A method for detecting cracks in a super high strength steel plate stamping for a new energy vehicle

Abstract

The application provides a kind of dark crack detection method for ultra-high strength steel plate stamping part of new energy vehicle, relates to stamping part detection technical field, comprising the following steps: obtaining the detection area and reference area of the stamping part to be measured;A group of same multi-frequency alternating excitation signals are injected into the detection area and reference area of the stamping part to be measured at the same time, and the first electrical response signal of the detection area under the multi-frequency alternating excitation signal and the second electrical response signal of the reference area under the multi-frequency alternating excitation signal are synchronously collected;Based on the first electrical response signal and the second electrical response signal, the conductivity difference of the detection area and the reference area under the same frequency point is calculated, and the conductivity difference-frequency curve including multiple frequency points is obtained;Based on the conductivity difference-frequency curve, it is judged whether there is a dark crack in the detection area.The detection method can realize quantitative evaluation on the degree of defect, meet the stringent requirements of new energy vehicle industry on quality consistency.

Description

Technical Field

[0001] This application relates to the field of stamping parts inspection technology, specifically to a method for detecting dark cracks in ultra-high strength steel plate stamping parts used in new energy vehicles. Background Technology

[0002] With the increasing demands for lightweighting and safety performance in new energy vehicles, ultra-high strength steel (UHSS) is being used more and more widely in vehicle body structural components. These parts are typically manufactured using a stamping process, during which the material undergoes severe plastic deformation, making it highly susceptible to developing microscopic cracks or localized thinning—known as "hidden cracks"—in areas of stress concentration. These hidden crack defects significantly reduce the fatigue strength and impact toughness of the parts, posing a potential threat to vehicle safety.

[0003] Currently, in industrial production, the detection of dark cracks in stamped parts still heavily relies on traditional methods, such as offline sampling for destructive testing (e.g., metallographic sectioning) or visual inspection by operators based on experience. These methods have significant limitations: destructive testing cannot achieve 100% inspection, is inefficient and costly; while manual visual inspection is highly subjective, has a high rate of missed detections, and is difficult to quantify the degree of defects, failing to meet the stringent quality consistency requirements of the new energy vehicle industry. Summary of the Invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, this application aims to provide a method for detecting hidden cracks in ultra-high strength steel plate stamping parts for new energy vehicles, so as to quantitatively assess the degree of defects and meet the stringent quality consistency requirements of the new energy vehicle industry; the detection method includes the following steps:

[0005] Obtain the inspection area and reference area of ​​the stamping part to be tested;

[0006] Simultaneously, a set of identical multi-frequency AC excitation signals are injected into the detection area and the reference area of ​​the stamping part under test, and the first electrical response signal of the detection area under the multi-frequency AC excitation signal and the second electrical response signal of the reference area under the multi-frequency AC excitation signal are acquired synchronously.

[0007] Based on the first electrical response signal and the second electrical response signal, the conductivity difference between the detection area and the reference area at the same frequency point is calculated to obtain a conductivity difference-frequency curve including multiple frequency points;

[0008] Based on the conductivity difference-frequency curve, it is determined whether there are dark cracks in the detection area.

[0009] According to the technical solution provided in this application, determining whether a dark crack exists in the detection area based on the conductivity difference-frequency curve includes the following steps:

[0010] Calculate the maximum absolute value of the conductivity difference within the first preset high-frequency band of the conductivity difference-frequency curve;

[0011] If the maximum absolute value of the conductivity difference is greater than or equal to the first preset threshold, then the conductivity difference corresponding to all the frequency points in the first preset high frequency band is extracted from the conductivity difference-frequency curve to form a dataset to be analyzed.

[0012] Calculate the arithmetic mean of all the conductivity differences in the dataset to be analyzed, and denot it as the average offset.

[0013] If the average offset is less than a preset defect threshold, it is determined that there is a dark crack in the detection area.

[0014] According to the technical solution provided in this application, after calculating the maximum value of the absolute value of the conductivity difference within the first preset high-frequency band of the conductivity difference-frequency curve, the method further includes the following steps:

[0015] If the maximum absolute value of the conductivity difference is less than the first preset threshold, then the phase difference of the AC signal between the detection area and the reference area at the same frequency point is calculated based on the first electrical response signal and the second electrical response signal.

[0016] A phase difference-frequency curve is generated based on the multiple frequency points and their corresponding AC signal phase differences;

[0017] Calculate the slope value of the phase difference-frequency curve within the second preset high-frequency band;

[0018] If the absolute value of the fitted slope is greater than the preset slope threshold, it is determined that there is a dark crack in the detection area.

[0019] According to the technical solution provided in this application, obtaining the detection area and reference area of ​​the stamping part to be tested includes the following steps:

[0020] A three-dimensional digital model of the stamping part to be tested is obtained. Based on the stamping forming simulation software, the forming process of the stamping part to be tested is simulated by computer to obtain the stamping simulation results.

[0021] From the stamping simulation results, the maximum principal strain value and thickness reduction rate of all regions on the stamped part under test after stamping are extracted;

[0022] The region that simultaneously satisfies that the maximum principal strain value is less than a preset strain threshold and the thickness reduction rate is less than a preset reduction threshold is determined as a candidate region.

[0023] From all the candidate regions, select the region that is physically far away from all potential defect detection regions and determine it as the baseline region.

[0024] According to the technical solution provided in this application, the potential defect detection area is the area on the stamping part to be tested where the maximum principal strain value is greater than or equal to the preset strain threshold, and / or the area where the thickness reduction rate is greater than or equal to the preset thinning threshold; all the potential defect detection areas constitute a potential defect detection area set;

[0025] Calculate the physical distance between the geometric center point of each candidate region and the geometric center point of each potential defect detection region to obtain multiple sets of distance values ​​corresponding to each candidate region;

[0026] For each candidate region, the minimum value is selected from the multiple sets of distance values ​​corresponding to it, and this value is taken as the minimum safe distance between the candidate region and the set of potential defect detection regions.

[0027] The minimum safe distance of all candidate regions is traversed, and the candidate region with the largest minimum safe distance is determined as the reference region.

[0028] According to the technical solution provided in this application, after obtaining the detection area and reference area of ​​the stamping part to be tested, the method further includes the following steps:

[0029] Verify that the reference area is free of defects;

[0030] The simultaneous injection of a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test includes the following steps:

[0031] If so, then simultaneously inject a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test.

[0032] According to the technical solution provided in this application, verifying whether the reference area is free of defects includes the following steps:

[0033] Select at least three verification points within and / or at the edge of the reference area;

[0034] Using the multi-frequency AC excitation signal, at least three of the verification points are scanned respectively, and the third electrical response signal of each verification point is collected;

[0035] Calculate the conductivity difference between any two verification points at the same frequency point to obtain multiple sets of verification difference-frequency curves;

[0036] If the fluctuation amplitude of all the verification difference-frequency curves within the preset frequency range is less than the preset verification threshold, then the verification result of the benchmark region is determined to be defect-free.

[0037] According to the technical solution provided in this application, the stamping part to be tested is the first piece after mold change or the first piece of a batch;

[0038] After determining whether there is a dark crack in the detection area based on the conductivity difference-frequency curve, the method further includes the following steps:

[0039] If a dark crack is found in the detection area of ​​the first piece or the first piece of the batch after the mold change, a mold repair and inspection prompt is generated, and the defect data of the dark crack defect is bound to the mold number of the mold used in the current production and stored in the defect database; the defect data includes the morphology of the dark crack, the location of the dark crack, and the characteristic data of the corresponding conductivity difference-frequency curve.

[0040] The defect database configuration is used to establish a mapping relationship library between dark crack features and mold damage types based on the historically stored multiple mold numbers and their associated defect data;

[0041] If a dark crack is detected again with a feature data similarity exceeding a preset similarity to a historical defect data in the mapping relationship library, the corresponding mold damage type and suggested repair location will be output.

[0042] According to the technical solution provided in this application, before simultaneously injecting a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part to be tested, the method further includes the following steps:

[0043] Obtain the material grade of the stamped part to be tested;

[0044] The excitation parameter list is retrieved and iterated to obtain the excitation parameters corresponding to the material grade of the stamping part to be tested, and these parameters are used as the target excitation parameters. The excitation parameters include frequency combination and voltage amplitude.

[0045] The excitation parameter list includes multiple optimal excitation parameters corresponding to different material grades. The optimal excitation parameters are determined through pre-experimentation, and the standard is to make the signal-to-noise ratio of the standard defect-free test block of the material corresponding to the material grade reach the optimal level.

[0046] The simultaneous injection of a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test includes the following steps:

[0047] Simultaneously, a set of multi-frequency AC excitation signals under the same target excitation parameters are injected into the detection area and the reference area of ​​the stamping part under test.

[0048] According to the technical solution provided in this application, obtaining the detection area and reference area of ​​the stamping part to be tested includes the following steps:

[0049] Query the defect database to obtain all the hidden crack location data of historical stamped parts produced with the same or similar molds as the stamped part to be tested, and form a set of empirical high-risk areas from all the locations of the hidden cracks.

[0050] The region obtained by taking the intersection of the set of potential defect detection regions and the set of empirical high-risk regions is taken as the detection region.

[0051] Compared with the prior art, the beneficial effects of this application are as follows:

[0052] I. Achieving efficient and accurate non-destructive online testing: By synchronously injecting the same multi-frequency AC excitation signal into the testing area and the reference area and acquiring the electrical response, a measurement can be completed within seconds. It is very suitable for integration into the production line to achieve fully automated online testing, which greatly improves the testing efficiency and avoids the lag of traditional offline testing.

[0053] II. Ultra-high sensitivity to micron-level dark cracks: Dark cracks in ultra-high strength steel cause subtle changes in its local electrical conductivity. This method, by constructing conductivity difference-frequency curves at multiple frequency points, can amplify and capture these subtle changes in electrical properties. It has a detection sensitivity for early, small dark crack defects (such as slight necking) that far exceeds that of manual visual inspection and traditional eddy current methods, greatly reducing the risk of missed detection.

[0054] Third, strong anti-interference ability and stable and reliable results: Real-time synchronous comparison is carried out using a reference area on the same part, which effectively eliminates the influence of common-mode interference caused by material batch fluctuations, changes in ambient temperature, and drift of excitation signals, making the test results more stable and reliable.

[0055] Fourth, it provides a data foundation for intelligent manufacturing and predictive maintenance: The conductivity difference-frequency curve output by this method is a quantified and storable digital signal, establishing a quality data archive for each product. This provides crucial data support for subsequent big data analysis, tracing the root causes of defects, and even realizing predictive maintenance of molds, aligning with the development direction of intelligent manufacturing in new energy vehicles. Attached Figure Description

[0056] Figure 1 The flowchart illustrates the steps of the method for detecting dark cracks in ultra-high strength steel plate stamped parts for new energy vehicles provided in this application. Detailed Implementation

[0057] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0058] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0059] Example 1

[0060] As mentioned in the background section, this application proposes a method for detecting dark cracks in ultra-high strength steel plate stamping parts for new energy vehicles, addressing the problems in the existing technology. Figure 1 As shown, it includes the following steps:

[0061] S1. Obtain the detection area and reference area of ​​the stamping part to be tested;

[0062] S2. Simultaneously inject a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part to be tested, and synchronously collect the first electrical response signal of the detection area under the multi-frequency AC excitation signal and the second electrical response signal of the reference area under the multi-frequency AC excitation signal.

[0063] Specifically, a multi-channel eddy current detector or a dedicated multi-frequency eddy current testing system is used. This system generates a set of pre-set multi-frequency AC signals. Two identical detection probes are used, one tightly coupled to the surface of the detection area and the other tightly coupled to the surface of the reference area. The key is to apply the excitation "simultaneously" and acquire the signals "synchronously." This eliminates common errors caused by factors such as ambient temperature drift and instrument instability, ensuring the validity of subsequent comparisons. The response signal acquired by the probes is a voltage signal containing complex information (real and imaginary parts, or amplitude and phase).

[0064] S3. Based on the first electrical response signal and the second electrical response signal, calculate the conductivity difference between the detection area and the reference area at the same frequency point to obtain a conductivity difference-frequency curve including multiple frequency points;

[0065] Specifically, the instrument's internal processor or host computer software processes the acquired raw voltage signal. Using a pre-calibrated algorithm model (e.g., based on impedance plane analysis or neural network inversion algorithm), the response signal at each frequency point is converted into the equivalent conductivity value of that region at that frequency. Then, for each frequency point Fi, the following calculation is made: Δσ(Fi) = σdetection(Fi) - σreference(Fi). Connecting all frequency points and their corresponding Δσ values ​​forms the conductivity difference-frequency curve.

[0066] S4. Based on the conductivity difference-frequency curve, determine whether there are dark cracks in the detection area.

[0067] Specifically, in the absence of defects, the material properties of the inspection area and the reference area should be very similar, the Δσ value should be close to 0 at all frequencies, and the curve should be a flat straight line close to zero with only minor random fluctuations. Once a hidden crack exists, the crack disrupts the distribution of the eddy current field, causing a change in the equivalent conductivity of the region. This change manifests differently at different frequencies, causing the Δσ value to deviate from 0 and exhibiting specific characteristic patterns on the curve (e.g., severe fluctuations or overall shifts at high frequencies). By analyzing these characteristic patterns, the presence of a hidden crack can be determined.

[0068] Specifically, the stamped parts to be tested are ultra-high strength steel plate stamped parts for new energy vehicles: These refer to components used in new energy vehicles (such as body structural parts, anti-collision parts, battery pack housings, etc.) obtained by applying pressure to ultra-high strength steel plates (usually steel grades with a tensile strength of over 1000 MPa) using a die to induce plastic deformation. They are characterized by high strength and light weight, but defects are easily generated during the stamping process due to uneven material flow and die wear. Hidden cracks: These are microscopic or submicroscopic crack defects, characterized by not penetrating the workpiece thickness and being located below the material surface or having extremely small surface openings, making them difficult to detect through conventional visual observation, optical inspection, or even some non-destructive testing methods (such as certain low-sensitivity eddy current testing). These cracks pose a potential safety hazard and may propagate during subsequent use, leading to sudden component failure. Inspection area: This refers to one or more localized areas on the stamped part to be tested that are suspected of having hidden cracks and require focused inspection. These areas are usually predetermined based on stamping simulation or historical defect data. Reference Region: This refers to the region, after analysis, that is confirmed to be highly defect-free and serves as a health reference standard. The selection of this region is crucial; its electrical response signal will be used as a benchmark for comparison with the signal from the detection region. Multi-Frequency AC Excitation Signal: This refers to a combination of sinusoidal AC signals containing multiple different frequencies (e.g., from tens of kHz to several MHz). The reason for using multiple frequencies is that defects at different depths or material properties respond differently to electromagnetic fields at different frequencies; comprehensively utilizing multi-frequency information can improve the comprehensiveness and accuracy of the detection. Electrical Response Signal: This refers to the signal reflecting the electromagnetic properties of the material, measured by a sensor (such as an eddy current probe) after the AC excitation signal is injected. It typically includes the amplitude and phase information of the signal, which are directly related to the material's intrinsic properties such as electrical conductivity and magnetic permeability. Conductivity Difference-Frequency Curve: This is a curve plotting the difference between the calculated conductivity of the detection region and the calculated conductivity of the reference region at various frequency points, as a function of frequency. The presence of dark cracks will change the equivalent conductivity of this region, thus exhibiting a characteristic on this curve.

[0069] This implementation achieves highly sensitive and quantitative detection of invisible surface cracks. Multi-frequency scanning acquires richer defect information; differential comparison significantly suppresses common-mode interference, improving the signal-to-noise ratio and detection reliability. This method is highly automated and suitable for rapid online or offline inspection on production lines. The underlying technical principles are eddy current effect and differential measurement. An AC excitation coil generates an alternating magnetic field, inducing eddy currents in the conductive workpiece. The amplitude and distribution of the eddy currents are affected by the material's conductivity. A hidden crack is equivalent to introducing a high-resistance barrier in the conductive path, hindering eddy current flow and altering the eddy current field distribution, thus being detected by the induction coil as an impedance change. Differential measurement cleverly compares the signal in the detection area with the signal in the reference area in real time, directly amplifying and extracting the minute changes caused by the defect, while canceling out interference signals generated by inherent material properties, lift-off effects, temperature changes, and other common factors, thereby significantly improving detection sensitivity.

[0070] In a preferred embodiment, determining whether a dark crack exists within the detection area based on the conductivity difference-frequency curve includes the following steps:

[0071] Calculate the maximum absolute value of the conductivity difference within the first preset high-frequency band of the conductivity difference-frequency curve;

[0072] If the maximum absolute value of the conductivity difference is greater than or equal to the first preset threshold, then the conductivity difference corresponding to all the frequency points in the first preset high frequency band is extracted from the conductivity difference-frequency curve to form a dataset to be analyzed.

[0073] Calculate the arithmetic mean of all the conductivity differences in the dataset to be analyzed, and denot it as the average offset.

[0074] If the average offset is less than a preset defect threshold, it is determined that there is a dark crack in the detection area.

[0075] Specifically, the first preset high-frequency band refers to a portion of the frequency band with higher frequencies (e.g., above 500kHz) within the multi-frequency excitation signal used. This high-frequency band is chosen because high-frequency eddy currents exhibit a significant skin effect, shallow penetration depth, and exceptional sensitivity to micro-cracks on and near the material surface. The maximum absolute value of the conductivity difference refers to the maximum value among the absolute values ​​of the conductivity differences at all frequency points within the specified high-frequency band. This value reflects the maximum deviation in conductivity difference between the detection area and the reference area. The dataset to be analyzed is the set of original conductivity differences (with signs) corresponding to all frequency points within the first preset high-frequency band from the previous step. The average offset is the arithmetic mean of all data in the dataset to be analyzed. It reflects the overall systematic deviation of the conductivity of the detection area relative to the reference area across the entire high-frequency band.

[0076] The specific implementation process is described below: The system software extracts the frequency from the generated conductivity difference-frequency curve, starting from F... high1 To F high2 The software first sets the high-frequency band data. It iterates through this band, taking the absolute value (|Δσ|) of the Δσ value at each point, and then finds the largest value, denoted as Max|Δσ|. Max|Δσ| is compared with a pre-defined first threshold (e.g., 0.5% IACS) calibrated through numerous experiments. If Max|Δσ| ≥ the first threshold, a large conductivity difference (positive or negative) at a high-frequency point indicates a strong local deviation between the electromagnetic response of the detection area at that specific frequency and the reference area. This is a strong anomalous signal. We need to determine whether it is a dangerous hidden crack (cause A) or a relatively harmless material inhomogeneity (cause B). The average offset is used as the criterion for this distinction. At this point, the software extracts the original Δσ values ​​(not absolute values) corresponding to all frequency points within the high-frequency band, forming a data array, i.e., the "dataset to be analyzed". The software performs a simple arithmetic mean calculation on all values ​​in the dataset to be analyzed and compares the arithmetic mean with another pre-set defect threshold (e.g., 0.1% IACS). If the arithmetic mean is less than the preset defect threshold, then a hidden crack is identified. The logic here is that a hidden crack is a very localized damage that causes strong local eddy current disturbances at certain high-frequency points (leading to a large Max|Δσ|). However, because it does not cause large-area material changes (such as inhomogeneous alloy composition or differences in heat treatment), the overall conductivity of the entire region does not change systematically (the arithmetic mean is very small). If it is a problem such as inhomogeneous material, it usually manifests as an overall shift (the arithmetic mean will also be larger).

[0077] Furthermore, after calculating the maximum absolute value of the conductivity difference within the first preset high-frequency band of the conductivity difference-frequency curve, the method further includes the following steps:

[0078] If the maximum absolute value of the conductivity difference is less than the first preset threshold, then the phase difference of the AC signal between the detection area and the reference area at the same frequency point is calculated based on the first electrical response signal and the second electrical response signal.

[0079] A phase difference-frequency curve is generated based on the multiple frequency points and their corresponding AC signal phase differences;

[0080] Calculate the slope value of the phase difference-frequency curve within the second preset high-frequency band;

[0081] If the absolute value of the fitted slope is greater than the preset slope threshold, it is determined that there is a dark crack in the detection area.

[0082] Specifically, AC signal phase difference: refers to the phase angle difference between the response signal induced in the detection area and the response signal in the reference area under the same frequency excitation. Phase information is highly sensitive to changes in the microstructure of materials (such as cracks). Phase difference-frequency curve: refers to the curve plotted as the calculated phase difference changes with frequency at various frequency points. Second preset high-frequency band: may be the same as the first preset high-frequency band, or it may be another high-frequency band optimized separately for phase analysis. Fitting slope value: the slope k of the straight line equation obtained by linearly fitting (such as the least squares method) the data points of the "phase difference-frequency curve" within the second preset high-frequency band. This slope value represents the rate of change of phase difference with frequency. Preset slope threshold: a critical value used to judge whether the slope is abnormal, determined through statistical analysis of defect-free samples.

[0083] The specific implementation process is described below: When Max|Δσ| < the first preset threshold, the conductivity difference does not show any particularly prominent peaks throughout the high-frequency band. From the perspective of amplitude / conductivity, the detection area and the reference area look similar. If only conductivity is used for judgment, it may lead to the conclusion of no defects. The absence of a strong amplitude signal does not mean 100% safety. There may be a more hidden defect: fine or closed dark cracks: some very fine, surface-closed, or slightly deeper cracks. They may not have a sufficiently large amplitude to disrupt eddies (amplitude / conductivity change is not significant), but they will significantly change the path and speed of eddies propagating in the material, thus affecting the phase of the signal. Phase information is extremely sensitive to changes in the microstructure of the material. When the amplitude signal is not strong, we urgently need to switch to the new, more sensitive dimension of phase to look for clues of defects. At this point, the system software extracts the phase angles θ_detect(Fi) and θ_reference(Fi) of the two signals at each frequency point from the stored original first and second electrical response signals (which are complex signals). Then, it calculates the phase difference at each frequency point: Δθ(Fi) = θ_detect(Fi) - θ_reference(Fi). The software plots a new curve with frequency as the x-axis and the calculated Δθ as the y-axis. The software selects a second preset high-frequency band and performs linear regression analysis on the [Fi, Δθ(Fi)] data points within this band to obtain a fitted straight line, and records the slope k of this line. The absolute value of the slope k, |k|, is calculated and compared with a preset slope threshold K_th. If |k| > K_th, it is determined that a dark crack exists in the detection area.

[0084] The principle behind this judgment lies in the fact that cracks alter the propagation path and energy dissipation mechanism of eddy currents within a material, thus affecting the phase delay of the electromagnetic field. For a healthy, homogeneous material, the phase change with frequency is gradual and predictable, with a small slope in the phase difference-frequency curve at high frequencies. When a dark crack exists, even if it doesn't cause a sufficiently large change in conductivity amplitude, it introduces additional inductive or capacitive effects (depending on the crack morphology), significantly altering the propagation characteristics of the electromagnetic field in the defect region. This change typically manifests as an increased rate of phase change with frequency, meaning the slope of the phase difference-frequency curve at high frequencies increases significantly (or decreases, in absolute terms). Therefore, even with small amplitude differences, an abnormal phase slope is sufficient to reveal the presence of a dark crack. This is equivalent to a comprehensive analysis of the same signal from both the amplitude and phase domains, ensuring the comprehensiveness of the detection.

[0085] In a preferred embodiment, obtaining the detection area and reference area of ​​the stamped part to be tested includes the following steps:

[0086] A three-dimensional digital model of the stamping part to be tested is obtained. Based on the stamping forming simulation software, the forming process of the stamping part to be tested is simulated by computer to obtain the stamping simulation results.

[0087] Specifically, engineers retrieve the 3D model file of the stamped part from the Product Data Management (PDM) system and import it into the stamping simulation software. Then, they set up a complete simulation environment in the software: defining the rigid body motion of the dies (punch, die, blank holder), specifying the material grade of the sheet metal (the software's material library must contain the material's precise performance parameters), and setting process parameters such as the friction coefficient and blank holder force. Finally, they submit the calculation task, which the software then solves.

[0088] From the stamping simulation results, the maximum principal strain value and thickness reduction rate of all regions on the stamped part under test after stamping are extracted;

[0089] Specifically, the maximum principal strain value refers to the maximum tensile strain value generated at a certain point in the sheet metal during the stamping process. It is an important formability indicator; excessive principal strain can lead to material thinning or even cracking. The thickness reduction rate refers to the ratio of the sheet metal's thickness at a certain point after stamping to its original thickness, usually expressed as a percentage. Thinning rate (%) = (1 - thickness after stamping / original thickness) * 100%. Severe thickness reduction is a precursor to cracking.

[0090] The region that simultaneously satisfies that the maximum principal strain value is less than a preset strain threshold and the thickness reduction rate is less than a preset reduction threshold is determined as a candidate region.

[0091] From all the candidate regions, select the region that is physically far away from all potential defect detection regions and determine it as the baseline region.

[0092] Furthermore, the potential defect detection area is the region on the stamping part under test where the maximum principal strain value is greater than or equal to the preset strain threshold, and / or the region where the thickness reduction rate is greater than or equal to the preset thinning threshold; all the potential defect detection areas constitute a potential defect detection area set;

[0093] Calculate the physical distance between the geometric center point of each candidate region and the geometric center point of each potential defect detection region to obtain multiple sets of distance values ​​corresponding to each candidate region;

[0094] For each candidate region, the minimum value is selected from the multiple sets of distance values ​​corresponding to it, and this value is taken as the minimum safe distance between the candidate region and the set of potential defect detection regions.

[0095] The minimum safe distance of all candidate regions is traversed, and the candidate region with the largest minimum safe distance is determined as the reference region.

[0096] Specifically, the system software obtains the coordinate information (usually meshed) of all candidate regions and potential defect detection regions from the simulation results. For each candidate region, the 3D coordinates (Xc, Yc, Zc) of its geometric center point are calculated. Then, each region in the set of potential defect detection regions is traversed, and the coordinates of its geometric center point are calculated. The distance between these two center points is calculated using the 3D spatial distance formula: Distance = In this way, each candidate region will obtain N distance values ​​corresponding to N risk regions. For the N distance values ​​obtained for each candidate region in the previous step, the smallest value is found by sorting or simply comparing them. This minimum value, D_min, is the minimum safe distance for that candidate region. It means that the nearest risk point to this candidate region is also outside of D_min. After all candidate regions have calculated their own D_min, the system compares all these D_min values. The candidate region with the largest D_min value is selected. That is, the "safe region farthest from the danger region among all safe regions" is selected. This region is finally determined as the baseline region.

[0097] The principle of this implementation is based on the spatial correlation of stamping defects. High-risk areas (such as strain concentration zones) are not only prone to failure themselves, but their surrounding areas may also be affected and in a sub-healthy state (e.g., with residual stress or microstructural changes). Therefore, the farther the reference area is from these high-risk areas, the higher its health. By quantifying the distance and finding its maximum value, the quality of the reference area is ensured.

[0098] In a preferred embodiment, after obtaining the detection area and reference area of ​​the stamped part to be tested, the method further includes the following steps:

[0099] Verify that the reference area is free of defects;

[0100] The simultaneous injection of a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test includes the following steps:

[0101] If so, then simultaneously inject a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test.

[0102] Furthermore, verifying whether the reference area is free of defects includes the following steps:

[0103] Select at least three verification points within and / or at the edge of the reference area;

[0104] Using the multi-frequency AC excitation signal, at least three of the verification points are scanned respectively, and the third electrical response signal of each verification point is collected;

[0105] Calculate the conductivity difference between any two verification points at the same frequency point to obtain multiple sets of verification difference-frequency curves;

[0106] If the fluctuation amplitude of all the verification difference-frequency curves within the preset frequency range is less than the preset verification threshold, then the verification result of the benchmark region is determined to be defect-free.

[0107] Specifically, the verification difference-frequency curve refers to the conductivity difference-frequency curve calculated between any two verification points during the verification phase. Its generation method is exactly the same as the curve in the main detection, only the comparison object has changed. The fluctuation amplitude refers to the degree to which the verification difference-frequency curve deviates from the zero line within a preset frequency range, which can be measured using statistical measures such as standard deviation and range. The preset verification threshold is a very small tolerance value because we are comparing theoretically identical regions. Based on the shape and size of the reference region, the system automatically selects at least three points distributed within and around its edges (e.g., one at the center and two at the diagonal edges). The detection probe moves sequentially to each verification point, injects the same multi-frequency AC excitation signal, and synchronously acquires the third electrical response signal at each point. The conductivity difference between any two points at the same frequency is calculated. For three points, C(3,2) = three sets of verification difference-frequency curves can be calculated. The fluctuation of all three sets of curves within a preset frequency range (usually the entire frequency range or the main frequency range) is analyzed. If all the data points of the curves are closely distributed around the zero point, and their fluctuation amplitude (such as the maximum absolute value of all data points) is less than a very strict preset verification threshold (this threshold is much smaller than the first preset threshold in the main detection), it indicates that the electrical characteristics between these three points are highly consistent, and the entire reference area is uniform and defect-free. If the fluctuation of any set of curves exceeds the tolerance, the verification fails.

[0108] This implementation takes into account that simulation is ultimately a theoretical prediction, and actual production may involve unsimulated factors (such as localized defects in the sheet metal, uneven lubrication, and unexpected mold wear), causing problems in theoretically safe areas. The verification process can capture such unexpected situations, preventing misjudgments or missed detections caused by using a flawed benchmark to inspect problematic areas. This improves the robustness and reliability of the entire inspection system, enabling it to rely not only on pre-simulation but also to form a closed-loop quality control system of prediction-verification-execution.

[0109] In a preferred embodiment, the stamped part to be tested is the first piece after mold change or the first piece of a batch;

[0110] After determining whether there is a dark crack in the detection area based on the conductivity difference-frequency curve, the method further includes the following steps:

[0111] If a dark crack is found in the detection area of ​​the first piece or the first piece of the batch after the mold change, a mold repair and inspection prompt is generated, and the defect data of the dark crack defect is bound to the mold number of the mold used in the current production and stored in the defect database; the defect data includes the morphology of the dark crack, the location of the dark crack, and the characteristic data of the corresponding conductivity difference-frequency curve.

[0112] The defect database configuration is used to establish a mapping relationship library between dark crack features and mold damage types based on the historically stored multiple mold numbers and their associated defect data;

[0113] If a dark crack is detected again with a feature data similarity exceeding a preset similarity to a historical defect data in the mapping relationship library, the corresponding mold damage type and suggested repair location will be output.

[0114] Specifically, mold damage type refers to the specific forms of damage that may occur to the mold, such as: drawbead wear, die surface scratches, increased concave corner radius, slight collapse of punch corner radius, and chrome plating peeling. Different types of damage will produce dark cracks with different locations, morphologies, and electromagnetic response characteristics on the workpiece.

[0115] The process is described below: After the production line produces the first workpiece of a batch or after mold change, it is automatically or manually sent to the inspection station. The inspection system performs inspection according to the steps described above. If a hidden crack is found in a certain inspection area, the process proceeds to this step: the system controller immediately generates a "Mold Repair and Inspection Notice" to notify the equipment maintenance engineer. Simultaneously, the system automatically reads the mold number of the current production line (which can be read via a barcode scanner), strongly correlates all defect data of the detected defects with this number, and packages and stores them in the central defect database, building a mapping relationship library: this is an offline data analysis process. As data accumulates in the database, data engineers or expert systems will use clustering algorithms (such as K-means) to classify the defect feature data. Each type of defect feature will be labeled and associated with the actual mold damage type ultimately confirmed by manual verification.

[0116] For example, historical data might show that all defects associated with the "minor collapse of the punch fillet" damage for "mold number A-123" exhibit a very high negative peak (a large and negative maximum value) in their conductivity difference-frequency curves around 800kHz, with a very small average offset. This forms a mapping rule. When the system discovers another hidden crack in subsequent inspections, it extracts the feature data of its curve in real time and performs similarity calculations (such as calculating Euclidean distance and cosine similarity) with all historical patterns in the mapping database. If the similarity with a pattern in the database exceeds a preset similarity (such as 95%), the system will no longer simply report "defect present," but will directly output a diagnostic conclusion: "Suspected mold damage type: punch fillet wear; suggested repair location: punch fillet at the workpiece [XX coordinates] on the mold."

[0117] In a preferred embodiment, before simultaneously injecting a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test, the method further includes the following steps:

[0118] Obtain the material grade of the stamping part to be tested; specifically, the material grade refers to the specific model of the ultra-high strength steel plate, such as DP980, MS1500, 22MnB5, etc.

[0119] The excitation parameter list is retrieved and iterated to obtain the excitation parameters corresponding to the material grade of the stamping part to be tested, and these parameters are used as the target excitation parameters. The excitation parameters include frequency combination and voltage amplitude.

[0120] The excitation parameter list includes multiple optimal excitation parameters corresponding to different material grades. The optimal excitation parameters are determined through pre-experimentation, and their standard is to make the signal-to-noise ratio of the standard defect-free test block of the material corresponding to the material grade reach the optimal level. Specifically, the standard defect-free test block is a standard sample made of the same material grade and confirmed by non-destructive testing to be absolutely free of any defects.

[0121] The simultaneous injection of a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test includes the following steps:

[0122] Simultaneously, a set of multi-frequency AC excitation signals under the same target excitation parameters are injected into the detection area and the reference area of ​​the stamping part under test.

[0123] Specifically, the detection system scans the barcode on the workpiece or obtains the material grade information of the current workpiece from the upstream MES system. The program then searches and matches a preset list of excitation parameters, finding the row of data that perfectly corresponds to the grade, and loads the frequency combination and voltage amplitude from that row as the target excitation parameters for this detection. The signal generator of the detection system generates a specific set of multi-frequency AC signals according to the frequency and voltage set in the target excitation parameters. This set of signals is then simultaneously applied to the probes in both the detection area and the reference area.

[0124] Specifically, the optimal excitation parameters are determined through preliminary experiments: this is an offline calibration process. Before the method is implemented, engineers must conduct parameter scanning tests using standard defect-free test blocks for each possible material grade. For example, different frequency combinations are explored while keeping the voltage constant; the voltage amplitude is varied while keeping the frequency constant. After each test, the signal-to-noise ratio (SNR) of the signal acquired in the defect-free region is calculated (usually using the standard deviation of the signal amplitude as a noise estimate). Finally, the set of parameters that maximizes the SNR is selected and added to the excitation parameter list.

[0125] In a preferred embodiment, obtaining the detection area and reference area of ​​the stamped part to be tested includes the following steps:

[0126] Query the defect database to obtain all the hidden crack location data of historical stamped parts produced with the same or similar molds as the stamped part to be tested, and form a set of empirical high-risk areas from all the locations of the hidden cracks.

[0127] The region obtained by taking the intersection of the set of potential defect detection regions and the set of empirical high-risk regions is taken as the detection region.

[0128] Specifically, the empirical high-risk area set refers to the collection of location data for hidden cracks that have actually occurred on all historical workpieces produced by the same (or similar) mold, obtained by querying the defect database. This is based on real-world defect cases. Before determining the detection area, the system program first automatically queries the defect database. Using the mold number of the current workpiece as the query condition, it retrieves the coordinates of all recorded hidden crack locations on all historical workpieces produced by that mold. The point cloud or region formed by all these location points constitutes the empirical high-risk area set. The system simultaneously loads the potential defect detection area set (theoretical high-risk area) obtained from the above stamping simulation. Subsequently, spatial set operations are performed to calculate the intersection of these two sets. This intersection area means that the simulation theory suggests this area is prone to problems, and historical data indeed proves that problems have occurred here multiple times. Finally, the system officially determines this intersection area as the detection area for this inspection.

[0129] This implementation method improves inspection efficiency: It avoids full scanning of the entire workpiece or all theoretically high-risk areas, focusing only on the highest-risk areas confirmed by both theoretical and empirical evidence, significantly reducing inspection time. Dynamic optimization: The inspection area evolves itself as production data accumulates. If a theoretically high-risk area has never experienced an actual defect, its priority may be reduced in subsequent inspections; conversely, if a new area begins to show frequent defects, it will immediately be included in the key inspection scope.

[0130] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. The above descriptions are only preferred embodiments of this application. It should be noted that due to the limitations of written expression, while there are objectively infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of this invention, and can also combine the above technical features in an appropriate manner. These improvements, modifications, changes, or combinations, or the direct application of the inventive concept and technical solution to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles, characterized in that, Includes the following steps: Obtain the inspection area and reference area of ​​the stamping part to be tested; Simultaneously, a set of identical multi-frequency AC excitation signals are injected into the detection area and the reference area of ​​the stamping part under test, and the first electrical response signal of the detection area under the multi-frequency AC excitation signal and the second electrical response signal of the reference area under the multi-frequency AC excitation signal are acquired synchronously. Based on the first electrical response signal and the second electrical response signal, the conductivity difference between the detection area and the reference area at the same frequency point is calculated to obtain a conductivity difference-frequency curve including multiple frequency points; Based on the conductivity difference-frequency curve, it is determined whether there are dark cracks in the detection area.

2. The method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 1, characterized in that: The determination of whether a dark crack exists within the detection area based on the conductivity difference-frequency curve includes the following steps: Calculate the maximum absolute value of the conductivity difference within the first preset high-frequency band of the conductivity difference-frequency curve; If the maximum absolute value of the conductivity difference is greater than or equal to the first preset threshold, then the conductivity difference corresponding to all the frequency points in the first preset high frequency band is extracted from the conductivity difference-frequency curve to form a dataset to be analyzed. Calculate the arithmetic mean of all the conductivity differences in the dataset to be analyzed, and denot it as the average offset. If the average offset is less than a preset defect threshold, it is determined that there is a dark crack in the detection area.

3. The method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 2, characterized in that: After calculating the maximum absolute value of the conductivity difference-frequency curve within the first preset high-frequency band, the method further includes the following steps: If the maximum absolute value of the conductivity difference is less than the first preset threshold, then the phase difference of the AC signal between the detection area and the reference area at the same frequency point is calculated based on the first electrical response signal and the second electrical response signal. A phase difference-frequency curve is generated based on the multiple frequency points and their corresponding AC signal phase differences; Calculate the slope value of the phase difference-frequency curve within the second preset high-frequency band; If the absolute value of the fitted slope is greater than the preset slope threshold, it is determined that there is a dark crack in the detection area.

4. The method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 1, characterized in that: The process of obtaining the detection area and reference area of ​​the stamped part to be tested includes the following steps: A three-dimensional digital model of the stamping part to be tested is obtained. Based on the stamping forming simulation software, the forming process of the stamping part to be tested is simulated by computer to obtain the stamping simulation results. From the stamping simulation results, the maximum principal strain value and thickness reduction rate of all regions on the stamped part under test after stamping are extracted; The region that simultaneously satisfies that the maximum principal strain value is less than a preset strain threshold and the thickness reduction rate is less than a preset reduction threshold is determined as a candidate region. From all the candidate regions, select the region that is physically far away from all potential defect detection regions and determine it as the baseline region.

5. The method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 4, characterized in that: The potential defect detection area is the area on the stamping part under test where the maximum principal strain value is greater than or equal to the preset strain threshold, and / or the area where the thickness reduction rate is greater than or equal to the preset thinning threshold; All the potential defect detection regions mentioned above constitute a set of potential defect detection regions; Calculate the physical distance between the geometric center point of each candidate region and the geometric center point of each potential defect detection region to obtain multiple sets of distance values ​​corresponding to each candidate region; For each candidate region, the minimum value is selected from the multiple sets of distance values ​​corresponding to it, and this value is taken as the minimum safe distance between the candidate region and the set of potential defect detection regions. The minimum safe distance of all candidate regions is traversed, and the candidate region with the largest minimum safe distance is determined as the reference region.

6. The method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 3, characterized in that: After obtaining the detection area and reference area of ​​the stamping part to be tested, the following steps are also included: Verify that the reference area is free of defects; The simultaneous injection of a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test includes the following steps: If so, then simultaneously inject a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test.

7. The method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 6, characterized in that: Verifying whether the reference area is free of defects includes the following steps: Select at least three verification points within and / or at the edge of the reference area; Using the multi-frequency AC excitation signal, at least three of the verification points are scanned respectively, and the third electrical response signal of each verification point is collected; Calculate the conductivity difference between any two verification points at the same frequency point to obtain multiple sets of verification difference-frequency curves; If the fluctuation amplitude of all the verification difference-frequency curves within the preset frequency range is less than the preset verification threshold, then the verification result of the benchmark region is determined to be defect-free.

8. The method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 5, characterized in that: The stamped part to be tested is the first piece after mold change or the first piece of a batch; After determining whether there is a dark crack in the detection area based on the conductivity difference-frequency curve, the method further includes the following steps: If a dark crack is found in the detection area of ​​the first piece or the first piece of the batch after the mold change, a mold repair and inspection prompt is generated, and the defect data of the dark crack defect is bound to the mold number of the mold used in the current production and stored in the defect database; the defect data includes the morphology of the dark crack, the location of the dark crack, and the characteristic data of the corresponding conductivity difference-frequency curve. The defect database configuration is used to establish a mapping relationship library between dark crack features and mold damage types based on the historically stored multiple mold numbers and their associated defect data; If a dark crack is detected again with a feature data similarity exceeding a preset similarity to a historical defect data in the mapping relationship library, the corresponding mold damage type and suggested repair location will be output.

9. The method for detecting hidden cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 1, characterized in that: Before simultaneously injecting a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test, the method further includes the following steps: Obtain the material grade of the stamped part to be tested; The excitation parameter list is retrieved and iterated to obtain the excitation parameters corresponding to the material grade of the stamping part to be tested, and these parameters are used as the target excitation parameters. The excitation parameters include frequency combination and voltage amplitude. The excitation parameter list includes multiple optimal excitation parameters corresponding to different material grades. The optimal excitation parameters are determined through pre-experimentation, and the standard is to make the signal-to-noise ratio of the standard defect-free test block of the material corresponding to the material grade reach the optimal level. The simultaneous injection of a set of identical multi-frequency AC excitation signals into the detection area and the reference area of ​​the stamping part under test includes the following steps: Simultaneously, a set of multi-frequency AC excitation signals under the same target excitation parameters are injected into the detection area and the reference area of ​​the stamping part under test.

10. The method for detecting dark cracks in ultra-high strength steel plate stamped parts for new energy vehicles according to claim 8, characterized in that: The process of obtaining the detection area and reference area of ​​the stamped part to be tested includes the following steps: Query the defect database to obtain all the hidden crack location data of historical stamped parts produced with the same or similar molds as the stamped part to be tested, and form a set of empirical high-risk areas from all the locations of the hidden cracks. The region obtained by taking the intersection of the set of potential defect detection regions and the set of empirical high-risk regions is taken as the detection region.

Images