Stress measurement system for spring fatigue condition

Abstract

The application relates to the technical field of function test, and discloses a stress measurement system for spring fatigue state, which comprises an excitation unit, a sensing unit and a signal processing unit. The excitation unit applies an alternating excitation magnetic field to the spring to generate a stress-modulated induced electromotive force signal. The sensing unit collects magnetic characteristic parameters by using a quadrature differential sensing structure. The signal processing unit obtains phase components in a tangential signal and a radial reference signal, calculates a dynamic spiral angle deflection based on a phase difference, determines a dynamic differential compensation coefficient to extract a tangential principal stress component and compensate for a magnetic flux leakage vector deviation. Through a space posture modulation and lift gap fluctuation suppression mechanism, the application realizes dynamic alignment of a sensor measurement main shaft and a spring stress tensor main shaft, effectively eliminates geometric deflection errors caused by large amplitude deformation, and ensures that the system has high signal-to-noise ratio and baseline stability under cyclic load.

Description

Technical Field

[0001] This invention relates to a stress measurement system for spring fatigue conditions, belonging to the field of functional testing technology. Background Technology

[0002] Currently, in the suspension systems of transport machinery and reciprocating energy storage systems for aviation, fatigue degradation monitoring of high-frequency, high-load springs is crucial for ensuring structural safety. The mainstream method for measuring spring stress is to use contact resistance strain gauges, which sense deformation and determine stress data by attaching them to the spring surface. This method has good measurement accuracy in static calibration and short-term testing. However, as testing enters the high-cycle fatigue stage with millions of cycles, the physical contact measurement method encounters a reliability bottleneck. Because the spring frequently undergoes large-stroke compression under alternating loads, the bonding interface of the strain gauge suffers cumulative damage due to the cyclic action of shear stress, which can easily lead to zero-point drift or physical peeling. This mismatch between the lifespan of the measurement system and the fatigue cycle of the measured object leads to the risk of failure in stress monitoring throughout its entire life cycle.

[0003] To avoid contact failure, this invention attempts to conduct non-contact measurement using the magnetoelastic effect. However, under real-world conditions, the heat generated by friction within the spring lattice is conducted to the surface due to physical inertia, causing the apparent thermal radiation signal to lag behind the evolution of the intrinsic magnetic state within the material. Static temperature compensation alone cannot resolve the spatiotemporal phase misalignment of thermomagnetic characteristics under high-frequency conditions. Furthermore, the spring's attitude deflection under large stroke compression, accompanied by helix angle changes, causes vector leakage in the induced magnetic circuit, making it difficult for existing magnetic measurement schemes to maintain signal fidelity under complex geometric deformations. There are also shortcomings in the software control layer; for example, Chinese invention patent CN112083059B discloses a filter... In addition to the method of removing interference from the top surface of the rail, the Spearman correlation coefficient between adjacent sensors and multi-axis components is calculated to identify and subtract the removal interference signal. This type of filtering algorithm is based on the topological regularity of the surface of the workpiece being tested and the consistent evolution space of the magnetic field. The spring fatigue process is accompanied by the evolution of the intrinsic magnetic state of the material and the deep coupling of the three-dimensional helix angle deflection. The dynamic displacement of the principal axis of the stress tensor is a physically deterministic vector distortion. The statistical correlation algorithm cannot establish the constitutive mapping relationship between the helix angle distortion and the leakage magnetic vector. Under extreme stroke, the peeling tangential principal stress component is blocked. The thermal inertia of the metal material causes the apparent thermal radiation signal to lag behind the evolution of the internal magnetic state of the material. The existing scheme lacks a mechanism for compensating for the spatiotemporal phase misalignment of thermomagnetic characteristics.

[0004] Therefore, how to determine the characteristic parameters that can decouple the heat conduction delay and compensate for spatial geometric distortion in order to obtain the true stress mapping in the high-frequency fatigue process has become the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: A stress measurement system for spring fatigue state, comprising: The excitation unit is used to apply an alternating excitation magnetic field to the spring under test, generating a stress-modulated induced electromotive force signal. The sensing unit includes a tangential measuring coil and a radial reference coil that are non-contactly arranged around the periphery of the spring under test. The tangential measuring coil and the radial reference coil form an orthogonal differential sensing structure for acquiring the magnetic characteristic parameters of the spring under test under cyclic load. The signal processing unit, connected to the sensing unit, is used to acquire the tangential signal output by the tangential measuring coil and the radial reference signal output by the radial reference coil, and to acquire the phase component of the tangential signal modulated by the principal stress of the spring under test. Specifically, the signal processing unit determines the dynamic helix angle deflection of the spring under test during the compression stroke based on the phase difference between the radial reference signal and the tangential signal, and determines the dynamic differential compensation coefficient according to the preset linear mapping relationship between the dynamic helix angle deflection and the compensation weight. The tangential signal is then weighted and subtracted using the dynamic differential compensation coefficient to extract the tangential principal stress component and compensate for the leakage magnetic vector deviation caused by the dynamic helix angle deflection.

[0006] Preferably, the excitation unit includes a piezoelectric ceramic vibrator and an excitation winding; the piezoelectric ceramic vibrator is used to generate high-frequency vibrations with a frequency range of 10kHz to 100kHz according to the frequency-modulated electrical signal output by the signal processing unit, so as to modulate the magnetic domain arrangement state inside the spring under test; the excitation winding is disposed at the end of the spring under test and is used to output a detection magnetic field whose frequency is controlled by feedback from the signal processing unit, so that the spring under test generates a stress-modulated leakage magnetic signal.

[0007] Preferably, the sensing unit further includes a shield; the tangential measuring coil and the radial reference coil are both disposed inside the shield to reduce the electromagnetic interference of the ambient magnetic field on the orthogonal differential sensing structure; the sensing axis of the tangential measuring coil is parallel to the axis of the spring being measured, and the sensing axis of the radial reference coil is perpendicular to the axis of the spring being measured.

[0008] Preferably, the signal processing unit includes a preamplifier circuit, a phase-sensitive detector module, and an arithmetic module; the preamplifier circuit is used to perform differential amplification processing on the tangential signal and the radial reference signal; the phase-sensitive detector module is used to extract the phase offset of the tangential signal relative to the reference excitation signal; and the arithmetic module is used to process the phase decoupling logic to compensate for the signal hysteresis caused by the thermal inertia of the spring under test.

[0009] Preferably, the calculation module uses the following formula to calculate the tangential principal stress components: ,in, These are the extracted tangential principal stress eigenvalues. The phase value of the tangential signal. Let w be the phase value of the radial reference signal, and w be the dynamic differential compensation coefficient. This represents the dynamic helix angle deflection.

[0010] Preferably, the signal processing unit is also used to monitor the amplitude of the radial reference signal and determine the lift-off gap offset between the sensing unit and the surface of the measured spring in real time based on the amplitude; the signal processing unit adjusts the gain coefficient of the tangential principal stress component according to the lift-off gap offset to suppress common-mode interference caused by system vibration.

[0011] Preferably, the excitation unit is also used to complete the magnetic field frequency sweep program; the signal processing unit is used to capture the energy peak point of the induced electromotive force signal and determine the corresponding frequency as the resonant operating frequency, while controlling the excitation unit to output the alternating excitation magnetic field at the resonant operating frequency.

[0012] Preferably, the system also includes an environmental monitoring module for acquiring temperature data around the spring under test; and a signal processing unit for inputting the temperature data into a preset temperature compensation model and outputting a correction value of the dynamic differential compensation coefficient to eliminate the influence of environmental temperature changes on the magnetic characteristic parameters.

[0013] Preferably, the signal processing unit is used to record the change sequence of the tangential principal stress components and calculate the zero-point drift rate within the continuous monitoring period; when the zero-point drift rate exceeds the preset damage judgment threshold, the signal processing unit outputs an alarm signal characterizing the fatigue damage of the tested spring.

[0014] Preferably, the system also includes a data output interface for generating evaluation data characterizing the residual life of the tested spring based on the tangential principal stress components and a preset degradation model, and inputting the evaluation data into the digital simulation model of the tested spring to realize the evolution analysis of the fatigue state of the tested spring.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. In stress measurement of springs under fatigue conditions, this invention achieves temporal phase alignment between the intrinsic thermal state and the surface radiative thermal state of the material, compensating for measurement lag distortion caused by the thermal conduction inertia of the metal. The invention uses a thermomagnetic decoupling processor to synchronously acquire the time derivative sequences of the radial reference coil output signal and the infrared temperature measurement array output signal. It then uses the sliding cross-correlation operation of the two derivative sequences to extract characteristic time delay parameters representing thermal conduction lag, and performs time-axis backtracking on the surface temperature signal accordingly. This mechanism solves the spatiotemporal isolation problem between the temperature rise in the core stress area of ​​the spring and the surface temperature measurement point during high-frequency fatigue, transforming the basis for correcting the reference permeability from the apparent lag temperature to the synchronous constitutive thermal state. This ensures that in the later stages of aging tests with tens of millions of cycles, the stress measurement data can still penetrate the material's thermal inertia and accurately map the deterioration trajectory of the internal stress.

[0016] 2. To achieve dynamic resonance alignment between the sensor's measurement spindle and the measured spring's force tensor spindle, eliminating geometric deflection errors caused by large-scale deformation, this invention utilizes the impedance phase difference generated at the orthogonal differential probe group by the probe component output by the excitation unit to calculate the dynamic helix angle deflection of the spring during the compression stroke in real time, and dynamically updates the differential weight coupling matrix based on this deflection. This mechanism upgrades the traditional static hardware differential to a dynamic logic differential modulated by spatial attitude, compensating for leakage magnetic vector caused by the physical flattening of the spring helix under extreme loads, and ensuring that the differential topology can always accurately extract the pure tangential principal stress component throughout the entire fatigue stroke.

[0017] 3. To achieve closed-loop suppression of lift-off gap fluctuations caused by high-frequency bouncing and improve the long-term baseline stability of the non-contact sensing architecture, this invention constructs an orthogonal differential sensing topology by non-contactly surrounding the spring under test with a tangential measurement coil and a radial reference coil. The high sensitivity of the radial signal to displacement fluctuations is used to cancel common-mode interference in the tangential signal. Since the radial magnetic circuit is theoretically not modulated by the axial principal stress, the processor automatically filters out random fluctuation noise in the gap between the probe and the spring surface caused by high-frequency vibration of the fatigue frame by performing differential operations. This allows the system to obtain a signal-to-noise ratio no lower than that of contact strain measurement without contacting the spring surface, avoiding the problem of premature failure of the test system caused by fatigue peeling of the adhesive layer in traditional sensors. Attached Figure Description

[0018] Figure 1 This is a diagram showing the hardware topology and orthogonal sensing structure of the spring fatigue stress monitoring system of the present invention; Figure 2 This invention integrates multi-dimensional feature decoupling signal processing logic and stress analysis block diagram.

[0019] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0021] A stress measurement system for spring fatigue state, comprising: The excitation unit is used to apply an alternating excitation magnetic field to the spring under test, generating a stress-modulated induced electromotive force signal. The sensing unit includes a tangential measuring coil and a radial reference coil that are non-contactly arranged around the periphery of the spring under test. The tangential measuring coil and the radial reference coil form an orthogonal differential sensing structure for acquiring the magnetic characteristic parameters of the spring under test under cyclic load. The signal processing unit, connected to the sensing unit, is used to acquire the tangential signal output by the tangential measuring coil and the radial reference signal output by the radial reference coil, and to acquire the phase component of the tangential signal modulated by the principal stress of the spring under test. Specifically, the signal processing unit determines the dynamic helix angle deflection of the spring under test during the compression stroke based on the phase difference between the radial reference signal and the tangential signal, and determines the dynamic differential compensation coefficient according to the preset linear mapping relationship between the dynamic helix angle deflection and the compensation weight. The tangential signal is then weighted and subtracted using the dynamic differential compensation coefficient to extract the tangential principal stress component and compensate for the leakage magnetic vector deviation caused by the dynamic helix angle deflection.

[0022] Preferably, the excitation unit includes a piezoelectric ceramic vibrator and an excitation winding; the piezoelectric ceramic vibrator is used to generate high-frequency vibrations with a frequency range of 10kHz to 100kHz according to the frequency-modulated electrical signal output by the signal processing unit, so as to modulate the magnetic domain arrangement state inside the spring under test; the excitation winding is disposed at the end of the spring under test and is used to output a detection magnetic field whose frequency is controlled by feedback from the signal processing unit, so that the spring under test generates a stress-modulated leakage magnetic signal.

[0023] Preferably, the sensing unit further includes a shield; the tangential measuring coil and the radial reference coil are both disposed inside the shield to reduce the electromagnetic interference of the ambient magnetic field on the orthogonal differential sensing structure; the sensing axis of the tangential measuring coil is parallel to the axis of the spring being measured, and the sensing axis of the radial reference coil is perpendicular to the axis of the spring being measured.

[0024] Preferably, the signal processing unit includes a preamplifier circuit, a phase-sensitive detector module, and an arithmetic module; the preamplifier circuit is used to perform differential amplification processing on the tangential signal and the radial reference signal; the phase-sensitive detector module is used to extract the phase offset of the tangential signal relative to the reference excitation signal; and the arithmetic module is used to process the phase decoupling logic to compensate for the signal hysteresis caused by the thermal inertia of the spring under test.

[0025] Preferably, the calculation module uses the following formula to calculate the tangential principal stress components: ,in, These are the extracted tangential principal stress eigenvalues. The phase value of the tangential signal. Let w be the phase value of the radial reference signal, and w be the dynamic differential compensation coefficient. This represents the dynamic helix angle deflection.

[0026] Preferably, the signal processing unit is also used to monitor the amplitude of the radial reference signal and determine the lift-off gap offset between the sensing unit and the surface of the measured spring in real time based on the amplitude; the signal processing unit adjusts the gain coefficient of the tangential principal stress component according to the lift-off gap offset to suppress common-mode interference caused by system vibration.

[0027] Preferably, the excitation unit is also used to complete the magnetic field frequency sweep program; the signal processing unit is used to capture the energy peak point of the induced electromotive force signal and determine the corresponding frequency as the resonant operating frequency, while controlling the excitation unit to output the alternating excitation magnetic field at the resonant operating frequency.

[0028] Preferably, the system also includes an environmental monitoring module for acquiring temperature data around the spring under test; and a signal processing unit for inputting the temperature data into a preset temperature compensation model and outputting a correction value of the dynamic differential compensation coefficient to eliminate the influence of environmental temperature changes on the magnetic characteristic parameters.

[0029] Preferably, the signal processing unit is used to record the change sequence of the tangential principal stress components and calculate the zero-point drift rate within the continuous monitoring period; when the zero-point drift rate exceeds the preset damage judgment threshold, the signal processing unit outputs an alarm signal characterizing the fatigue damage of the tested spring.

[0030] Preferably, the system also includes a data output interface for generating evaluation data characterizing the residual life of the tested spring based on the tangential principal stress components and a preset degradation model, and inputting the evaluation data into the digital simulation model of the tested spring to realize the evolution analysis of the fatigue state of the tested spring.

[0031] Example 1: In the field monitoring of high-frequency high-load fatigue deterioration of suspension springs in the chassis system of a transport equipment, the spring under test is in a continuous vibration state under cyclic alternating load. The internal lattice friction of the material generates heat and induces nonlinear reversal of the magnetic domains inside the material. Due to the thermal inertia of the metal material, the surface thermal radiation signal captured by the infrared temperature measurement array lags behind the evolution of the intrinsic magnetic state inside the material on the time axis. In addition, the spring generates a dynamic flattening of the helix angle during the compression stroke, which causes a dynamic angle between the principal axis of the sensing coil and the principal axis of the force tensor and induces leakage of the leakage magnetic vector.

[0032] The excitation unit outputs a driving current containing an alternating component in the first frequency band and a detection component in the second frequency band to the spring under test. An orthogonal differential probe group positioned around the spring collects stress-modulated magnetic characteristic parameters. The thermomagnetic decoupling processor simultaneously acquires the time derivative sequences of the low-frequency thermal drift signal output from the infrared thermometer array and the time derivative sequence of the radial induced electromotive force envelope. The characteristic time delay parameters characterizing thermal conduction hysteresis are extracted using a sliding cross-correlation operation on the two time derivative sequences. The thermomagnetic decoupling processor is configured with a ring-shaped FIFO buffer with a capacity of 5000 sampling points, and the sliding cross-correlation operation has a time window... The aperture width is set to 2048 sampling points, and the window sliding step size is 1024 sampling points, i.e., the overlap rate is 50%. After detecting the time delay parameter corresponding to the maximum correlation coefficient of 1.35s, the signal processing unit accurately retrieves historical surface temperature data from the buffer 1350 milliseconds ago and uses it as the reference value for the current permeability correction. This solves the phase misalignment problem between internal magnetic domain flipping and surface temperature rise during 20Hz high-frequency fatigue. Based on this, the surface temperature signal is time-axis advanced-shifted, and a synchronous compensation temperature aligned with the internal magnetic state is output to correct the reference permeability. Based on the anisotropic tensor characteristics of the magnetoelastic effect, the axial principal stress induces nonlinear flipping of the tangential magnetic domains. The physical projection component of the principal stress on the orthogonal radial section approaches zero. The signal processing unit performs differential operations to filter out random fluctuation noise in the probe-spring surface gap caused by high-frequency vibrations of the fatigue frame. The dynamic compensation module extracts the impedance phase difference of the second-band detection component at the measuring coil and the reference coil to determine the dynamic helix angle deflection of the spring during its compression stroke. And update the differential weight coupling matrix. Using this matrix to measure the tangential induced electromotive force With radial induced electromotive force Complete dynamic differential calculations controlled by spatial attitude.

[0033] The system outputs absolute principal stresses The calculation formula is as follows: ,in, For the compensated absolute principal stress, This represents the amplitude of the tangential induced electromotive force. This represents the amplitude of the radial induced electromotive force. The differential weighted coupling matrix is ​​controlled by the dynamic helix angle deflection. The reference permeability is corrected by time phase backtracking. By converting the time parameter and spatial deflection, which characterize thermal conduction hysteresis, into correction factors, the systematic phase misalignment error accumulated by material physical thermal inertia and geometric deformation during cyclic loading is eliminated, thus establishing the fidelity of the full life cycle aging test data.

[0034] Example 2: In a laboratory verification environment for suspension spring fatigue deterioration, the test platform consisted of a hydraulic fatigue testing machine, a multi-channel infrared thermal radiation acquisition matrix, and a magnetic field excitation sensing module. To verify the linearity and long-term stability of the stress measurement system under full-range operating conditions, the test sample was a 60Si2Mn spring steel suspension spring with a wire diameter of 12mm and a mean diameter of 80mm. The sampling period in the signal acquisition procedure was... The setting depends on the bandwidth of the high-frequency components output by the excitation unit. The sampling frequency needs to be at least 10 times the highest frequency of the monitored signal (100kHz) to ensure phase detection accuracy. Simultaneously, to balance the processor's computational load, the decay time constant of the signal's autocorrelation function is calculated. Determined as ,in, The signal sampling period is defined as follows: During the experiment, broadband electromagnetic noise with a signal-to-noise ratio of 20dB is superimposed on the sensing front end, and 50Hz power grid interference harmonics are introduced to simulate the complex background interference conditions of the industrial electromagnetic environment, in order to examine the feature extraction capability of the system under non-ideal signal input.

[0035] The excitation unit applied a 50kHz high-frequency vibration excitation to the spring in the test group. The infrared thermal radiation acquisition matrix measured the initial equilibrium temperature of the spring surface to be 25.4℃. When the fatigue testing machine entered the 20Hz alternating loading stage, the surface temperature rose due to the frictional heating of the internal lattice of the material. The observed peak of the original thermal radiation signal lagged behind the stress peak by 1.32s on the time axis, reflecting the thermal inertial hysteresis of the conduction from the internal stress state to the surface thermal state. If phase backtracking correction is not applied at this time, the stress amplitude error calculated by the system reaches 18.2%. According to Fourier's law of heat conduction and the relationship between the metal crystal structure and the thermal equilibrium temperature of the spring, the stress amplitude error will reach 18.2%. The internal frictional heat generation model of the spring is used. The intrinsic heat source excited by the high-frequency alternating load in the core force region of the spring is conducted to the surface with a physical time constant. This causes the apparent thermal radiation signal to lag behind the evolution of the internal intrinsic magnetic state. Based on this physical law, the system establishes a time-domain mapping compensation model. The thermomagnetic decoupling processor obtains the time derivative sequence of the radial induced electromotive force envelope and the temperature time derivative sequence. A sliding cross-correlation function with a length of 2048 sampling points is calculated. The time delay parameter corresponding to the maximum value of the cross-correlation coefficient is detected to be 1.35 s. Based on this, the temperature sequence is shifted forward by 1.35 s, and the obtained reference permeability is corrected. The phase difference with real-time stress converges to within 0.5ms, where, The reference permeability is the phase-corrected value, and T is the real-time temperature. Within the temperature rise range of 25.4℃ to 85.6℃, the linear correlation coefficient of the magnetic signal after phase backtracking correction increased from 0.82 to 0.99, confirming that thermomagnetic coupling interference was suppressed.

[0036] When examining the geometric deflection effect caused by a large compression stroke, the spring load was gradually increased from 0N to 8000N. When the load exceeded the performance inflection point of 5500N, the spring helix angle dynamically flattened from the initial 5.2° to 4.15°. Due to the change in the geometric relationship between the sensing coil and the spring axis, the measurement error in the control group without weight compensation increased from 2.4% to 14.8%. The experimental group used a signal processing unit to extract the phase difference between the radial reference signal and the tangential signal. According to the principle of orthogonal magnetic circuit vector decomposition, when the main axis of the induction coil and the spatial leakage magnetic vector undergo relative deflection, the phase offset of the output signal of the orthogonal differential coil has a monotonic mapping relationship with the spatial tilt angle. The signal processing unit obtains the real-time phase difference between the tangential signal and the radial reference signal, and then calculates the phase difference using the formula... Determine the dynamic helix angle deflection. ,in, This represents the real-time phase difference. The dimensionless constant C is the structural coupling coefficient determined based on the no-load calibration. The signal processing unit calls the pre-stored linear relationship. The dynamic differential compensation coefficient w is determined, where constants a and b are the slope and intercept offsets pre-determined based on the reference material. The calculation module follows the formula. Calculate the eigenvalues ​​of the tangential principal stresses. ,in, The phase value of the tangential signal. The radial reference signal phase value is used as the basis for feature value extraction based on the phase component. Compensation for leakage magnetic vector deviation, and the calculated dynamic helix angle deflection. It is 4.18°, where, This represents the dynamic helix angle deflection; the system updates the differential weight coupling matrix in real time based on this. The coefficients of each order in the figure are obtained by adjusting the tangential induced electromotive force. With radial induced electromotive force Processing weighted differences, where, The differential weight coupling matrix is... This represents the amplitude of the tangential induced electromotive force. This represents the amplitude of the radial induced electromotive force. The dynamic deflection angle; under a stroke of 8000N, the zero-point drift of the synthesized magnetic characteristic parameter is maintained within 0.8% of the full scale. In the full-life test of 10 million fatigue loading cycles, the stress measurement baseline of the control group exhibits a nonlinear drift of 12.5% ​​with material fatigue aging, while the experimental group, through dynamic weight compensation and thermal phase alignment mechanism, has a maximum deviation of 1.62% in its measurement baseline throughout the entire cycle. This demonstrates that the system has resolved the measurement inaccuracy contradiction caused by geometric distortion and thermal accumulation.

[0037] Example 3: In a large-scale suspension spring fatigue aging test production line, the sensing units at each station are affected by the slight difference in installation gap and the uneven distribution of stray magnetic fields in the environment, which causes the tangential and radial magnetic flux coupling efficiency of the orthogonal differential sensing structure in the initial state to be different. If the stress is calculated using a general weighting coefficient, the measurement deviation will occur due to the difference in hardware mutual inductance parameters. Moreover, after the system has been running continuously for more than 10 million cycles, the inductive heat loss generated by the excitation unit and the accumulation of residual magnetism of the shielding cover induce nonlinear drift of the sensor baseline.

[0038] The sensor uses an in-situ calibration process to determine the coefficients of the differential weight coupling matrix. Under zero-load initial conditions, the excitation unit outputs the probe component, and the signal processing unit obtains the initial induced voltage amplitude ratio between the tangential measurement coil and the radial reference coil. This initial voltage amplitude ratio is then used as a weight bias value and injected into the differential weight coupling matrix. The time window length L for the sliding cross-correlation operation is determined based on the material's thermal diffusivity and sampling frequency. By calculating the shortest characteristic time for heat conduction to the surface, L is set to 2048 sampling points. When the preset cycle count threshold is reached, the timeliness reconstruction module detects the zero-point deviation vector under static no-load conditions. If the magnitude of the zero-point deviation vector exceeds 0.5% of the principal stress range, the time average of the current no-load signal is used to update the time. The initial constant term.

[0039] The system follows the mapping rules. Calculate and output absolute principal stress ;in, The difference weight coupling matrix is ​​dynamically updated with the spiral angle. For the compensated absolute principal stress, This represents the amplitude of the tangential induced electromotive force. This represents the amplitude of the radial induced electromotive force. The real-time reference permeability is L, and the sampling window length is L. The measurement consistency error of each test station is within 0.72%, and the physical alignment of the stress baseline is maintained during the fatigue test of continuous operation for 500 hours, thereby verifying the service stability of the system in a complex industrial cluster environment.

[0040] Example 4: In the scenario of fatigue strength testing of alloy springs from different metallurgical batches, the differences in residual austenite content and carbon equivalent distribution among materials from different furnaces cause fluctuations in the sensitivity of magnetic characteristic parameters to stress changes. Before the alternating loading is initiated, a stepped load from 0N to 10000N is applied to the reference sample of the same batch using a static calibration machine. The signal processing unit simultaneously acquires the differential induced voltage component output by the orthogonal differential sensing structure. Based on the stress value and induced voltage data sequence, least squares linear regression is performed to determine the slope operator S and intercept term B in the mapping rule f(·), where f(·) is the stress mapping function, S is the stress conversion sensitivity, and B is the zero-position compensation constant. This represents the amplitude of the tangential induced electromotive force. This represents the amplitude of the radial induced electromotive force. The differential weight coupling matrix is... The dynamic deflection angle is determined, and the defined function parameters are stored in the arithmetic module register to complete the analytical parameter calibration of specific material characteristics. The structural coupling coefficient is obtained by the static alignment process of the system at the initial power-on stage: by installing a mechanical scale turntable with a step of 0.5 degrees at the two ends of the spring, the slope of the phase difference change in the simulated deflection range from 0 degrees to 10 degrees is recorded. The linear slope term in the dynamic differential compensation coefficient is obtained by fitting the first-order least squares method with 10 sets of standard samples with known stress gradients. The intercept offset is calibrated to zero by the initial induced voltage of 50mV measured by the sensing unit under zero stress. All calibration coefficients are stored in non-volatile memory in 16-bit floating-point format, and the calibration load step is accurate to 500N.

[0041] When the measurement system is deployed at a test station where there are differences in assembly clearance, the lift-off clearance between the sensing unit and the surface of the spring being measured is... This causes a change in the magnetic reluctance of the magnetic circuit and induces signal amplitude attenuation. The system then initiates pre-calibration, and the excitation unit outputs a swept-frequency detection signal with a frequency range of 10kHz to 100kHz to the excitation winding. The signal processing unit acquires the real and imaginary parts of the complex impedance trajectory of the sensing coil at different frequencies and matches the physical distance according to the complex impedance phase angle and the preset lift-off height model. ,in, To measure the vertical lift distance of the sensing unit relative to the surface of the measured spring, the feedback gain of the preamplifier circuit is adjusted synchronously, and the signal amplitude entering the thermomagnetic decoupling processor is normalized to eliminate the quantity-effect relationship deviation caused by mechanical assembly errors.

[0042] Example 5: In the suspension spring testing environment, the system faces eddy current loss interference generated by the electromagnetic shielding structure. Before monitoring is activated, the signal processing unit controls the excitation unit to transmit a sweep frequency detection current in 5kHz steps within the range of 10kHz to 100kHz and collect the induced electromotive force sequence generated by the orthogonal differential sensing structure. By calculating the ratio of signal power to background noise power at each frequency point, the signal-to-noise ratio (SNR) response distribution curve is determined, and the frequency point corresponding to the peak value of the SNR response distribution curve is locked as the optimal monitoring frequency. and at the excitation signal frequency equal Under certain conditions, the piezoelectric ceramic exciter is driven to generate high-frequency mechanical vibration. The system load is near the material's yield limit, causing the internal magnetization to enter the nonlinear saturation range. The computation module initiates a nonlinear analytical calibration process to apply a stepped pressure from 80% to 110% of the rated load to the reference sample. Simultaneously, the synthesized magnetic characteristic parameter sequence is recorded, and a nonlinear compensation lookup table is constructed using a cubic spline interpolation algorithm. During real-time measurement, the system retrieves the nonlinear correction operator from the lookup table based on the current synthesized electromotive force amplitude. And it is used as a correction factor in the mapping rule to calculate the absolute principal stress. It eliminates nonlinear analytical errors caused by large loads and outputs stress evolution data in the material fatigue critical region.

[0043] The phase-sensitive detector module in the signal processing unit receives the stress-modulated tangential signal and inputs it into the first signal terminal of the multiplier. The arithmetic module synchronously generates an in-phase reference signal with a frequency consistent with and orthogonal to the fundamental frequency of the excitation signal. Orthogonal reference signal The induced electromotive force sequence input from the first signal terminal is then input to the second signal terminal of the multiplier. The phase-sensitive detection module performs a product logic operation on the induced electromotive force sequence input from the first signal terminal with two orthogonal reference signals. The product result is then sent to a low-pass filter circuit to filter out second harmonic and higher-order frequency components, obtaining the in-phase component voltage that characterizes the complex vector of magnetic characteristic parameters. With orthogonal component voltage ,in, The real part of the signal, For the imaginary part of the signal, For in-phase reference components, As orthogonal reference components, the signal processing unit calculates... and The arctangent function of the ratio is used to determine the phase component modulated by the principal stress. ,in, The signal phase angle is used to complete the physical mapping of the stress characteristic vector from the time domain waveform to the phase characteristic domain.

[0044] In the orthogonal differential sensing structure, the tangential measuring coil and the radial reference coil are spatially distributed to satisfy the orthogonal geometric constraint of their axes. The effective sensing axis of the tangential measuring coil is parallel to the longitudinal axis of the spring being measured, while the effective sensing axis of the radial reference coil passes through the geometric center of the spring and is perpendicular to the longitudinal axis. During deployment, a laser alignment collimator is used to adjust the geometric centers of the two coils to a preset coincidence point, ensuring that the initial spatial alignment deviation is within 0.05mm. This allows the induction outputs of the orthogonal differential sensing structure to cancel each other out in the presence of no load and with the helix angle in its initial state, thereby establishing a physical zero-point reference to extract the phase difference signal generated by the change in the spring's geometric topology during the compression stroke, which is then used to determine the dynamic helix angle deflection. The computation provides the original input with vector determinism.

[0045] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A stress measurement system for spring fatigue state, characterized in that, include: The excitation unit is used to apply an alternating excitation magnetic field to the spring under test, generating a stress-modulated induced electromotive force signal. The sensing unit includes a tangential measuring coil and a radial reference coil that are non-contactly arranged around the periphery of the spring under test. The tangential measuring coil and the radial reference coil form an orthogonal differential sensing structure for acquiring the magnetic characteristic parameters of the spring under test under cyclic load. The signal processing unit, connected to the sensing unit, is used to acquire the tangential signal output by the tangential measuring coil and the radial reference signal output by the radial reference coil, and to acquire the phase component of the tangential signal modulated by the principal stress of the spring under test. Specifically, the signal processing unit determines the dynamic helix angle deflection of the spring under test during the compression stroke based on the phase difference between the radial reference signal and the tangential signal, and determines the dynamic differential compensation coefficient according to the preset linear mapping relationship between the dynamic helix angle deflection and the compensation weight. The tangential signal is then weighted and subtracted using the dynamic differential compensation coefficient to extract the tangential principal stress component and compensate for the leakage magnetic vector deviation caused by the dynamic helix angle deflection.

2. The stress measurement system for spring fatigue state according to claim 1, characterized in that, The excitation unit includes a piezoelectric ceramic vibrator and an excitation winding. The piezoelectric ceramic vibrator is used to generate high-frequency vibrations with a frequency range of 10kHz to 100kHz according to the frequency-modulated electrical signal output by the signal processing unit, so as to modulate the magnetic domain arrangement state inside the spring under test. The excitation winding is set at the end of the spring under test and is used to output a detection magnetic field whose frequency is controlled by the feedback of the signal processing unit, so that the spring under test generates a stress-modulated leakage magnetic signal.

3. The stress measurement system for spring fatigue state according to claim 1, characterized in that, The sensing unit also includes a shield; the tangential measuring coil and the radial reference coil are both located inside the shield to reduce electromagnetic interference from the ambient magnetic field on the orthogonal differential sensing structure; the sensing axis of the tangential measuring coil is parallel to the axis of the spring being measured, and the sensing axis of the radial reference coil is perpendicular to the axis of the spring being measured.

4. The stress measurement system for spring fatigue state according to claim 1, characterized in that, The signal processing unit includes a preamplifier circuit, a phase-sensitive detector module, and an arithmetic module; the preamplifier circuit is used to perform differential amplification processing on the tangential signal and the radial reference signal. The phase-sensitive detection module is used to extract the phase offset of the tangential signal relative to the reference excitation signal; the arithmetic module is used to process the phase decoupling logic to compensate for the signal hysteresis caused by the thermal inertia of the spring under test.

5. A stress measurement system for spring fatigue state according to claim 4, characterized in that, The calculation module uses the following formula to calculate the tangential principal stress components: ,in, These are the extracted tangential principal stress eigenvalues. The phase value of the tangential signal. Let w be the phase value of the radial reference signal, and w be the dynamic differential compensation coefficient. This represents the dynamic helix angle deflection.

6. A stress measurement system for spring fatigue state according to claim 1, characterized in that, The signal processing unit is also used to monitor the amplitude of the radial reference signal and determine the lift-off gap offset between the sensing unit and the surface of the measured spring in real time based on the amplitude; the signal processing unit adjusts the gain coefficient of the tangential principal stress component according to the lift-off gap offset to suppress common-mode interference caused by system vibration.

7. A stress measurement system for spring fatigue state according to claim 1, characterized in that, The excitation unit is also used to complete the magnetic field frequency sweep program; the signal processing unit is used to capture the energy peak point of the induced electromotive force signal and determine the corresponding frequency as the resonant operating frequency, while controlling the excitation unit to output the alternating excitation magnetic field at the resonant operating frequency.

8. A stress measurement system for spring fatigue state according to claim 1, characterized in that, The system also includes an environmental monitoring module for acquiring temperature data around the spring under test; and a signal processing unit for inputting the temperature data into a preset temperature compensation model and outputting the correction value of the dynamic differential compensation coefficient to eliminate the influence of environmental temperature changes on magnetic characteristic parameters.

9. A stress measurement system for spring fatigue state according to claim 1, characterized in that, The signal processing unit is used to record the change sequence of the tangential principal stress components and calculate the zero-point drift rate within the continuous monitoring period. When the zero-point drift rate exceeds the preset damage judgment threshold, the signal processing unit outputs an alarm signal that indicates fatigue damage to the tested spring.

10. A stress measurement system for spring fatigue state according to claim 1, characterized in that, The system also includes a data output interface, which is used to generate evaluation data characterizing the residual life of the tested spring based on the tangential principal stress components and a preset degradation model, and input the evaluation data into the digital simulation model of the tested spring to realize the evolution analysis of the fatigue state of the tested spring.

Images