A test method for evaluating the resistance of reinforced plastics to the penetration of chemical solvents

Abstract

The present application relates to the technical field of material performance testing, and discloses a kind of test methods for evaluating the chemical solvent penetration resistance of reinforced plastic, comprising the following steps: constructing inverse transfer function to compensate platform lag, using historical strain data to predict the starting time of microcrack opening in the next cycle, combining platform inherent delay to calculate instruction sending time, based on stiffness attenuation rate calculated in real time to adaptively adjust the best disturbance frequency, using inverse transfer function to generate modified disturbance instruction and superimposed injection, deducting fluid viscous damping work previously calibrated from single cycle data to obtain effective dissipation density, determining failure end point based on the second derivative change of effective dissipation density sequence.The present application realizes the time synchronization of pressure wave and crack opening through prediction feedforward strategy, maintains deep penetration driving force using adaptive disturbance, and eliminates fluid damping interference through energy decoupling mechanism, improving the authenticity and result accuracy of stress-assisted solvent penetration test.

Description

Technical Field

[0001] This invention relates to the field of material performance testing technology, specifically a test method for evaluating the chemical solvent penetration resistance of reinforced plastics. Background Technology

[0002] In the field of robotics, especially for heavy-duty hydraulic robots operating in complex environments, their hydraulic power transmission systems often employ glass fiber or carbon fiber reinforced engineering plastic tubing to meet the requirements of lightweight design and corrosion resistance. During service, these power lines not only withstand high-frequency, high-pressure pulsed mechanical stress, but their inner walls also require prolonged contact with permeable chemical solvents (such as glycol-based hydraulic oil). This stress-solvent coupling environment causes the solvent to penetrate the material matrix more rapidly under alternating stress, leading to resin swelling and interfacial debonding, thereby affecting the robot's operational reliability.

[0003] To assess the service life of such materials, existing testing techniques largely rely on traditional pressure pulse tests or static immersion fatigue tests. However, in simulation tests targeting high dynamic response conditions, the physical response lag between the controller issuing commands and the actual establishment of the target pressure field inside the tubular sample is limited by the bulk modulus of the hydraulic oil in the electro-hydraulic servo loading system and the inertial mass of the mechanical actuator. This lag causes the fluid pressure peaks to fail to accurately align with the moment when microcracks open in the material, resulting in stress-assisted penetration test conditions deviating from real-world conditions and failing to effectively simulate the physical process by which high-frequency alternating stress promotes solvent penetration into microcracks.

[0004] Meanwhile, existing testing methods typically employ constant loading frequency and waveform parameters throughout the entire lifespan. As chemical solvents penetrate deeper into the reinforced plastic, the resin matrix undergoes a plasticizing effect, leading to changes in the material's dynamic stiffness and damping characteristics. If a fixed perturbation frequency is maintained, the hydrodynamic excitation will struggle to adapt to the changes in the material's pore structure, resulting in insufficient penetration driving force in the later stages of the test and failing to accurately reflect the deep damage evolution of the material under stiffness decay conditions.

[0005] Furthermore, in fluid-structure coupling tests involving high-viscosity chemical solvents, the viscous frictional heat generated by the reciprocating flow of the fluid within the tube leads to energy dissipation. Traditional damage assessment methods often directly calculate the dissipated energy based on the total hysteresis loop area. This approach fails to separate the fluid dissipation caused by viscous damping from the irreversible work generated by microscopic interface damage in the material. Consequently, fluid dynamics factors interfere with the identification of debonding signals at the fiber-matrix interface, making it difficult to accurately determine the failure endpoint. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a test method for evaluating the chemical solvent penetration resistance of reinforced plastics. This method solves the problems in existing technologies, such as the misalignment of pressure wave and crack opening timing due to system physical response lag, the inability of fixed frequency loading to adapt to material stiffness decay leading to decreased penetration driving efficiency, and the inability to accurately identify the material interface damage failure endpoint due to fluid viscosity damping interference.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] This invention provides a test method for evaluating the chemical solvent penetration resistance of reinforced plastics. The method is based on a fluid-mechanical coupling test platform and includes the following steps:

[0009] While maintaining the basic static pressure state, input a random pressure command and collect the measured internal pressure. Construct an inverse transfer function based on the relationship between the random pressure command and the measured internal pressure.

[0010] Historical strain data of tubular specimens are collected using a laser scanning micrometer to estimate the start time of microcrack opening in the next cycle, and the command transmission time is calculated in combination with the inherent delay of the platform.

[0011] The stiffness attenuation rate of the current cycle is calculated based on real-time data collected by the laser scanning micrometer and pressure sensor. The optimal perturbation frequency matching the current damage state is calculated according to the perturbation frequency adaptive formula. The correction perturbation command is generated using the inverse transfer function and sent to the electro-hydraulic servo valve at the command sending time.

[0012] The measured internal pressure and radial strain data are extracted from the collected single-cycle data. The pre-calibrated fluid viscosity damping work is deducted, and the effective dissipation density is calculated using the modified dissipation energy formula.

[0013] The second derivative of the effective dissipation density over time is calculated. The failure endpoint is determined based on the change in the second derivative. The number of cycles at the failure endpoint is then used as the media resistance life of the reinforced plastic.

[0014] Furthermore, the construction of the inverse transfer function based on the relationship between the random pressure command and the measured internal pressure specifically includes: synchronously acquiring the command voltage generated by the random pressure command and the measured internal pressure fed back by the pressure sensor; using the Fourier transform algorithm to process the command voltage and the measured internal pressure to calculate the forward transfer function reflecting the frequency response characteristics of the platform; performing the inverse operation on the forward transfer function to generate the inverse transfer function and storing it, which serves as a pre-filter for subsequent high-frequency disturbance commands.

[0015] Furthermore, during the inverse operation of the forward transfer function, a lower amplitude threshold is introduced for judgment: when the magnitude of the forward transfer function is greater than the lower amplitude threshold, the reciprocal is directly calculated; when the magnitude of the forward transfer function is less than or equal to the lower amplitude threshold, the gain of the inverse function at the corresponding frequency point is limited to a preset saturation value; wherein, the lower amplitude threshold is preset based on the system background noise level, and the preset saturation value is preset based on the numerical stability requirements of the test platform.

[0016] Furthermore, the calculation of the starting time of microcrack opening in the next cycle specifically includes: applying a basic load to the tubular sample by driving an electro-hydraulic servo valve, while simultaneously collecting historical strain data of the tubular sample for a set number of cycles using a laser scanning micrometer; constructing a time series prediction model using the historical strain data, projecting the waveform characteristics of the complete cycle in the historical strain data onto the future on the time axis to reconstruct the predicted strain curve for the next cycle; searching in the predicted strain curve for the time point when the value first crosses the opening threshold, and marking this time point as the starting time; wherein, the opening threshold is a critical strain value determined based on the stress-strain constitutive relationship of the reinforced plastic material.

[0017] Furthermore, the calculation logic for calculating the command transmission time in conjunction with the platform's inherent delay is as follows: the command transmission time is equal to the predicted start time of the microcrack opening in the next cycle minus the platform's inherent delay; where the platform's inherent delay is obtained by converting the phase lag of the forward transfer function at the base load frequency.

[0018] Furthermore, calculating the optimal perturbation frequency to match the current damage state based on the perturbation frequency adaptive formula specifically includes: extracting stress and strain data in the linear segment of the current cycle along the rising edge of the base load based on real-time data collected by the laser scanning micrometer and pressure sensor; calculating the current dynamic stiffness modulus; and calculating the stiffness attenuation rate in combination with the initial dynamic stiffness modulus; substituting the stiffness attenuation rate into the perturbation frequency adaptive formula, which defines that the optimal perturbation frequency is proportional to the preset initial frequency and inversely proportional to the logarithmic function term containing the stiffness attenuation rate; wherein, the perturbation frequency adaptive formula includes a frequency attenuation coefficient, which is preset based on the kinematic viscosity of the chemical solvent to be tested and the fiber content of the reinforcing plastic; and the initial dynamic stiffness modulus is calculated based on the average data of the initial stage of the test.

[0019] Furthermore, the process of generating a corrected disturbance command using the inverse transfer function and superimposing it onto the electro-hydraulic servo valve at the command transmission time specifically includes: generating an original disturbance command in the form of a sine wave based on the optimal disturbance frequency; calling the stored inverse transfer function to preprocess and correct the amplitude and phase of the original disturbance command to generate a corrected disturbance command; and when the clock of the test platform controller reaches the command transmission time, superimposing the corrected disturbance command onto the low-frequency trapezoidal wave base load command sequence and sending it to the electro-hydraulic servo valve.

[0020] Furthermore, the calculation of effective dissipation density using the modified dissipation energy formula specifically includes: extracting measured internal pressure and radial strain data synchronously collected by a pressure sensor and a laser scanning micrometer within a single cycle; differentiating the radial strain data to obtain the radial strain rate; performing closed-loop integration on the data sequence extracted from the collected single-cycle data to calculate the total dissipation energy, and subtracting the pre-calibrated fluid viscous damping work to obtain the effective dissipation density; wherein, the pre-calibrated fluid viscous damping work is determined based on the hysteresis loop area measured under the same test conditions of a rigid standard tube.

[0021] Furthermore, determining the failure endpoint based on the change in the second derivative value, and using the number of cycles at the failure endpoint as the media resistance life of the reinforced plastic, specifically includes: real-time monitoring of the second derivative value, calculating the absolute difference between the second derivative value of the current sampling cycle and the second derivative value of the previous sampling cycle; when the absolute difference is greater than a preset failure threshold, determining that the solvent front has broken through the surface resin-rich region and entered the fiber interface layer, sending a stop command to the electro-hydraulic servo fatigue testing machine, and recording the current number of cycles as the media resistance life of the reinforced plastic; wherein, the failure threshold is preset according to a statistical dynamic setting method, which includes extracting a stable interval of second derivative data in the initial stage of the test as a background noise sample, calculating the standard deviation of the data within the background noise sample, and setting the failure threshold as a preset multiple of the standard deviation.

[0022] Furthermore, the calculation of the second derivative of the time series of effective dissipation density also includes: smoothing the time series using the moving average filtering method before performing the second derivative operation; the second derivative operation is specifically performed using the three-point central difference formula.

[0023] This invention provides a test method for evaluating the chemical solvent penetration resistance of reinforced plastics. It has the following beneficial effects:

[0024] 1. This invention solves the problem of misalignment between pressure wave and crack opening timing caused by the physical response lag in electro-hydraulic servo systems by establishing a strain gating strategy based on periodic predictive feedforward. This method utilizes historical data to predict the microcrack opening time and combines this with the platform's inherent delay to send commands in advance, ensuring that the high-frequency fluid pressure wave accurately acts at the moment of maximum molecular chain segment gap or microcrack opening. This achieves time synchronization between control commands and physical actions, improving the accuracy of stress-assisted solvent penetration test results.

[0025] 2. This invention addresses the failure of traditional fixed-frequency tests during the material softening stage by injecting adaptive micro-perturbations based on the material's damage state. This method monitors the material's stiffness decay rate in real time and dynamically adjusts the perturbation frequency according to the nonlinear rheological relationship of the fluid in the porous medium. This achieves dynamic matching between the fluid perturbation frequency and the material's mechanical properties, ensuring an effective deep-penetration driving force throughout the material's entire lifespan and guaranteeing the fluid medium's penetration into the micropores.

[0026] 3. This invention addresses the challenge of determining the true damage characteristics of materials affected by fluid viscous damping by constructing an energy dissipation decoupling analysis mechanism and a failure determination method based on the second derivative. This method extracts the effective dissipation density sensitive to chemical penetration by subtracting a pre-calibrated fluid viscous damping work, and utilizes the abrupt change in its second derivative to identify interface failure points, thus achieving real-time determination of the failure endpoint in environments with strong fluid-mechanical coupling. Attached Figure Description

[0027] Figure 1 This is a flowchart of a test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to the present invention;

[0028] Figure 2 A flowchart illustrating the adaptive micro-perturbation based on material damage state for the present invention;

[0029] Figure 3 This is a graph showing the evolution of the effective dissipation density and the failure determination process of the present invention. Detailed Implementation

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Please see the appendix Figure 1 This invention provides a test method for evaluating the chemical solvent penetration resistance of reinforced plastics, which is performed based on a fluid-mechanical coupling test platform.

[0032] The fluid-mechanical coupling testing platform is built upon an electro-hydraulic servo fatigue testing machine. An electro-hydraulic servo valve, featuring high-frequency response, is integrated into the platform to drive a hydraulic cylinder. The transmission medium inside the hydraulic cylinder is replaced with the test chemical solvent. The tubular specimen is prepared using injection molding, with both ends connected to the fluid circuit of the hydraulic cylinder via high-pressure flanges, allowing the test chemical solvent to directly fill the interior of the tubular specimen. A pressure sensor, a flush diaphragm structure, is installed at the inlet of the high-pressure flange. Its pressure-sensing diaphragm is in direct contact with the fluid and has no dead space, used to acquire hysteresis-free measured internal pressure. A laser scanning micrometer is vertically positioned outside the axis of the tubular specimen to acquire real-time changes in the specimen's outer diameter and convert them into radial strain. A constant-temperature oil bath circulation system is connected to the hydraulic power source to maintain a constant temperature for the test chemical solvent.

[0033] The method includes the following steps:

[0034] Step S1 involves constructing the fluid dynamics frequency response calibration and compensation model. Before formal loading, while maintaining the baseline static pressure, a random pressure command with a frequency range covering 10Hz to 100Hz is input to the electro-hydraulic servo valve. Simultaneously, the command voltage generated by the random pressure command and the measured internal pressure fed back by the pressure sensor are acquired.

[0035] The Fourier transform algorithm is used to process the command voltage and the measured internal pressure to calculate the forward transfer function, which reflects the frequency response characteristics of the platform. The forward transfer function is then inverted to generate the inverse transfer function, which is stored. This inverse transfer function serves as a pre-filter for subsequent high-frequency disturbance commands, used to eliminate viscous damping and phase lag in the transmission of high-frequency fluids in the pipeline.

[0036] Step S2: Establish a strain gating strategy based on periodic predictive feedforward. A basic load, a low-frequency trapezoidal wave, is applied to the tubular specimen by driving an electro-hydraulic servo valve. Simultaneously, the strain gating of the tubular specimen is acquired in real time using a laser scanning micrometer. Historical strain data for each cycle. A time series prediction model is constructed using the historical strain data, and this model is used to estimate the start time when the strain in the next cycle exceeds a preset opening threshold.

[0037] The pre-calibrated inherent platform latency is read, and the difference is calculated by combining it with the start time to obtain the command transmission time that can offset the physical lag. This time is then set as the disturbance trigger point for the next cycle. The formula for calculating the command transmission time is as follows:

[0038] ;

[0039] In the formula: To compensate for the physical delay in command transmission; This is to predict the start time of microcrack opening in the next cycle; This refers to the inherent latency of the platform, specifically the inherent delay time of the platform's physical response.

[0040] Step S3: Inject adaptive micro-perturbations based on the material damage state. Calculate the stiffness decay rate for the current cycle based on real-time data collected by a laser scanning micrometer and pressure sensor. Substitute the stiffness decay rate into the adaptive perturbation frequency formula to calculate the optimal perturbation frequency matching the current damage state. The adaptive perturbation frequency formula is as follows:

[0041] ;

[0042] In the formula: For the first The optimal perturbation frequency that the platform should inject within each load cycle; The preset initial frequency; This is the frequency attenuation coefficient; The operator for natural logarithms; For the first Stiffness decay rate over a load cycle.

[0043] The original disturbance command is generated based on the optimal disturbance frequency, and the stored inverse transfer function is called to preprocess and correct the amplitude and phase of the original disturbance command to generate a corrected disturbance command. When the clock of the test platform's controller reaches the command transmission time, the corrected disturbance command is superimposed and sent to the electro-hydraulic servo valve, thereby generating pressure oscillation at the moment when the microcracks inside the tubular sample open.

[0044] Step S4: Analyze the damage characteristics based on energy dissipation decoupling. Extract the measured internal pressure and historical strain data synchronously acquired within a single cycle, and differentiate the historical strain data to obtain the radial strain rate. Perform closed-loop integration on the extracted data sequence to calculate the total dissipated energy, subtract the pre-calibrated fluid viscous damping work, and use the modified dissipated energy formula to calculate the effective dissipation density. The modified dissipated energy formula is as follows:

[0045] ;

[0046] In the formula: For the first Effective dissipation density of material per unit volume within one load cycle; To complete a load cycle Closed-loop integral operators within; For a moment The measured internal pressure collected by a pressure sensor; For a moment The radial strain rate, i.e., the derivative of radial strain with respect to time; For a moment Radial strain acquired by a laser scanning micrometer; For time, it is the differential variable; This is the pre-calibrated fluid viscosity damping work.

[0047] Step S5: Determine the failure endpoint. Arrange the continuously calculated effective dissipation densities into a time series, and perform a second-order derivative operation on the series to obtain the second-order derivative value. Monitor the second-order derivative value in real time and calculate the absolute difference between the second-order derivative value of the current sampling period and the second-order derivative value of the previous sampling period. When the absolute difference is greater than the preset failure threshold, it is determined that the solvent front has broken through the surface resin-rich region and entered the fiber interface layer. At the same time, a stop command is sent to the electro-hydraulic servo fatigue testing machine, and the current number of cycles is recorded as the media resistance life of the reinforced plastic.

[0048] In this embodiment, step S1 constructs a mathematical model that can accurately describe and inversely compensate for the physical transmission process from the controller command to the actual pressure inside the tubular sample. Due to the inherent fluid viscosity damping, frictional resistance along the pipeline, and compressibility of the hydraulic oil column in the transmission of the tested chemical solvent (such as fuel or coolant), the actual pressure waveform acting on the inner wall of the sample will inevitably exhibit amplitude attenuation and phase lag compared to the controller command waveform. Without active compensation, especially under high-frequency loading conditions, this physical distortion will cause the test stress level to deviate from the preset value, thus affecting the accuracy of the life assessment. Therefore, it is necessary to analyze the frequency response characteristics of the platform through the calibration process and construct an inverse model. The specific implementation logic of this step is as follows:

[0049] In sub-step S101, to obtain the full-frequency response characteristics, a wideband excitation needs to be applied to the system. While maintaining the baseline static pressure, a random pressure command with a frequency range covering 10Hz to 100Hz is input to the electro-hydraulic servo valve. The baseline static pressure should be set higher than the saturated vapor pressure of the chemical solvent under test at the test temperature, for example, set to 0.2MPa to 0.5MPa, to ensure the pipeline is in a fully pre-tensioned state and to prevent fluid cavitation.

[0050] Random pressure commands are preferably generated using band-limited white noise signals, whose power spectral density remains uniformly distributed within the aforementioned frequency bands of interest. The selection of 10Hz to 100Hz as the calibration frequency band is based on a comprehensive analysis of the microporous response characteristics of reinforced plastic materials under solvent permeation induction and the linear frequency response bandwidth of commercial high-frequency hydraulic servo valves. Random signal excitation within this frequency band can simultaneously elicit the platform's dynamic response at all key frequency points. Compared to traditional sinusoidal frequency sweep methods, this shortens calibration time and avoids equipment fatigue caused by prolonged resonance.

[0051] In sub-step S102, based on the clock synchronization mechanism, the command voltage generated by the random pressure command and the measured internal pressure fed back by the pressure sensor are synchronously acquired. To ensure the integrity of time-domain and frequency-domain information, the sampling frequency of the controller's data acquisition channel should be set to 5 to 10 times the highest frequency of interest (100Hz) according to the Nyquist sampling theorem, for example, set to 1kHz, to effectively suppress frequency domain aliasing.

[0052] The acquired discrete-time domain signal is converted into a frequency domain signal using the Fourier transform algorithm: the command voltage in the time domain is converted into the input spectrum in the frequency domain. The measured internal pressure in the time domain is converted into the output spectrum in the frequency domain. Based on this, the ratio of the output spectrum to the input spectrum is calculated using complex division, yielding the positive transfer function that reflects the platform's inherent frequency response characteristics. This function It is in complex form, and its modulus is Physically, this characterizes the pressure amplitude gain ratio at that frequency, while its argument... Physically, it characterizes the phase delay of the pressure wave at this frequency.

[0053] In sub-step S103, to achieve active control compensation, the forward transfer function needs to be modified. Perform the inverse operation to generate the inverse transfer function. Its basic calculation logic is to take the reciprocal of the forward transfer function, that is... .

[0054] To avoid the singularity problem caused by the denominator approaching zero in numerical calculations (i.e., when the response amplitude at certain frequency points is extremely small, it can lead to calculation overflow or noise amplification), this embodiment introduces a lower amplitude threshold. Make a judgment: when When, directly calculate the reciprocal; when When the frequency point is at that frequency, the inverse function gain is limited to a preset saturation value or kept at the value of the previous frequency point.

[0055] in, The value of is typically set to 3 to 5 times the system background noise level, for example, set to 0.01 (normalized amplitude) to ensure the numerical stability of the inverse model. The generated inverse transfer function... In a physical sense, this constitutes a digital predistortion filter, which is stored in the controller in the form of lookup tables or polynomial coefficients. The preset saturation value is pre-set based on the numerical stability requirements of the test platform.

[0056] In subsequent steps, any target pressure command issued must first be processed by the convolution of the filter to pre-amplify the command signal in amplitude to compensate for fluid damping and pre-lead in phase to compensate for transmission lag, thereby ensuring that the pressure waveform acting on the inner wall of the tubular sample can reproduce the expected characteristics with high fidelity.

[0057] In this embodiment, step S2 constructs a control strategy to solve the technical problem of the physical response lag and the asynchronous time axis of the dynamic opening window of microcracks in the electro-hydraulic servo control loop.

[0058] Due to the bulk modulus of the hydraulic transmission medium (such as chemical solvents) and the inertial mass of the mechanical actuator, there is an inherent time phase difference between the command signal issued by the controller and the actual establishment of the target pressure field inside the hydraulic cylinder. If a traditional threshold triggering logic based on real-time sensor data is used, when the controller detects that the strain signal exceeds the threshold and immediately issues a disturbance command, the pressure wave, after transmission delay, may miss the optimal time window for maximum crack opening and highest permeability, resulting in a significant reduction in the effectiveness of stress-assisted penetration. Therefore, this step introduces a time series prediction algorithm. By deeply analyzing the waveform characteristics of historical periodic data, the timing of the next cycle's action can be locked in advance, achieving feedforward control. The specific implementation logic of this step is as follows:

[0059] In sub-step S201, as the basis for data accumulation and feature extraction, the electro-hydraulic servo valve is driven to apply a basic load to the tubular sample, and the high-frequency sampling capability of the laser scanning micrometer is used to acquire data on the front of the tubular sample in real time. Historical strain data for each cycle .

[0060] As a preferred embodiment, the applied base load is set as a low-frequency trapezoidal wave (e.g., a frequency of 0.1 Hz to 1 Hz with a duty cycle of 50%). Compared to a continuously varying sine wave, the trapezoidal wave possesses a defined and stable high-pressure holding plateau period. This plateau period provides sufficient time for stress relaxation of polymer chain segments and full opening of microcracks, thereby creating a more stable physical channel for deep penetration of solvent molecules. Regarding the number of historical periods... The selection is usually set to 5 to 10 (e.g., take...). The determination of this value follows the principle of balancing signal-to-noise ratio and computing power.

[0061] On the one hand, it is necessary to accumulate a sufficient amount of data to filter out random noise interference in a single measurement through averaging and to establish periodic features with high confidence.

[0062] On the other hand, it is necessary to avoid excessively long data queues from consuming the controller's memory resources and increasing the real-time computing load, so as to ensure the real-time performance of the control cycle.

[0063] Historical strain data collected It is structured and stored in the controller's circular buffer as an input source for subsequent prediction models.

[0064] In sub-step S202, historical strain data within the buffer are utilized. Construct a time series forecasting model and use the model to estimate the strain in the next period. Exceeding the preset opening threshold The start time Considering the controlled periodicity of the base load in this testing method, and to reduce algorithm complexity and improve stability in the industrial setting, the time series prediction model preferably employs a periodic mapping extrapolation method. Its specific computational logic is as follows:

[0065] The controller locks onto the waveform characteristics of the most recent complete cycle in the historical data, based on the preset cycle of the base load. Projecting the waveform characteristics onto the future along the time axis, i.e., predicting the value... This allows for the reconstruction of the predicted strain curve for the next cycle. Of course, if computing power allows, an autoregressive moving average model can be used for more precise fitting and prediction.

[0066] Based on this, the controller generates the prediction curve. During monotonic retrieval, the search value first crosses the opening threshold. The time point is determined and marked as the start time. If the maximum value of the predicted curve over the entire period is still less than... (Indicating that the material has not yet entered the microcrack propagation stage or the basic load is insufficient), the controller will remain silent in the next cycle and will not trigger any disturbances to avoid ineffective energy injection.

[0067] Among them, the opening threshold The threshold is determined based on the stress-strain constitutive relationship of reinforced plastic materials, and is physically defined as the critical strain value at which the material transitions from the linear elastic deformation stage to the nonlinear viscoelastic deformation stage or the microcrack initiation stage. As a specific example, it is typically taken as 10% to 20% of the material's elongation at break; for chopped glass fiber reinforced polypropylene, this threshold can be specifically set to 0.5%. This threshold setting ensures that high-frequency perturbations are precisely introduced only in high-permeability states where the spacing between molecular chain segments in the material matrix increases or interfacial microcracks open.

[0068] In sub-step S203, the platform-inherent latency obtained through calibration in step S1 is read. Combined with the predicted start time Perform difference calculation to obtain the instruction sending time. And set it as the disturbance trigger point for the next cycle.

[0069] Among them, inherent delay Characterized by the physical time difference between the issuance of a digital command by the controller and the establishment of fluid pressure on the inner wall of the sample, its value is expressed through a forward transfer function. The phase lag at the base load frequency is calculated. Command transmission time. The calculation follows the formula:

[0070] ;

[0071] In the formula: To compensate for the physical delay in command transmission, this moment must be located on the timeline before the actual crack opening moment; The predicted start time of the next cycle of microcrack opening represents the absolute time when the expected physical event of microcrack opening occurs. This refers to the inherent latency of the platform, specifically the inherent delay time of the platform's physical response.

[0072] The instruction transmission time calculated based on this formula The controller predicts before the crack is about to open. A high-frequency disturbance command is issued in advance, second by second. After a physical transmission delay of one second, the pressure wave front just occurred at the opening of the microcrack. The system accurately reaches and acts on the inner wall of the sample at all times, thereby achieving precise spatiotemporal synchronization between the control logic (digital domain) and physical actions (fluid domain), maximizing the effect of stress-assisted permeation. The filtering and smoothing algorithms involved in the above time series processing can be implemented by those skilled in the art with reference to relevant digital signal processing technical standards; these are well-known technologies in the field and will not be elaborated upon here.

[0073] See appendix Figure 2 In this embodiment, step S3 establishes a micro-perturbation injection mechanism that can be dynamically adjusted as the material properties evolve, in order to solve the problem of fixed frequency perturbation failing during the material softening stage.

[0074] As the penetration depth of the chemical solvent increases, the microstructure of the reinforced plastic material transforms from a dense state to a swollen state, and its viscoelastic characteristics evolve. In the initial stage of testing, the material matrix is ​​dense, and high-frequency perturbation helps the fluid overcome the capillary forces of the micropores. However, in the later stages of testing, the solvent plasticizing effect leads to a decrease in matrix modulus and an increase in porosity. If the perturbation frequency is maintained at too high a frequency, the fluid pressure wave will be absorbed by the material's high damping characteristics and converted into heat energy, resulting in insufficient driving force for deep penetration. Therefore, this step dynamically adjusts the injection frequency by monitoring the changes in the material's dynamic stiffness in real time to achieve real-time impedance matching between the fluid dynamic excitation and the material's mechanical state. The specific implementation logic of this step is as follows:

[0075] In sub-step S301, the stiffness decay rate for the current cycle is calculated based on real-time data collected by the laser scanning micrometer and pressure sensor. As a preferred data processing method, extracting the current data... The stress and strain data for each period along a linear segment of the base load increase (e.g., the pressure range from 30% to 80%) are used to fit the slope of this data segment using the least squares method to obtain the current dynamic stiffness modulus. .

[0076] Simultaneously, retrieve the initial dynamic stiffness modulus from the initial testing phase (e.g., the average value of the first 50 cycles). To ensure the numerical stability of the algorithm, it is necessary to perform calculations before computation. Perform non-zero verification:

[0077] like If the noise level is less than the preset sensor noise threshold (e.g., 0.1 MPa), it is considered an installation malfunction and an alarm is triggered. If the noise level is not less than the preset sensor noise threshold, the installation is considered normal.

[0078] If the installation is successful, proceed according to the formula. Calculate the attenuation rate. Stiffness attenuation rate. Physically, it comprehensively characterizes the degree of microcrack propagation caused by matrix softening due to solvent penetration and mechanical fatigue.

[0079] Based on this, the calculated stiffness attenuation rate Substituting into the adaptive perturbation frequency formula, the optimal perturbation frequency matching the current damage state is calculated. This formula employs a logarithmic decay model to simulate the nonlinear rheological relationship of permeation resistance in porous media as the equivalent pore size increases. The adaptive formula for the perturbation frequency is as follows:

[0080] ;

[0081] In the formula: For the first The optimal perturbation frequency that the platform should inject within each load cycle; The preset initial frequency; This is the frequency attenuation coefficient; The operator for natural logarithms; For the first Stiffness decay rate over a load cycle.

[0082] Among them, the preset initial frequency It is set based on the natural frequency of the undamaged material and the upper limit of the frequency response bandwidth of the hydraulic servo valve, usually set to 30Hz to 60Hz (e.g., 50Hz), and this frequency should avoid the natural resonant frequency of the pipeline system to prevent the generation of uncontrollable pressure standing waves.

[0083] Frequency attenuation system This adjustment factor, set based on the kinematic viscosity of the chemical solvent being tested and the fiber content of the reinforcing plastic, is used to control the sensitivity of frequency decrease as stiffness decays. Higher solvent viscosity or higher fiber content increases the sensitivity of fluid flow resistance within the micropores to frequency. A larger value indicates that the frequency needs to be reduced more rapidly to accommodate the increased fluid resistance. Based on engineering experience, for a combination of ethylene glycol coolant and glass fiber reinforced nylon, It can be set to 2.0 to 3.0.

[0084] In sub-step S302, based on the calculated optimal perturbation frequency... Generate the initial perturbation command. This initial perturbation command is a standard sine wave signal. To avoid excessive perturbation stress leading to unexpected fatigue fracture of the material, its amplitude is preferably set to 5% to 10% of the basic load amplitude. Subsequently, the inverse transfer function stored in step S1 is called. The original perturbation command is preprocessed and corrected to generate a corrected perturbation command. The specific processing logic is as follows:

[0085] In the frequency domain, the spectrum of the original perturbation command and the inverse transfer function are compared. Perform complex multiplication; or convolve the time series of the original perturbation command with the impulse response of the inverse transfer function in the time domain. This operation mathematically achieves signal predistortion, i.e., for the platform at a specific frequency. The amplitude attenuation is pre-compensated for, and the phase lag is pre-leaded to ensure that the pressure waveform acting on the inner wall of the tubular sample meets the set requirements in both amplitude and phase.

[0086] In sub-step S303, the superimposed transmission operation is performed. This occurs when the internal clock reaches the instruction transmission time calculated in step S2. At that time, the generated correction disturbance command is superimposed on the command sequence of the low-frequency trapezoidal wave base load and sent to the electro-hydraulic servo valve.

[0087] Due to the time of instruction transmission The physical lag is already included in the lead time, and the hydraulic energy driven by the corrected disturbance command is exactly when the microcrack in the tubular sample opens (i.e., when the opening threshold is reached). The solvent instantly reaches the interior of the sample and continues to act throughout the high-tension plateau period. This superposition creates a composite stress field of quasi-static high tension and high-frequency fretting pressure inside the sample. The high tension forces open the gaps between molecular chain segments, while the fluid excitation effect generated by the high-frequency fretting pressure disrupts the boundary layer of the solvent on the micropore walls, thereby accelerating the directional migration of the solvent into the deeper layers of the material.

[0088] In this embodiment, step S4 constructs an energy decoupling analysis mechanism to separate the effective energy dissipation caused purely by chemical penetration and interface damage from the total input energy, so as to eliminate the interference of fluid dynamic factors on the material performance evaluation.

[0089] In a fluid-solid coupled testing environment, the total energy injected by the platform into the tubular sample is actually consumed by three parts: viscous frictional heat generated by the reciprocating flow of the fluid, viscoelastic damping heat of the material itself, and irreversible damage work caused by solvent penetration breaking chemical bonds or interfacial debonding. Since the tubular sample is filled with a high-viscosity chemical solvent (such as an ethylene glycol mixture), under alternating pressure, the viscous damping work generated by the fluid shear flow often dominates. If the evaluation is directly based on the total hysteresis loop area, the fluid viscous dissipation component will mask the fine damage characteristics of the material itself, leading to failure of lifetime prediction.

[0090] Therefore, this step extracts the effective dissipation density index, which is highly sensitive to chemical permeation, by subtracting the fluid viscous dissipation component. The specific implementation logic of this step is as follows:

[0091] In sub-step S401, the measured internal pressure synchronously acquired within a single cycle is extracted. Compared with historical strain data To ensure the closed-loop nature of energy calculations, the extracted data segments must cover a complete load cycle. Furthermore, the data sequence needs to undergo time-domain alignment processing (e.g., determining the time offset by maximizing the cross-correlation function) to eliminate phase misalignment caused by differences in sensor response time. Subsequently, historical strain data... Differential processing is performed to obtain the radial strain rate. .

[0092] Considering that directly performing differential operations on discrete sampled data would amplify high-frequency quantization noise, a five-point cubic smoothing algorithm is preferred for numerical differentiation. To balance denoising effectiveness and waveform fidelity, the filter window width should be set to 1% to 2% of the total number of sampling points per cycle (and be an odd number). For example, for 1000 sampling points per cycle, the window width can be set to 11 or 13. This process suppresses measurement noise while preserving the true characteristics of the strain rate waveform at its peak, ensuring the accuracy of subsequent fluid damping calculations.

[0093] In sub-step S402, based on the principle of energy conservation, the total dissipated energy is calculated by closed-loop integration of the extracted data sequence, and the pre-calibrated fluid viscous damping work is subtracted. The effective dissipation density is calculated using the modified dissipation energy formula. Among them, the integral term Geometrically, this represents the area of ​​the pressure-strain hysteresis loop; physically, it characterizes the total energy density injected into a unit volume of material by the plateau during that period. This is achieved by subtracting the effects of purely fluid factors. The remaining This refers to the hysteresis loss, which characterizes the evolution of the material's microstructure. The modified dissipation energy formula is as follows:

[0094] ;

[0095] In the formula: For the first Effective dissipation density of material per unit volume within one load cycle; To complete a load cycle Closed-loop integral operators within; For a moment The measured internal pressure collected by a pressure sensor; For a moment The radial strain rate, i.e., the derivative of radial strain with respect to time; For a moment Radial strain acquired by a laser scanning micrometer; For time, it is the differential variable; This is the pre-calibrated fluid viscosity damping work.

[0096] For the fluid viscous damping work involved in the formula It must be determined through calibration experiments with controlled variables. The specific implementation method is as follows:

[0097] Before the formal test, a rigid standard tube with the same inner diameter as the tubular specimen (its elastic modulus should be more than 100 times greater than that of the tubular specimen, such as a stainless steel tube) was used to replace the plastic specimen and connected to the platform. Loading was performed while maintaining the fluid medium type, temperature, loading waveform, and frequency exactly the same as in the formal test. At this time, due to the small deformation of the rigid tube, the material damping is negligible, and the measured hysteresis loop area mainly originates from internal fluid shear and wall friction. This stable value was recorded as... .

[0098] In addition, to prevent numerical calculation errors from causing If a non-physical negative value appears, a non-negative constraint needs to be added to the calculation logic: if the calculation result is less than 0, it should be forcibly set to 0.

[0099] The effective dissipation density is calculated using this formula. This directly quantifies the cumulative damage of reinforced plastics under coupling effects. When solvent molecules penetrate the resin matrix, causing molecular chain relaxation (plasticizing effect), or when solvent wedges into the fiber / matrix interface, causing interfacial debonding (friction effect), the inelastic internal friction within the material increases, manifesting as... The value shows a non-linear increasing trend with the number of cycles. Compared with simple stiffness decay, this index is more sensitive to early micro-damage.

[0100] In this embodiment, step S5 establishes a failure identification mechanism based on damage evolution dynamics to accurately determine the critical state of the material transitioning from steady-state penetration to interface failure.

[0101] In the early and middle stages of chemical solvent infiltration, the performance degradation of reinforced plastics is mainly dominated by the physical swelling of the resin matrix. This process exhibits a gradual diffusion behavior, reflected in the effective dissipation density. Typically, the growth rate exhibits linear or low-rate exponential growth characteristics. However, when the solvent front penetrates the resin-rich layer and contacts the reinforcing fibers (such as glass fibers), it disrupts the interfacial coupling agent between the fibers and the matrix, triggering rapid debonding and microcrack interconnection under capillary action. This topological abrupt change in the microstructure manifests as a step change in the dissipation rate in macroscopic energy dissipation. Therefore, this step identifies this physical inflection point by monitoring the abrupt change in the acceleration of energy dissipation. The specific implementation logic of this step is as follows:

[0102] In sub-step S501, the effective dissipation density obtained from step S4 is continuously calculated. Based on load cycle count The order of these elements forms a time series.

[0103] Considering the presence of high-frequency random noise in the experimental data, direct difference operations could amplify the noise and lead to misjudgments. Therefore, as a preferred data preprocessing method, a moving average filter is used to smooth the time series, with a recommended sliding window length of 5 to 10 periods. Subsequently, the second derivative is calculated on the smoothed series to obtain the second derivative value. The specific numerical calculations can be achieved using the three-point central difference formula: .

[0104] The second derivative value Physically, it characterizes the rate of change of energy dissipation in a material. During the matrix swelling stage, the damage rate is relatively constant, and the second derivative value approaches zero or remains at a low baseline level; however, at the instant of interface failure, the dissipation rate changes abruptly, causing the second derivative value to jump.

[0105] In sub-step S502, the output second derivative value is monitored in real time. And calculate the absolute difference between the second derivative value of the current sampling period and the second derivative value of the previous sampling period. This indicator quantifies the degree of fluctuation in damage acceleration, and its calculation formula is as follows: Then, the absolute difference was... Compared with the preset failure threshold Compare them.

[0106] To ensure that the decision-making logic adapts to different material systems and testing environments, the failure threshold is... A statistically based, dynamic setting method is used, rather than fixed empirical values.

[0107] In practice, the controller automatically extracts a stable interval (e.g., cycles 100 to 500) of second-order derivative data from the initial stage of the test as background noise samples, and calculates the standard deviation of the data within this interval. Failure threshold Set as a multiple of that standard deviation, i.e. .coefficient According to statistics 3 The criteria are typically set to 3 to 5 (e.g., 4), meaning that when the current change exceeds 4 times the background noise level, a non-random structural mutation is considered to have occurred.

[0108] When the absolute difference is detected Greater than the preset failure threshold At this point, it is determined that the solvent penetration front has broken through the surface resin-rich zone and deeply penetrated the fiber interface layer, indicating that the material has entered the irreversible interface failure stage. Immediately, a stop command is sent to the electro-hydraulic servo fatigue testing machine to interrupt the hydraulic loading, and the current load cycle count is recorded. Number of loops. This is defined as the media resistance life of the reinforced plastic under the current stress level and solvent environment. Compared with the traditional weight change rate method (which requires interruption of the test for weighing) or residual strength method (which is a destructive test), this determination method achieves in-situ, continuous identification of the life end, improving test efficiency and data consistency.

[0109] See appendix Figure 3 To further illustrate the technical principles and implementation process of the present invention, the following application examples are provided:

[0110] In this application embodiment, the object under test is a core component of the hydraulic power transmission assembly of a certain type of heavy-duty bomb disposal robot, namely a glass fiber reinforced nylon high-pressure pipeline (diameter: (12mm × 2mm). The robot is designed to operate in an environment involving high-concentration ethylene glycol-based hydraulic oil, and the pipeline must withstand high-frequency pulse pressure during bomb disposal operations. To accurately assess the actual service life of this pipeline under the coupled effects of ethylene glycol solvent penetration and pulsed mechanical stress, the test method of this invention was employed.

[0111] Before the test began, a standard stainless steel pipe of the same specification was first connected to the fluid-mechanical coupling test platform. Under the conditions of a test temperature of 80℃ and a base load frequency of 1Hz, the fluid viscous damping work of the platform was calibrated and obtained. 3.5×10 4 J / m 3 At the same time, the inherent physical response delay time of the platform was calibrated. It takes 0.03 seconds.

[0112] Entering the testing phase, in the... During the next load cycle, the controller executes the strain gating strategy in step S2. Based on the historical strain data from the previous five cycles, the time series prediction model calculates the opening of the pipe wall microcracks in the next cycle (i.e., the strain exceeds the opening threshold). The start time of ) This is the 1.250th second of the cycle. According to the formula... The controller calculates the command transmission time that can compensate for physical lag. The controller issues a disturbance command precisely at 1.220 seconds to ensure that the pressure wave acts accurately on the moment the microcrack opens at 1.250 seconds.

[0113] Next, the adaptive micro-perturbation calculation in step S3 is performed. At this time, the dynamic stiffness modulus of the material is monitored by a laser scanning micrometer. From the initial The stiffness decreased to 2125 MPa. First, the stiffness attenuation rate was calculated. :

[0114] ;

[0115] Then, this attenuation rate is substituted into the adaptive formula for the disturbance frequency. A preset initial frequency is set. Frequency attenuation coefficient (Based on ethylene glycol viscosity settings). Calculate the optimal perturbation frequency. :

[0116] ;

[0117] Based on the calculation results, the controller injects a correction disturbance command with a frequency of 37.05Hz during this cycle to match the fluid impedance characteristics under the current 15% stiffness decay.

[0118] During the test, at the [number]th Near the next cycle, the energy decoupling analysis in step S4 is performed. The controller performs closed-loop integration on the single-cycle data and calculates the total dissipated energy (i.e., the total hysteresis loop area) as 5.2 × 10⁻⁶. 4 J / m 3 By using the modified dissipated energy formula, the pre-calibrated fluid viscous damping work is subtracted. (Set to 3.5×10) 4 J / m 3 ), calculate effective dissipation density :

[0119] ;

[0120] This value characterizes the irreversible dissipation within the current cycle caused purely by ethylene glycol permeation and interfacial damage. For example... Figure 3 As shown by the solid line, during the stable phase before failure (approximately 84,000 to 85,000 cycles), the effective dissipation density remains at 1.7 × 10⁻⁶. 4 J / m 3 The benchmark level.

[0121] Finally, the failure determination in step S5 is performed. The controller monitors in real time. The value of the second derivative. For example... Figure 3 As shown by the dashed line, before the 85,000th cycle, the absolute difference of the second derivative remained below the failure threshold (background noise level). However, at the 85,020th cycle, rapid debonding occurred due to the solvent front contacting the glass fiber interface, causing an inflection point in the effective dissipation density curve and a sharp increase. This resulted in a sudden surge in the calculated absolute difference of the second derivative to 800 J / m². 3 This value clearly exceeded... Figure 3 The preset failure threshold shown is 200 J / m 3 ,based on (Criterion setting). Based on this, it is determined that the material has experienced interface failure, the test is immediately stopped, and the media resistance life is recorded as 85020 cycles.

[0122] To verify the effectiveness of the present invention, it was compared with two existing general testing methods. The experimental subjects were all hydraulic pipelines from the same batch, and the test medium was a mixture of ethylene glycol and water at 80°C.

[0123] Control group A (static immersion method): The sample is immersed in the medium and taken out at regular intervals for tensile testing. A 50% decrease in strength is used as the failure criterion.

[0124] Control group B (fixed frequency fatigue method): A traditional pulse testing machine is used to apply pulse pressure at a fixed frequency (50Hz), without considering physical hysteresis compensation, and pipeline rupture and leakage are used as the failure criterion.

[0125] Experimental group (method of the present invention): The adaptive micro-perturbation and energy decoupling determination method in the above application examples is adopted.

[0126] Table 1: Comparison of experimental results for different testing methods

[0127] Test group Failure determination time / life Microscopic characterization of damage patterns (SEM analysis) Evaluation of the accuracy of results Control group A 1200 hours (equivalent to an extremely long period) Only the surface resin swelled, while the fiber interface remained intact. Failure hysteresis: Static penetration is slow, making it impossible to reproduce the coupling damage in actual robot operation, and overestimating the material life. Control group B 150,000 cycles The sample underwent macroscopic brittle fracture with a smooth fracture surface. False failure: Due to excessively high frequency and lack of hysteresis compensation, the fluid mainly oscillates in the center of the pipeline and fails to effectively penetrate the microcracks. The material fractures primarily due to mechanical fatigue, rather than failure caused by media penetration. experimental group 85020 cycles Significant fiber-matrix interface debonding was observed, and solvent traces penetrated to 80% of the wall thickness. Precise failure analysis: It accurately reproduces the interface failure mode caused by stress-assisted penetration. The lifespan data is in high agreement with the failure statistics of the robot in the field (approximately 80,000-90,000 operating conditions).

[0128] Experimental conclusion:

[0129] Table 1 shows that the present invention successfully induced real medium penetration damage in a short time through strain gating strategy (eliminating hysteresis) and adaptive micro-perturbation (impedance matching). Meanwhile, the determination method based on effective dissipation density can identify interface failure risk approximately 40% earlier than macroscopic rupture (control group B), providing an early warning indicator for preventative maintenance of the robot.

Claims

1. A test method for evaluating the chemical solvent penetration resistance of reinforced plastics, characterized in that, Includes the following steps: While maintaining the basic static pressure state, input a random pressure command and collect the measured internal pressure. Construct an inverse transfer function based on the relationship between the random pressure command and the measured internal pressure. Historical strain data of tubular specimens are collected using a laser scanning micrometer to estimate the start time of microcrack opening in the next cycle, and the command transmission time is calculated in combination with the inherent delay of the platform. The stiffness attenuation rate of the current cycle is calculated based on real-time data collected by the laser scanning micrometer and pressure sensor. The optimal perturbation frequency matching the current damage state is calculated according to the perturbation frequency adaptive formula. The inverse transfer function is used to generate a correction perturbation command, which is superimposed and sent to the electro-hydraulic servo valve at the time of command transmission. The measured internal pressure and radial strain data are extracted from the collected single-cycle data. The pre-calibrated fluid viscosity damping work is deducted, and the effective dissipation density is calculated using the modified dissipation energy formula. The second derivative of the time series of the effective dissipation density is calculated, the failure endpoint is determined based on the change of the second derivative, and the number of cycles at the failure endpoint is taken as the media resistance life of the reinforced plastic.

2. The test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 1, characterized in that, The steps for constructing the inverse transfer function based on the relationship between the random pressure command and the measured internal pressure specifically include: While maintaining the basic static pressure, the random pressure command is input to the electro-hydraulic servo valve, and the command voltage generated by the random pressure command and the measured internal pressure fed back by the pressure sensor are collected simultaneously. The command voltage and the measured internal pressure are processed using the Fourier transform algorithm to calculate the positive transfer function that reflects the frequency response characteristics of the platform. The forward transfer function is inverted to generate the inverse transfer function, which is then stored. The inverse transfer function serves as a pre-filter for subsequent high-frequency disturbance commands.

3. The test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 2, characterized in that, The step of performing the inverse operation on the forward transfer function also includes a step of introducing a lower limit threshold for judgment: When the magnitude of the forward transfer function is greater than the lower limit threshold of the amplitude, the inverse transfer function is obtained by directly calculating the reciprocal. When the magnitude of the forward transfer function is less than or equal to the lower limit threshold of the amplitude, the gain of the inverse function at the corresponding frequency point is limited to a preset saturation value to obtain the inverse transfer function; The lower limit of amplitude is preset based on the system background noise level, and the preset saturation value is preset based on the numerical stability requirements of the test platform.

4. The test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 3, characterized in that, The specific steps for calculating the start time of microcrack opening in the next cycle include: The electro-hydraulic servo valve is driven to apply a basic load to the tubular specimen, while a laser scanning micrometer is used to collect the historical strain data of the tubular specimen for a set number of cycles. A time series prediction model is constructed using the historical strain data. The waveform characteristics of the complete cycle in the historical strain data are projected onto the future on the time axis to reconstruct the predicted strain curve for the next cycle. The search is performed on the predicted strain curve to find the time point at which the value first crosses the opening threshold, and this time point is marked as the starting time. The opening threshold is a critical strain value determined based on the stress-strain constitutive relationship of the reinforced plastic material.

5. The test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 4, characterized in that, The calculation logic for calculating the command transmission time based on the platform's inherent delay includes: The command sending time is equal to the predicted start time of the next cycle microcrack opening minus the inherent platform delay; The inherent delay of the platform is calculated by converting the phase lag of the forward transfer function at the base load frequency.

6. The test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 4, characterized in that, The specific steps for calculating the optimal perturbation frequency to match the current damage state based on the adaptive perturbation frequency formula include: Based on the real-time data collected by the laser scanning micrometer and pressure sensor, the stress and strain data of the linear segment of the rising edge of the base load in the current period are extracted, the current dynamic stiffness modulus is calculated, and the stiffness attenuation rate is calculated in combination with the initial dynamic stiffness modulus. Substituting the stiffness attenuation rate into the perturbation frequency adaptive formula, the perturbation frequency adaptive formula defines that the optimal perturbation frequency is proportional to the preset initial frequency and inversely proportional to the logarithmic function term containing the stiffness attenuation rate. The perturbation frequency adaptive formula includes a frequency attenuation coefficient, which is preset based on the kinematic viscosity of the chemical solvent to be tested and the fiber content of the reinforcing plastic. The initial dynamic stiffness modulus is calculated based on the average data from the initial stage of the test.

7. The test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 1, characterized in that, The specific steps of generating a correction disturbance command using the inverse transfer function and superimposing it onto the electro-hydraulic servo valve at the command transmission time include: Generate a sinusoidal original disturbance command based on the optimal disturbance frequency; The stored inverse transfer function is invoked to preprocess and correct the amplitude and phase of the original disturbance command, thereby generating the corrected disturbance command; When the clock of the test platform controller reaches the instruction transmission time, the correction disturbance instruction is superimposed on the basic load instruction sequence of the low-frequency trapezoidal wave and sent to the electro-hydraulic servo valve.

8. The test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 1, characterized in that, The specific steps for calculating the effective dissipation density using the modified dissipation energy formula include: Extract the measured internal pressure and radial strain data synchronously collected by the pressure sensor and laser scanning micrometer within a single cycle, and perform differential processing on the radial strain data to obtain the radial strain rate; The total dissipated energy is calculated by performing closed-loop integration on the data sequence extracted from the collected single-cycle data, and the pre-calibrated fluid viscous damping work is subtracted to obtain the effective dissipation density. The pre-calibrated fluid viscosity damping work is determined based on the hysteresis loop area measured under the same test conditions of a rigid standard tube.

9. The test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 1, characterized in that, The steps of determining the failure endpoint based on the change in the second derivative value, and using the number of cycles at the failure endpoint as the media resistance life of the reinforced plastic, specifically include: The second derivative value is monitored in real time, and the absolute difference between the second derivative value of the current sampling period and the second derivative value of the previous sampling period is calculated. When the absolute difference is greater than the preset failure threshold, it is determined that the solvent front has broken through the surface resin-rich area and entered the fiber interface layer. A stop command is sent to the electro-hydraulic servo fatigue testing machine, and the current number of cycles is recorded as the medium resistance life of the reinforced plastic. The failure threshold is preset based on a statistical dynamic setting method. The dynamic setting method includes extracting the second derivative data of a stable interval at the beginning of the test as a background noise sample, calculating the standard deviation of the data within the background noise sample, and setting the failure threshold as a multiple of the standard deviation.

10. A test method for evaluating the chemical solvent penetration resistance of reinforced plastics according to claim 9, characterized in that, The step of calculating the second derivative of the time series of the effective dissipation density further includes: Before performing the second-order derivative operation, the time series is smoothed using the moving average filtering method; The second-order derivative operation is specifically performed using the three-point central difference formula for numerical calculation.

Images