A stamping forming quality on-line monitoring device and method based on multi-channel acoustic emission signals
Patent Information
- Application Number
- CN202610705101.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-21
- Publication Date
- 2026-09-25
AI Technical Summary
[0005]本发明的目的在于提供一种基于多通道声发射信号的冲压成形质量在线监测装置及方法，以克服冲压成形过程中成形区域封闭、成形质量难以实时探测以及损伤演化过程难以准确识别的问题，实现对冲压成形质量的在线监测与损伤定位
1. 本发明通过在凸模、凹模等不同结构位置布设多通道声发射传感器阵列，从不同声波传播路径同步采集冲压成形过程中板材变形与损伤产生的声发射信号，实现对成形封闭区域内损伤信息的多路径感知。与单一传感器或单一安装位置相比，本发明能够更全面地表征成形过程状态，并对早期损伤迹象实现更高灵敏度的捕获，从而提高监测结果的完整性与可靠性；
Smart Images
Figure CN122814756A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of stamping quality monitoring, and specifically to an online stamping quality monitoring device and method based on multi-channel acoustic emission signals. Background Technology
[0002] Metal stamping technology is a crucial component of modern industrial manufacturing, widely applied in the automotive, aerospace, and home appliance industries. During stamping, sheet metal blanks undergo complex elasto-plastic deformation within a very short time under die constraints, characterized by high forming speed, drastic load changes, and highly transient deformation. Simultaneously, due to the typically enclosed structure of stamping dies, the actual deformation state and localized damage behavior of the sheet metal are difficult to directly observe during the forming process. This makes quality defects such as wrinkling, excessive thinning, and cracking in stamping sudden and insidious. Once these defects occur, they are often irreparable by subsequent processes, easily leading to workpiece scrap, die damage, or even production interruption, causing significant economic losses and safety hazards for enterprises.
[0003] To address the aforementioned issues, current stamping production commonly employs post-forming quality inspection for quality control. Specifically, after the production department processes sample workpieces according to predetermined process parameters, they are transferred to the quality inspection department. The dimensional accuracy, surface quality, and mechanical properties of the workpieces are assessed manually or using measuring equipment, and the process parameters are adjusted based on the inspection results. Furthermore, some research attempts to indirectly monitor the stamping forming process using process signals such as stamping force, displacement, or strain, aiming to reflect the overall trend of the forming process to some extent.
[0004] However, most of the aforementioned methods are post-processing detection or indirect evaluation methods. These methods are not only complex and slow to respond, but also struggle to effectively identify transient local deformation instability and micro-damage evolution during the stamping process. Once quality defects occur, irreversible forming failures often result. Furthermore, monitoring methods based on single-process signals have limited information dimensions, failing to comprehensively reflect the complex deformation characteristics of different regions and stages during stamping, thus limiting the accuracy and reliability of forming quality assessment. Therefore, how to achieve real-time, comprehensive, and highly sensitive quality monitoring of the stamping process without affecting the stamping production cycle remains a pressing technical problem for those skilled in the art. Summary of the Invention
[0005] The purpose of this invention is to provide an online monitoring device and method for stamping quality based on multi-channel acoustic emission signals, so as to overcome the problems of closed forming area, difficulty in real-time detection of forming quality and difficulty in accurate identification of damage evolution process during stamping, and realize online monitoring and damage location of stamping quality.
[0006] To achieve the above objectives, the present invention adopts the following technical solution.
[0007] This invention provides an online monitoring device for stamping quality based on multi-channel acoustic emission signals, including stamping equipment, acoustic emission sensors, preamplifiers, multi-channel acoustic emission detectors, and industrial control computers.
[0008] The stamping equipment includes a die and a punch arranged opposite each other, and a blank holder disposed between the die and the punch. The die, the blank holder and the punch are mounted on a frame to stamp and form sheet metal.
[0009] The acoustic emission sensors are installed at different structural positions on the stamping equipment to collect multi-channel acoustic emission signals generated by the deformation and damage of the sheet metal during the stamping process.
[0010] The preamplifier is connected to the acoustic emission sensor and is used to amplify the acoustic emission signal.
[0011] The multi-channel acoustic emission detector is connected to a preamplifier for synchronous acquisition and analysis of the amplified multi-channel acoustic emission signals.
[0012] The industrial control computer is connected to a multi-channel acoustic emission detector to monitor the workpiece forming quality online during the stamping process based on the analysis results of the multi-channel acoustic emission signals. When an abnormal forming state is detected, alarm information and parameter information related to the forming process adjustment are output.
[0013] Based on the above technical solutions, the present invention may have the following further specific options or optimizations.
[0014] Specifically, the acoustic emission sensor includes: a first type of acoustic emission sensor, installed on the structural surface of the punch near the forming contact area of the sheet metal, for collecting forming acoustic emission signals propagating through the punch structure and reaching the first type of acoustic emission sensor; a second type of acoustic emission sensor, installed on the structural surface of the die near the forming contact area of the sheet metal, for collecting forming acoustic emission signals propagating through the die structure and reaching the second type of acoustic emission sensor; and a third type of acoustic emission sensor, installed on the stamping equipment support structure away from the forming contact area of the sheet metal, for collecting acoustic emission signals generated during the operation of the stamping equipment.
[0015] The first type of acoustic emission sensor and the second type of acoustic emission sensor each have at least three sensors arranged in a non-collinear manner on the punch and the die, and the sensors on the punch side and the die side are spatially distributed in a corresponding manner to form a multi-channel acoustic emission sensor spatial array.
[0016] The third type of acoustic emission sensor serves as a protective sensor, used to collect acoustic emission or background vibration signals generated during the operation of the stamping equipment, and as a noise reference signal to correct or eliminate interference in the forming acoustic emission signals collected by the first and second type of acoustic emission sensors, thereby improving the consistency and reliability of the forming acoustic emission signals.
[0017] Specifically, the acoustic emission sensors are all fixedly installed using magnetic clamps, and a contact layer made of coupling agent is provided between the magnetic clamps and the contact surface of the mold.
[0018] Specifically, the multi-channel acoustic emission detector has multiple signal acquisition channels and includes, in sequence, a main amplification module for amplifying signals, an A / D data conversion module for converting A / D data into digital signals, and a digital signal processing module for analyzing and processing digital signals. The acoustic emission detector is connected to a display terminal.
[0019] Secondly, the present invention provides an online monitoring method for stamping quality based on multi-channel acoustic emission signals, which is applied to a monitoring device, and the specific steps are as follows.
[0020] S10: By deploying a multi-channel acoustic emission sensor array at a preset position on the stamping die, the multi-channel acoustic emission signals generated by the sheet metal during the stamping process are collected in real time.
[0021] S20: Perform time synchronization and forming cycle segmentation preprocessing on the acquired multi-channel acoustic emission signals, and correct the acoustic emission signals based on the background noise signal through variational mode decomposition.
[0022] S30: Perform event detection on the preprocessed acoustic emission signal and extract acoustic emission characteristic parameters to characterize the stamping quality under multi-channel consistency constraints.
[0023] S40: Based on the pre-established correspondence between metal damage evolution and acoustic emission characteristic parameters, the acoustic emission characteristic parameters are analyzed to identify the forming quality state of the stamped part.
[0024] S50: After detecting an abnormal forming state, the damage source is spatially located based on the time delay difference of the arrival time of the multi-channel acoustic emission signals, and the location information of the area where the damage occurred is output.
[0025] S60: Input the monitoring results and location information of the abnormal forming state into the mold zone blank holder force control system to realize the adaptive matching and adjustment of blank holder force parameters in each zone of the mold.
[0026] In one possible implementation of the second aspect, step S20 involves time synchronization, forming cycle segmentation, and preprocessing of the acquired multi-channel acoustic emission signals, and correction of the acoustic emission signals based on background noise signals. This includes: performing unified time synchronization on the continuously acquired multi-channel acoustic emission signals according to the stamping stroke information of the stamping equipment and the timeline recorded by the third type of acoustic emission sensor, and dividing the waveform signals into time segments such as "feed-contact-forming-return" according to the stamping cycle; performing bandpass filtering and baseline correction on the segmented waveform signals, and performing noise cancellation processing using the background noise signal acquired by the third type of acoustic emission sensor as a reference; specifically, performing VMD (variational mode decomposition) decomposition processing on the segmented waveform signals, including...
[0027] 1) Multi-channel acoustic emission waveform signals for each time segment (feed, contact, forming, return) of each stamping cycle. Bandpass filtering is performed to remove low-frequency vibrations and high-frequency electromagnetic interference from the equipment. Subsequently, baseline correction is applied to the filtered signal. This baseline correction includes mean removal, detrending, or estimating the baseline using a sliding window and subtracting it from the original signal, so that the signal mean returns to zero baseline during event-free periods, resulting in a preprocessed signal. .in Indicates the channel number. 2) Stream the background noise waveform acquired by the third type of acoustic emission sensor. As a reference noise signal, it is compared with The same bandpass filtering and baseline correction process yields... This ensures that the reference noise and the target channel maintain consistency in frequency band and baseline definition. 3) Preprocessing of the signal for each channel. Perform VMD decomposition to satisfy the following constraints:
[0028] K eigenmode components were obtained and its corresponding center frequency It also includes the intrinsic mode components of the reference noise signal. and its center frequency .
[0029] To minimize the sum of the bandwidths of each mode near its respective center frequency, the alternating direction multiplier method is preferably used to iteratively update the modal components and center frequencies until the convergence threshold is met. 4) The similarity between each modal component of each target channel and each modal component of the reference noise is evaluated to determine the set of noise-dominant modes. 5) The noise-dominant modes are removed or canceled from the target channels to obtain the denoised reconstructed signal. This serves as the input signal for event detection and feature extraction in step S30.
[0030] In step S30: For input signal A threshold-triggered mechanism is used for event detection. For each detected acoustic emission event, its start time is extracted. Termination time Duration D
[0031] and event energy
[0032] Furthermore, non-forming-related interference events are eliminated through multi-channel time consistency constraints. Among them, forming events must be triggered by at least a preset number N (≥3) channels within the forming section, and their frequency band energy distribution must meet a preset frequency band proportion threshold. For events that pass the consistency constraint, multi-dimensional feature parameters for characterizing the stamping quality are further extracted, including time-domain features (peak amplitude, duration, event energy, number of counts), frequency-domain features (dominant frequency, center frequency, spectral bandwidth), entropy features, and multi-channel arrival time difference, to construct a multi-dimensional feature vector for characterizing the stamping quality state. .
[0033] In step S40: By conducting standard uniaxial tensile tests on stamped metal materials, mechanical response data and acoustic emission signals were simultaneously acquired. Based on the stress-strain curve, the material deformation and damage evolution process was divided into four stages: elastic deformation, uniform plastic deformation, strengthening deformation, and necking and fracture. Figure 3 As shown, the acoustic emission events and their characteristic parameters corresponding to each stage are labeled to form a labeled sample set. ,in For the first Multidimensional feature vectors of an event Label its stage; Simultaneously, the acoustic emission characteristic parameters (amplitude, energy, duration, center frequency, spectral entropy, VMD modal energy ratio, multi-channel energy distribution, etc.) within each damage stage were statistically analyzed to obtain the characteristic distribution patterns of each stage; based on this, the stages were further analyzed. eigenmean vector With covariance An estimate is made, and the discrimination score is used as the identification result; Furthermore, by setting up sensor arrays with different propagation distances, parameters such as the amplitude, energy, and arrival time of the first wave of acoustic emission events are obtained under known propagation path conditions. Based on the statistical results, a quantitative attenuation relationship model between acoustic emission characteristic parameters and propagation distance is established. Acoustic emission energy attenuation has two types: viscous attenuation and distance attenuation. This attenuation occurs during sound wave propagation. Viscous attenuation is based on the time dimension and is caused by the viscosity of the medium, while distance attenuation is based on the spatial dimension and is affected by the distance between the acoustic emission point and the measurement point. The attenuation of elastic waves is... The value is expressed quantitatively as follows: when the energy is The wave attenuates to its original value after traveling a distance of one wavelength. hour, The value is defined as follows:
[0034] When the AE wave propagates a distance D, its velocity With frequency components The amplitude will be from decay to :
[0035] In one feasible implementation, viscous decay and distance decay are characterized by combining them with an equivalent decay coefficient to obtain an equivalent decay model for engineering correction. Before determining the forming quality status, the amplitude and energy of each channel event are normalized and corrected by distance based on the quantitative attenuation relationship model to reduce the discrimination bias caused by the difference in distance between the sensor and the sound source. The multidimensional feature vector extracted during the stamping process is input into the corresponding relationship model to determine the forming quality state of the stamped part and obtain the initial forming quality analysis results. In step S50: When the forming quality state discrimination result meets the abnormal forming criterion, the spatial localization process of the damage source of the abnormal acoustic emission event is triggered. Spatial localization of the damage source is based on the time delay difference of arrival time of multi-channel acoustic emission signals. Specifically, this includes: extracting the arrival time of the first wave of each channel, calculating the arrival time difference between channels, establishing a localization equation by combining the pre-calibrated propagation velocity and the geometric rules of the mold-plate, and solving for the initial position of the damage source in the mold coordinate system. Considering the differences in propagation paths and the initial wave picking deviation caused by threshold triggering, the arrival time of the initial wave is compensated and corrected based on the attenuation relationship model and the material propagation velocity calibration results to obtain the final damage source location.
[0036] When multiple acoustic emission sources exist simultaneously during the stamping process, the simultaneous localization of multiple damage sources can be achieved by clustering and separating multi-channel acoustic emission events in terms of time, frequency, and spatial characteristics.
[0037] Furthermore, the present invention can also be used to assist in the adjustment of stamping process parameters based on the abnormal forming state type and the spatial distribution information of damage sources, so as to improve the stability of the stamping process and the consistency of forming quality.
[0038] The present invention achieves the following technical effects compared to the prior art. 1. This invention utilizes a multi-channel acoustic emission sensor array deployed at different structural locations such as the punch and die to simultaneously acquire acoustic emission signals generated by sheet metal deformation and damage during the stamping process from different sound wave propagation paths, achieving multi-path perception of damage information within the formed closed area. Compared to a single sensor or a single installation location, this invention can more comprehensively characterize the forming process state and achieve higher sensitivity in capturing early signs of damage, thereby improving the completeness and reliability of monitoring results; 2. This invention introduces a third type of acoustic emission sensor as a protective sensor, which is used to collect background acoustic emission or vibration signals generated during equipment operation, and to correct or eliminate interference in the formed acoustic emission signal, effectively reducing the impact of equipment noise on monitoring results and improving the signal-to-noise ratio and consistency of monitoring signals; 3. This invention simultaneously acquires mechanical response and acoustic emission signals through a standard uniaxial tensile test, and establishes a correspondence model between the metal damage evolution stages and acoustic emission characteristic parameters, so that the damage identification basis formed has mechanistic consistency and interpretability; 4. Based on damage identification, this invention further constructs a quantitative attenuation relationship model between acoustic emission characteristic parameters and propagation distance. This model is then used to perform distance normalization correction on amplitude-related characteristics such as amplitude / energy of multi-channel acoustic emission signals, and to compensate for the arrival time of the first wave. This effectively reduces the impact of propagation path differences, installation condition differences, and attenuation characteristics on quality judgment and location calculation. 5. This invention, through a multi-channel acoustic emission sensor array and its spatial arrangement, enables the simultaneous localization of multiple damage sources during the stamping process, even if multiple acoustic emission sources exist, based on the separation and clustering of events in the time, frequency, and spatiotemporal feature dimensions, thereby enhancing the applicability of the monitoring method under complex forming conditions. 6. By acquiring the location of damage and its evolution process in real time, this invention provides quantifiable data support for optimizing the stamping process and controlling the consistency of forming quality. It can also be used to assist in adjusting the stamping process parameters, thereby improving the stability of the stamping process. Attached Figure Description
[0039] Figure 1 This is a schematic diagram of the stamping quality monitoring device provided by the present invention; Figure 2 This is a schematic diagram of the standard tensile test AE detection process of the present invention; Figure 3 This is a schematic diagram showing the acquisition results of the AE amplitude-mechanical signal and the division of material damage evolution stages in the tensile test of the present invention. Figure 4 This is a top view schematic diagram of the arrangement position of the acoustic emission sensor of the present invention; Figure 5 This is a schematic diagram of the acoustic emission control multi-channel system of the present invention; Figure 6 A flowchart illustrating the steps of an online monitoring method for stamping quality based on multi-channel acoustic emission signals provided by this invention; The correspondence between the reference numerals and components in the attached drawings is as follows: 10-stamping equipment, 11-die, 12-pressure ring, 13-punch, 21-first type acoustic emission sensor, 22-second type acoustic emission sensor, 23-third type acoustic emission sensor, 30-amplifier, 40-multi-channel acoustic emission detector, 50 industrial computer. Detailed Implementation
[0040] To make the objectives and advantages of this invention clearer, the invention will be specifically described below with reference to embodiments. It should be understood that the following text is merely used to describe one or more specific embodiments of the invention and does not strictly limit the scope of protection specifically claimed by the invention.
[0041] As used herein, the terms “parallel,” “perpendicular,” and the like are not limited to their strict geometric definitions, but include tolerance for reasonable and inconsistent machining or human errors.
[0042] This embodiment takes the sheet metal deep drawing and stamping process as an example, and uses multi-channel acoustic emission signals to monitor the deformation and damage status of the sheet metal during the stamping process online.
[0043] An online monitoring device and method for stamping quality based on multi-channel acoustic emission signals This embodiment provides an online monitoring device for stamping quality based on multi-channel acoustic emission signals, such as... Figure 1 , Figure 4 and Figure 5 As shown, it includes a stamping device 10, an acoustic emission sensor, a preamplifier 30, a multi-channel acoustic emission detector 40, and an industrial control computer 50.
[0044] The stamping equipment 10 includes a die 11 and a punch 13 arranged opposite to each other, and a pressure ring 12 disposed between the die 11 and the punch 13. The die 11, the pressure ring 12 and the punch 13 are mounted together on the frame for stamping and forming sheet metal.
[0045] The acoustic emission sensors are installed at different structural positions on the stamping equipment 10 to collect acoustic emission signals generated by the deformation and damage of the sheet metal during the stamping process. Specifically, the acoustic emission sensors include a first type of acoustic emission sensor 21, a second type of acoustic emission sensor 22, and a third type of acoustic emission sensor 23.
[0046] The first type of acoustic emission sensor 21 is installed on the structural surface of the punch 13 near the sheet metal forming contact area, and is used to collect the forming acoustic emission signal that propagates through the punch structure and reaches the first type of acoustic emission sensor 21; the second type of acoustic emission sensor 22 is installed on the structural surface of the die 11 near the sheet metal forming contact area, and is used to collect the forming acoustic emission signal that propagates through the die structure and reaches the second type of acoustic emission sensor 22; the third type of acoustic emission sensor 23 is installed on the stamping equipment bearing structure away from the sheet metal forming contact area, and is used to collect the acoustic emission signal or background vibration signal generated during the operation of the stamping equipment.
[0047] The specific operations for deploying a multi-channel acoustic emission sensor array include: The first step is to determine the mounting reference for the die. The die cavity is used as the upper die (2), located above the punch. Its top forming base surface is used as the die cavity mounting reference surface. The intersection of the die cavity mounting reference surface and the outer side of the die cavity is defined as the upper die mounting reference line. The second step is to determine the punch installation reference. The punch, as the lower die (5), is located below the die, and its bottom forming base surface is used as the punch mounting reference surface; If the outer edge of the forming structure inside the punch is completely inside the outer edge of the forming structure inside the die, then the boundary line formed by the intersection of the punch mounting reference surface and the outer side surface shall be used as the punch mounting reference line. If the outer edge of the forming structure inside the punch extends beyond the outer edge of the forming structure inside the die, then the intersection of the outer side of the die and the punch mounting reference surface shall be used as the punch mounting reference line. The third step is to install a multi-channel acoustic emission sensor. In this embodiment, at least three of the first type of acoustic emission sensor 21 and the second type of acoustic emission sensor 22 are arranged in a non-collinear manner on the punch 13 and the die 11, and the acoustic emission sensors on the punch side and the die side are symmetrically or correspondingly distributed in space, thereby forming a multi-channel acoustic emission sensor spatial array for realizing the synchronous acquisition and spatial positioning of multi-path acoustic emission signals.
[0048] The third type of acoustic emission sensor 23 serves as a protective sensor, and the signal it collects is used as a noise reference signal to correct or eliminate interference in the shaped acoustic emission signals collected by the first type of acoustic emission sensor 21 and the second type of acoustic emission sensor 22, so as to reduce the impact of equipment operating noise on the monitoring results and improve the consistency and reliability of the shaped acoustic emission signals.
[0049] In this embodiment, each acoustic emission sensor is fixedly mounted on the surface of the mold or stamping equipment support structure by a magnetic clamp. A contact layer made of coupling agent is provided between the magnetic clamp and the mounting surface to ensure effective coupling of the acoustic emission signal between the sensor and the structure surface.
[0050] The preamplifier 30 is connected to each acoustic emission sensor and is used to amplify the acquired acoustic emission signals. The multi-channel acoustic emission detector 40 is connected to the preamplifier 30 and is used to synchronously acquire and analyze the amplified multi-channel acoustic emission signals. Specifically, the multi-channel acoustic emission detector 40 includes a main amplification module, an A / D data conversion module, and a digital signal processing module arranged in sequence, and is connected to a display terminal, as shown in the schematic diagram. Figure 1 As shown.
[0051] The industrial control computer 50 is communicatively connected to the multi-channel acoustic emission detector 40, which is used to further analyze and process the multi-channel acoustic emission signals, realize online monitoring of the forming quality of the workpiece during the stamping process, and output alarm information and parameter information related to the forming process adjustment when an abnormal forming state is detected.
[0052] Based on the above-described device, this embodiment also provides a method for online monitoring of stamping quality based on multi-channel acoustic emission signals, such as... Figure 5 As shown, it includes the following steps: S10: By deploying a multi-channel acoustic emission sensor array at a preset position on the stamping die, the multi-channel acoustic emission signals generated by the sheet metal during the stamping process are collected in real time.
[0053] S20: Perform time synchronization and forming cycle segmentation preprocessing on the acquired multi-channel acoustic emission signals, and correct the acoustic emission signals based on the background noise signal through variational mode decomposition; Specifically, based on the stamping stroke information of the stamping equipment and the timeline recorded by the third type of acoustic emission sensor 23, the continuously acquired multi-channel acoustic emission signals are synchronized in a unified time and divided into waveform signals in time segments such as "feed-contact-forming-return" according to the stamping cycle; To ensure the stability of the segment boundaries, the characteristic points of the stamping stroke signal (such as top dead point / bottom dead point, contact point or pressure surge point) can be used as the basis for segmentation, and the signals of each cycle can be aligned with a uniform length for subsequent multi-channel consistency analysis. Bandpass filtering and baseline correction are performed on the segmented waveform signal, and noise cancellation processing is performed using the background noise signal collected by the third type acoustic emission sensor as a reference. Specifically, the segmented waveform signal is decomposed using VMD (Variational Mode Decomposition), including: 1) Multi-channel acoustic emission waveform signals for each time segment (feed, contact, forming, return) of each stamping cycle. Bandpass filtering is performed to remove low-frequency vibrations and high-frequency electromagnetic interference from the equipment. Subsequently, baseline correction is applied to the filtered signal. This baseline correction includes mean removal, detrending, or estimating the baseline using a sliding window and subtracting it from the original signal, so that the signal mean returns to zero baseline during event-free periods, resulting in a preprocessed signal. .in Indicates the channel number. 2) Stream the background noise waveform acquired by the third type of acoustic emission sensor. As a reference noise signal, it is compared with The same bandpass filtering and baseline correction process yields... This ensures that the reference noise and the target channel maintain consistency in frequency band and baseline definition. 3) Preprocessing of the signal for each channel. Perform VMD decomposition to satisfy the constraints.
[0054] K eigenmode components were obtained and its corresponding center frequency It also includes the intrinsic mode components of the reference noise signal. and its center frequency .
[0055] To minimize the sum of the bandwidths of each mode near its respective center frequency, the alternating direction multiplier method is preferably used to iteratively update the modal components and center frequencies until the convergence threshold is met. 4) The similarity between each modal component of each target channel and each modal component of the reference noise is evaluated to determine the set of noise-dominant modes. 5) The noise-dominant modes are removed or canceled from the target channels to obtain the denoised reconstructed signal. This serves as the input signal for event detection and feature extraction in step S30.
[0056] S30: Perform event detection on the preprocessed acoustic emission signal and extract acoustic emission characteristic parameters to characterize the stamping quality under multi-channel consistency constraints; For input signal A threshold-triggered mechanism is used for event detection. For each detected acoustic emission event, its start time is extracted. Termination time Duration D
[0057] and event energy
[0058] Furthermore, non-forming-related interference events are eliminated through multi-channel time consistency constraints. Among them, forming events must be triggered by at least a preset number N (≥3) channels within the forming section, and their frequency band energy distribution must meet a preset frequency band proportion threshold. For events that pass the consistency constraint, multi-dimensional feature parameters for characterizing the stamping quality are further extracted, including time-domain features (peak amplitude, duration, event energy, number of counts), frequency-domain features (dominant frequency, center frequency, spectral bandwidth), entropy features, and multi-channel arrival time difference, to construct a multi-dimensional feature vector for characterizing the stamping quality state.
[0059] S40: Based on the pre-established correspondence between metal damage evolution and acoustic emission characteristic parameters, the acoustic emission characteristic parameters are analyzed to identify the forming quality state of the stamped part.
[0060] Specifically, the correspondence is established by conducting a standard uniaxial tensile test on the target stamping material, and the calibration rules and attenuation correction parameters of the acoustic emission characteristic parameters are obtained simultaneously during the establishment process. To ensure that the correspondence can be stably transferred to the stamping forming condition, the following points need to be noted in the tensile test: 1) Specimen and loading conditions: The material grade, sampling direction, thickness and surface condition of the tensile specimen should be consistent with the actual stamped sheet, and the tensile rate or strain rate should be selected within a comparable range. Preferably, no-load and fixture friction background test should be performed before formal calibration to eliminate acoustic emission caused by fixture friction or frame vibration. 2) Consistency of acoustic emission acquisition: The installation method, coupling agent type and coating thickness, and clamp preload of the acoustic emission sensor in the tensile test should be consistent with the installation conditions in the stamping monitoring as much as possible; the sampling rate, preamplifier gain, and event definition parameters (such as PDT / HDT / HLT waveform segmentation parameters) of the acquisition system should be consistent or recorded between different tests so that they can be traced when the model is called up. 3) Sensor placement: such as Figure 2 As shown, the specimen is clamped between tensile fixtures, and at least two acoustic emission sensors are symmetrically arranged on both sides of the gauge length of the specimen. The sensors form different propagation distances with the expected damage area along the specimen axis. At least one acoustic emission sensor is arranged in the clamping area at both ends of the specimen, so that the sound wave propagation path includes a coupling propagation process. 4) Stage Division and Labeling: Mechanical response data are collected simultaneously during the tensile test. Based on the stress-time curve, the material deformation and damage evolution process is divided into elastic deformation stage, uniform plastic deformation stage, strengthening deformation stage, and necking and fracture stage. The acoustic emission events and characteristic parameters corresponding to each stage are labeled to form a labeled sample set. ,in For the first Multidimensional feature vectors of an event Label its stage; Simultaneously, the acoustic emission characteristic parameters (amplitude, energy, duration, center frequency, spectral entropy, VMD modal energy ratio, multi-channel energy distribution, etc.) within each damage stage were statistically analyzed to obtain the characteristic distribution patterns of each stage, such as... Figure 3As shown, mechanical response signals and acoustic emission signals were simultaneously acquired during the uniaxial tensile test. Based on the stress-time curve characteristics, the material deformation and damage evolution process was divided into stages I to IV. The amplitude distribution and event density of acoustic emission signals showed significant differences in different stages. Stage I corresponds to the initial loading and elastic deformation stage, with fewer acoustic emission events and a stable performance. Stage II corresponds to the uniform plastic deformation stage, with deformation throughout the gauge length and a significant increase in the number of acoustic emission events, but low-amplitude events predominate. Stage III corresponds to the strengthening deformation stage, with an increased acoustic emission event density and an amplitude distribution that begins to migrate to higher ranges, indicating enhanced deformation non-uniformity and intensified acoustic emission activity. Stage IV corresponds to the necking and fracture stage, with high-amplitude events concentrated and reaching their highest amplitude within a short period, which can serve as a significant acoustic emission characteristic of severe material damage or failure.
[0061] Based on this, the stages eigenmean vector With covariance Estimate the discriminant features Calculate which stage it belongs to. Similarity / probability score: The highest scorer will be used as the stage result. 5) Propagation Attenuation Calibration: Acoustic emission energy attenuation has two types: viscous attenuation and distance attenuation. This attenuation occurs during sound wave propagation. Viscous attenuation is based on the time dimension and is caused by the viscosity of the medium, while distance attenuation is based on the spatial dimension and is affected by the distance between the acoustic emission point and the measurement location. Elastic wave attenuation is... The value is expressed quantitatively as follows: when the energy is The wave attenuates to its original value after traveling a distance of one wavelength. hour, The value is defined as follows:
[0062] When the AE wave propagates a distance D, its velocity With frequency components The amplitude will be from decay to :
[0063] In one feasible implementation, viscous decay and distance decay are characterized by a combined equivalent decay coefficient. To reduce the impact of sensor-sound source distance differences on stamping discrimination and positioning, sensor arrangement conditions with different propagation paths can be set during tensile testing to obtain the amplitude of acoustic emission events. ,energy Using parameters such as the arrival time of the first wave, a quantitative attenuation relationship model between acoustic emission characteristic parameters and propagation distance was established by fitting the model. In one feasible implementation, the amplitude and energy can be fitted using exponential or power-law decay. For example, before determining the stamping quality state, the amplitude and energy correlation characteristics of each channel event during the stamping process can be corrected by distance normalization based on this decay relationship model. Taking the exponential model as an example, for the first... Channel, estimated propagation distance is The event, according to the reference distance Revised to:
[0064] When necessary, the arrival time of the first wave is compensated and corrected to reduce the discrimination bias and positioning error caused by the difference in propagation path.
[0065] In this step, the multidimensional feature vector extracted during the stamping process is input into the above correspondence model or the pre-trained damage recognition model to obtain the forming quality status judgment result of the stamped part; S50: When an abnormal forming state is detected, the damage source is spatially located based on the time delay difference of the arrival time of the multi-channel acoustic emission signals, and the location information of the area where the damage occurred is output.
[0066] Specifically, the arrival time of the first wave of the acoustic emission signal from each channel is extracted, the arrival time difference between channels is calculated, and the pre-calibrated equivalent propagation speed is combined. In addition to the geometric relationship between the mold and the sheet metal, a damage source coordinate system is established. The equation for locating the unknown TDOA (Time Difference of Arrival) is as follows:
[0067] Among them, the The coordinates of the sensor in the mold coordinate system are: ; The initial position of the damage source in the mold coordinate system is obtained by solving the problem. Combined with the attenuation relationship model and the propagation velocity calibration results, the arrival time of the first wave is compensated and corrected to obtain the final position of the damage source. When multiple acoustic emission sources exist simultaneously during the stamping process, the simultaneous localization of multiple damage sources can be achieved by clustering and separating multi-channel acoustic emission events in terms of time, frequency, and spatial characteristics.
[0068] S60: Input the monitoring results of the abnormal forming state and the location information of the damage source into the mold zone blank holder force control system to realize the adaptive matching and adjustment of the blank holder force parameters of each zone of the mold.
[0069] Specifically, the blank holder force parameter can be matched and adjusted according to the damage type, damage location and damage severity; and the location information of the damage location can be used to assist in the adjustment of stamping process parameters to improve the stability of the stamping process and the consistency of forming quality.
[0070] The online monitoring method for stamping quality described in this embodiment can be applied to the monitoring of forming quality during sheet metal deep drawing, bulging, or composite stamping processes.
[0071] The above description is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained in this invention are implemented according to conventional methods in the art unless otherwise specified or limited.
Claims
1. A stamping forming quality online monitoring device based on multi-channel acoustic emission signals, characterized in that, It includes a stamping machine (10), an acoustic emission sensor, a preamplifier unit (30), a multi-channel acoustic emission detector (40), and an industrial control computer (50). The stamping equipment (10) includes a die (11), a punch (13) and a blank holder (12). The die (11), punch (13) and blank holder (12) are mounted on the frame for stamping and forming sheet metal blanks. The acoustic emission sensor is installed on the stamping equipment (10) to collect multi-channel acoustic emission signals generated by the deformation and damage of the sheet metal blank during the stamping process; The preamplifier unit (30) is connected to the acoustic emission sensor and is used to amplify the acoustic emission signal; The multi-channel acoustic emission detector (40) is connected to the preamplifier unit (30) for synchronous acquisition and analysis of the amplified multi-channel acoustic emission signal; The industrial control computer (50) is connected to the multi-channel acoustic emission detector (40) for online monitoring of the workpiece forming quality during the stamping process based on the analysis results of the multi-channel acoustic emission signal, and outputs alarm information and parameter information related to forming process adjustment when an abnormal forming state is detected.
2. The online monitoring device for stamping quality based on multi-channel acoustic emission signals according to claim 1, characterized in that, The acoustic emission sensor includes: The first type of acoustic emission sensor (21) is installed on the structural surface of the punch (13) facing the forming contact area of the sheet metal blank, and is used to collect the forming acoustic emission signal that propagates through the punch structure and reaches the first type of acoustic emission sensor (21). The second type of acoustic emission sensor (22) is installed on the structural surface of the die (11) facing the forming contact area of the sheet metal blank, and is used to collect the forming acoustic emission signal that propagates through the die structure and reaches the second type of acoustic emission sensor (22). The third type of acoustic emission sensor (23) is installed on the bearing structure of the stamping equipment away from the forming contact area of the sheet metal blank, and is used to collect the background noise signal generated during the operation of the stamping equipment. At least three of the first type of acoustic emission sensors (21) are arranged in a non-collinear manner on the punch (13), and at least three of the second type of acoustic emission sensors (22) are arranged in a non-collinear manner on the die (11). The first type of acoustic emission sensors (21) and the second type of acoustic emission sensors (22) are arranged correspondingly. The background noise signal collected by the third type of acoustic emission sensor (23) is used to suppress noise or eliminate interference from the shaped acoustic emission signals collected by the first type of acoustic emission sensor (21) and the second type of acoustic emission sensor (22).
3. The online monitoring device for stamping quality based on multi-channel acoustic emission signals according to claim 2, characterized in that, The acoustic emission sensor is mounted on the surface of the stamping equipment bearing structure via a magnetic mounting base. A contact layer composed of an acoustic coupling medium layer is provided between the magnetic mounting base and the mounting surface. The multi-channel acoustic emission detector (40) has multiple signal acquisition channels. The multi-channel acoustic emission detector (40) includes a main amplification module, an analog-to-digital conversion module, and a digital signal processing module arranged in sequence. The multi-channel acoustic emission detector (40) is connected to a display terminal.
4. A method for online monitoring of stamping quality based on multi-channel acoustic emission signals, characterized in that, Includes the following steps: S10: Install acoustic emission sensors at preset positions on the stamping equipment to collect multi-channel acoustic emission signals generated by the sheet metal blank during the stamping process in real time. S20: Perform time synchronization and forming cycle segmentation preprocessing on the acquired multi-channel acoustic emission signals, and perform noise suppression processing on the acoustic emission signals based on the background noise signal through variational mode decomposition; S30: Perform event detection on the preprocessed acoustic emission signal and extract acoustic emission characteristic parameters to characterize the stamping quality under multi-channel consistency constraints; S40: Based on the pre-established correspondence model between the damage evolution process of metal materials and acoustic emission characteristic parameters, the extracted acoustic emission characteristic parameters are analyzed to identify the forming quality state of the stamped parts. S50: After detecting an abnormal forming state, the damage source is spatially located based on the time delay difference of the arrival time of the multi-channel acoustic emission signals, and the location information of the area where the damage occurred is output. S60: The monitoring results of the abnormal forming state and the location information of the damage source are transmitted to the die partition blank holder force control system to drive the control system to adaptively match and adjust the blank holder force parameters of each partition.
5. The online monitoring method for stamping quality based on multi-channel acoustic emission signals according to claim 4, characterized in that, Step S20 includes: synchronizing the continuously acquired multi-channel acoustic emission signals according to the stamping stroke information of the stamping equipment and the time stamp signal recorded by the third type of acoustic emission sensor (23), and dividing them into corresponding waveform signal segments according to the stamping cycle; performing bandpass filtering and baseline correction processing on each waveform signal segment; and performing VMD decomposition processing on the segmented waveform signals to obtain the input signal. .
6. The online monitoring method for stamping quality based on multi-channel acoustic emission signals according to claim 4, characterized in that, Step S30 includes: processing the input signal A threshold-triggered mechanism is used for event detection. For each detected acoustic emission event, its start time is extracted. Termination time The event duration (D) and event energy (E) are considered. Non-forming-related interference events are eliminated through multi-channel time consistency constraints. A valid forming event must satisfy the following conditions: at least a preset number of N signal channels are triggered simultaneously within the forming section, and the frequency band energy distribution of the event meets a preset frequency band proportion threshold, N≥3. For acoustic emission events that pass the multi-channel consistency constraint test, multi-dimensional feature parameters for characterizing stamping forming quality are further extracted. Based on the extracted multi-dimensional feature parameters, a multi-dimensional feature vector for characterizing the stamping forming quality state is constructed. .
7. The online monitoring method for stamping quality based on multi-channel acoustic emission signals according to claim 4, characterized in that, Step S40 includes: Standard uniaxial tensile tests were conducted on stamped metal materials, and mechanical response data and acoustic emission signals were collected simultaneously. Based on the stress-strain curve, the material deformation and damage evolution process was divided into stages, and the acoustic emission events and characteristic parameters corresponding to each stage were labeled to form a labeled sample set. ,in For the first Multidimensional feature vectors of an event Label its stage; Statistical analysis of acoustic emission characteristic parameters at each damage stage yielded the characteristic distribution patterns for each stage; [further details about the stages are needed for accurate translation.] eigenmean vector With covariance An estimate is made, and the discrimination score is used as the identification result; Sensor arrays with different propagation distances were set up, and parameters of acoustic emission events were obtained under known propagation path conditions. Based on the statistical results, a quantitative attenuation relationship model between acoustic emission characteristic parameters and propagation distance was established. Before determining the forming quality status, the amplitude correlation characteristics of each channel event are corrected by distance normalization based on the quantitative attenuation relationship model to reduce the discrimination bias caused by the difference in sensor-sound source distance. The multidimensional feature vectors extracted during the stamping process are input into the corresponding relationship model to determine the forming quality state of the stamped part and obtain the initial forming quality analysis results.
8. The online monitoring method for stamping quality based on multi-channel acoustic emission signals according to claim 7, characterized in that, Step S50 includes: When the forming quality state discrimination result meets the abnormal forming criterion, the damage source spatial localization process for the abnormal acoustic emission event is triggered. Spatial localization of damage sources is based on the time delay difference of arrival time of multi-channel acoustic emission signals; Based on the attenuation relationship model and the material propagation velocity calibration results, the arrival time of the first wave is compensated and corrected to obtain the final damage source location. When multiple acoustic emission sources exist simultaneously during the stamping process, the simultaneous localization of multiple damage sources can be achieved by clustering and separating multi-channel acoustic emission events in terms of time, frequency, and spatial characteristics.
9. The online monitoring method for stamping quality based on multi-channel acoustic emission signals according to claim 4, characterized in that, In step S60, when the monitoring results of the abnormal forming state and the location information of the damage source are transmitted to the mold partition blank holder force control system, the blank holder force parameters are adjusted according to the identified damage type, damage location and damage severity. The damage type includes at least wrinkling, local excessive thinning and cracking.
10. The online monitoring method for stamping quality based on multi-channel acoustic emission signals according to claim 4, characterized in that, The method is applied to the monitoring of forming quality during sheet metal deep drawing, bulging, or composite stamping processes.