A method for testing the tensile resistance of a modified mpp cable conduit material

Abstract

This invention discloses a tensile testing method for modified MPP cable conduit materials, belonging to the field of material mechanical property testing technology. The method includes: acquiring initial tensile response data of the modified MPP cable conduit material under a preset strain rate; performing time-frequency domain decomposition processing on the initial tensile response data to extract transient oscillation components and steady-state creep components; constructing a viscoelastic equivalent model of the modified MPP cable conduit material based on the transient oscillation components and steady-state creep components, and using this model to calculate the dynamic stiffness degradation curve during the tensile process; determining the tensile strength evaluation index based on the inflection point characteristics of the dynamic stiffness degradation curve and the accumulated energy of the acoustic emission signal of the material before tensile fracture. This invention achieves a more accurate characterization of the tensile properties of modified MPP cable conduit materials.

Description

Technical Field

[0001] This invention relates to the field of material mechanical property testing technology, specifically to a tensile testing method for modified MPP cable conduit materials. Background Technology

[0002] In the quality inspection and performance evaluation of modified MPP cable conduit materials, tensile strength is a key indicator for measuring their load-bearing capacity. Existing tensile testing methods typically rely on a universal testing machine to stretch standard specimens at a constant rate to obtain stress-strain curves, and directly identify the maximum stress value on the curve as the tensile strength of the material. This phenomenological testing method is simple to operate and the data processing is relatively intuitive. However, modified MPP materials are typical viscoelastic polymers, and their tensile mechanical response is not a single elastic or plastic behavior, but a complex response containing multiple components such as transient elastic deformation, delayed elastic deformation, and irreversible viscous flow deformation. Existing methods simply correlate force and elongation during the tensile process, failing to distinguish the deformation mechanisms of the material at different time scales during loading. The macroscopic characteristics presented by the tensile curve are essentially the result of the coupling and superposition of multiple microscopic deformation mechanisms. Traditional testing only captures the final characterizing limit stress, completely ignoring the dynamic evolution process of the material's internal stiffness continuously degrading over time under sustained load, and the real impact of the gradual accumulation of microscopic damage on macroscopic failure. The single tensile strength value obtained in this way is insufficient to reflect the true resistance reserve of a material under long-term use or complex loading history, easily leading to an overestimation of material durability or an omission of potential failure risks. To overcome these limitations, it is necessary to solve the problem of effectively separating different time response components from macroscopic tensile response data and quantitatively describing the stiffness degradation process. Simultaneously, it is also necessary to address how to organically integrate the critical transition characteristics of stiffness degradation with the energy release process characterizing the accumulation of microscopic damage within the material, in order to establish tensile performance evaluation parameters that better reflect the nature of damage evolution in viscoelastic materials. Summary of the Invention

[0003] The purpose of this invention is to provide a tensile testing method for modified MPP cable conduit materials. By performing time-frequency domain decomposition on the initial tensile response data to extract transient oscillation components and steady-state creep components, a viscoelastic equivalent model is constructed to calculate the dynamic stiffness degradation curve. The tensile strength evaluation index is determined by combining the acoustic emission cumulative energy and the stiffness degradation inflection point characteristics, thereby solving the problem that existing methods cannot reflect the viscoelastic stiffness degradation process and micro-damage accumulation of materials.

[0004] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a tensile testing method for modified MPP cable conduit materials. This method achieves a comprehensive evaluation of the tensile strength of the material by synergistically analyzing the mechanical response and acoustic emission characteristics of the modified MPP cable conduit material during the tensile process. This invention is achieved through the following technical solution: acquiring the initial tensile response data of the modified MPP cable conduit material under test at a preset strain rate; performing time-frequency domain decomposition processing on the acquired initial tensile response data to extract the transient oscillation component characterizing the instantaneous elastic oscillation behavior of the material and the steady-state creep component characterizing the long-term viscous flow behavior of the material; constructing a viscoelastic equivalent model that can accurately describe the viscoelastic mechanical behavior of the material based on the extracted transient oscillation component and steady-state creep component, and using the viscoelastic equivalent model to calculate the dynamic stiffness degradation curve of the modified MPP cable conduit material under test during the entire tensile process. This curve reflects the continuous decrease in macroscopic stiffness caused by the evolution of microstructural damage within the material; based on the inflection point characteristics of the dynamic stiffness degradation curve, which marks the critical moment when the material transitions from slow damage to accelerated failure, and combined with the accumulated energy of the acoustic emission signal accumulated by the modified MPP cable conduit material under test before tensile fracture, determining a tensile strength evaluation index that can quantitatively evaluate the tensile properties of the material.

[0005] As a preferred technical solution of the present invention, the process of obtaining initial tensile response data is as follows: the modified MPP cable conduit material to be tested is processed into a dumbbell-shaped standard specimen, and multiple resistance strain gauges are pasted axially on the surface of its gauge length section. The specimen is installed in the fixture of a universal testing machine, the tensile rate is set to a preset strain rate and loading is started. The tensile force value and axial strain value are collected synchronously to form a time series data pair, thereby obtaining initial tensile response data containing complete elastoplastic deformation information.

[0006] As a preferred technical solution of the present invention, the process of performing time-frequency domain decomposition on the initial tensile response data specifically involves: resampling the axial strain value sequence at equal time intervals to obtain a uniformly sampled strain sequence; performing empirical mode decomposition to obtain multiple intrinsic mode function components; calculating the instantaneous frequency mean of each component; superimposing the components with instantaneous frequency mean higher than a preset high-frequency threshold to obtain a transient oscillation component, which corresponds to the rapid stretching vibration of polymer chain segments inside the material; superimposing the components with instantaneous frequency mean lower than a preset low-frequency threshold to obtain an intermediate trend component; and then performing polynomial fitting to obtain a steady-state creep component, which reflects the irreversible slippage of molecular chains in the amorphous region of the material.

[0007] As a preferred technical solution of the present invention, the process of constructing a viscoelastic equivalent model and calculating the dynamic stiffness degradation curve specifically involves mapping the transient oscillation component and the steady-state creep component to the Laplace domain to obtain the corresponding frequency domain functions. An initial viscoelastic network model is constructed using multiple spring elements and multiple damping elements. The spring modulus and damping coefficient of the high-frequency branch of the model are determined based on the transient oscillation frequency domain function to characterize the material's rapid response characteristics. The spring modulus and damping coefficient of the low-frequency branch of the model are determined based on the steady-state creep frequency domain function to characterize the material's slow rheological characteristics, thereby obtaining the viscoelastic equivalent model with determined parameters. In the Laplace domain, this model is used to calculate the dynamic storage modulus and dynamic loss modulus at different stretching time points. The ratio of these two moduli is used to obtain the dynamic stiffness degradation curve that varies with stretching time. This curve accurately depicts the process of the material stiffness gradually decaying from its initial intact state to near fracture.

[0008] In the above technical solution, the preferred initial viscoelastic network model adopts a three-parameter solid model containing two Maxwell branches and one independent spring branch in parallel. This configuration can simultaneously take into account the high-frequency elastic response and low-frequency viscous flow of the modified MPP material, and has the advantages of simple structure and clear physical meaning.

[0009] As a preferred technical solution of the present invention, the process of determining the tensile strength evaluation index specifically involves: calculating the second derivative of the dynamic stiffness degradation curve; finding the zero-crossing point on the obtained second derivative curve; marking the tensile time corresponding to the zero-crossing point within the effective time interval as the inflection point time; and taking the slope of the tangent line of the dynamic stiffness degradation curve at the inflection point time as the inflection point feature, which quantifies the degree of stiffness decay of the material at the instant it enters the accelerated damage stage; during the tensile loading process, using an acoustic emission sensor to continuously collect the elastic wave signal released by the generation and expansion of micro-damage inside the material; obtaining the cumulative energy curve after full-wave rectification and integration; extracting the total cumulative energy value before the inflection point time and the incremental cumulative energy value after the inflection point time from the curve, the former reflecting the total energy accumulated during the damage incubation period and the latter characterizing the sudden energy released during the damage instability and expansion period; using the value of the inflection point feature, the total cumulative energy value, and the incremental cumulative energy value as input features, feeding them into a pre-trained nonlinear regression model, and outputting the tensile strength evaluation index from the model, which achieves multi-dimensional feature fusion and overcomes the problem of large dispersion in single index evaluation.

[0010] As a further improvement to the above technical solution, the method for obtaining inflection point characteristics is as follows: cubic spline interpolation is performed on discrete data points on the dynamic stiffness degradation curve to obtain a smooth and continuous curve function, and its first-order derivative and second-order derivative are obtained. The zero-crossing position that makes the first-order derivative reach a local minimum value is obtained by numerical root search within the domain of the second-order derivative function, which is taken as the inflection point time. The first-order derivative value at this position is taken as the tangent slope. This method can stably extract the accurate inflection point reflecting the transformation of the material damage mechanism from the original discrete data.

[0011] As a specific implementation of the acoustic emission signal processing of the present invention, an acoustic emission sensor is attached to the outer surface of the gauge length of a dumbbell-shaped standard specimen and connected to an acoustic emission acquisition system. After setting the sampling frequency and trigger threshold, the original acoustic emission waveform data of the entire tensile process is recorded. The original waveform is rectified by full wave and integrated over continuous time to obtain the cumulative energy curve. The cumulative energy value corresponding to the inflection point time on the cumulative energy curve is directly read as the total cumulative energy value, and the difference between the cumulative energy value at the end of the curve and the total cumulative energy value is taken as the incremental cumulative energy value. This processing method preserves the complete temporal evolution information of the acoustic emission energy.

[0012] As another key technical aspect of this invention, the training process of the nonlinear regression model involves obtaining multiple sets of calibration samples of modified MPP cable conduit material with known tensile strength values, performing the aforementioned steps on each set of calibration samples to obtain the corresponding inflection point feature sample values, total cumulative energy sample values, and incremental cumulative energy sample values, thereby constructing a training dataset containing input features and output labels. The training set is then trained using a support vector regression algorithm, and the optimal penalty parameters and kernel function parameters are selected through cross-validation to obtain a nonlinear regression model capable of accurately predicting the tensile strength of unknown samples.

[0013] As a complete application loop of the technical solution of this invention, after obtaining the tensile strength evaluation index, the index is compared with the preset strength grading threshold to determine the strength level of the modified MPP cable conduit material to be tested; according to the strength level, the corresponding target extrusion molding process parameters are retrieved from the pre-established process parameter database and output to the display terminal to guide the adjustment of production equipment; at the same time, the tensile strength evaluation index, inflection point characteristics, total cumulative energy value and incremental cumulative energy value are associated and stored and a test report is generated to provide data support for product batch quality traceability and continuous process optimization.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: The initial tensile response data were subjected to time-frequency domain decomposition. The axial strain value sequence was resampled at equal time intervals and subjected to empirical mode decomposition to adaptively separate intrinsic mode function components with different instantaneous frequency characteristics. Based on the correspondence between the instantaneous frequency mean and preset high-frequency and low-frequency thresholds, the high-frequency components were superimposed to obtain transient oscillation components, reflecting the rapid stress wave propagation and local resonance response generated by the material under tensile impact excitation. The low-frequency trend term was fitted with a polynomial to obtain steady-state creep components, characterizing the slow creep deformation controlled by the viscous flow of the material. The transient oscillation components and steady-state creep components were mapped to the Laplace domain to obtain transient oscillation frequency domain functions and steady-state creep frequency domain functions, respectively. A viscoelastic equivalent model was constructed using multiple spring and damping elements, and the spring modulus and damping coefficient of the high-frequency and low-frequency branches in the model were determined based on the two frequency domain functions. Subsequently, the dynamic storage modulus and dynamic loss modulus at different tensile time points were calculated in the Laplace domain, and the ratio of the two was used to form a dynamic stiffness degradation curve. This method enables the extraction and quantitative reconstruction of the time-varying decay process of material viscoelastic stiffness from single tensile test data. The continuous stiffness drop trajectory recorded by the dynamic stiffness degradation curve breaks through the obscuring effect of the viscoelastic nature by the traditional single maximum stress index, providing continuous observational evidence for capturing the inflection point of material performance degradation under time-series loading. Inflection point features are extracted from the dynamic stiffness degradation curve and fused with the energy information accumulated before and after the inflection point time by the acoustic emission signal. The inflection point time is located by the zero-crossing position of the second derivative and the local minimum condition, and the tangent slope at this moment is used as the inflection point feature, reflecting the critical node of the change in stiffness decay rate. Simultaneously, acoustic emission waveforms during the complete tensile process are acquired in real time on a dumbbell-shaped standard specimen. The accumulated energy curve is obtained through full-wave rectification and continuous integration, extracting the total accumulated energy value before the inflection point time and the incremental accumulated energy value after the inflection point time. The total accumulated energy before the inflection point time corresponds to the micro-damage incubation energy in the slow stiffness degradation stage, while the incremental accumulated energy after the inflection point time corresponds to the damage propagation energy in the accelerated stiffness drop stage. The inflection point characteristic values ​​and these two types of acoustic emission energy parameters are input into a pre-trained nonlinear regression model to output a tensile strength evaluation index. This index simultaneously contains information on three aspects: the stiffness transition sensitivity within the material's viscoelastic framework, the total energy absorbed during the damage incubation period, and the energy increment released during the rapid damage propagation period. Therefore, it can truly reflect the comprehensive ability of modified MPP materials to resist damage nucleation and propagation during tensile testing, avoiding a one-sided evaluation of the complex failure behavior coupled with multiple deformation processes. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 This is a flowchart of the tensile strength testing method for modified MPP cable conduit materials; Figure 2 This is a flowchart of the process for obtaining initial tensile response data for modified MPP cable conduit materials; Figure 3 This is a flowchart of the extraction process for transient oscillation components and steady-state creep components from tensile response data; Figure 4 This is a flowchart of the determination of the viscoelastic equivalent model and the generation of the dynamic stiffness degradation curve based on frequency domain component fitting. Figure 5 This is a flowchart of a tensile strength evaluation method based on dynamic stiffness and acoustic emission; Figure 6 This is a flowchart of the tensile strength assessment and process parameter matching for modified MPP cable conduits based on a nonlinear regression model. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0018] See Figure 1 This invention provides a tensile testing method for modified MPP cable conduit materials. The method acquires the initial tensile response data of the material under test at a preset strain rate, performs time-frequency domain decomposition on the initial tensile response data, extracts transient oscillation components and steady-state creep components, constructs a viscoelastic equivalent model of the material under test using the two extracted components, and calculates the dynamic stiffness degradation curve during the tensile process using the viscoelastic equivalent model. Then, based on the inflection point characteristics of the dynamic stiffness degradation curve and the accumulated energy of the acoustic emission signal collected by the material under test before tensile fracture, the tensile strength evaluation index of the modified MPP cable conduit material under test is determined.

[0019] Example 1: In specific implementation, please refer to Figure 2The process of obtaining the initial tensile response data of the modified MPP cable conduit material under the preset strain rate is as follows: The modified MPP cable conduit material to be tested is processed into a dumbbell-shaped standard specimen. The dumbbell-shaped standard specimen has two clamping ends and a gauge length section located between the two clamping ends, the cross-sectional area of ​​which remains constant along its entire length. Multiple resistance strain gauges are attached to the surface of the gauge length section along the axial direction of the dumbbell-shaped standard specimen. The sensitive grid direction of each strain gauge is parallel to the axial direction of the dumbbell-shaped standard specimen, and the strain gauges are arranged circumferentially at intervals on the surface of the gauge length section. The dumbbell-shaped standard specimen is mounted in the fixture of a universal testing machine. One clamping end of the dumbbell-shaped standard specimen is clamped by the upper fixture of the universal testing machine, and the other clamping end is clamped by the lower fixture of the universal testing machine. The major axis of the dumbbell-shaped standard specimen coincides with the loading axis of the universal testing machine.

[0020] In the control interface of the universal testing machine, input the preset strain rate value. The preset strain rate is the rate of increase of axial strain in the gauge length section of the dumbbell-shaped standard specimen during the tensile process. The value of the preset strain rate is selected based on the engineering application load conditions of the modified MPP cable conduit material. After setting, start the universal testing machine to apply tensile loading to the dumbbell-shaped standard specimen. The crossbeam of the universal testing machine moves upward at the linear velocity corresponding to the preset strain rate, applying an axial tensile load to the dumbbell-shaped standard specimen.

[0021] During tensile loading, the force sensor built into the universal testing machine acquires the tensile force applied to the dumbbell-shaped standard specimen at constant acquisition time intervals. Simultaneously, multiple resistance strain gauges acquire the axial strain values ​​on the gauge length surface at the same acquisition time intervals as the force sensor. The acquisition clocks of the force sensor and the multiple resistance strain gauges are synchronized via a hardware trigger signal. For each acquisition time interval, a tensile force value recorded by the force sensor within that interval is paired with an axial strain value recorded by the multiple resistance strain gauges within the same interval, forming a data pair. All data pairs generated from the start to the end of the tensile loading process are arranged in chronological order of acquisition time; the resulting sequence is the initial tensile response data.

[0022] Example 2: In specific implementation, please refer to Figure 3The axial strain value sequence is extracted from the initial tensile response data, which consists of multiple time-arranged data pairs. Each data pair contains a tensile force value and an axial strain value. The axial strain values ​​from all data pairs are extracted in chronological order to form the axial strain value sequence. The time interval between adjacent data points in the axial strain value sequence depends on the acquisition time interval set by the force sensor and multiple resistance strain gauges during the acquisition process. When the acquisition time interval is not constant, the axial strain value sequence is a non-uniform sampling sequence.

[0023] The axial strain value sequence is resampled at equal time intervals to determine the target resampling time interval, which is set as the average of the time intervals between all adjacent time points in the axial strain value sequence. Cubic spline interpolation is selected as the resampling interpolation method. Using the time points in the axial strain value sequence as interpolation nodes and the axial strain values ​​as the function values ​​corresponding to the interpolation nodes, a cubic spline interpolation function is constructed, passing through all interpolation nodes. Within the start and end time range of the axial strain value sequence, starting from the initial time point, a sequence of equally spaced resampling time points is generated sequentially with the target time interval as the step size. Each resampling time point in the resampling time point sequence is substituted into the cubic spline interpolation function to calculate the axial strain value corresponding to each resampling time point. All resampling time points and their corresponding axial strain values ​​constitute a uniformly sampled strain sequence.

[0024] Empirical mode decomposition (EMD) is performed on the uniformly sampled strain sequence. The decomposition process is as follows: All local maxima and minima in the uniformly sampled strain sequence are identified. An upper envelope sequence is constructed using cubic spline interpolation, with all local maxima as nodes. A lower envelope sequence is constructed using cubic spline interpolation, with all local minima as nodes. The average values ​​of the upper and lower envelope sequences at each sampling point are calculated to obtain the instantaneous mean sequence. The instantaneous mean sequence is subtracted from the uniformly sampled strain sequence to obtain the candidate intrinsic mode function (EMF) sequence. The candidate EMF sequence is then judged to meet the EMF screening stopping criteria. The screening stopping criteria include two criteria: the first criterion is that the absolute value of the difference between the number of extrema and the number of zero-crossings in the candidate EMF sequence is less than or equal to 1; the second criterion is that the symmetry deviation between the upper and lower envelopes of the candidate EMF sequence is less than a preset symmetry deviation tolerance value. If the candidate intrinsic mode function (IMF) sequence does not meet the screening stopping condition, the candidate IMF sequence is used as a new input sequence. The steps of identifying local maxima and local minima, constructing upper and lower envelope sequences, calculating the instantaneous mean sequence, and subtracting the instantaneous mean sequence are repeated until the obtained candidate IMF sequence meets the screening stopping condition. The candidate IMF sequence that meets the screening stopping condition is taken as an IMF component. The obtained IMF component is subtracted from the uniformly sampled strain sequence to obtain the residual sequence. The residual sequence is used as a new sequence to be decomposed. The above empirical mode decomposition steps are repeated to extract the next IMF component until the residual sequence becomes a monotonic sequence or the amplitude of the residual sequence is lower than a preset amplitude threshold. The decomposition is then stopped, and multiple IMF components are obtained.

[0025] Calculate the instantaneous frequency mean of each intrinsic mode function (IMF) component. For each IMF component, perform a Hilbert transform to obtain its corresponding Hilbert transform result. Construct an analytic signal by taking the IMF component as its real part and its corresponding Hilbert transform result as its imaginary part. Calculate the instantaneous phase function from the analytic signal; the instantaneous phase function represents the argument of the analytic signal. Take the first derivative of the instantaneous phase function with respect to the time variable to obtain the instantaneous frequency curve. Calculate the arithmetic mean of the values ​​of the instantaneous frequency curve over the entire time axis, and use this arithmetic mean as the instantaneous frequency mean of that IMF component.

[0026] Preset high-frequency and low-frequency thresholds are used. The preset high-frequency threshold is set to 0.3 times the Nyquist frequency of the uniformly sampled strain sequence, which is half the sampling frequency of the uniformly sampled strain sequence. The preset low-frequency threshold is set to 0.02 times the sampling frequency of the uniformly sampled strain sequence. The instantaneous mean frequency of each intrinsic mode function component is compared with the preset high-frequency threshold. All intrinsic mode function components with an instantaneous mean frequency higher than the preset high-frequency threshold are selected. All selected intrinsic mode function components are then superimposed point-by-point at time points. The resulting time-series signal is the transient oscillation component.

[0027] The instantaneous frequency mean of each intrinsic mode function (IMF) component is compared with a preset low-frequency threshold. IMF components with an instantaneous frequency mean lower than the threshold are selected. These selected IMF components are then superimposed point-by-point at time points to obtain the intermediate trend component. A polynomial fitting process is performed on the intermediate trend component using the least squares method. The order of the fitting polynomial is set to third, with the time points of the intermediate trend component as the independent variable and the amplitude of the intermediate trend component as the dependent variable. The coefficients of the third-order polynomial are solved to obtain the third-order polynomial function. Substituting the time points of the intermediate trend component into the third-order polynomial function, the resulting smooth curve is the steady-state creep component.

[0028] Example 3: In specific implementation, please refer to Figure 4 The process of mapping the transient oscillatory component and the steady-state creep component to the Laplace domain is as follows: Perform a Laplace transform on the transient oscillatory component; the integral kernel of the Laplace transform is... The upper limit of integration is positive infinity, and the lower limit of integration is zero. Substituting the time-domain function of the transient oscillation component into the Laplace transform formula and integrating, we obtain the frequency-domain function of the transient oscillation. The frequency-domain function of the transient oscillation is based on complex variables. A function of the independent variable, a complex variable The real part of the variable represents the decay factor, a complex variable. The imaginary part represents the angular frequency. Performing the same Laplace transform operation on the steady-state creep component, substituting the time-domain function of the steady-state creep component into the Laplace transform formula and integrating, yields the steady-state creep frequency domain function. The steady-state creep frequency domain function is also in terms of complex variables. It is a function with the independent variable.

[0029] An initial viscoelastic network model is constructed using multiple spring elements and multiple damping elements. The mechanical relationship of the spring elements is that stress is proportional to strain, with the proportionality coefficient being the spring modulus; the mechanical relationship of the damping elements is that stress is proportional to strain rate, with the proportionality coefficient being the damping coefficient. Each Maxwell branch consists of one spring element and one damping element connected in series. The initial viscoelastic network model includes two Maxwell branches and one independent spring branch connected in parallel. The two Maxwell branches are denoted as the first Maxwell branch and the second Maxwell branch, respectively, and the independent spring branch is denoted as the independent spring branch. The spring element in the first Maxwell branch has a first spring modulus, and the damping element in the first Maxwell branch has a first damping coefficient; the spring element in the second Maxwell branch has a second spring modulus, and the damping element in the second Maxwell branch has a second damping coefficient; the spring element in the independent spring branch has a third spring modulus. The two ends of the first Maxwell branch, the second Maxwell branch, and the independent spring branch are all connected between the same input node and the same output node, and are subjected to the same axial strain.

[0030] The method for determining the spring modulus and damping coefficient of the high-frequency branch in the initial viscoelastic network model based on the transient oscillation frequency domain function is as follows: the first Maxwell branch is designated as the high-frequency branch, which corresponds to the viscoelastic response characteristics of the transient oscillation component in a higher frequency range. In the Laplace domain, the total complex modulus of the initial viscoelastic network model is expressed as a Laplace variable. The function is expressed as the sum of the complex moduli of the parallel branches. The high-frequency band is defined as the range from the angular frequency corresponding to a preset high-frequency threshold to the angular frequency corresponding to the Nyquist frequency. Multiple discrete frequency points are selected within the high-frequency band, and the complex values ​​of the transient oscillation frequency domain function at each discrete frequency point are calculated, resulting in a sequence of measured complex moduli of the transient oscillation frequency domain function within the high-frequency band. Using the first spring modulus and the first damping coefficient as undetermined parameters, the Levonberg-Marquardt nonlinear least squares optimization algorithm is employed to curve-fit the calculated values ​​of the complex modulus expression of the initial viscoelastic network model within the high-frequency band with the sequence of measured complex modulus values ​​of the transient oscillation frequency domain function. Iteration stops when the cost function value drops below a preset convergence tolerance. The obtained first spring modulus value is taken as the spring modulus of the high-frequency branch, and the obtained first damping coefficient value is taken as the damping coefficient of the high-frequency branch. The cost function is defined as the sum of the squares of the magnitudes of the differences between the calculated and measured values ​​at all discrete frequency points.

[0031] The method for determining the spring modulus and damping coefficient of the low-frequency branch in the initial viscoelastic network model based on the steady-state creep frequency domain function is as follows: The second Maxwell branch is designated as the low-frequency branch, corresponding to the viscoelastic response characteristics of the steady-state creep component in a lower frequency range. The low-frequency band is defined as the range from zero frequency to the angular frequency corresponding to a preset low-frequency threshold. Multiple discrete frequency points are selected within the low-frequency band, and the complex values ​​of the steady-state creep frequency domain function at each discrete frequency point are calculated to obtain a sequence of measured values ​​of the complex modulus of the steady-state creep frequency domain function within the low-frequency band. Using the second spring modulus and the second damping coefficient as undetermined parameters, the Levenberg-Marquardt nonlinear least squares optimization algorithm is used to curve fit the calculated value of the complex modulus expression of the initial viscoelastic network model in the low-frequency range with the measured value sequence of the complex modulus of the steady-state creep frequency domain function. The iteration stops when the cost function value drops below the preset convergence tolerance. The value of the second spring modulus obtained at this time is taken as the spring modulus of the low-frequency branch, and the value of the second damping coefficient obtained at this time is taken as the damping coefficient of the low-frequency branch.

[0032] The third spring modulus of the independent spring branch is determined by comparing it with the asymptotic behavior of the steady-state creep component over a long time interval. The asymptotic strain value of the steady-state creep component is extracted as the time approaches the maximum tensile time. The ratio of the asymptotic strain value to the asymptotic stress value of the tensile force at the corresponding time is taken as the long-term compliance. The difference between the reciprocal of the long-term compliance and the sum of the spring moduli of the high-frequency branch and the low-frequency branch is taken as the third spring modulus of the independent spring branch. The initial viscoelastic network model after determining the first spring modulus, first damping coefficient, second spring modulus, second damping coefficient, and third spring modulus is used as the viscoelastic equivalent model.

[0033] Using a viscoelastic equivalent model, the dynamic storage modulus and dynamic loss modulus of the modified MPP cable conduit material under test at different tensile time points are calculated in the Laplace domain. Based on the ratio of the dynamic storage modulus to the dynamic loss modulus, the formula for the dynamic stiffness degradation curve is obtained as follows:

[0034] in, Indicates the first step in the tensile loading process. Each stretching time point This is the index number of the stretching time point. The value is from A positive integer representing the total number of stretching time points; Indicates the stretching time point The corresponding dynamic stiffness value; Indicates the stretching time point Corresponding dynamic energy storage modulus; Indicates the stretching time point The corresponding dynamic loss modulus; Indicates the time point of stretching Corresponding equivalent frequency, equivalent frequency The value is determined by the preset strain rate and stretching time point. The dynamic storage modulus is obtained by dividing the axial strain value of the dumbbell-shaped standard specimen by the constant time. It is the equivalent frequency Converting to angular frequency and then inputting into the complex modulus expression of the viscoelastic equivalent model, the value obtained by taking the real part of the complex modulus expression; dynamic loss modulus. This value is obtained by taking the imaginary part of the complex modulus expression. The dynamic stiffness values ​​corresponding to all stretching time points are then calculated according to the stretching time points. The order in which they are arranged forms a dynamic stiffness degradation curve.

[0035] Example 4: In specific implementation, please refer to Figure 5 The process of calculating the second derivative curve of the dynamic stiffness degradation curve is as follows: The dynamic stiffness degradation curve consists of multiple discrete data points, each containing a stretching time point and a corresponding dynamic stiffness value. Cubic spline interpolation is performed on all discrete data points on the dynamic stiffness degradation curve, using all stretching time points as interpolation nodes and all dynamic stiffness values ​​as the function values ​​corresponding to the interpolation nodes. A cubic spline interpolation function passing through all interpolation nodes is constructed. This cubic spline interpolation function is denoted as the smoothed continuous curve function, which uses stretching time as the independent variable and dynamic stiffness value as the dependent variable. The first derivative of the smoothed continuous curve function with respect to the stretching time variable is obtained. The second derivative of the smoothed continuous curve function with respect to the stretching time variable is obtained. The function value curve of the second derivative function over the entire stretching time range covered by the dynamic stiffness degradation curve is taken as the second derivative curve.

[0036] The zero-crossing points are located on the second derivative curve. The tension time corresponding to each zero-crossing point is marked as the inflection point time, and the slope of the tangent line to the dynamic stiffness degradation curve at the inflection point time is used as the inflection point characteristic. Specifically, within the domain of the second derivative function (which is the effective time interval of the dynamic stiffness degradation curve), the left endpoint of the effective time interval is the start time of the tension loading, and the right endpoint is the end time. Within the effective time interval, a numerical root-finding method is used to search for root nodes where the second derivative is equal to zero. The Brunt method is employed, with the initial search interval divided sequentially from the left endpoint to the right endpoint. Intervals with opposite signs for every two adjacent discrete values ​​of the second derivative are designated as intervals containing root nodes. Root nodes where the second derivative is equal to zero are obtained through iterative approximation. The root nodes located within the effective time interval of the dynamic stiffness degradation curve among all the searched root nodes are designated as zero-crossing points. If multiple zero-crossing points are found, the zero-crossing point that causes the first derivative to reach a local minimum is selected as the inflection point time. The local minimum is determined as follows: within the neighborhood of each zero-crossing point, if the value of the first derivative to the left of the zero-crossing point is positive and the value to the right is negative, then the zero-crossing point corresponds to a local maximum; conversely, if the value to the left is negative and the value to the right is positive, then the zero-crossing point corresponds to a local minimum. The zero-crossing point that causes the first derivative to reach a local minimum is selected, and the stretching time corresponding to this zero-crossing point is marked as the inflection point time. The value of the first derivative at the inflection point time is used as the tangent slope, which is the inflection point characteristic.

[0037] During tensile loading, an acoustic emission sensor continuously collects the acoustic emission signals of the modified MPP cable conduit material under test. The acoustic emission signals are rectified and integrated to obtain a cumulative energy curve. The total cumulative energy value before the inflection point and the incremental cumulative energy value after the inflection point are extracted from the cumulative energy curve. Specifically, the acoustic emission sensor is attached to the outer surface of the gauge length of a dumbbell-shaped standard specimen made from the modified MPP cable conduit material under test. The attachment method involves filling the gap between the acoustic emission sensor and the outer surface of the gauge length with a vacuum grease coupling agent, and then fixing the acoustic emission sensor to the outer surface of the gauge length with an elastic clamp. The output of the acoustic emission sensor is connected to the acoustic emission acquisition system via a coaxial cable. The acoustic emission acquisition system includes a preamplifier and an analog-to-digital converter. In the control interface of the acoustic emission acquisition system, the sampling frequency is set to five times the resonant frequency of the acoustic emission sensor, and the trigger threshold is set to three times the effective value of the background noise of the acoustic emission acquisition system. The acoustic emission acquisition system is started to record the original acoustic emission waveform data during the entire tensile loading process. The original acoustic emission waveform data is a discrete sampling sequence of voltage amplitude changing with time.

[0038] The original acoustic emission waveform data undergoes full-wave rectification, which involves taking the absolute value of each sample value from the original acoustic emission waveform data to obtain the rectified waveform data. Continuous integration is then performed on the rectified waveform data along the time axis to obtain the cumulative energy curve. The integral formula for the cumulative energy curve is:

[0039] in, This indicates the upper limit of the integration operation. This is the index number at the upper limit time. The value range is from the start time of the tensile loading. until the end of the tensile loading ; Indicates the time from the start of the stretch loading. Reaching the upper limit The cumulative energy value; Indicates the integral variable The original acoustic emission waveform data and voltage amplitude at that moment; This represents the absolute value of the voltage amplitude, i.e., the waveform data after rectification. The numerical value at any given time; Representing the integral variable The differential; The tensile loading start time is the moment when the force sensor and acoustic emission acquisition system synchronously begin data acquisition. In the discrete implementation, a composite trapezoidal integral method is used for the rectified waveform data, starting from the tensile loading start time. Initially, the product of the rectified waveform data and the time step is accumulated sequentially for each sampling time step to obtain the cumulative energy value corresponding to each sampling moment. All sampling moments and their corresponding cumulative energy values ​​constitute the cumulative energy curve.

[0040] Find the cumulative energy value corresponding to the inflection point time on the cumulative energy curve. The inflection point time is the stretching time determined in the previous steps. Find the sampling point on the cumulative energy curve whose time is equal to the inflection point time, and take the cumulative energy value corresponding to this sampling point as the total cumulative energy value before the inflection point time. Record the cumulative energy value at the end of the cumulative energy curve as the end-point cumulative energy value, which is the time when the stretching load ends. The corresponding cumulative energy value. Subtract the cumulative energy value at the endpoint from the total cumulative energy value, and use the difference as the incremental cumulative energy value after the inflection point time.

[0041] The numerical values ​​of the inflection point features, the total accumulated energy value before the inflection point time, and the incremental accumulated energy value after the inflection point time are input into a pre-trained nonlinear regression model, which outputs the tensile strength evaluation index. The input feature vector of the nonlinear regression model consists of three components: the first component is the value of the tangent slope, the second component is the total accumulated energy value before the inflection point time, and the third component is the incremental accumulated energy value after the inflection point time. The pre-trained nonlinear regression model uses the input feature vector as input, performs internal nonlinear mapping calculations, and outputs a scalar value, which is the tensile strength evaluation index.

[0042] Example 5: In practical implementation, the process of establishing a pre-trained nonlinear regression model is as follows: (See attached document) Figure 6 Multiple sets of calibration samples of modified MPP cable conduit material with known tensile strength values ​​were obtained. Each set of calibration samples was identical to the modified MPP cable conduit material to be tested in terms of raw material ratio, molding process parameters, and dimensional specifications. The known tensile strength values ​​of each set of calibration samples were measured and recorded using standard tensile testing methods. For each set of calibration samples, the following steps were performed: acquiring initial tensile response data; time-frequency domain decomposition processing to extract transient oscillation components and steady-state creep components; constructing a viscoelastic equivalent model and calculating the dynamic stiffness degradation curve; calculating the second derivative of the dynamic stiffness degradation curve to find the inflection point time and extract the tangent slope; and acquiring acoustic emission signals and calculating the cumulative energy curve. This yielded the tangent slope sample value, the total cumulative energy sample value before the inflection point time, and the incremental cumulative energy sample value after the inflection point time for each set of calibration samples. The tangent slope sample value is used as the inflection point feature sample value. The inflection point feature sample value, the total cumulative energy sample value, and the incremental cumulative energy sample value constitute a set of input feature vectors. The known tensile strength value corresponding to this set of calibrated samples is used as the output label.

[0043] Construct a training dataset. Each training sample in the training dataset contains an input feature vector and a corresponding output label. The first dimension of the input feature vector is the inflection point feature sample value, the second dimension is the total cumulative energy sample value, and the third dimension is the incremental cumulative energy sample value. Divide all training samples into a training subset and a validation subset according to a set ratio, with the ratio of the number of training subsets to the number of validation subsets being four to one.

[0044] The support vector regression algorithm is used to train the training dataset. The core architecture of the support vector regression algorithm is as follows: the input layer receives 3D input feature vectors, and a kernel function maps the 3D input feature vectors from the original 3D feature space to a high-dimensional feature space. A linear regression function is then constructed in the high-dimensional feature space. The expression for the linear regression function is: ,in, This represents the input feature vector. , These are the inflection point feature sample values. This represents the total cumulative energy sample value. For incremental cumulative energy sample values; Represents the input feature vector The image vector in the high-dimensional feature space after kernel function mapping; This represents the weight vector in the high-dimensional feature space. The dimension of the feature space is the same as the dimension of the high-dimensional feature space. Indicates the bias term, bias term It is a scalar; Represents the weight vector With image vector The dot product operation.

[0045] The training objective of the Support Vector Regression algorithm is to find the weight vectors that satisfy the following formula. and bias terms The optimal value:

[0046] in, This represents the sample number of the training sample. This represents the total number of training samples in the training subset; Represents the weight vector The square of the L2 norm; This represents the penalty parameter. The degree to which the model penalizes the fitting error of the training samples, the penalty parameter. The value range is positive real numbers, and the penalty parameter The selection method is to use cross-validation to select candidate values ​​from the set. Selected from; Indicates the first The lower bound relaxation variables for each training sample. Indicates the first The upper bound of the relaxation variables for each training sample. and All are non-negative real numbers; Let represent the sum of the upper and lower bound slack variables for all training samples. The constraint is: , , , ,in Indicates the first The output label of each training sample is the known tensile strength value. Indicates the first The input feature vector of each training sample This represents the insensitive loss width parameter. The value is set to one-tenth of the standard deviation of all output labels in the training subset.

[0047] The kernel function used is the radial basis function (RBF) kernel function, and the expression for the RBF kernel function is: ,in, and These represent two different input feature vectors. and The index identifier for the input feature vector; Represents the input feature vector With input feature vector The square of the Euclidean distance between them; Represents kernel function parameters, kernel function parameters The kernel function parameters determine the range of influence of a single training sample on the radial basis function value. The selection method is to use cross-validation to select candidate values ​​from the set. Selected from; Represented by natural constant An exponential function with base 0.

[0048] Penalty parameters were selected using cross-validation. and kernel function parameters The process is as follows: For each parameter combination in the candidate sets of penalty parameters and kernel function parameters, perform five-fold cross-validation on the training subset. Divide the training subset into five disjoint subsets. Each time, select four subsets as the training set to train the support vector regression model, and use the remaining subset as the validation set to calculate the root mean square error (RMSE) of the model's predictions. The average of the RMSEs obtained from the five validations is used as the cross-validation evaluation value for that parameter combination. The parameter combination with the smallest cross-validation evaluation value is selected as the optimal penalty parameter. and optimal kernel function parameters With optimal penalty parameters and optimal kernel function parameters The optimization problem is resolved on the entire training subset to obtain the weight vector. and bias terms The final value will contain the final weight vector. Final bias term Optimal kernel function parameters The support vector regression model corresponding to the input feature vectors of all support vectors in the training subset is used as a pre-trained nonlinear regression model.

[0049] For a new input feature vector, the output of the pre-trained nonlinear regression model is calculated as follows: ,in This represents the input feature vector corresponding to the modified MPP cable conduit material to be tested; This represents the tensile strength evaluation index output by the model; Indicates the index number of the support vector; This represents the total number of support vectors; and They represent the first The Lagrange multipliers corresponding to each support vector; Indicates the first The input feature vectors corresponding to the support vectors; Represents the input feature vector With input feature vector Radial basis kernel function values ​​between; This represents the final bias term.

[0050] After the tensile strength assessment index is output by the pre-trained nonlinear regression model, the tensile strength assessment index is compared with preset strength grading thresholds. The preset strength grading thresholds are an increasing numerical sequence, including a first grading threshold, a second grading threshold, and a third grading threshold. The value of the first grading threshold is set to 0.8 times the minimum tensile strength value specified in the modified MPP cable conduit material standard, the value of the second grading threshold is set to 1.0 times the minimum tensile strength value specified in the modified MPP cable conduit material standard, and the value of the third grading threshold is set to 1.2 times the minimum tensile strength value specified in the modified MPP cable conduit material standard. When the tensile strength assessment index is less than the first grading threshold, the strength grade of the modified MPP cable conduit material to be tested is determined to be unqualified; when the tensile strength assessment index is greater than or equal to the first grading threshold and less than the second grading threshold, the strength grade of the modified MPP cable conduit material to be tested is determined to be qualified; when the tensile strength assessment index is greater than or equal to the second grading threshold and less than the third grading threshold, the strength grade of the modified MPP cable conduit material to be tested is determined to be good; when the tensile strength assessment index is greater than or equal to the third grading threshold, the strength grade of the modified MPP cable conduit material to be tested is determined to be excellent.

[0051] The target extrusion molding process parameters corresponding to the strength grade are retrieved from a pre-established process parameter database. The pre-established process parameter database is a relational data table containing four columns: strength grade, extrusion temperature, screw speed, and traction speed. The strength grade column stores four strength grade values: unacceptable, acceptable, good, and excellent. The extrusion temperature column stores the setpoint vector of the extruder temperature for each section corresponding to each strength grade. The screw speed column stores the setpoint screw speed for each strength grade. The traction speed column stores the setpoint traction speed for each strength grade. The retrieval method for the target extrusion molding process parameters is as follows: using the currently determined strength grade as the query condition, records in the process parameter database whose strength grade column values ​​equal the current strength grade are matched. The values ​​of the extrusion temperature, screw speed, and traction speed columns from the matched records are then extracted as the target extrusion molding process parameters.

[0052] The target extrusion molding process parameters are output to a display terminal, which is an LCD screen connected to the control computer. The display shows the current test number, the tensile strength assessment index value, a text description of the strength grade, and the values ​​of each parameter in the target extrusion molding process parameters. Simultaneously, the tensile strength assessment index, the inflection point characteristic value, the total cumulative energy value before the inflection point time, and the incremental cumulative energy value after the inflection point time are written to the test record database according to an associated storage format. Each record in the test record database contains six fields: test number, test timestamp, tensile strength assessment index, inflection point characteristic, total cumulative energy value, and incremental cumulative energy value. After the test is completed, the complete record of this test is read from the test record database, generating a test report containing the tensile strength assessment index, inflection point characteristic, total cumulative energy value, incremental cumulative energy value, and strength grade. The test report is saved to local storage media in PDF format.

[0053] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for tensile testing of modified MPP cable conduit material, characterized in that, include: Obtain the initial tensile response data of the modified MPP cable conduit material under test at a preset strain rate; The initial tensile response data is subjected to time-frequency domain decomposition to extract the transient oscillation component and steady-state creep component from the initial tensile response data; Based on the transient oscillation component and the steady-state creep component, a viscoelastic equivalent model of the modified MPP cable conduit material to be tested is constructed, and the dynamic stiffness degradation curve of the modified MPP cable conduit material to be tested during the tensile process is calculated using the viscoelastic equivalent model. Based on the inflection point characteristics of the dynamic stiffness degradation curve, and combined with the accumulated energy of the acoustic emission signal of the modified MPP cable conduit material before tensile fracture, the tensile strength evaluation index of the modified MPP cable conduit material is determined.

2. The tensile strength testing method for modified MPP cable conduit material according to claim 1, characterized in that, The acquisition of initial tensile response data of the modified MPP cable conduit material under a preset strain rate includes: The modified MPP cable conduit material to be tested is processed into a dumbbell-shaped standard specimen, and multiple resistance strain gauges are attached axially to the surface of the gauge length section of the dumbbell-shaped standard specimen. The dumbbell-shaped standard specimen is installed in the fixture of the universal testing machine, the tensile rate of the universal testing machine is set to the preset strain rate, and the universal testing machine is started to apply tensile loading to the dumbbell-shaped standard specimen. During the tensile loading process, the tensile force value is synchronously collected by the force sensor of the universal testing machine, and the axial strain value of the gauge length is synchronously collected by the multiple resistance strain gauges. The tensile force value and the axial strain value collected at the same time are combined into a data pair. After arranging all the data pairs in chronological order, the initial tensile response data is obtained.

3. The tensile strength testing method for modified MPP cable conduit material according to claim 1, characterized in that, The step of performing time-frequency domain decomposition on the initial tensile response data to extract the transient oscillation component and steady-state creep component from the initial tensile response data includes: The axial strain value sequence in the initial tensile response data is resampled at equal time intervals to obtain a uniformly sampled strain sequence, and empirical mode decomposition is performed on the uniformly sampled strain sequence to obtain multiple intrinsic mode function components. The instantaneous frequency mean of each intrinsic mode function component is calculated, and the intrinsic mode function components with an instantaneous frequency mean higher than a preset high-frequency threshold are superimposed to obtain the transient oscillation component. The intrinsic mode function components whose instantaneous frequency mean is lower than the preset low frequency threshold are superimposed to obtain the intermediate trend component. The intermediate trend component is then subjected to polynomial fitting to obtain the steady-state creep component.

4. The tensile strength testing method for modified MPP cable conduit material according to claim 1, characterized in that, The step involves constructing a viscoelastic equivalent model of the modified MPP cable conduit material under test based on the transient oscillation component and the steady-state creep component, and using the viscoelastic equivalent model to calculate the dynamic stiffness degradation curve of the modified MPP cable conduit material under test during tensile stress, including: By mapping the transient oscillation component and the steady-state creep component to the Laplace domain, the transient oscillation frequency domain function and the steady-state creep frequency domain function are obtained respectively. An initial viscoelastic network model is constructed using multiple spring elements and multiple damping elements. The spring modulus and damping coefficient of the high-frequency branch in the initial viscoelastic network model are determined according to the transient oscillation frequency domain function, and the spring modulus and damping coefficient of the low-frequency branch in the initial viscoelastic network model are determined according to the steady-state creep frequency domain function. The initial viscoelastic network model with determined parameters is used as the viscoelastic equivalent model. Using the viscoelastic equivalent model, the dynamic storage modulus and dynamic loss modulus of the modified MPP cable conduit material under test are calculated in the Laplace domain at different tensile time points. Based on the ratio of the dynamic storage modulus to the dynamic loss modulus, the dynamic stiffness degradation curve is obtained.

5. The tensile strength testing method for modified MPP cable conduit material according to claim 4, characterized in that, The initial viscoelastic network model is a three-parameter solid model consisting of two Maxwell branches and one independent spring branch connected in parallel.

6. The tensile strength testing method for modified MPP cable conduit material according to claim 1, characterized in that, The tensile strength evaluation index of the modified MPP cable conduit material under test is determined based on the inflection point characteristics of the dynamic stiffness degradation curve and the accumulated energy of the acoustic emission signal before tensile fracture, including: The second derivative of the dynamic stiffness degradation curve is calculated to obtain the second derivative curve. The zero-crossing point is found on the second derivative curve. The tension time corresponding to the zero-crossing point is marked as the inflection point time. The slope of the tangent of the dynamic stiffness degradation curve at the inflection point time is used as the inflection point feature. During the tensile loading process, an acoustic emission sensor is used to continuously collect the acoustic emission signal of the modified MPP cable conduit material under test during the tensile process. The acoustic emission signal is rectified and integrated to obtain a cumulative energy curve. The total cumulative energy value before the inflection point time and the incremental cumulative energy value after the inflection point time are extracted from the cumulative energy curve. The numerical values ​​of the inflection point feature, the total cumulative energy value before the inflection point time, and the incremental cumulative energy value after the inflection point time are input into a pre-trained nonlinear regression model, and the tensile strength evaluation index is output by the nonlinear regression model.

7. The tensile strength testing method for modified MPP cable conduit material according to claim 6, characterized in that, The process of calculating the second derivative of the dynamic stiffness degradation curve to obtain the second derivative curve, finding the zero-crossing point on the second derivative curve, marking the tension time corresponding to the zero-crossing point as the inflection point time, and using the slope of the tangent line of the dynamic stiffness degradation curve at the inflection point time as the inflection point feature includes: Cubic spline interpolation is performed on the discrete data points on the dynamic stiffness degradation curve to obtain a smooth continuous curve function, and the first and second derivatives of the smooth continuous curve function are obtained. Within the domain of the second derivative function, the root nodes where the second derivative function is equal to zero are searched by numerical root-finding method, and the root nodes located within the effective time interval of the dynamic stiffness degradation curve are taken as the zero-crossing points. The zero-crossing position that causes the first derivative function to obtain a local minimum value is selected from the zero-crossing positions as the inflection point time, and the value of the first derivative function at the inflection point time is taken as the tangent slope.

8. The tensile strength testing method for modified MPP cable conduit material according to claim 6, characterized in that, During the tensile loading process, an acoustic emission sensor continuously collects the acoustic emission signals of the modified MPP cable conduit material under test during the tensile process. The acoustic emission signals are rectified and integrated to obtain a cumulative energy curve. The total cumulative energy value before the inflection point time and the incremental cumulative energy value after the inflection point time are extracted from the cumulative energy curve, including: The acoustic emission sensor is attached to the outer surface of the gauge length of the dumbbell-shaped standard specimen made of the modified MPP cable conduit material to be tested, and the output end of the acoustic emission sensor is connected to the acoustic emission acquisition system. The sampling frequency and trigger threshold of the acoustic emission acquisition system are set, and the acoustic emission acquisition system is started to record the original acoustic emission waveform data throughout the tensile loading process. The original acoustic emission waveform data is subjected to full-wave rectification to obtain rectified waveform data, and the rectified waveform data is continuously integrated on the time axis to obtain the cumulative energy curve. Find the cumulative energy value corresponding to the inflection point time on the cumulative energy curve, take the cumulative energy value as the total cumulative energy value before the inflection point time, and take the difference between the cumulative energy value at the end of the cumulative energy curve and the total cumulative energy value as the incremental cumulative energy value after the inflection point time.

9. The tensile strength testing method for modified MPP cable conduit material according to claim 6, characterized in that, Before inputting the numerical value of the inflection point feature, the total cumulative energy value before the inflection point time, and the incremental cumulative energy value after the inflection point time into a pre-trained nonlinear regression model, and before the nonlinear regression model outputs the tensile strength evaluation index, the method further includes: Multiple sets of calibration samples of modified MPP cable conduit material with known tensile strength values ​​are obtained. The method steps of any one of claims 1 to 8 are performed on each set of calibration samples to obtain the inflection point characteristic sample value, the total cumulative energy sample value before the inflection point time, and the incremental cumulative energy sample value after the inflection point time for each set of calibration samples. Using the inflection point feature sample value, the total cumulative energy sample value, and the incremental cumulative energy sample value as input features, and the known tensile strength value as the output label, a training dataset is constructed. The training dataset is trained using the support vector regression algorithm, and the penalty parameter and kernel function parameter are selected through cross-validation to obtain the pre-trained nonlinear regression model.

10. A method for tensile testing of modified MPP cable conduit material according to claim 6, characterized in that, After the tensile strength assessment index is output from the nonlinear regression model, the method further includes: The tensile strength evaluation index is compared with a preset strength grading threshold to determine the strength grade of the modified MPP cable conduit material to be tested, and the target extrusion molding process parameters corresponding to the strength grade are retrieved from a pre-established process parameter database based on the strength grade. The target extrusion molding process parameters are output to the display terminal, and the tensile strength evaluation index, the inflection point characteristic, the total cumulative energy value and the incremental cumulative energy value are associated and stored in the test record database to generate a test report containing the tensile strength evaluation index.

Images