the thermoelectric property detection stage, a pulsed thermal excitation method is used to apply a short-term constant amplitude heat flow to the metal sheet and measure its temperature recovery curve and potential change rate. The microstructure of the material is determined based on the settling time of the thermoelectric response and the voltage change gradient. In the hysteresis property detection stage, the area of ​​the hysteresis loop and the coercivity value of the material are measured synchronously with a Hall sensor array through low-frequency alternating magnetic field excitation. The bainitic structure is distinguished from other microstructure types based on the magnetic energy loss and the trend of magnetic susceptibility change.

[0053] The determination of coarse grain characteristics is accomplished using micron-level scanning electron microscopy combined with in-plane structural homogeneity analysis. High-magnification imaging is used to acquire images of the microscopic grain distribution on the material surface, and grayscale mean square error calculation and structural spectrum analysis are employed to assess the uniformity and orientation consistency of the grain arrangement. If the in-plane structural homogeneity is low or the grain boundary spacing exceeds a set threshold, the material is determined to have coarse grain characteristics.

[0054] In the splitting path control stage, this invention employs a nonlinear speed segmented control method, dividing the punching process into three stages: a pre-acceleration stage, a stress-locking stage, and a fracture-progression stage. The pre-acceleration stage establishes initial shear stress and ensures stable contact between the punch and the material; the stress-locking stage advances at a constant speed to stabilize the stress concentration zone and maintain energy accumulation in the shear band; the fracture-progression stage rapidly increases the punch speed after the system detects a critical fracture signal, causing the crack to propagate instantaneously along a predetermined path to form a smooth fracture surface. Segmented control can adjust the punch displacement curve in real time through a servo system to achieve coordinated matching between speed and material response.

[0055] Before punching, a pulsed thermal disturbance is applied to the metal sheet to be tested to obtain its thermoelectric properties. This pulsed thermal disturbance involves periodic, short-duration thermal energy input to the surface of the metal sheet via a thermal excitation device, causing the material to undergo a controllable temperature rise and fall process within a very short time. The system simultaneously records the voltage response changes during this process, and analyzes the internal thermoelectric coupling characteristics of the material by measuring the relationship between the rate of change of potential difference and the thermal excitation time. Bainitic microstructures, due to their high dislocation density and non-uniform carbon atom distribution, exhibit a fast thermoelectric response rate and short settling time, while pearlitic microstructures, due to their regular layered structure and continuous conductive paths, have a lower voltage response rate and a longer recovery time. By comparing the slope and stable range of the potential change curves before and after pulsed heating, the two microstructure types can be accurately distinguished, achieving non-destructive identification of the bainitic and pearlitic phases.

[0056] After completing the thermoelectric property tests, the system further verifies the material using hysteresis response behavior. The magnetic detection section employs a low-frequency magnetic field scanning device to establish an alternating magnetic field with controllable amplitude around the material. A Hall effect sensor array is used to collect real-time data on the correlation between magnetization and the applied magnetic field, plotting hysteresis loops and calculating their area and coercivity. Bainitic structures, due to internal dislocation concentration and high residual stress, exhibit wide hysteresis loop areas and high coercivity, while pearlitic structures have narrower hysteresis loops and lower coercivity. The system comprehensively determines the material's microstructure based on the changes in hysteresis loop area and the location of the coercivity peak. This method does not require damaging the material's surface structure during detection, enabling rapid acquisition of microstructure differences before punching, achieving high-precision material classification, and providing accurate input for adaptive punching control strategies.

[0057] During bainitic phase identification, the system simultaneously acquires thermoelectric response signals under pulsed thermal perturbation and hysteresis curve data under low-frequency magnetic field scanning, and performs joint feature matching on both. The thermoelectric signal provides information on the material's thermal conductivity rate and potential change rate, while the hysteresis curve reflects the material's magnetization behavior and energy dissipation characteristics. The system processes the two sets of data synchronously and establishes matching criteria within the signal feature space, determining the microstructure type by comparing the coupling relationship between the peak potential change rate and the hysteresis loop area. Under thermal perturbation conditions, bainitic structures exhibit rapid potential recovery and high magnetic energy loss, with signal characteristics showing a sharp peak and wide loop coupling morphology; pearlitic structures, on the other hand, exhibit slow potential change and small hysteresis loop area, with signals showing a gentle voltage gradient and narrow toroidal magnetic response. Through this dual-signal coupling feature matching method, bainitic phase microstructures can be rapidly identified without destructive sampling, thereby achieving high-precision material classification.

[0058] In the stage of determining coarse grain characteristics, micron-level scanning electron microscopy is used to observe the microstructure of the metal sheet surface. The system's field-of-view resolution is controlled within the range of 0.3~0.5µm to ensure clear resolution of grain boundaries. Through multi-frame image acquisition and regional averaging, the system extracts the gray-level distribution information of the sample surface and uses the gray-level root mean square error (RMSE) as an indicator of structural homogeneity for quantitative analysis. A smaller MSE value indicates uniform grain distribution and consistent orientation, resulting in higher structural stability; while a larger MSE value indicates uneven grain size distribution and significant orientation deviation, which is a characteristic of coarse grains. This image statistical analysis method can effectively reflect the microstructure evolution of the metal sheet during rolling or heat treatment, providing a reliable basis for judging potential shearing unevenness and fracture displacement during punching.

[0059] In the material microstructure analysis stage before blanking, the system uses a two-dimensional Fourier transform algorithm to perform spectral analysis on scanning electron microscopy images, extracting the spatial distribution characteristics of grain boundaries from the gray-scale matrix of the material surface structure. The directional components in the Fourier spectrum are used to calculate the grain boundary directional distribution density index, which reflects the concentration and uniformity of grain orientation. When the distribution density curve has a single-peak narrow band shape, it indicates that the grain orientation is uniform and the microstructure is homogeneous; if the density distribution shows a multi-peak or broadband diffusion shape, it indicates that the grain arrangement is disordered and there are coarse orientation deviations. Through this method, the homogeneity of the internal structure can be quantitatively evaluated, and the index parameters can be input into the blanking control module to provide a basis for material classification and path strategy.

[0060] Once the system identifies coarse grain characteristics in the material, it automatically adjusts the die clearance control strategy to compensate for the risk of irregular fracture zones caused by differences in grain size. The die clearance adjustment module appropriately increases or decreases the cutting edge spacing based on the identification results, rebalancing the stress distribution. For materials with large grains and uneven microstructure, the clearance adjustment biases towards decreasing to limit local shear displacement and suppress fracture zone propagation; while for materials with moderate grain size distribution, a standard clearance is maintained to ensure stable release of punching energy. Through clearance correction, the irregularity of the fracture surface in the coarse grain region can be significantly reduced.

[0061] During the punching control stage, the system further implements phased and graded control of the punch feed speed. The control speed in the pre-acceleration phase is set at 20%–40% of the target punching speed. This stage primarily aims to induce initial plastic stress in the material. By appropriately increasing the punch speed, stress in the shear zone is rapidly concentrated, forming a stable plastic deformation zone, thus providing favorable energy transfer conditions for the subsequent stress locking and fracture propulsion stages. After the pre-acceleration phase, the material is in a stable stress state, and the punching process enters the stress locking and fracture propulsion stage, achieving a smooth transition of energy input overall.

[0062] In the nonlinear speed segmented control of the punching path, the stress-locking stage is the key stage for achieving stable fracture zone formation. The control strategy in this stage is constant speed maintenance. The system stabilizes the punch feed speed through a servo control module, ensuring that the stress field within the shear band remains in equilibrium close to the yield limit. Constant speed control avoids abrupt stress concentration or localized plastic zone diffusion caused by speed fluctuations, allowing the material to form a continuous slip band and shear layer at the microscopic level, thereby promoting energy accumulation and microstructural stability before fracture. While maintaining constant speed control, the system monitors the shear force variation curve in real time, synchronously acquiring punching force signals through a force sensor and displacement encoder, and calculating its fluctuation amplitude and rate of change. When the detected shear force fluctuation exceeds the set stability threshold or exhibits a nonlinear abrupt change trend, the system determines that the material has entered the critical fracture zone. At this point, the stress concentration region has reached a critical state, microcracks begin to initiate and propagate along the shear plane, providing conditions for the rapid advancement of the subsequent fracture process.

[0063] Upon detecting entry into the critical fracture zone, the system immediately switches to the control strategy for the fracture advance stage. The servo drive module rapidly increases the punch acceleration, ensuring the shear zone completes the fracture stage traversal within a very short time. The punch acceleration in the fracture advance stage is set to no less than 10~15 m / s² to achieve high energy density input within a limited stroke, allowing the crack to propagate rapidly along a predetermined path, avoiding fracture delay and tearing. High-acceleration entry effectively shortens fracture time, reduces material rebound and burr formation, and keeps the fracture surface smooth and flat. The entire advance process is monitored in real time. When the shear force drops sharply and the displacement curve reaches a stable plateau, the control module automatically reduces the feed rate to enter the unloading stage, completing the punching cycle.

[0064] Example 1:

[0065] Combined with appendix Figure 4 This embodiment uses a 1.2mm thick cold-rolled bainitic steel sheet as an example to conduct multi-directional pre-stretch disturbance tests before punching. The steel sheet, after preliminary metallographic examination, exhibits a clear rolling texture orientation, with significant differences in grain arrangement along the rolling direction and perpendicular to it, indicating potential anisotropic shear response characteristics. To accurately identify shear-sensitive regions in different directions, a dual-pivot stretching device was used in the experiment to sequentially apply intermittent pre-stretch excitation to the steel sheet along the X-axis of the rolling direction and the Y-axis of the perpendicular direction.

[0066] In the initial stage of the test, the tensile device applied a micro-tension of 0.05 mm to the steel sheet for 0.2 s, followed by a 0.5 s static interval to observe its natural recovery. Thereafter, the tensile amplitude increased in a stepwise manner, increasing by 0.03 mm each time, until the maximum disturbance amplitude reached 0.17 mm. During each cycle, laser displacement sensors were fixed to the upper and lower surfaces of the steel sheet, respectively, with a sampling frequency set to 10 kHz to synchronously record displacement changes during the loading and unloading phases. Through high-precision synchronous data analysis, a complete elastic recovery path curve could be obtained.

[0067] During the experiment, the maximum displacement hysteresis along the X-axis was measured to be 6.4 µm, with a recovery time constant of 0.11 s and an average slope change of 1.8 N / mm during the stress unloading phase. The maximum displacement hysteresis along the Y-axis was 9.7 µm, with a recovery time constant of 0.16 s and a stress unloading slope change of 2.6 N / mm. The comparison shows that the hysteresis amplitude and recovery time in the Y-axis direction are significantly greater than those in the X-axis direction, indicating a slower stress release rate and weaker structural compatibility in the vertical direction. This characteristic suggests that the material is more prone to forming localized plastic concentration zones in the vertical direction, which are potential shear-sensitive areas.

[0068] To further quantify the judgment results, the system employs a multi-parameter weighted decision criterion for comprehensive assessment. Based on experimental data, the combined ratio of deformation rate and microcrack indications is calculated as follows:

[0069]

[0070] in, The average deformation rate per unit time is 0.85 mm / s; The change rate of microcrack density per unit area is taken as 0.014 mm⁻² / s. The combined ratio was calculated. The system's preset threshold is... ,when The material was determined to be shear-shifted. Based on the results, the vertical direction significantly exceeded the threshold, confirming that this direction is the principal shear-shifted direction, while the rolling direction was below the threshold, exhibiting strain concentration characteristics.

[0071] To verify the validity of the classification results, different path control strategies were applied to the material in subsequent punching tests. A high-speed splitting path was used along the rolling direction with a punch feed speed of 85 mm / s, resulting in a bright band ratio of 71.4% in the formed cross-section. A low-speed sliding shear path was used along the vertical direction with a punch feed speed of 42 mm / s and a die clearance adjusted to 0.035 mm, reducing the burr height to 14 µm. The comparison results show that the pre-stretching disturbance and multi-parameter judgment method established according to this invention can accurately identify the differences in the directional shear response of the metal sheet and guide the selection of the punching path, improving the cross-sectional flatness by approximately 22% and reducing burr generation by approximately 38%.

[0072] A 1.5 mm thick and 25 mm wide sample was selected for comparative experiments. After pre-stretching disturbance, the experiment entered the elastic recovery path analysis stage. Data was simultaneously acquired using a high-frequency laser displacement sensor and a surface micro-strain imaging system to analyze the material's response characteristics during loading and unloading. The system sampling frequency was 20 kHz, and the measurement period was set to 0.6 s. The loading stage... When uninstallation is complete The key parameters recorded during the analysis are as follows: The time displacement lag is the largest. At the same moment, the intensity of the micro-strain disturbance is Elastic recovery time constant Change in stress unloading slope Microcrack density variation function per unit area Material surface roughness factor .

[0073] According to the integral function definition proposed in this invention, the shear response capability index The calculation formula is as follows:

[0074]

[0075] To facilitate calculation, the parameters within the time interval are discretized, with each time step being 0.05s, and a time interval of 0.05s is set. , , It changes linearly with time within the interval. The time mean is calculated as follows: , , Substitute the integral expression and calculate using the approximate method of definite integral:

[0076]

[0077] Substitute the values:

[0078]

[0079] First, calculate the denominator: .

[0080] Then calculate the molecule: .

[0081] Dividing the two, we get: .

[0082] Next, calculate the inverse hyperbolic cosine term: ,Depend on achievable .

[0083] Substituting the results, we get:

[0084]

[0085] Set system response threshold The calculated values ​​for this experimental sample The material is determined to be of the shear offset type, and the sliding shear path control mode should be adopted.

[0086] A subsequent comparative experiment was conducted on another batch of steel sheets of the same material, and under the same loading conditions, it was found that their microcrack density was lower. Substituting into the formula, we get The values ​​were significantly below the threshold, indicating strain concentration, and a splitting path control method should be adopted. Different path strategies were used for the two groups of materials during actual punching: in the sliding shear path, the punch feed speed was 40 mm / s, and the die clearance was 0.03 mm; in the splitting path, the punch speed was increased to 85 mm / s, and the rake angle strengthening coefficient was set to 1.6. The final test results showed that the burr height of the sliding shear path specimen was 11 µm, and the bright band accounted for 78%; the cross-sectional regularity of the splitting path specimen was 93%, and the collapse angle width was controlled within 0.09 mm.

[0087] Example 2:

[0088] Combined with appendix Figure 5 Based on Example 1, this example selects three groups of cold-rolled steel sheet samples with a thickness of 1.2 mm. Sample A is bainitic steel, sample B is pearlitic steel, and sample C is low-carbon steel with obvious coarse grain characteristics. The microstructure of each sample was verified. The average grain size of the bainitic sample was approximately 1.3 µm, the pearlitic sample was 0.9 µm, and the coarse-grained sample was 2.8 µm.

[0089] Before the punching test, based on the tissue identification and classification process of this invention, the system identified samples A and C as shear-shifting materials, and sample B as strain-concentrated. The punching test used a servo precision punching system with a punch diameter of 12 mm and an initial die clearance of 0.04 mm. Under sliding shear path control, the system automatically adjusted the die clearance reduction based on the change in shear band width, with a clearance adjustment speed of 1 µm / ms and a maximum reduction of 0.012 mm. Simultaneously, a nonlinear velocity segmented control strategy was adopted in the splitting path, dividing the punch motion into three stages: a pre-acceleration stage with a speed of 30% of the target speed, a stress-locking stage maintaining a constant speed of 60 mm / s, and a breaking advance stage with an acceleration set at 12 m / s², to ensure the formation of a stable stress concentration induction band.

[0090] During the experiment, punching force and displacement signals were acquired in real time using a dual-channel dynamic sampling system at a sampling frequency of 25 kHz. The force channel signal was low-pass filtered, and the displacement signal was smoothed using a moving average filtering algorithm. The rate of change of force was obtained by calculating the derivative of the punching force-displacement curve. and velocity volatility Set the first threshold. Second threshold .when and When both occur simultaneously, the system determines that the punching process has entered the critical stage of fracture and triggers the adaptive control module to automatically adjust the punch speed or gap parameters.

[0091] Taking sample A as an example, monitoring data shows that the peak value of the shear force change rate reaches The punching speed fluctuation rate is The critical fracture condition was met. The system triggered dynamic adjustment, reducing the punch speed from 58 mm / s to 52 mm / s and simultaneously reducing the die clearance by 0.008 mm, resulting in a more stable energy input during the punching process. The final measured cross-section of sample A showed a bright band ratio of 81.5%, a burr height of 10.8 µm, and a fracture surface flatness error of 0.04 mm.

[0092] Sample B exhibits significant strain concentration characteristics in the splitting path control, with its highest shear force change rate being only [missing information]. The velocity fluctuation rate is Below the threshold range, the system maintains its high-speed inrush mode. The acceleration of 12.5 m / s² during the fracture propagation phase causes the fracture to quickly penetrate, forming a regular cross-section with a bright band ratio of 76.8% and a collapse angle width of 0.06 mm.

[0093] Sample C, due to its coarse grains, exhibits uneven stress distribution, resulting in frequent fluctuations in its shear force derivative, with peak values ​​reaching as high as [missing value]. The system repeatedly identified the critical stage and triggered adjustments accordingly. On the second trigger, the die clearance was reduced by 0.01 mm and the punch speed was reduced by 5 mm / s. Ultimately, the proportion of bright bands on the sample's cross-section increased to 74.3%, and the burr height was controlled at 13 µm.

[0094] Thermoelectric and magnetic response tests and path control optimization experiments were conducted on cold-rolled steel sheets with three different microstructures before punching. The selected samples were numbered D1 (bainitic steel), D2 (pearlitic steel), and D3 (coarse-grained low-carbon steel), all with a thickness of 1.5 mm and a surface roughness Ra of 0.38 µm.

[0095] In the tissue identification phase, a combined method of pulsed thermal perturbation and hysteresis response was employed. Thermoelectric testing used a micro-thermal excitation system to apply a single-pulse heat flux to the sample, with a heat flux density of 45 kW / m² and a pulse duration of 0.25 s. The system recorded temperature changes and potential responses via thermocouples and a synchronous voltage acquisition module. Signal analysis revealed the voltage change rate of sample D1. The voltage change rate reached 5.7 mV / s with a recovery time of only 0.43 s; the voltage change rate of sample D2 was 2.9 mV / s with a recovery time of 0.85 s; the voltage change rate of sample D3 was 4.1 mV / s with a recovery time of 0.61 s. According to the preset threshold, if... Furthermore, the recovery time was less than 0.6 s, indicating a bainitic structure. The results showed that sample D1 clearly met the characteristics of bainite, while samples D2 and D3 did not meet the criteria.

[0096] Magnetic response tests were then conducted. The system employed a low-frequency magnetic field scanning method, applying a maximum magnetic flux density of 1.2T to the samples under a 10Hz magnetic field. The magnetization change was measured using a Hall effect sensor array, hysteresis loops were plotted, and the loop area and coercivity were calculated. Sample D1 had a hysteresis loop area of ​​4.8 × 10⁻³ T·A / m and a coercivity of 310 A / m; sample D2 had a loop area of ​​2.1 × 10⁻³ T·A / m and a coercivity of 180 A / m; and sample D3 had a loop area of ​​3.9 × 10⁻³ T·A / m and a coercivity of 270 A / m. Based on empirical rules, the high dislocation density of bainitic structures results in high magnetic energy loss and high coercivity. Therefore, the magnetic properties of sample D1 were consistent with the thermoelectric results, confirming its bainitic structure. D2 was determined to be pearlite, and D3, due to its large grain size and relatively wide hysteresis loops, was classified as a coarse-grained structure.

[0097] After material classification, the punching path control stage begins. For shear-shifting materials D1 and D3, a nonlinear segmented control mode for the splitting path is adopted; for D2, a constant-speed sliding shear path control is used. The system sets a three-stage speed curve on the servo punching machine. The pre-acceleration stage has a speed of 35% of the target speed, set at 50 mm / s, lasting 0.05 s, used to stabilize contact and induce initial plastic stress; the stress-locking stage maintains a constant speed of 80 mm / s for 0.1 s to achieve stable energy accumulation; the breaking advance stage sets the punch acceleration at 13 m / s² for 0.02 s to ensure rapid crack penetration. The speed curve is described by an exponential smoothing function to avoid mechanical impact.

[0098] During the punching process, a dual-channel sensing system synchronously monitors the punching force and displacement signals at a sampling frequency of 25 kHz. For sample D1, the shear force was detected to rise to 9.3 kN and remain stable in the stress-locked section. When the punching section started, the force value instantly dropped to 5.2 kN, indicating rapid fracture completion. The proportion of bright bands on the fracture surface was 83.6%, and the burr height was 9.4 µm. Sample D2, due to the constant speed mode, showed a smooth rise in the punching force curve to 7.8 kN, with a fracture time extended to 0.21 s. The proportion of bright bands on the fracture surface was 76.1%, and the burr height was 12.6 µm. Sample D3 experienced slightly higher stress fluctuations in the punching section due to coarse grains, with a peak punching force change rate of 1.15 × 10⁻⁶. 4 The system automatically adjusts the acceleration to reduce it to 11.5 m / s², resulting in a final cross-section with a bright band ratio of 78.9% and a fracture surface flatness error of 0.05 mm.

[0099] Three groups of metal sheet samples were selected: E1 (bainitic steel), E2 (coarse-grained low-carbon steel), and E3 (mixed microstructure steel). All samples were 1.3 mm thick with a surface roughness of Ra 0.35 µm and were mechanically polished to eliminate the influence of the oxide film. The experiment aimed to accurately classify the material microstructure type through combined thermoelectric and magnetic feature matching and microscopic image spectral analysis, and to adaptively correct the die clearance and punch motion curve based on the identification results, thereby improving the cross-sectional smoothness and the stability of the bright band.

[0100] During the bainitic phase identification process, the system simultaneously acquires thermoelectric response signals and hysteresis curve shape characteristics, and performs joint feature matching analysis. Taking sample E1 as an example, a pulsed thermal disturbance with a heat flux density of 40 kW / m² and a pulse duration of 0.3 s is applied, and the voltage change rate is measured. The recovery time was 0.42 s. Simultaneously, a low-frequency magnetic field scan was performed at 12 Hz with a maximum magnetic flux density of 1.1 T, yielding the hysteresis loop area. The coercivity is 320 A / m. After inputting the thermoelectric signal and hysteresis loop data into the matching algorithm, the system calculates the signal morphological correlation coefficient. The value is above the threshold of 0.85, indicating a bainitic structure. Sample E3 has a correlation coefficient of only 0.62, a significant delay in the thermoelectric response peak, and a narrow hysteresis loop, indicating a mixed structure.

[0101] In determining the coarse grain characteristics, scanning electron microscopy was used on samples E2 and E3, with a field-of-view resolution of 0.4 µm and an image sampling area of ​​200 × 200 µm². After multi-point acquisition, the root mean square error (S_g) of the image grayscale was calculated, where E2's... E3 By combining two-dimensional Fourier transform with spectral analysis of grain boundary distribution, directional distribution density indices are extracted. E2 E3 The system-defined criteria for determining grain homogeneity are as follows: and The time frame is defined as a coarse grain feature. The results show that E2 exhibits a significant coarse grain arrangement, while the structure of E3 is relatively uniform.

[0102] After identification, the system automatically corrects the punching parameters based on the results. For bainitic steel E1, a splitting path control is selected and a three-stage velocity curve is set. The pre-acceleration stage velocity is 30% of the target velocity, i.e., 48 mm / s, lasting 0.04 s; the stress-locking stage velocity is constant at 80 mm / s, lasting 0.12 s; and the breaking advance stage punch acceleration is 13 m / s², lasting 0.02 s. For coarse-grained steel E2, a sliding shear path is adopted, with an initial die clearance of 0.04 mm, which is adjusted in real time according to stress fluctuations in the fracture zone, with a maximum reduction of 0.012 mm.

[0103] In the punching test, the system monitored shear force and displacement signals through dual-channel sampling. Sample E1 exhibited a peak shear force of 9.1 kN, with a stress lock-in phase fluctuation of less than 3%, a rapid decrease to 5.0 kN during the breakthrough phase, a fracture time of 0.018 s, a bright band ratio of 85.7% on the fracture surface, and a burr height of 9.2 µm. Sample E2, due to its coarse grains, showed a peak shear force derivative of 1.18 × 10⁻⁶. 4 The system automatically reduced the punch speed to 55 mm / s and narrowed the gap by 0.01 mm, resulting in a final bright band ratio of 78.4% and a burr height of 12.6 µm. Sample E3 adopted a splitting path mode, but due to the mixed structure, the stress fluctuation was large. The system determined that the critical fracture zone appeared 0.015 s earlier and automatically accelerated the acceleration of the protrusion segment to 15 m / s², resulting in a bright band of 80.2% and a collapse angle width of 0.07 mm.

[0104] High-precision classification of bainitic phases can be achieved by matching the characteristics of thermoelectric and magnetic response signals. Combined with gray-scale statistics and Fourier spectrum analysis of scanning micrographs, the distribution of coarse grains and structural homogeneity can be accurately determined. After implementing adaptive correction of mold gap and speed based on the above identification results, the cross-sectional flatness of the three groups of samples improved by an average of 0.05 mm, the proportion of bright bands increased by 17.8%, and the burr height decreased by an average of 33%.

[0105] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for controlling the surface quality of stamped metal sheets, characterized in that, include: Before punching, a micro-plastic excitation is applied to the metal sheet, including small-amplitude hammering or pre-stretching disturbance, to obtain the instantaneous recovery parameters of the material under low-energy deformation. The parameters include deformation recovery rate, displacement hysteresis or microcrack indications, which are used to determine the shear response state of the metal sheet. The deformation stability of a material under the current stress field is determined based on its shear response state, and it is classified as either shear displacement type or strain concentration characteristic. In the blanking control process, for shear offset type materials, a low-speed feed and a sliding shear path with reduced clearance are used; for strain concentrated type materials, a high-speed punching and a splitting path with enhanced front angle fracture are used. During the punching process, the derivative changes of the punching force and displacement curves are monitored to identify the critical fracture stage and dynamically adjust the punch feed speed. The fracture morphology is controlled by accelerating or decelerating. At the end of the punching process, the edge springback behavior during the punch return stage is detected. If there is a warping trend, the end rolling correction action is activated. The obtuse angle punch head or conical structure is combined with the material springback kinetic energy to achieve secondary shaping of the edge area.

2. The method for controlling the surface quality of metal sheet punching according to claim 1, characterized in that, The pre-stretching perturbation includes intermittent stretching in multiple directions to identify differences in shear response due to material anisotropy; the acquisition of the instantaneous recovery parameters includes synchronously recording the elastic recovery path before and after the perturbation using a laser displacement sensor; the judgment of the shear response state adopts a multi-parameter weighted decision criterion, including the combined ratio of deformation rate and microcrack indications.

3. The method for controlling the surface quality of metal sheet punching according to claim 2, characterized in that, The intermittent stretching in multiple directions includes tensile excitation along at least two principal axes along the material rolling direction and the perpendicular direction, used to identify directional shear-sensitive areas; wherein the time interval and amplitude of the tensile disturbance increase in a stepwise manner to observe the change law of the material's response in different directions under different stress levels; the laser displacement sensor is deployed on the upper and lower surfaces of the material to realize synchronous monitoring of in-plane and out-of-plane elastic recovery paths.

4. The method for controlling the surface quality of metal sheet punching according to claim 2, characterized in that, The elastic recovery path analysis includes evaluating the maximum displacement hysteresis, recovery time constant, and stress unloading slope change; the evaluation of microcrack signs in the multi-parameter weighted decision criterion is performed by a material surface micro-strain imaging system, combined with grayscale perturbation to identify crack initiation locations; the joint ratio is the ratio of the average deformation rate per unit area to the microcrack density change rate, used to quantify shear response capability.

5. The method for controlling the surface quality of metal sheet punching according to claim 1, characterized in that, The shear offset material includes steel types with bainitic phase structure or coarse grain characteristics; the reduction in die gap in the sliding shear path is dynamically adjusted during the punching process to match the change in shear band width, wherein the high-speed punching in the splitting path adopts nonlinear speed segmented control to form a stress concentration induction band; wherein the monitoring of the derivative of the punching force and displacement curve uses a dual-channel dynamic sampling system for filtering; the identification of the fracture critical stage is based on the combined judgment that the shear force change rate exceeds a first threshold and the punching speed fluctuation exceeds a second threshold.

6. The method for controlling the surface quality of metal sheet punching according to claim 5, characterized in that, The identification of the bainitic phase structure is determined by the response behavior of thermoelectric properties or hysteresis curves; the determination of coarse grain characteristics is based on micron-level scanning electron microscopy and in-plane structural homogeneity analysis; the nonlinear velocity segmented control of the splitting path includes three stages: pre-acceleration stage, stress locking stage, and fracture breakthrough stage.

7. The method for controlling the surface quality of metal sheet punching according to claim 6, characterized in that, The determination of the thermoelectric properties is based on applying pulsed thermal disturbance to the material before punching and simultaneously measuring the voltage response change rate to distinguish the structural differences between bainite and pearlite; the determination of the response behavior of the hysteresis curve is obtained by a Hall sensor array under low-frequency magnetic field scanning to detect the hysteresis loop area and coercivity characteristics of the material.

8. The method for controlling the surface quality of metal sheet punching according to claim 6, characterized in that, The thermoelectric response signal and the shape of the hysteresis curve obtained during the bainitic phase identification process are combined for feature matching; while the field resolution of the scanning electron microscopy imaging in the grain coarseness feature determination is not less than 0.5µm, and the structural homogeneity is analyzed by combining the root mean square error of the image grayscale distribution.

9. The method for controlling the surface quality of metal sheet punching according to claim 6, characterized in that, The in-plane structure homogeneity analysis includes extracting the grain boundary direction distribution density index based on two-dimensional Fourier transform to quantify the uniformity of coarse grain arrangement. After identifying coarse grains, the die gap control strategy is modified to compensate for the risk of irregular fracture zones; the control speed of the pre-acceleration section is 20% to 40% of the punching target speed, which is used to induce the initial plastic stress of the material.

10. The method for controlling the surface quality of metal sheet punching according to claim 6, characterized in that, The control strategy for the stress-locking section is to maintain a constant speed, while the system detects shear force fluctuations in real time to determine whether it has entered the critical fracture zone; and the acceleration of the punch in the fracture advance section is not less than 10~15m / s², so that the shear zone passes through the fracture stage.

Images