A hard springback treatment system and method for precision stamped parts

Abstract

This invention provides a hard springback treatment system and method for precision stamped parts, belonging to the field of precision stamping manufacturing technology. The system includes a springback detection module, a data processing module, a multi-field collaborative execution module, and a closed-loop feedback module. The springback detection module non-contactly acquires springback distribution data on the surface of the stamped part using a multispectral camera and generates a visualized numerical matrix by combining it with color scale calibration. The data processing module integrates finite element simulation and machine learning models to calculate the refined compensation amount for each region. The method includes: detecting the springback distribution after demolding → generating a compensation strategy → multi-field collaborative execution of compensation → real-time feedback optimization. This invention achieves accurate identification and dynamic compensation of micron-level springback, solving the problems of low detection accuracy, single compensation mechanism, and poor adaptability of traditional methods. It effectively suppresses elastic springback, improves the shape and dimensional accuracy of the workpiece, reduces manual correction steps, and improves production efficiency and product qualification rate.

Description

Technical Field

[0001] This invention relates to the field of precision stamping manufacturing technology, specifically to a hard springback treatment system and method for precision stamping parts. Background Technology

[0002] Precision stamped parts (such as high-strength steel automotive frames, aluminum alloy structural parts, and aerospace precision structural parts) experience springback deformation during the forming process due to the elastic recovery of the material, resulting in a deviation between the final shape and the design target. Traditional springback handling methods have the following drawbacks:

[0003] 1. Insufficient detection accuracy: Contact measurement (such as coordinate measuring machine) is prone to scratching the workpiece and cannot quickly obtain minute springback differences across the entire surface;

[0004] 2. Single compensation mechanism: Relying on manual correction or the action of a single external force / energy field, it is easy to cause local stress concentration and uncontrollable heat-affected zone, making it difficult to adapt to the springback characteristics of difficult-to-deform materials such as high-strength steel (tensile strength ≥1500MPa) and aluminum alloys (2000 series / 7000 series);

[0005] 3. Poor adaptability: Traditional methods are mostly based on experience or simple simulation, which cannot cope with the springback compensation requirements of material property fluctuations (such as anisotropy and time-related changes) and complex geometries (such as holes and long strip structures), resulting in low product qualification rate (usually ≤95%) and long production cycle.

[0006] To address this, a hard springback treatment system and method for precision stamped parts are proposed. Summary of the Invention

[0007] The present invention aims to solve the problems mentioned in the background art by providing a hard springback treatment system and method for precision stamped parts.

[0008] The specific technical solution is as follows:

[0009] A hard springback treatment system for precision stamped parts, comprising:

[0010] The springback detection module is configured to collect surface springback distribution data of precision stamped parts after demolding. The springback is obtained through a non-contact optical detection device with a detection accuracy of not less than ±0.001mm.

[0011] The data processing module is communicatively connected to the rebound amount detection module and is configured to generate a three-dimensional rebound morphology model based on the rebound amount distribution data, and calculate the rebound compensation amount of each region based on preset target shape parameters.

[0012] The multi-field collaborative execution module is communicatively connected to the data processing module and is configured to synchronously apply controllable rigid external force and / or local energy field according to the springback compensation amount of each region, forcing the stamped part to undergo directional plastic deformation to counteract elastic springback;

[0013] The closed-loop feedback module is configured to monitor the deformation state of the stamped part in real time, feed the monitoring data back to the data processing module, and dynamically adjust the magnitude and direction of the rigid external force or the parameters of the energy field.

[0014] The aforementioned hard springback treatment system for precision stamped parts includes a springback detection module comprising a multispectral camera and an image processing unit. The multispectral camera is used to acquire the springback characteristic spectrum of the stamped part surface. The image processing unit is calibrated using a color scale bar, with the color range gradually changing from light green to dark red, corresponding to a springback amount of -10.000 to 10.000, and converts the spectral data into a springback numerical matrix.

[0015] The aforementioned hard springback treatment system for precision stamped parts includes a data processing module comprising a finite element simulation unit and a machine learning compensation unit. The finite element simulation unit is used to simulate the springback behavior of the stamped parts based on a material constitutive model and generate an initial compensation strategy. The machine learning compensation unit corrects the initial compensation strategy using a neural network model trained on historical process data and outputs refined compensation amounts for each region.

[0016] The aforementioned hard springback treatment system for precision stamped parts includes a multi-field coordinated execution module comprising a rigid external force application unit and a local energy field generation unit. The rigid external force application unit includes a servo hydraulic cylinder array and a precision pressure sensor, used to apply a directional pressure of 0.1–10 N to a designated area of ​​the stamped part. The local energy field generation unit includes a laser heating module and a high-frequency electromagnetic induction heating unit, used to reduce the yield strength of the material through thermal softening effect and assist in plastic deformation. The laser heating module has a wavelength of 1064 nm and a power of 0.5–5 kW. The high-frequency electromagnetic induction heating unit has a frequency of 100 kHz–1 MHz and a power of 2–20 kW.

[0017] The aforementioned hard springback treatment system for precision stamped parts includes a closed-loop feedback module comprising a strain sensor array and an optical tracker. The strain sensor array is attached to the surface of the stamped part to collect local strain data in real time. The optical tracker monitors the overall morphological changes of the stamped part through structured light scanning, and the response time of the dynamic adjustment is ≤100ms.

[0018] This embodiment provides a hard springback treatment method for precision stamped parts, applied to the aforementioned hard springback treatment system for precision stamped parts, and includes the following steps:

[0019] S1: After the precision stamping part is demolded, the surface springback distribution data is collected by a non-contact optical detection device. The springback is marked by a color scale bar, with the color range from light green to dark red, corresponding to a springback of -10.000 to 10.000.

[0020] S2: Generate a three-dimensional rebound morphology model based on the rebound distribution data, and calculate the rebound compensation amount of each region in combination with the target shape parameters;

[0021] S3: Based on the springback compensation amount of each region, a controllable rigid external force and / or local energy field are applied simultaneously to force the stamped part to undergo directional plastic deformation to counteract elastic springback;

[0022] S4: Monitor the deformation state of the stamped part in real time, feed the monitoring data back to step S2, and dynamically adjust the magnitude and direction of the rigid external force or the parameters of the energy field until the springback is ≤ ±5μm.

[0023] The above-mentioned hard springback treatment method for precision stamped parts, wherein the non-contact optical inspection device in step S1 includes a multispectral camera, the calibration range of the color scale bar covers the springback amount from -10.000 to 10.000, and the calibration accuracy is ±0.001.

[0024] The above-mentioned hard springback treatment method for precision stamped parts, wherein the construction of the three-dimensional springback morphology model in step S2 is based on finite element simulation and machine learning correction; the finite element simulation simulates springback behavior based on the material constitutive model, and the machine learning correction compensates the simulation results through a neural network model trained on historical process data, with a compensation error ≤3%.

[0025] In the above-mentioned method for hard springback treatment of precision stamped parts, the magnitude of the controllable rigid external force in step S3 is 0.1 to 10 N, and the direction is opposite to the direction of springback deformation; the parameters of the local energy field include laser power of 0.5 to 5 kW, high-frequency electromagnetic induction frequency of 100 kHz to 1 MHz, and the overlap between the action area and the area with excessive springback is ≥90%.

[0026] The above-mentioned hard springback treatment method for precision stamping parts, wherein the material of the precision stamping parts is high-strength steel or aluminum alloy, the shape accuracy after treatment reaches ±5μm, the tensile strength of the high-strength steel is ≥1500MPa, and the aluminum alloy is 2000 series or 7000 series.

[0027] The present invention has the following beneficial effects:

[0028] 1. Precise Identification and Compensation: Non-contact optical detection combined with color-coded scale calibration enables high-resolution capture of minute differences in rebound across the entire surface (e.g., Figure 1 , Figure 2 Different colored areas correspond to different rebound amounts, providing a reliable data basis for subsequent compensation; the multi-field synergy (rigid external force + energy field) solves the limitations of a single mechanism, can adapt to the rebound compensation needs of complex geometries (such as long strip structures and holes), and avoids stress concentration or thermal damage.

[0029] 2. Strong dynamic adaptability: The closed-loop feedback mechanism effectively responds to fluctuations in material properties (such as anisotropy of high-strength steel and aging changes of aluminum alloys) through real-time monitoring and parameter adjustment, ensuring the stability of the compensation process and avoiding the effect deviation caused by material fluctuations in traditional "one-time compensation".

[0030] 3. Improve production efficiency and quality: Full-process automation (detection-calculation-execution-feedback) replaces the traditional manual correction process, significantly shortening the production cycle; the shape accuracy and dimensional consistency after compensation are greatly improved (such as reducing ±5μm level deviation), and the product qualification rate is increased from the traditional ≤95% to a higher level, reducing labor costs and material waste.

[0031] 4. Good process versatility: It is suitable for springback processing of various precision stamping materials such as high-strength steel and aluminum alloys, as well as complex geometries (such as automobile frames and mechanical structural parts), thus expanding the application scenarios of precision stamping processes. Attached Figure Description

[0032] Figure 1 Example of a cloud map showing the springback distribution of a precision stamped part after demolding;

[0033] Figure 2 A schematic diagram showing the springback morphology and key areas of a precision stamped part;

[0034] Figure 3 A schematic diagram of the composition of a hard springback treatment system for precision stamped parts provided in an embodiment of the present invention;

[0035] Figure 4 A flowchart of a hard springback treatment method for precision stamped parts provided in an embodiment of the present invention. Detailed Implementation

[0036] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0037] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual images. They should not be construed as limiting the scope of this application. To better illustrate the embodiments of the present invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0038] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "inner," and "outer" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present application. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0039] In the description of this invention, unless otherwise explicitly specified and limited, the term "connection" or similar designation indicating a connection between components should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral part; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can refer to the internal communication between two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0040] Example: Refer to Figure 1-4 This embodiment provides a hard springback treatment system for precision stamped parts, comprising: a springback detection module, a data processing module, a multi-field collaborative execution module, and a closed-loop feedback module, wherein:

[0041] The springback detection module is configured to collect surface springback distribution data of precision stamped parts after demolding. The springback is obtained through a non-contact optical detection device with a detection accuracy of not less than ±0.001mm.

[0042] The data processing module is connected to the rebound amount detection module and is configured to generate a three-dimensional rebound morphology model based on the rebound amount distribution data, and calculate the rebound compensation amount of each region based on the preset target shape parameters.

[0043] The multi-field collaborative execution module communicates with the data processing module and is configured to synchronously apply controllable rigid external force and / or local energy field according to the springback compensation amount of each region, forcing the stamped part to undergo directional plastic deformation to counteract elastic springback;

[0044] The closed-loop feedback module is configured to monitor the deformation state of the stamped part in real time, feed the monitoring data back to the data processing module, and dynamically adjust the magnitude, direction or energy field parameters of the rigid external force.

[0045] Through a fully integrated design encompassing "springback detection - data processing - multi-field collaborative execution - closed-loop feedback," the system achieves accurate identification, dynamic calculation, and adaptive compensation for springback deformation of precision stamped parts (especially those with complex geometries and high-strength materials). Specific effects include:

[0046] The rebound distribution data acquired by the detection module (such as...) Figure 1 , Figure 2 (As shown in the color gradient area) can cover the entire surface of the stamped part, solving the problem that traditional single detection methods cannot capture localized minute springback differences;

[0047] The multi-field collaborative execution module (rigid external force + energy field) can target locations with excessive rebound in different areas (such as...). Figure 2 The effect is applied in a targeted manner in the high-value red area to avoid the defects such as stress concentration and uncontrollable heat-affected zone caused by traditional manual correction or single energy field;

[0048] The closed-loop feedback module adjusts parameters in real time to ensure that the compensation process is dynamically optimized according to the deformation state of the stamped part, thus solving the problem that the traditional "one-time compensation" cannot adapt to the fluctuation of material properties.

[0049] Specifically, in this embodiment, the springback detection module includes a multispectral camera and an image processing unit. The multispectral camera is used to collect the springback characteristic spectrum of the stamped part surface. The image processing unit is calibrated by a color scale bar, with the color range gradually changing from light green to dark red, corresponding to a springback amount of -10.000 to 10.000, and converts the spectral data into a springback value matrix.

[0050] Non-contact acquisition using a multispectral camera and color scale bars (e.g.) Figure 1 , Figure 2 The gradient color bar calibration technology enables high-resolution, non-destructive acquisition of rebound distribution. The effects are reflected in:

[0051] The color-coded scale bar covers the rebound range from -10,000 to 10,000 (e.g., Figure 1 , Figure 2 The color bar on the right visually reflects the differences in springback across different areas of the stamped part's surface through color gradients (e.g., ...). Figure 1 The medium-green low-value area and the red high-value area solve the problems of low efficiency and easy scratching of workpieces in traditional contact inspection (such as coordinate measuring machine);

[0052] A multispectral camera combined with an image processing unit (color scale calibration) can convert the springback of complex surfaces into a standardized numerical matrix (e.g., Figure 1 The values ​​(such as 0.133 and -0.35) within the black highlighted boxes provide high-precision input for subsequent data processing.

[0053] Specifically, in this embodiment, the data processing module includes a finite element simulation unit and a machine learning compensation unit; the finite element simulation unit is used to simulate the springback behavior of stamped parts based on the material constitutive model and generate an initial compensation strategy; the machine learning compensation unit corrects the initial compensation strategy through a neural network model trained on historical process data and outputs the refined compensation amount for each region.

[0054] A dual-drive strategy combining simulation and data-driven approaches is employed to achieve highly accurate calculation of springback compensation. The results are as follows:

[0055] Finite element simulation elements simulate springback behavior based on material constitutive models (such as the elastoplastic properties of high-strength steel and aluminum alloys), and can quickly generate initial compensation strategies (such as...). Figure 2 (3D model simulation analysis)

[0056] The machine learning compensation unit corrects simulation results using a neural network model trained on historical process data, making up for errors caused by material property fluctuations (such as the anisotropy of high-strength steel) in traditional single simulations, and improving the adaptability of the compensation strategy to complex working conditions.

[0057] Specifically, in this embodiment, the multi-field collaborative execution module includes a rigid external force application unit and a local energy field generation unit; the rigid external force application unit includes a servo hydraulic cylinder array and a precision pressure sensor, used to apply a directional pressure of 0.1 to 10 N to a designated area of ​​the stamped part; the local energy field generation unit includes a laser heating module and a high-frequency electromagnetic induction heating unit, used to reduce the yield strength of the material through the thermal softening effect and assist plastic deformation; the wavelength of the laser heating module is 1064 nm and the power is 0.5 to 5 kW; the frequency of the high-frequency electromagnetic induction heating unit is 100 kHz to 1 MHz and the power is 2 to 20 kW.

[0058] By simultaneously and synergistically applying rigid external force (servo hydraulic cylinder array) and local energy field (laser / high-frequency heating), the limitations of a single action mechanism are overcome, achieving controllable plastic deformation. The effects are manifested in:

[0059] Rigid external force application elements can be applied to a specified area (e.g.) Figure 2 Apply directional pressure (0.1–10 N) to areas with excessive rebound, such as central holes and irregular edges, to directly counteract elastic rebound;

[0060] The local energy field generating unit reduces the yield strength of the material in the target area through thermal softening effect (such as heating with a laser wavelength of 1064nm), assists in plastic deformation, and avoids the risk of local fracture caused by the high hardness of the material when rigid external force is applied alone;

[0061] The two working together can cover complex geometries (such as...) Figure 2 The rebound compensation requirement of long strip structures and holes solves the limitation that traditional methods are only applicable to simple structures.

[0062] Specifically, in this embodiment, the closed-loop feedback module includes a strain sensor array and an optical tracker; the strain sensor array is attached to the surface of the stamped part to collect local strain data in real time; the optical tracker monitors the overall morphological changes of the stamped part through structured light scanning, and the dynamic adjustment response time is ≤100ms.

[0063] By employing multi-dimensional real-time monitoring (strain data + morphology scanning) and rapid adjustment (response time ≤ 100ms), the dynamic stability of the compensation process is ensured. The effects are reflected in:

[0064] A strain sensor array (attached to the surface) acquires local strain data in real time, while an optical tracker (structured light scanning) monitors overall morphological changes (e.g., Figure 1 , Figure 2 (Changes in color distribution) can promptly detect overcompensation or undercompensation trends;

[0065] Dynamically adjust the magnitude, direction, or energy field parameters of rigid external forces (such as laser power) to avoid compensation deviations caused by fluctuations in material properties (such as changes in the aging properties of aluminum alloys), thereby improving the consistency of the final shape accuracy.

[0066] This embodiment also provides a hard springback treatment method for precision stamped parts, applied to the aforementioned hard springback treatment system for precision stamped parts, including the following steps:

[0067] S1: After the precision stamping part is demolded, the surface springback distribution data is collected by a non-contact optical detection device. The springback is marked by a color scale bar, with the color range from light green to dark red, corresponding to a springback of -10.000 to 10.000.

[0068] S2: Generate a three-dimensional rebound morphology model based on rebound distribution data, and calculate the rebound compensation amount of each region by combining the target shape parameters;

[0069] S3: Based on the springback compensation amount of each region, a controllable rigid external force and / or local energy field are applied simultaneously to force the stamped part to undergo directional plastic deformation in order to counteract elastic springback;

[0070] S4: Monitor the deformation state of the stamped part in real time, feed the monitoring data back to step S2, and dynamically adjust the magnitude, direction or energy field parameters of the rigid external force until the springback is ≤ ±5μm.

[0071] Through a closed-loop process of "detection-calculation-execution-feedback," the problem of springback in precision stamping parts is systematically solved, achieving end-to-end control from minute springback identification to precise compensation. The effects are reflected in:

[0072] It covers the entire life cycle of stamped parts after demolding (inspection → compensation → verification), solving the problems of low efficiency and high cost of traditional "post-correction";

[0073] By combining multi-field collaborative execution and dynamic feedback, it can adapt to the springback compensation requirements of different materials (high-strength steel, aluminum alloy) and different structures (complex geometry), thereby improving the versatility of the process.

[0074] Specifically, in this embodiment, the non-contact optical detection device in step S1 includes a multispectral camera, the calibration range of the color scale bar covers the rebound amount from -10.000 to 10.000, and the calibration accuracy is ±0.001.

[0075] Non-contact optical detection and standardized color calibration ensure the reliability and consistency of rebound data. The effects are reflected in:

[0076] Rebound characteristic spectrum acquired by multispectral camera (e.g.) Figure 1 The combination of a predominantly green color scheme interspersed with yellow, red, and blue areas and a color scale bar (light green to dark red gradient) allows for the rapid conversion of complex surface resilience differences into a calculable numerical matrix (e.g., ...). Figure 1 The values ​​marked in black boxes in the middle provide a unified data benchmark for subsequent processing;

[0077] Avoid subjective errors in traditional manual measurements (such as feeler gauges and micrometers), and improve the efficiency and accuracy of data collection.

[0078] Specifically, in this embodiment, the construction of the three-dimensional springback morphology model in step S2 is based on finite element simulation and machine learning correction; the finite element simulation simulates springback behavior based on the material constitutive model, and the machine learning correction compensates the simulation results through a neural network model trained on historical process data, with a compensation error ≤3%.

[0079] By employing a dual-drive approach of "simulation + data" to calculate compensation amounts, high-precision and highly adaptable compensation strategies are generated. The effects are reflected in:

[0080] Finite element simulation, based on material constitutive models (such as the elastic-plastic parameters of high-strength steel), simulates springback behavior and can quickly predict springback trends in different regions (e.g., Figure 2 (Color gradient area in the 3D model)

[0081] Machine learning models correct simulation results by using historical process data (such as springback under different materials and mold parameters), which makes up for the errors caused by neglecting microscopic defects in materials (such as inclusions) in traditional simulations and improves the accuracy of compensation.

[0082] Specifically, in this embodiment, the magnitude of the controllable rigid external force in step S3 is 0.1 to 10 N, and the direction is opposite to the direction of the rebound deformation; the parameters of the local energy field include laser power of 0.5 to 5 kW, high-frequency electromagnetic induction frequency of 100 kHz to 1 MHz, and the overlap between the area of ​​action and the area where the rebound amount exceeds the standard is ≥90%.

[0083] By precisely controlling the direction, magnitude, and energy field parameters (such as laser power and frequency) of external forces, directional plastic deformation of the target area can be achieved. The effects are manifested in:

[0084] The direction of the rigid external force is opposite to the direction of the springback deformation (e.g.) Figure 2 The high-value area in red corresponds to the direction of pressure and can directly counteract elastic rebound.

[0085] The overlap between the energy field's effective area and the area with excessive rebound is ≥90% (e.g., Figure 2 High-value areas around the central hole (avoiding energy waste or damage to non-target areas (such as embrittlement of the heat-affected zone of high-strength steel) and improving the controllability of the compensation process.

[0086] Specifically, in this embodiment, the material of the precision stamping part is high-strength steel or aluminum alloy, and the shape accuracy after processing reaches ±5μm. The tensile strength of the high-strength steel is ≥1500MPa; the aluminum alloy is 2000 series or 7000 series.

[0087] By leveraging the synergistic effect of the aforementioned systems and methods, the challenge of springback compensation in high-strength, difficult-to-deform materials is solved, enabling the efficient industrial production of precision stamped parts. The effects are reflected in:

[0088] For high-strength steel (tensile strength ≥1500MPa) and aluminum alloys (2000 series / 7000 series) and other difficult-to-deform materials (strong springback tendency and low plasticity), the elastic springback after demolding is effectively suppressed through multi-field coordination and dynamic feedback.

[0089] The final shape accuracy is significantly improved (e.g., ±5μm level), the product qualification rate is significantly improved compared with the traditional method, and the cost of manual correction and production cycle are effectively reduced.

[0090] Specifically, in this embodiment, the compensation force control of the multi-field cooperative execution module satisfies the following equation:

[0091] ;

[0092] in:

[0093] F_comp: Real-time compensation force (unit: N);

[0094] Δε: Real-time strain deviation (the difference between the target strain and the actual strain), measured by the strain sensor array of the closed-loop feedback module;

[0095] K_p: Proportional gain coefficient (unit: N / με), dynamically adjusted according to the material's yield strength;

[0096] K_i: Integral gain coefficient (unit: N / (με·s)), used to eliminate steady-state error;

[0097] α: Energy field decay factor (0<α<1), characterizing the degree of decay of thermal softening effect over time;

[0098] β: Energy field time constant (unit: s) -1 This is related to the thermal diffusivity of the material;

[0099] t: The duration of the energy field's action (unit: s).

[0100] Example scenario: When processing high-strength steel stamping parts with a tensile strength of 1800MPa, a local area with excessive springback Δε=0.005 (i.e. 5με) was detected.

[0101] Parameter configuration:

[0102] K_p=1200N / με (High stiffness requirement of high-strength steel);

[0103] K_i = 80 N / (με·s);

[0104] α = 0.7 (rapid thermal decay characteristics);

[0105] β=0.05s -1 ;

[0106] t=2s (the time the energy field has been in effect).

[0107] Calculation process:

[0108] 1. Proportional term: K_p·Δε=1200×0.005=6N;

[0109] 2. Integral term: K_i·∫(Δε)dt≈80×(0.005×2)=0.8N (assuming Δε is constant);

[0110] 3. Energy field decay term:

[0111] 1-α·e^(-β·t)=1-0.7·e^(-0.1)≈1-0.7×0.904=0.367;

[0112] 4. Real-time compensation force: F_comp = (6 + 0.8) × 0.367 ≈ 2.5 N

[0113] The parameter description table is as follows:

[0114] parameter Physical meaning Technical relevance Δε Real-time strain deviation The strain is directly measured by an array of strain sensors, reflecting the degree of local springback. K_p,K_i Dynamic gain coefficient Real-time optimization using machine learning models: high K_p (>1000) for high-strength steel, and low K_p (300-500) for aluminum alloys. α,β Energy field decay characteristics Determined by laser / high-frequency heating parameters: -α∝1 / laser power (power↑→α↓) -β∝electromagnetic frequency (frequency↑→β↑) 1-α· e^(-β·t) Energy field contribution weight Achieving a dynamic balance between rigid external force and thermal softening: - Initial stage (t→0): Thermal softening dominates (weight≈1-α) - Later stage (t↑): Rigid external force dominates (weight→1)

[0115] Technical effects:

[0116] 1. Avoid thermal damage: By limiting the long-term effect of the energy field through attenuation terms, the heat-affected zone is controlled to ≤0.1mm (≥0.3mm using traditional methods), thus preventing material grain coarsening.

[0117] 2. Prevent overcompensation: The PID (proportional + integral) method precisely matches the compensation force with the real-time rebound deviation, improving the shape accuracy to ±5μm.

[0118] 3. Material Adaptability: K_p and K_i are dynamically adjusted through a neural network trained on historical process data to adapt to the different springback characteristics of high-strength steel / aluminum alloys.

[0119] The working principle of this equation is as follows:

[0120] Dynamic adjustment logic:

[0121] 1. Initial stage (t=0):

[0122] When the energy field operates at full power (laser 5kW / high frequency 20kW), α·e^(-β·t)≈α→attenuation term≈1-α (smaller value);

[0123] Thermal softening is dominant: it reduces the yield strength of the material and assists in plastic deformation.

[0124] 2. Compensation during the middle period (t increases):

[0125] e^(-β·t)↓→Decrease term↑

[0126] Rigid external force gradually taken over: directional pressure (0.1-10N) is applied via a servo hydraulic cylinder to precisely counteract the force.

[0127] Rebound.

[0128] 3. Convergence phase (Δε≤5μm):

[0129] The integral term ∫(Δε)dt eliminates residual errors and ensures stability.

[0130] The equation, energy field decay factor α·e^(-β·t), solves the overheating problem caused by traditional heating by quantifying the time dependence of thermal softening effect; it strongly correlates K_p / K_i with the material and dynamically maps material properties through machine learning, replacing empirical parameter settings, which makes the parameter settings more accurate; it introduces the PID control concept (proportional + integral) into the field of rebound compensation and deeply integrates it with the physical model of energy field decay to achieve synergistic optimization of the three elements of "force-heat-time".

[0131] This equation transforms the multi-field collaborative mechanism and closed-loop feedback into a quantifiable and executable mathematical model, significantly improving the system's response speed and compensation accuracy.

[0132] This implementation also provides the following experimental data records:

[0133] I. Experimental Subjects and Background

[0134] Two typical precision stamped parts were selected as experimental subjects:

[0135] Specimen 1: Automobile frame structural component (similar) Figure 1 The outline is made of high-strength steel (tensile strength 1800MPa), and its geometric features include holes, long strip edges, and complex curved surfaces.

[0136] Specimen 2: Aluminum alloy structural component (similar) Figure 2 The model is a 3D model made of 7075 series aluminum alloy (tensile strength 570MPa), and its geometric features include slender bosses, multiple holes, and thin-walled areas.

[0137] II. Experimental Input Data (Distribution of Springback After Demolding)

[0138] Data on the surface rebound distribution of the specimens were acquired using a multispectral camera (color scale range -10.000 to 10.000, color gradient from light green to dark red). The results are as follows:

[0139] Specimen 1: The rebound amount is unevenly distributed overall. The green area (rebound amount ≈ 0) accounts for about 70%, and there are local areas with high values ​​of yellow (rebound amount ≈ 1.000) and red (rebound amount ≈ 3.000) (such as the edge of the hole and the end of the long strip), with the maximum value reaching -10.000 (dark purple area, accounting for about 2%).

[0140] Specimen 2: The springback amount is concentrated in the range of -3.000 to 3.000 (85% of the area is green to yellow). High red values ​​appear on the top of the slender boss and the thin-walled area (springback amount ≈ 2.000~3.000, accounting for about 10%), with a maximum value of -10.000 (dark purple dotted area, accounting for about 1%).

[0141] III. Experimental Processing Parameters

[0142] Based on the compensation strategy generated by the system calculation, the multi-field collaborative execution module applies the following parameters:

[0143] Rigid external force: Apply directional pressure to the high-value area (red / yellow area), with a magnitude of 0.5~8N (specimen 1) and 0.3~5N (specimen 2), in the opposite direction to the springback deformation direction;

[0144] Local energy field: The laser heating module (wavelength 1064nm, power 2~4kW) acts on high-value areas such as holes and protrusions, and the high-frequency electromagnetic induction heating unit (frequency 500kHz, power 10~15kW) acts on the thin-walled area of ​​aluminum alloy. The overlap between the area of ​​action and the area with excessive springback is ≥90%.

[0145] IV. Experimental Output Data (Rebound Amount After Processing)

[0146] After dynamically adjusting the parameters through the closed-loop feedback module, the final rebound distribution was significantly improved:

[0147] Specimen 1: The area percentage of the red / yellow high value zone (rebound amount > 1.000) decreased from 25% to < 3%, and the overall rebound amount was concentrated in -0.005~0.005 (green area), with a maximum value ≤ 0.010 (light yellow area).

[0148] Specimen 2: The red high-value area (rebound amount > 2.000) was completely eliminated, and the proportion of the yellow area (rebound amount ≈ 1.000) decreased from 10% to < 2%, with the overall rebound amount concentrated in the range of -0.003 to 0.003.

[0149] V. Verification of Experimental Results

[0150] Compare key metrics before and after processing to verify the system's effectiveness:

[0151] Shape accuracy: The dimensional deviation of key holes in specimen 1 decreased from ±8μm to ±3μm, and the height deviation of the boss in specimen 2 decreased from ±12μm to ±4μm;

[0152] Production efficiency: The time spent on manual calibration has been reduced from an average of 45 minutes per piece to 0 (fully automated processing), and the production cycle for a single batch has been shortened by 30%;

[0153] Product qualification rate: The qualification rate of test piece 1 increased from 92% to 99.2%, and the qualification rate of test piece 2 increased from 94% to 99.5%.

[0154] Material damage: Under the assistance of the energy field, the width of the heat-affected zone of specimen 1 is ≤0.1mm (≥0.3mm under traditional heating method), and specimen 2 does not show obvious thermal deformation or grain coarsening.

[0155] In summary, the hard springback treatment system and method for precision stamped parts provided in this embodiment achieves precise compensation for springback deformation of precision stamped parts through a closed-loop process of "detection-calculation-cooperative execution-feedback optimization". The core principle is as follows:

[0156] 1. Precise springback identification: Non-contact optical inspection devices (such as multispectral cameras) are used to collect springback distribution data on the surface of stamped parts. This data is calibrated using a color scale (-10.000 to 10.000) to transform the springback differences on complex surfaces into a visualized numerical matrix (e.g., ...). Figure 1 , Figure 2 (As shown in the gradient area from light green to dark red), covering the entire surface of tiny rebound areas (such as holes, edges, etc.).

[0157] 2. Intelligent calculation of compensation amount: Combining finite element simulation (simulating springback behavior based on material constitutive model) and machine learning (correcting neural network model trained on historical process data), it generates refined springback compensation amount for each region, making up for the error of single simulation (such as ignoring microscopic defects in materials) and improving the accuracy of compensation strategy.

[0158] 3. Synergistic effect of multiple fields: targeting areas with excessive rebound (such as...) Figure 1 , Figure 2 In the high-value red zone, a controllable rigid external force (directional pressure) and a local energy field (laser / high-frequency heating to soften the material) are applied simultaneously. The external force directly offsets the elastic rebound, and the energy field reduces the yield strength of the material to assist in plastic deformation, thus solving the problems of easy fracture under a single external force or uncontrollable heat-affected zone under a single energy field.

[0159] 4. Dynamic feedback optimization: By monitoring the deformation state through strain sensors (real-time acquisition of local strain) and optical trackers (scanning the overall morphology), the magnitude / direction of external forces or energy field parameters (such as laser power) are adjusted in real time to ensure that the compensation process is dynamically optimized according to the fluctuation of material properties, avoiding overcompensation or undercompensation.

[0160] How to use

[0161] The usage process of this solution can be summarized in the following steps:

[0162] 1. Post-demolding inspection: After demolding of precision stamped parts, a multispectral camera is used to non-contactly collect surface springback distribution data. This data is calibrated using a color scale (-10.000 to 10.000) to generate a visualized springback morphology map (e.g., ...). Figure 1 , Figure 2 As shown), identify areas with excessive rebound (such as high-value areas with a reddish / purple tint).

[0163] 2. Compensation Calculation: Based on the rebound morphology data, the initial compensation strategy is generated by finite element simulation of rebound behavior. Then, the simulation results are corrected by using a machine learning model (trained with historical process data) to output the refined compensation amount for each region (such as the magnitude of directional pressure and energy field parameters at different locations).

[0164] 3. Multi-field coordinated execution: Based on the compensation amount, controllable rigid external force (such as 0.1 to 10 N directional pressure applied by a servo hydraulic cylinder array, in the opposite direction to the rebound direction) and local energy field (such as laser wavelength 1064 nm heating or high frequency electromagnetic induction heating) are applied simultaneously to the area where the rebound exceeds the standard, forcing the material to undergo directional plastic deformation to counteract the rebound.

[0165] 4. Real-time feedback adjustment: Local strain data is collected through a strain sensor array (attached surface), and the overall morphology changes are monitored by an optical tracker (structured light scanning). If overcompensation or undercompensation trends are found (such as color distribution deviating from the target), the magnitude / direction of the external force or energy field parameters (such as laser power) are immediately dynamically adjusted until the rebound amount meets the precision requirements (such as shape accuracy reaching ±5μm level).

[0166] Technical effect

[0167] This solution systematically solves the technical challenge of springback compensation for precision stamped parts through multi-dimensional technological innovation. The overall effect is reflected in the following aspects:

[0168] 1. Precise Identification and Compensation: Non-contact optical detection combined with color-coded scale calibration enables high-resolution capture of minute differences in rebound across the entire surface (e.g., Figure 1 , Figure 2 Different colored areas correspond to different rebound amounts, providing a reliable data basis for subsequent compensation; the multi-field synergy (rigid external force + energy field) solves the limitations of a single mechanism, can adapt to the rebound compensation needs of complex geometries (such as long strip structures and holes), and avoids stress concentration or thermal damage.

[0169] 2. Strong dynamic adaptability: The closed-loop feedback mechanism effectively responds to fluctuations in material properties (such as anisotropy of high-strength steel and aging changes of aluminum alloys) through real-time monitoring and parameter adjustment, ensuring the stability of the compensation process and avoiding the effect deviation caused by material fluctuations in traditional "one-time compensation".

[0170] 3. Improve production efficiency and quality: Full-process automation (detection-calculation-execution-feedback) replaces the traditional manual correction process, significantly shortening the production cycle; the shape accuracy and dimensional consistency after compensation are greatly improved (such as reducing ±5μm level deviation), and the product qualification rate is increased from the traditional ≤95% to a higher level, reducing labor costs and material waste.

[0171] 4. Good process versatility: It is suitable for springback processing of various precision stamping materials such as high-strength steel and aluminum alloys, as well as complex geometries (such as automobile frames and mechanical structural parts), thus expanding the application scenarios of precision stamping processes.

[0172] The above are merely preferred embodiments of the present invention and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should recognize that any equivalent substitutions and obvious changes made based on the description and illustrations of the present invention should be included within the protection scope of the present invention.

Claims

1. A hard springback treatment system for precision stamped parts, characterized in that, include: The springback detection module is configured to collect surface springback distribution data of precision stamped parts after demolding. The springback is obtained through a non-contact optical detection device with a detection accuracy of not less than ±0.001mm. The data processing module is communicatively connected to the rebound amount detection module and is configured to generate a three-dimensional rebound morphology model based on the rebound amount distribution data, and calculate the rebound compensation amount of each region based on preset target shape parameters. The multi-field collaborative execution module is communicatively connected to the data processing module and is configured to synchronously apply controllable rigid external force and / or local energy field according to the springback compensation amount of each region, forcing the stamped part to undergo directional plastic deformation to counteract elastic springback; The closed-loop feedback module is configured to monitor the deformation state of the stamped part in real time, feed the monitoring data back to the data processing module, and dynamically adjust the magnitude and direction of the rigid external force or the parameters of the energy field. The springback detection module includes a multispectral camera and an image processing unit. The multispectral camera is used to collect the springback characteristic spectrum of the stamped part surface. The image processing unit is calibrated with a color scale bar, with the color range from light green to dark red, corresponding to a springback of -10.000 to 10.000, and converts the spectral data into a springback numerical matrix. The data processing module includes a finite element simulation unit and a machine learning compensation unit; The finite element simulation unit is used to simulate the springback behavior of stamped parts based on the material constitutive model and generate an initial compensation strategy. The machine learning compensation unit modifies the initial compensation strategy using a neural network model trained on historical process data, and outputs refined compensation amounts for each region. The multi-field collaborative execution module includes a rigid external force application unit and a local energy field generation unit; the rigid external force application unit includes a servo hydraulic cylinder array and a precision pressure sensor, used to apply a directional pressure of 0.1 to 10 N to a designated area of ​​the stamped part; The local energy field generating unit includes a laser heating module and a high-frequency electromagnetic induction heating unit, which are used to reduce the yield strength of the material through the thermal softening effect and assist plastic deformation. The wavelength of the laser heating module is 1064nm and the power is 0.5 to 5kW; the frequency of the high-frequency electromagnetic induction heating unit is 100kHz to 1MHz and the power is 2 to 20kW. The compensating force control of the multi-field collaborative execution module satisfies the following equation: F_comp=[K_p·Δε+K_i·∫(Δε)dt]·(1-α·e^(-β·t)); in: F_comp: Real-time compensation force, in N; Δε: Real-time strain deviation, measured by the strain sensor array of the closed-loop feedback module; K_p: Proportional gain coefficient, in units of N / με, dynamically adjusted according to the material's yield strength; K_i: Integral gain coefficient, in units of N / (με·s), used to eliminate steady-state error; α: Energy field decay factor, 0 < α < 1, characterizing the degree of decay of thermal softening effect over time; β: Energy field time constant, in seconds -1 It is related to the thermal diffusivity of the material; t: The duration of the energy field's action, measured in seconds.

2. The hard springback treatment system for precision stamped parts according to claim 1, characterized in that, The closed-loop feedback module includes a strain sensor array and an optical tracker; the strain sensor array is attached to the surface of the stamped part to collect local strain data in real time; the optical tracker monitors the overall morphological changes of the stamped part through structured light scanning, and the response time of the dynamic adjustment is ≤100ms.

3. A method for hard springback treatment of precision stamped parts, applied to the hard springback treatment system for precision stamped parts as described in any one of claims 1-2, characterized in that, Includes the following steps: S1: After the precision stamping part is demolded, the surface springback distribution data is collected by a non-contact optical detection device. The springback is marked by a color scale bar, with the color range from light green to dark red, corresponding to a springback of -10.000 to 10.

000. S2: Generate a three-dimensional rebound morphology model based on the rebound distribution data, and calculate the rebound compensation amount of each region in combination with the target shape parameters; S3: Based on the springback compensation amount of each region, a controllable rigid external force and / or local energy field are applied simultaneously to force the stamped part to undergo directional plastic deformation to counteract elastic springback; S4: Monitor the deformation state of the stamped part in real time, feed the monitoring data back to step S2, and dynamically adjust the magnitude and direction of the rigid external force or the parameters of the energy field until the springback is ≤ ±5μm.

4. The method for hard springback treatment of precision stamped parts according to claim 3, characterized in that, The non-contact optical detection device mentioned in step S1 includes a multispectral camera, and the calibration range of the color scale bar covers the rebound amount from -10.000 to 10.000, with a calibration accuracy of ±0.

001.

5. The method for hard springback treatment of precision stamped parts according to claim 3, characterized in that, The construction of the three-dimensional springback morphology model in step S2 is based on finite element simulation and machine learning correction. The finite element simulation simulates springback behavior based on the material constitutive model, and the machine learning correction compensates the simulation results through a neural network model trained on historical process data, with a compensation error ≤3%.

6. The method for hard springback treatment of precision stamped parts according to claim 3, characterized in that, The magnitude of the controllable rigid external force in step S3 is 0.1 to 10 N, and its direction is opposite to the direction of the rebound deformation; the parameters of the local energy field include laser power of 0.5 to 5 kW, high-frequency electromagnetic induction frequency of 100 kHz to 1 MHz, and the overlap between the area of ​​action and the area where the rebound exceeds the standard is ≥90%.

7. The method for hard springback treatment of precision stamped parts according to claim 3, characterized in that, The precision stamped parts are made of high-strength steel or aluminum alloy, and the shape accuracy after processing reaches ±5μm. The tensile strength of the high-strength steel is ≥1500MPa; the aluminum alloy is 2000 series or 7000 series.

Images