A micro-magnetic detection calibration method for mechanical properties of ferromagnetic material structure
By using spatiotemporal registration and multi-level calibration network processing in micromagnetic detection technology, the problem of fuzzy correlation between magnetic signals and stress states in ferromagnetic materials is solved, and more accurate mechanical property calibration is achieved.
Patent Information
- Application Number
- CN202610173928.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-06
- Publication Date
- 2026-05-29
Smart Images
Figure CN122108825A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of ferromagnetic material testing technology, specifically a micromagnetic testing and calibration method for the structural mechanical properties of ferromagnetic materials. Background Technology
[0002] Current methods for testing the structural mechanical properties of ferromagnetic materials rely on inverting the stress state using micromagnetic response signals. These methods typically only collect time-series magnetic signal data without simultaneously correlating it with the dynamic distribution of the internal stress field. Conventional techniques often limit micromagnetic detection to macroscopic signal analysis, treating the magnetic response and stress field as independent information. This lack of precise temporal and spatial correspondence between the two leads to ambiguity in the relationship between the magnetic signal and the actual stress state.
[0003] Existing methods for processing micromagnetic signals often employ overall filtering or statistical averaging to separate noise from the effective signal, without considering the detailed analysis of the material's microstructural characteristics. In real materials, uneven grain boundary distribution introduces random magnetic noise, which, when superimposed on stress-induced domain reversal signals, cannot be distinguished at the grain scale using conventional methods. This results in the stress-dominant signal being masked by noise, reducing calibration accuracy. To address these issues, it is necessary to establish a synchronous correlation between the micromagnetic response and the stress field to clarify the physical field correspondence. Simultaneously, it is crucial to analyze the magnetic signal components at the grain scale to remove grain boundary noise, thereby achieving accurate extraction of the stress-dominant signal. Summary of the Invention
[0004] The purpose of this invention is to provide a micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials, so as to solve the problems mentioned in the background art.
[0005] To achieve the above objectives, the present invention provides a micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials, the method comprising:
[0006] Obtain the original micromagnetic response sequence of the material to be calibrated under stress; Construct a digital mirror of the material's internal stress field that is time-synchronized with the original micromagnetic response sequence; The original micromagnetic response sequence is spatiotemporally registered frame by frame with the digital image of the internal stress field of the material to generate a registered micromagnetic stress coupling data stream. Based on the material microstructure map, the registered micro-magnetic stress coupled data stream is decoupled at the grain scale according to the magnetic domain flipping characteristics, and the magnetic noise substrate caused by grain boundaries and the magnetic signal principal component dominated by stress are separated. Based on a preset material mechanical property degradation path model, the principal components of the stress-dominated magnetic signal are resampled in a path-dependent manner to form a standardized magnetomechanical evolution trajectory. The standardized magnetomechanical evolution trajectory is input into a pre-trained multi-level calibration network, and magnetic feature fragments corresponding to the material yielding, hardening and fracture stages are extracted layer by layer. Based on the morphology and intensity parameters of the magnetic feature segments, dynamic pattern matching is performed in the performance spectrum of the reference material. Based on the results of dynamic pattern matching, a discretized calibration index is generated for the mechanical property state of the material to be calibrated; Based on the discretized calibration index, the corresponding set of micromagnetic signal compensation factors and mechanical performance mapping function are activated and assembled, and finally an executable calibration instruction package integrating the set of micromagnetic signal compensation factors and mechanical performance mapping function is output.
[0007] Preferably, obtaining the original micromagnetic response sequence of the material to be calibrated under stress specifically includes: A mechanical load varying according to a preset load spectrum is applied to the material to be calibrated; Simultaneously collect data on the vector magnetic field intensity changes, domain wall migration rates, and Barkhausen noise energy distribution of the surface and near-surface regions of the material to be calibrated. The collected vector magnetic field intensity changes, domain wall migration rates, and Barkhausen noise energy distributions are aligned with timestamps and encoded into the original micromagnetic response sequence.
[0008] Preferably, the construction of a digital mirror of the material's internal stress field, synchronized in time with the original micromagnetic response sequence, specifically includes: Based on the geometry, boundary conditions, and preset load spectrum of the material to be calibrated, a quasi-static stress field solution based on the finite element method is performed. During the solution process, the real-time elastoplastic constitutive relation of the material to be calibrated is introduced to calculate the data of the spatially discrete stress tensor field evolution over time. The data of the spatially discrete stress tensor field evolving over time are reconstructed into a digital mirror image of the internal stress field of the material, which has the same time baseline and spatial reference frame as the original micromagnetic response sequence.
[0009] Preferably, the step of performing frame-by-frame spatiotemporal registration of the original micromagnetic response sequence with the digital image of the internal stress field of the material to generate a registered micromagnetic stress coupling data stream specifically includes: Using the equivalent stress cloud map in the digital mirror of the internal stress field of the material as a spatial reference, the magnetic signals of each measurement point in the original micromagnetic response sequence are transformed and interpolated in spatial coordinates. For each time frame, the transformed and interpolated magnetic signal value is correlated point-to-point with the equivalent stress value corresponding to the time frame to form a data point with four-dimensional attributes: position, time, magnetic signal value, and stress value. The data points with four-dimensional attributes of position, time, magnetic signal value, and stress value of all time frames are concatenated in chronological order to generate the registered micro-magnetic stress coupling data stream.
[0010] Preferably, the step of decoupling the registered micro-magnetic stress coupled data stream based on the material microstructure map at the grain scale using magnetic domain flipping characteristics specifically includes: Obtain the electron backscatter diffraction pattern or metallographic pattern of the material to be calibrated, and identify and locate the grain boundary and grain orientation distribution; Based on the grain boundary and grain orientation distribution, in the registered micromagnetic stress coupling data stream, each data point with four-dimensional attributes of position, time, magnetic signal value, and stress value is marked with its corresponding grain and distance from the nearest grain boundary. A magnetic noise transfer function based on grain boundary distance is established, and the magnetic noise substrate caused by grain boundaries is filtered out from the registered micromagnetic stress coupling data stream using the magnetic noise transfer function based on grain boundary distance. The remaining signal components are the stress-dominated principal components of the magnetic signal.
[0011] Preferably, the path-dependent resampling of the stress-dominated principal component of the magnetic signal based on a preset material mechanical property degradation path model specifically includes: The preset material mechanical property degradation path model defines the nonlinear mapping relationship between magnetic parameters and mechanical state parameters from the initial intact state of the material to the final fracture failure. The principal component of the stress-dominated magnetic signal is non-uniformly resampled along the mapping curve defined by the preset material mechanical property degradation path model in the dimension of mechanical state parameters. This ensures that the resampled data points are evenly distributed across the key stages of the preset material mechanical property degradation path model, thereby forming the standardized magnetomechanical evolution trajectory.
[0012] Preferably, the step of inputting the standardized magnetomechanical evolution trajectory into a pre-trained multi-level calibration network to extract magnetic feature fragments corresponding to the material's yielding, hardening, and fracture stages layer by layer specifically includes: The pre-trained multi-level calibration network includes a yield feature sensing layer, a hardening feature sensing layer, and a fracture feature sensing layer connected in sequence. The yield feature sensing layer extracts the feature region where the magnetic signal first shows an inflection point after the linear increase of stress from the initial part of the standardized magnetomechanical evolution trajectory, as the yield magnetic feature segment. The hardening feature sensing layer extracts the feature region where the rate of change of magnetic signal changes is adjusted in stages from the part of the standardized magnetomechanical evolution trajectory that continues from the yield inflection point, and uses it as a hardening magnetic feature segment. The fracture feature sensing layer extracts the characteristic regions where the magnetic signal exhibits severe instability or sudden drop from the end part of the standardized magnetomechanical evolution trajectory, as fracture magnetic feature segments.
[0013] Preferably, the step of performing dynamic pattern matching in the reference material property spectrum based on the morphology and intensity parameters of the magnetic feature fragments specifically includes: Calculate the inflection point sharpness of the yield magnetic feature segment, the slope change spectrum of the hardening magnetic feature segment, and the instability precursor amplitude of the fracture magnetic feature segment; The calculated inflection point sharpness, slope change spectrum, and instability precursor amplitude are used as query vectors; similarity search based on dynamic time warping algorithm is performed on the benchmark material property spectrum that stores the correspondence between known material mechanical properties and standard magnetic characteristics; Find one or more baseline performance trajectories that best match the query vector in terms of evolutionary morphology, and return the material mechanical property identifiers corresponding to these baseline performance trajectories.
[0014] Preferably, the step of generating a discretized calibration index for the mechanical property state of the material to be calibrated based on the results of dynamic pattern matching specifically includes: The material mechanical property identifiers returned by dynamic pattern matching are statistically analyzed and fused. If the matching results are concentrated on a single identifier, then the single identifier is directly used as the discretization index; If the matching results are scattered across multiple identifiers, a weighted voting process is performed based on the similarity scores of each identifier during the matching process. The identifier with the highest number of votes or its weighted combination is then encoded into the composite discrete labeling index.
[0015] Preferably, the step of activating and assembling the corresponding micromagnetic signal compensation factor set and mechanical performance mapping function according to the discretized calibration index, and finally outputting an executable calibration instruction package integrating the micromagnetic signal compensation factor set and mechanical performance mapping function, specifically includes: In the preset compensation factor library, a set of correction parameters for factors such as grain boundary noise, stress concentration effect, and detection lift-off effect are retrieved according to the discretization calibration index and assembled into the micromagnetic signal compensation factor set. In the pre-set mapping function library, a set of mathematical relationships that convert specific magnetic parameters into mechanical performance indicators such as yield strength, tensile strength, and elongation are retrieved according to the discretization calibration index and assembled into the mechanical performance mapping function; The set of micromagnetic signal compensation factors and the mechanical property mapping function, together with the discretization calibration index itself, are encapsulated into an executable calibration instruction package. The executable calibration instruction package is used to configure the equipment and algorithms for subsequent micromagnetic detection and performance determination of similar materials.
[0016] Compared with the prior art, the beneficial effects of the present invention are: A digital mirror of the material's internal stress field, synchronized with the original micromagnetic response sequence, is constructed. The original micromagnetic response sequence and this digital mirror are then spatiotemporally registered frame-by-frame to generate a micromagnetic stress-coupled data stream. This creates a dynamic, frame-by-frame correlation between the magnetic signal and the stress field in both time and space. Each frame of the magnetic signal sequence corresponds one-to-one with the information in the stress field mirror at both spatial location and time point. Changes in the magnetic signal can be directly mapped to the instantaneous distribution of the corresponding stress field, eliminating misalignment caused by spatiotemporal misalignment. The physical connection between the magnetic signal and the stress field becomes a continuous and clearly defined synchronous process, making the interpretation of the magnetic signal closer to the real mechanical environment.
[0017] Based on the material's microstructure map, the registered micro-magnetic stress coupled data stream is decoupled at the grain scale to separate the magnetic domain flipping characteristics, thus separating the magnetic noise substrate caused by grain boundaries from the principal component of the magnetic signal dominated by stress. Utilizing the detailed information on grain morphology and distribution in the microstructure map, the differences in magnetic domain flipping behavior are examined at the grain scale. The disordered magnetic disturbances caused by interface mismatch at grain boundaries are separated as the noise substrate, while the ordered magnetic domain changes driven by stress are retained as the principal component. This allows the effective mechanical response information in the magnetic signal to escape the masking effect of grain boundary noise, making the correspondence between magnetic domain flipping characteristics and stress effects more direct and pure. Attached Figure Description
[0018] Figure 1 This is a schematic diagram illustrating the working principle of the micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials described in this invention. Figure 2 A flowchart for obtaining the original micromagnetic response sequence; Figure 3 A flowchart for generating micro-magnetic stress coupling data stream; Figure 4 A bar chart comparing the magnetic characteristic parameters of the material to be calibrated and the reference material; Figure 5 This is a thermodynamic diagram showing the spatiotemporal evolution of magnetic stress coupling strength in ferromagnetic materials. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] Please see Figure 1 This invention provides a micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials. The method includes: simultaneously acquiring the micromagnetic response sequence and digital mirror image of the internal stress field of the material under stress, achieving spatiotemporal registration of the magnetic-stress signals; then decoupling the contributions of grain boundaries and stress to the magnetic signals by combining the material's microstructure spectrum; and resampling the magnetic signals based on a material mechanical property degradation path model to form a standardized magnetomechanical evolution trajectory. A multi-level calibration network is used to extract magnetic feature fragments corresponding to the material's yielding, hardening, and fracture stages. Finally, dynamic pattern matching is performed on the benchmark material property spectrum to generate a discretized calibration index, and an executable calibration instruction package integrating a micromagnetic signal compensation factor and a mechanical property mapping function is output.
[0021] In one embodiment of the present invention, see [reference] Figure 2 Mechanical loads varying according to a preset load spectrum are applied to the material to be calibrated. Simultaneously, changes in vector magnetic field intensity, domain wall migration rate, and Barkhausen noise energy distribution are acquired at the material surface and near-surface regions. The acquired signals are aligned by timestamps and encoded into an original micromagnetic response sequence. Based on the material's geometry, boundary conditions, and the preset load spectrum, a quasi-static stress field solution based on the finite element method is performed. The real-time elastoplastic constitutive relation of the material is introduced, and the data on the time evolution of the spatially discrete stress tensor field are calculated. This data is then reconstructed into a digital mirror image of the material's internal stress field, synchronized with the original micromagnetic response sequence.
[0022] In practice, a mechanical load varying according to a preset load spectrum is applied to the material to be calibrated. This preset load spectrum is typically defined as a time-load sequence, such as a load history including linear loading, holding, and unloading phases. The load application device applies force or displacement conforming to the preset load spectrum to a specific area or the entire material to be calibrated. The loading process requires precise load control and deterministic timing to produce reproducible changes in the material's mechanical state. Simultaneously, the vector magnetic field strength changes, domain wall movement rates, and Barkhausen noise energy distribution on and near the surface of the material are acquired. This process is achieved through an array of micro-magnetic sensors arranged on the material surface. In practice, the vector magnetic field strength changes are measured using high-precision fluxgate sensors or Hall effect sensor arrays; the domain wall movement rate is indirectly obtained by analyzing the time-domain characteristics of the Barkhausen noise signal; and the Barkhausen noise energy distribution is calculated by frequency domain energy integration after acquiring the raw noise signal using a dedicated probe. All sensor acquisition clocks are synchronized through a unified time server to ensure data time consistency.
[0023] The collected vector magnetic field strength changes, domain wall migration rates, and Barkhausen noise energy distributions are aligned with a unified timestamp and encoded into a raw micromagnetic response sequence. In some embodiments, the encoding process involves standardizing the data format, such as combining multi-channel data from different sensors according to a fixed data structure, like a multidimensional time series matrix. In a specific implementation, each row of this matrix corresponds to a sampling time, and each column corresponds to the value of a micromagnetic measurement parameter at a certain spatial measurement point, thus forming a raw micromagnetic response sequence containing complete spatiotemporal information. It is understood that the raw micromagnetic response sequence is the foundational data for all subsequent analyses, and the accuracy of its time reference is crucial.
[0024] Based on the geometry, boundary conditions, and preset load spectrum of the material to be calibrated, a quasi-static stress field solution based on the finite element method is performed. In practice, a finite element geometric model of the material to be calibrated must first be established, accurately reflecting its actual three-dimensional shape. The boundary conditions must be consistent with the actual constraints of the material during mechanical testing, such as fixed support surfaces or symmetry plane constraints. The load conditions are directly converted from the preset load spectrum into force or displacement boundary conditions on the corresponding nodes or surfaces of the finite element model. During the solution process, a real-time elastoplastic constitutive relation of the material to be calibrated is introduced. This constitutive relation describes the nonlinear relationship between stress and strain in the material and is typically obtained through standard mechanical property tests. The data of the spatially discrete stress tensor field evolving over time is reconstructed into a digital mirror image of the material's internal stress field with the same time baseline and spatial reference frame as the original micromagnetic response sequence. In some embodiments, the reconstruction operation includes temporal interpolation and spatial mapping. Temporal interpolation ensures that the output time sequence of the finite element solution is perfectly aligned with the time sequence of the micromagnetic acquisition. Spatial mapping involves mapping the stress data at finite element mesh nodes or integration points to the actual spatial measurement coordinates of the micromagnetic sensor using interpolation methods. In practice, for each micromagnetic measurement point, the equivalent stress value at a series of time points can be calculated using the following interpolation formula:
[0025] in: Indicates time Sensor spatial location The interpolated equivalent stress value at the point, It is the number of finite element nodes participating in the interpolation. It is a shape function at the sensor position The location corresponds to the node. The value, It is time Time node The equivalent stress.
[0026] In one embodiment of the present invention, see [reference] Figure 3 Using the equivalent stress cloud map in the digital mirror of the internal stress field of the material as a spatial reference, the magnetic signals of each measurement point in the original micromagnetic response sequence are transformed and interpolated in space. For each time frame, the transformed and interpolated magnetic signal values are correlated point-to-point with the equivalent stress values corresponding to that time frame to form data points with four-dimensional attributes of position, time, magnetic signal value, and stress value. The data points of all time frames are concatenated in time order to generate a registered micromagnetic stress coupling data stream.
[0027] In practical implementation, the equivalent stress cloud map in the digital mirror of the material's internal stress field is used as the spatial reference. The equivalent stress cloud map, obtained through post-processing of the finite element analysis, visually displays the magnitude and distribution of the equivalent stress at various points within the material at a specific time in a color cloud map format. In practice, it is necessary to first extract the equivalent stress values and their corresponding spatial coordinates for each time frame from the digital mirror of the material's internal stress field. These coordinates are typically located at nodes or integration points of the finite element model, forming a spatially discrete stress data field. Spatial coordinate transformation and interpolation are then performed on the magnetic signals at each measurement point in the original micromagnetic response sequence. The magnetic signals in the original micromagnetic response sequence are attached to the actual physical measurement positions of the micromagnetic sensor array. These position coordinates need to be transformed to a spatial reference system consistent with the finite element model, based on a unified global coordinate system. In some embodiments, the coordinate transformation involves translation, rotation, and scaling operations. These operation parameters are obtained by pre-calibrating the sensor layout and the finite element model. After the coordinate transformation is completed, the coordinates of the micromagnetic measurement points are unified with the coordinate system of the finite element model.
[0028] For the transformed micromagnetic measurement points, they may not completely coincide with the finite element nodes in space. Therefore, spatial interpolation is required to obtain the stress values corresponding to these point positions. In some embodiments, an interpolation method based on finite element shape functions is used, or a more general spatial interpolation algorithm, such as the inverse distance weighted method or the Kriging method, is employed. It can be understood that the purpose of interpolation is to assign each micromagnetic measurement point an equivalent stress value derived from a digital mirror of the material's internal stress field that precisely corresponds to its spatial position. For each time frame, the transformed and interpolated magnetic signal value is correlated point-to-point with the equivalent stress value corresponding to that time frame. In specific implementations, this means creating a data record for each micromagnetic measurement point at each sampling time. This data record contains four core attributes: the three-dimensional spatial coordinates of the measurement point, the timestamp of the sampling time point, the magnetic signal value of that point at that time, and the equivalent stress value of that point at that time obtained through interpolation.
[0029] Data points with four-dimensional attributes (position, time, magnetic signal value, and stress value) from all time frames are concatenated in chronological order to generate a registered micromagnetic stress coupling data stream. In practice, this data stream can be organized as a multi-dimensional data array or database table. Each row represents a unique observation with a specific "position-time" combination, and each column represents an observation attribute. This micromagnetic stress coupling data stream serves as the foundational data structure for subsequent feature decoupling and analysis. To more accurately handle spatially unevenly distributed measurement points, the interpolation process can employ shape function interpolation based on local finite element units. For a sensor point located within a specific quadrilateral element, its interpolated equivalent stress can be calculated using the nodal stress values and shape functions of that element. In one example, for the two-dimensional case, the formula is expressed as:
[0030] in: It is the interpolation equivalent force of the sensor point (s). It is the quadrilateral unit number The shape function of each node. These are the coordinates of the sensor point (s) in the natural coordinate system of the cell. This is the first in this unit. The equivalent stress value of each node. Optionally, for three-dimensional solid elements, the shape function and the number of nodes will be increased accordingly. Through the above steps, the original spatially discrete micromagnetic measurement data is transformed and associated with regular or semi-regular data points that match the continuous stress field, thereby realizing frame-by-frame spatiotemporal registration of micromagnetic signals and the internal stress state of the material. In a specific example scenario, a low-carbon steel plate subjected to uniaxial tension is tested. Before registration, the magnetic signal distribution map and the stress cloud map are offset in shape; after coordinate transformation and interpolation registration, the high value region of the magnetic signal and the high value region of the stress show a high degree of overlap in spatial position, directly verifying the effectiveness of the registration operation.
[0031] In one embodiment of the present invention, the electron backscatter diffraction pattern or metallographic pattern of the material to be calibrated is acquired, and the grain boundaries and grain orientation distribution are identified and located. Each data point in the registered micro-magnetic stress-coupled data stream is labeled with its corresponding grain and its distance from the nearest grain boundary. A magnetic noise transfer function based on grain boundary distance is established, and this function is used to filter out the magnetic noise substrate caused by grain boundaries. The remaining signal component is the stress-dominated principal component of the magnetic signal. Based on the nonlinear mapping relationship between magnetic parameters and mechanical state parameters defined by a preset material mechanical property degradation path model, the stress-dominated principal component of the magnetic signal is non-uniformly resampled along the mechanical state parameter dimension. This ensures that the resampled data points are uniformly distributed across the key stages of the material mechanical property degradation path model, forming a standardized magnetomechanical evolution trajectory.
[0032] In specific implementations, noise originating from the material's microstructure and the response originating from the stress state are separated from the micromagnetic signal. The signal is then standardized according to the material degradation path to obtain the electron backscatter diffraction (EPD) pattern or metallographic pattern of the material to be calibrated. The EPD pattern is obtained using the diffraction system of a scanning electron microscope, while the metallographic pattern is obtained by observing the polished and etched sample surface using an optical microscope. Digital image processing technology is used to identify and locate the grain boundaries and grain orientation distribution in the patterns. Grain boundaries are identified as continuous lines or boundaries where the orientation difference between adjacent grains exceeds a set threshold. Each grain is assigned a unique identifier and its average crystallographic orientation. In some embodiments, the grain boundary coordinates and grain identification information are converted into a digital raster map with the same coordinate origin and scaling ratio as the registered micromagnetic stress coupled data stream. This map establishes a precise correspondence between the material's micro-geometry and macroscopic spatial coordinates.
[0033] Based on the grain boundary and grain orientation distribution, in the registered micromagnetic stress coupling data stream, each data point with four-dimensional attributes (position, time, magnetic signal value, and stress value) is labeled with its corresponding grain and its distance to the nearest grain boundary. In specific implementation, for the three-dimensional spatial coordinates of the data point, the grain number of the coordinate point is directly read by querying the grain identifier digital raster map. Simultaneously, the shortest spatial distance from the coordinate point to all grain boundary lines is calculated as the distance value to the nearest grain boundary. Optionally, when the grain boundary data is stored in the form of vector line segments, the distance calculation adopts the shortest distance algorithm from point to line segment; when stored in the form of high-resolution pixel map, a distance transformation algorithm can be used to quickly generate the distance field from each coordinate point to the nearest grain boundary, and the grain identifier and distance value are added as two new attribute fields to each data point in the registered micromagnetic stress coupling data stream.
[0034] A magnetic noise transfer function based on grain boundary distance is established. This function is used to filter out the magnetic noise substrate caused by grain boundaries from the registered micro-magnetic stress coupled data stream. The grain boundary distance-based magnetic noise transfer function characterizes the quantitative relationship between the intensity of background magnetic noise caused by microstructural factors such as grain boundary defects and dislocation pile-ups in the magnetic signal and its spatial location. In some embodiments, this function is obtained by measuring the Barkhausen noise RMS value or hysteresis loop coercivity at different grain boundary distances in a fully annealed state without residual stress and performing curve fitting. In specific implementations, its mathematical form can be expressed as:
[0035] in: This indicates that at a distance of 0.5 meters from the grain boundary At the location of the magnetic noise substrate caused by grain boundaries, the quantization amplitude is... It is the amplitude coefficient reflecting the strength of grain boundary pinning to magnetic domains. It is a characteristic attenuation length that describes the range of magnetic noise influence. This represents the uniform background magnetic noise level far from grain boundaries. For each data point in the registered micromagnetic stress coupling data stream, this is determined based on its marked distance from the nearest grain boundary. Calculate the corresponding The value is subtracted from the original magnetic signal value of that data point. value.
[0036] Based on a pre-defined material mechanical property degradation path model, path-dependent resampling is performed on the principal components of the magnetic signal dominated by stress. The pre-defined material mechanical property degradation path model describes the theoretical or empirical evolution relationship between representative magnetic parameters and representative mechanical state parameters in the whole process of material from elastic deformation, yielding, plastic hardening to fracture failure in the form of curves or functions. Along the mapping curve defined by the preset material mechanical property degradation path model, the principal component of the magnetic signal dominated by stress is non-uniformly resampled in the mechanical state parameter dimension. In specific implementation, firstly, the stress values of all data points and the magnetic signal values after noise filtering are extracted from the registered micro-magnetic stress coupling data stream to form a series of scatter pairs. The preset material mechanical property degradation path model provides a reference curve. Resampling is to select a series of sampling points on the mechanical state parameter axis of the reference curve. For each selected mechanical state parameter sampling value, all scatter pairs are found in its neighborhood, and the magnetic signal values corresponding to these scatter points are weighted averaged or interpolated and assigned to the sampling point. This makes the resampled data points uniformly distributed in various key stages of the preset material mechanical property degradation path model, such as the elastic segment, near the yield point, the uniform plastic hardening segment, the necking initiation point, and the unstable fracture segment, thereby forming a standardized magnetomechanical evolution trajectory in which the magnetic signal evolves smoothly with the mechanical state parameter. In a specific example scenario, the analysis of cold-rolled low-carbon steel specimens revealed that after filtering out grain boundary noise, the principal component of the stress-dominated magnetic signal exhibited a nonlinear upward trend that was initially slow and then accelerated during the tensile process. Meanwhile, the original magnetic signal was superimposed with significant grain boundary scattering fluctuations near the yield stage. After resampling based on the degradation path set by the standard stress-strain curve, the obtained standardized trajectory showed a significantly improved consistency with the curve shape predicted by the classical magnetic model.
[0037] In one embodiment of the present invention, a standardized magnetic evolution trajectory is input into a pre-trained multi-level calibration network. This network includes a yield feature sensing layer, a hardening feature sensing layer, and a fracture feature sensing layer connected sequentially. The yield feature sensing layer extracts a feature region from the initial part of the trajectory where the magnetic signal first reaches an inflection point after a linear increase in stress, as the yield magnetic feature segment. The hardening feature sensing layer extracts a feature region from the part following the yield inflection point where the rate of change of the magnetic signal undergoes a phased adjustment, as the hardening magnetic feature segment. The fracture feature sensing layer extracts a feature region from the end part of the trajectory where the magnetic signal exhibits severe instability or a sudden drop, as the fracture magnetic feature segment. The inflection point sharpness of the yield magnetic feature segment, the slope change spectrum of the hardening magnetic feature segment, and the instability precursor amplitude of the fracture magnetic feature segment are calculated and used as query vectors. A similarity search based on a dynamic time warping algorithm is performed on a benchmark material property spectrum to find the benchmark performance trajectory that best matches the query vector in terms of evolutionary morphology, and the corresponding material mechanical property identifier is returned.
[0038] In practice, key magnetic feature segments are extracted from the standardized trajectory and pattern matching is performed to identify material properties. The standardized magnetic evolution trajectory is input into a pre-trained multi-level calibration network. This pre-trained multi-level calibration network is a deep neural network model with a hierarchical structure. Its weight parameters have been trained to convergence using a large amount of "standardized magnetic evolution trajectory - mechanical property identifier" data of known material properties. The multi-level calibration network contains a yield feature sensing layer, a hardening feature sensing layer, and a fracture feature sensing layer connected in sequence. Each layer consists of operations such as convolution, pooling, and activation functions, specifically designed to capture local morphological features in the magnetic signal evolution curve related to specific mechanical stages.
[0039] The yield feature sensing layer extracts the characteristic region where the magnetic signal first reaches an inflection point after a linear increase in stress from the initial part of the standardized magnetic evolution trajectory as the yield magnetic feature segment. In specific implementation, the yield feature sensing layer scans the initial segment of the trajectory through a sliding time window, calculates the curvature change of the data within the window, and determines the yield inflection point when the curvature continuously changes from a positive value to a negative value and exceeds a set threshold. The interval containing specific data points before and after this inflection point is then extracted as the yield magnetic feature segment. The hardening feature sensing layer extracts the characteristic region where the rate of change of the magnetic signal undergoes a phased adjustment from the portion of the standardized magnetic evolution trajectory following the yield inflection point as the hardening magnetic feature segment. It can be understood that the hardening feature sensing layer focuses on the plateau or turning point of the rate of increase of the magnetic signal after the yield point. This layer defines the boundary of the hardening magnetic feature segment by analyzing the local extrema and stability of the first derivative of the trajectory. The fracture feature sensing layer extracts the characteristic regions where the magnetic signal exhibits severe instability or sudden drop from the end of the standardized magnetomechanical evolution trajectory as fracture magnetic feature segments. In some embodiments, the fracture feature sensing layer monitors the standard deviation of the signal amplitude or the energy of the high-frequency components at the end of the trajectory. When these indicators rise sharply and exceed the critical value, the segment before instability is truncated to form a fracture magnetic feature segment.
[0040] The inflection point sharpness of the yield magnetic feature segment, the slope change spectrum of the hardening magnetic feature segment, and the instability precursor amplitude of the fracture magnetic feature segment are calculated. Inflection point sharpness is defined as the absolute value of the second derivative of the yield magnetic feature segment at the inflection point. The slope change spectrum is obtained by piecewise linear fitting of the hardening magnetic feature segment, extracting the slope values of each segment and calculating their statistical distribution characteristics. The instability precursor amplitude is quantified by calculating the ratio of the average amplitude of several data points before the severe instability point of the fracture magnetic feature segment to the overall average amplitude. These calculated inflection point sharpness, slope change spectrum, and instability precursor amplitude are used as query vectors. The query vector is a multi-dimensional feature array used to search for similar patterns in the database. A similarity search based on a dynamic time warping algorithm is performed on a benchmark material property spectrum map storing the correspondence between known material mechanical properties and standard magnetic features. The benchmark material property spectrum map is a database where each record contains a mechanical property identifier of a known material and its corresponding standard magnetic feature evolution trajectory or feature vector.
[0041] The similarity search process based on the dynamic time warping algorithm involves aligning the query vector with the standard feature vector stored in each record of the benchmark material property spectrum and calculating the distance. The dynamic time warping algorithm can handle sequences of different lengths or local scaling, find the optimal matching path, and calculate the minimum cumulative distance, which serves as the similarity metric. It finds one or more benchmark property trajectories that best match the query vector in terms of evolutionary morphology and returns the corresponding material mechanical property identifiers. In some embodiments, "best match" refers to all records with a cumulative distance less than a preset threshold, or the top K records sorted by distance from smallest to largest. The returned material mechanical property identifiers include the material grade, yield strength range, and tensile strength value. For a clearer representation of part of the benchmark material property spectrum, refer to Table 1, which provides a simplified example structure.
[0042] Table 1: Example Record Table of Performance Spectra of Reference Materials
[0043] In practice, the similarity between the query vector and the standard feature parameters in Table 1 can be calculated using a weighted Euclidean distance formula:
[0044] in: This represents the weighted distance between the query vector and a reference record in the spectrogram. These represent the inflection point sharpness, slope change spectrum variance, and instability precursor amplitude of the query vector, respectively. These represent the corresponding standard feature parameters in the reference record. These are weighting coefficients pre-set based on the contribution of each feature to performance discrimination. Optionally, distance calculation can also be performed directly on the original feature segment sequence using a dynamic time warping algorithm. The specific implementation of the dynamic time warping algorithm involves processing the extracted yield magnetic feature segments, hardening magnetic feature segments, and fracture magnetic feature segments directly as time series data. For each feature segment, the algorithm constructs a cumulative distance matrix to align the query sequence with the standard feature sequence stored in the benchmark material performance spectrum. The dynamic time warping algorithm calculates the local distance between corresponding data points in the two sequences point by point and uses dynamic programming principles to find an optimal curvature path that minimizes the cumulative distance along that path. This minimum cumulative distance is the similarity metric, used to evaluate the degree of matching between the query sequence and the benchmark sequence in terms of evolutionary morphology. Finally, the most similar benchmark performance trajectory is selected based on this distance value.
[0045] See Figure 4This is a bar chart comparing the magnetic characteristic parameters of the material to be calibrated and the reference materials. It compares three magnetic characteristic parameters of the material to be calibrated with three types of reference materials: Q235, Q345, and 304 stainless steel. This chart is a core tool in the "feature quantization and pattern matching" stage of micro-magnetic testing calibration, visually demonstrating the similarity of the magnetic characteristics between the material to be calibrated and the reference materials; it provides prior information for subsequent Dynamic Time Warping (DTW) matching, narrowing down the matching range of reference materials; and it supports the generation of the "discrete calibration index," assisting in determining the mechanical property identifier of the material to be calibrated. By comparing the magnetic characteristic parameters of the material to be calibrated and the reference materials, the most similar reference material category can be quickly identified, reducing the computational load of subsequent Dynamic Time Warping (DTW) matching and improving calibration efficiency.
[0046] In one embodiment of the present invention, the material mechanical property identifiers returned by dynamic pattern matching are statistically analyzed and fused. If the matching results are concentrated on a single identifier, it is directly used as the discretization calibration index. If the matching results are scattered across multiple identifiers, a weighted vote is performed based on the similarity scores of each identifier, and the identifier with the highest number of votes or its weighted combination is encoded as a composite discretization calibration index. Based on the discretization calibration index, correction parameters for factors such as grain boundary noise, stress concentration effect, and detection lift-off are retrieved from a pre-set compensation factor library and assembled into a micro-magnetic signal compensation factor set. Simultaneously, mathematical relationships that convert specific magnetic parameters into mechanical property indicators such as yield strength, tensile strength, and elongation are retrieved from a pre-set mapping function library and assembled into a mechanical property mapping function. The micro-magnetic signal compensation factor set, the mechanical property mapping function, and the discretization calibration index are encapsulated into an executable calibration instruction package.
[0047] In practice, a defined calibration index is generated based on the pattern matching results, and the final executable configuration package is output. The material mechanical property identifiers returned by dynamic pattern matching are statistically analyzed and fused. The material mechanical property identifiers returned by dynamic pattern matching are a list containing multiple candidate identifiers and their corresponding similarity scores. In practice, the statistical operation includes calculating the frequency of each unique material mechanical property identifier in the returned list, while the fusion operation involves a comprehensive analysis of these identifiers and their associated confidence levels. If the matching results are concentrated on a single identifier, that single identifier is directly used as the discretized calibration index. The criteria for "concentrated on a single identifier" can be: the frequency of the identifier's occurrence exceeds a preset threshold, or the similarity score of the identifier is significantly higher than other identifiers.
[0048] If the matching results are scattered across multiple identifiers, a weighted voting process is performed based on the similarity scores of each identifier during the matching process. The identifier with the highest number of votes, or its weighted combination, is encoded as a composite discretized labeling index. In some embodiments, the weighted voting is implemented by normalizing the similarity score of each returned identifier, converting it into weights, and then summing all the weights for the same identifier to obtain its total number of votes. The material mechanical property identifier with the highest total number of votes is selected as the discretized labeling index. Optionally, when the advantage of the identifier with the highest number of votes is not significant, a weighted combination method can be used to generate a composite discretized labeling index, for example, represented by a vector containing multiple identifiers and their probability weights.
[0049] Based on the discretized calibration index, the corresponding set of micromagnetic signal compensation factors and the mechanical property mapping function are activated and assembled. In a pre-set compensation factor library, a set of correction parameters for factors such as grain boundary noise, stress concentration effect, and detection lift-off effect are retrieved according to the discretized calibration index and assembled into a micromagnetic signal compensation factor set. In specific implementations, the pre-set compensation factor library is a relational database or lookup table, using material identifiers or performance ranges as primary keys, storing a series of correction coefficients obtained through extensive experimental calibration or theoretical derivation for different materials. For example, for grain boundary noise, the compensation factor includes a scaling factor to correct the characteristic attenuation length δ in the magnetic noise transfer function; for the lift-off effect, it includes a signal attenuation compensation coefficient that varies with the lift-off distance. The retrieval process performs exact matching or fuzzy query based on the discretized calibration index, extracting and packaging all matched correction parameters into a structured dataset, i.e., the micromagnetic signal compensation factor set. In some embodiments, for composite discretized calibration indexes, the micromagnetic signal compensation factor set can be assembled by weighted averaging of the compensation factors corresponding to each component identifier.
[0050] In a pre-built mapping function library, a set of mathematical relationships that convert specific magnetic parameters into mechanical performance indicators such as yield strength, tensile strength, and elongation are retrieved based on the discretization calibration index. These relationships are then assembled into mechanical performance mapping functions. The pre-built mapping function library stores quantitative conversion models between magnetic characteristics and mechanical performance indicators for different material categories. These models are linear regression equations, polynomials, or more complex nonlinear functions. The retrieval process also relies on the discretization calibration index, and the extracted mathematical relationships are encapsulated into a set of callable functions or calculation formulas, forming the mechanical performance mapping functions. The set of micromagnetic signal compensation factors and the mechanical performance mapping functions, along with the discretization calibration index itself, are encapsulated into an executable calibration instruction package. This executable calibration instruction package is a structured file, such as using JSON, XML, or binary format, containing all the configuration information required for subsequent real-time signal compensation and performance inversion by micromagnetic detection equipment or analysis algorithms. In practical implementation, the encapsulation operation writes the micromagnetic signal compensation factor set, mechanical property mapping function, and discretized calibration index as different data blocks into a file, while attaching necessary metadata such as version number, generation timestamp, and applicable scope description. It can be understood that the executable calibration instruction package is the final output of the method of this invention, used to directly configure the equipment and algorithms for subsequent micromagnetic detection and performance determination of similar materials. In a specific example scenario, the discretized calibration index generated after analyzing an unknown steel plate is "Q345". Based on this, the retrieved and assembled micromagnetic signal compensation factor set includes a grain boundary noise attenuation length correction coefficient of 1.1 and a lift-off compensation coefficient of 0.95; the mechanical property mapping function includes the yield strength calculation formula "YS=145+185". "α", this information is packaged into an instruction package file named "Calibration_Q345_v1.2.cfg". After loading this file, the detection system can use the embedded parameters and formulas to accurately interpret the micro-magnetic signals and evaluate the mechanical properties of similar steel plates. In the scenario of generating a composite discretized calibration index through weighted voting, the final value of the index can be calculated using the following formula:
[0051] in: The numerical representation of the composite discretized index generated by the encoding. It refers to the number of different material mechanical property labels that participated in the voting. It is the first A material mechanical property identifier, It is a function that maps material identifiers to their standard numerical codes. It is the first The normalized weight of an identifier is obtained by processing the similarity scores of all returned instances of that identifier using the softmax function.
[0052] See Figure 5 This is a spatiotemporal evolution thermogram of the magnetic stress coupling intensity of a ferromagnetic material. It shows the dynamic change of the magnetic stress coupling intensity at different spatial locations over time during the stress process. The difference in coupling intensity at different locations reflects the spatial non-uniformity of stress distribution within the material. This graph is a core tool in the "micro-magnetic stress coupling data flow analysis" stage of micro-magnetic testing of ferromagnetic materials. It intuitively presents the spatiotemporal correlation between stress and magnetic response, helps to locate stress concentration areas within the material, and divides the evolution stages of the material's mechanical state (loading rise stage, peak stage, damage decline stage). This provides a temporal basis for subsequent decoupling of magnetic domain reversal characteristics; it supports the identification of the material's mechanical property degradation path and is a key carrier for "spatiotemporal registration data visualization" in micro-magnetic testing calibration.
[0053] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0054] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A micromagnetic testing and calibration method for the structural mechanical properties of ferromagnetic materials, characterized in that, include: Obtain the original micromagnetic response sequence of the material to be calibrated under stress; Construct a digital mirror of the material's internal stress field that is time-synchronized with the original micromagnetic response sequence; The original micromagnetic response sequence is spatiotemporally registered frame by frame with the digital image of the internal stress field of the material to generate a registered micromagnetic stress coupling data stream. Based on the material microstructure map, the registered micromagnetic stress coupled data stream is decoupled at the grain scale according to the magnetic domain flipping characteristics, and the magnetic noise substrate caused by grain boundaries and the magnetic signal principal component dominated by stress are separated. Based on a preset material mechanical property degradation path model, the principal components of the stress-dominated magnetic signal are resampled in a path-dependent manner to form a standardized magnetomechanical evolution trajectory. The standardized magnetomechanical evolution trajectory is input into a pre-trained multi-level calibration network, and magnetic feature fragments corresponding to the material yielding, hardening and fracture stages are extracted layer by layer. Based on the morphology and intensity parameters of the magnetic feature segments, dynamic pattern matching is performed in the performance spectrum of the reference material. Based on the results of dynamic pattern matching, a discretized calibration index is generated for the mechanical property state of the material to be calibrated; Based on the discretized calibration index, the corresponding set of micromagnetic signal compensation factors and mechanical performance mapping function are activated and assembled, and finally an executable calibration instruction package integrating the set of micromagnetic signal compensation factors and mechanical performance mapping function is output.
2. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 1, characterized in that, The acquisition of the original micromagnetic response sequence of the material to be calibrated under stress specifically includes: A mechanical load varying according to a preset load spectrum is applied to the material to be calibrated; Simultaneously collect data on the vector magnetic field intensity changes, domain wall migration rates, and Barkhausen noise energy distribution of the surface and near-surface regions of the material to be calibrated. The collected vector magnetic field intensity changes, domain wall migration rates, and Barkhausen noise energy distributions are aligned with timestamps and encoded into the original micromagnetic response sequence.
3. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 2, characterized in that, The construction of a digital mirror of the material's internal stress field, synchronized with the original micromagnetic response sequence, specifically includes: Based on the geometry, boundary conditions, and preset load spectrum of the material to be calibrated, a quasi-static stress field solution based on the finite element method is performed. During the solution process, the real-time elastoplastic constitutive relation of the material to be calibrated is introduced to calculate the data of the spatially discrete stress tensor field evolution over time. The data of the spatially discrete stress tensor field evolving over time are reconstructed into a digital mirror image of the internal stress field of the material, which has the same time baseline and spatial reference frame as the original micromagnetic response sequence.
4. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 3, characterized in that, The step of performing frame-by-frame spatiotemporal registration of the original micromagnetic response sequence with the digital image of the internal stress field of the material to generate a registered micromagnetic stress coupling data stream specifically includes: Using the equivalent stress cloud map in the digital mirror of the internal stress field of the material as a spatial reference, the magnetic signals of each measurement point in the original micromagnetic response sequence are transformed and interpolated in spatial coordinates. For each time frame, the transformed and interpolated magnetic signal value is correlated point-to-point with the equivalent stress value corresponding to the time frame to form a data point with four-dimensional attributes: position, time, magnetic signal value, and stress value. The data points with four-dimensional attributes of position, time, magnetic signal value, and stress value of all time frames are concatenated in chronological order to generate the registered micro-magnetic stress coupling data stream.
5. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 4, characterized in that, The process of decoupling the registered micro-magnetic stress coupled data stream based on the material's microstructure map at the grain scale using magnetic domain flipping characteristics specifically includes: Obtain the electron backscatter diffraction pattern or metallographic pattern of the material to be calibrated, and identify and locate the grain boundary and grain orientation distribution; Based on the grain boundary and grain orientation distribution, in the registered micromagnetic stress coupling data stream, each data point with four-dimensional attributes of position, time, magnetic signal value, and stress value is marked with its corresponding grain and distance from the nearest grain boundary. A magnetic noise transfer function based on grain boundary distance is established, and the magnetic noise substrate caused by grain boundaries is filtered out from the registered micromagnetic stress coupling data stream using the magnetic noise transfer function based on grain boundary distance. The remaining signal components are the stress-dominated principal components of the magnetic signal.
6. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 5, characterized in that, The path-dependent resampling of the stress-dominated principal component of the magnetic signal, based on the preset material mechanical property degradation path model, specifically includes: The preset material mechanical property degradation path model defines the nonlinear mapping relationship between magnetic parameters and mechanical state parameters from the initial intact state of the material to the final fracture failure. The principal component of the stress-dominated magnetic signal is non-uniformly resampled along the mapping curve defined by the preset material mechanical property degradation path model in the dimension of mechanical state parameters. This ensures that the resampled data points are evenly distributed across the key stages of the preset material mechanical property degradation path model, thereby forming the standardized magnetomechanical evolution trajectory.
7. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 6, characterized in that, The process of inputting the standardized magnetic evolution trajectory into a pre-trained multi-level calibration network to extract magnetic feature fragments corresponding to the material's yielding, hardening, and fracture stages layer by layer specifically includes: The pre-trained multi-level calibration network includes a yield feature sensing layer, a hardening feature sensing layer, and a fracture feature sensing layer connected in sequence. The yield feature sensing layer extracts the feature region where the magnetic signal first shows an inflection point after the linear increase of stress from the initial part of the standardized magnetomechanical evolution trajectory, as the yield magnetic feature segment. The hardening feature sensing layer extracts the feature region where the rate of change of magnetic signal changes is adjusted in stages from the part of the standardized magnetomechanical evolution trajectory that continues from the yield inflection point, and uses it as a hardening magnetic feature segment. The fracture feature sensing layer extracts the characteristic regions where the magnetic signal exhibits severe instability or sudden drop from the end part of the standardized magnetomechanical evolution trajectory, as fracture magnetic feature segments.
8. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 7, characterized in that, The dynamic pattern matching in the reference material property spectrum based on the morphology and intensity parameters of the magnetic feature segments specifically includes: Calculate the inflection point sharpness of the yield magnetic feature segment, the slope change spectrum of the hardening magnetic feature segment, and the instability precursor amplitude of the fracture magnetic feature segment; The calculated inflection point sharpness, slope change spectrum, and instability precursor amplitude are used as query vectors; similarity search based on dynamic time warping algorithm is performed on the benchmark material property spectrum that stores the correspondence between known material mechanical properties and standard magnetic characteristics; Find one or more baseline performance trajectories that best match the query vector in terms of evolutionary morphology, and return the material mechanical property identifiers corresponding to these baseline performance trajectories.
9. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 8, characterized in that, Based on the results of dynamic pattern matching, a discrete calibration index is generated for the mechanical property state of the material to be calibrated, specifically including: The material mechanical property identifiers returned by dynamic pattern matching are statistically analyzed and fused. If the matching results are concentrated on a single identifier, then the single identifier is directly used as the discretization index; If the matching results are scattered across multiple identifiers, a weighted voting process is performed based on the similarity scores of each identifier during the matching process. The identifier with the highest number of votes or its weighted combination is then encoded into the composite discrete labeling index.
10. The micromagnetic detection and calibration method for the structural mechanical properties of ferromagnetic materials according to claim 9, characterized in that, The step involves activating and assembling the corresponding set of micromagnetic signal compensation factors and the mechanical performance mapping function based on the discretized calibration index, and finally outputting an executable calibration instruction package that integrates the set of micromagnetic signal compensation factors and the mechanical performance mapping function, specifically including: In the preset compensation factor library, a set of correction parameters for factors such as grain boundary noise, stress concentration effect, and detection lift-off effect are retrieved according to the discretization calibration index and assembled into the micromagnetic signal compensation factor set. In the pre-set mapping function library, a set of mathematical relationships that convert specific magnetic parameters into mechanical performance indicators such as yield strength, tensile strength, and elongation are retrieved according to the discretization calibration index and assembled into the mechanical performance mapping function; The set of micromagnetic signal compensation factors and the mechanical property mapping function, together with the discretization calibration index itself, are encapsulated into an executable calibration instruction package. The executable calibration instruction package is used to configure the equipment and algorithms for subsequent micromagnetic detection and performance determination of similar materials.