A test method for sealing performance of check valve under high pressure difference working condition
By constructing a high-pressure differential impact flow field in the check valve, collecting dynamic process parameters, and establishing a sealing correlation model, the problem of difficulty in evaluating the dynamic sealing performance of the check valve under high-pressure differential conditions is solved, and the accurate quantification of dynamic sealing performance and prediction of potential leakage rate are realized.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHUANGHENG VALVE
- Filing Date
- 2026-05-11
- Publication Date
- 2026-08-04
AI Technical Summary
Existing technologies struggle to simulate and quantify the dynamic sealing performance of check valves under high pressure differential conditions, especially the impact force that causes damage and leakage to the sealing surface when the medium flows back at high speed. Furthermore, traditional testing methods cannot accurately assess the dynamic leakage rate.
By constructing a high-pressure differential impact flow field, simulating the dynamic impact closing process, collecting the valve disc motion state, the valve disc and valve seat contact state, and transient acoustic signals, establishing a sealing correlation model, and combining machine learning algorithms to evaluate dynamic sealing performance and quantify leakage rate.
It enables accurate evaluation and quantification of the dynamic sealing performance of check valves under high pressure differential conditions, predicts potential leakage rates, improves the accuracy and reliability of testing, and reduces maintenance risks.
Smart Images
Figure CN122505484A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of sealing check valve technology, and specifically to a method for testing the sealing performance of a check valve under high pressure differential conditions. Background Technology
[0002] Most current standards for check valves specify relatively static sealing tests. During the test, the valve is usually pre-positioned in the closed position and then pressurized. However, in real-world operating conditions, high-pressure differential check valve failures often occur at the moment of dynamic closure. As the medium flows back at high speed, the valve disc violently impacts the valve seat under the enormous impact force and pressure differential. The instantaneous local pressure and vibration generated by this effect are the main causes of plastic deformation, fine cracks, and eventual leakage on the sealing surface. Furthermore, this extreme condition cannot be reproduced by general static or relatively static tests.
[0003] Furthermore, during the movement of the valve disc in some check valves, it is crucial to determine whether the closing trajectory of the valve disc aligns perfectly with the center of the valve seat, and whether any deviation from this alignment affects the sealing test results. The instantaneous change in the closing state is also an issue not systematically addressed in existing sealing tests. Moreover, under high pressure differentials, the leakage rate of minor leaks in some check valves is typically assumed to be within a relatively standard allowable range. However, the impact on the entire valve control system cannot be ignored, especially in cases of steam leakage. Existing methods for detecting such leaks, such as bubble methods, pressure drop methods, or volumetric measurement methods, have limited accuracy under ultra-high pressures, making it difficult to capture and quantify these "micro-leaks." Furthermore, simulating actual media conditions in valve testing is challenging, creating potential maintenance hazards for long-term operation. Summary of the Invention
[0004] The purpose of this invention is to provide a method for testing the sealing performance of a check valve under high pressure differential conditions, thereby solving the following technical problems: How to construct a valve sealing performance test process under dynamic impact conditions to accurately predict the valve sealing performance under high pressure differential conditions.
[0005] The objective of this invention can be achieved through the following technical solutions: A method for testing the sealing performance of a check valve under high pressure differential conditions. The check valve includes a valve body, valve disc, valve seat, valve cavity, guide mechanism, operating mechanism, and rocker arm. The test method includes: S1. Construct a high pressure differential impact flow field in the valve cavity between the inlet and outlet of the check valve body, and simulate the dynamic impact closing process; S2. Synchronously collect process parameters within the target time period before and after the dynamic impact closing process, and collect the leakage rate of the check valve body in the closed state; process parameters include valve disc motion state parameters, valve disc and valve seat contact state parameters, and transient acoustic signals; S3. Construct a check valve sealing correlation model by performing correlation analysis on process parameters and leakage rate, and determine whether the correlation results are qualified based on the check valve sealing correlation model; S4. Based on the correlation analysis results, call the check valve sealing performance evaluation model to evaluate and quantify the dynamic sealing performance of the check valve under high pressure differential impact.
[0006] 2. The method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 1, characterized in that step S1, simulating the dynamic impact closing process, includes: S11. Install the check valve into the test pipeline and position its valve disc at the specified opening degree. S12. A stable flow field is provided to the valve body inlet through a positive impact flow path to simulate the valve's working state; S13. Through the reverse impact flow path containing a high-pressure accumulator and an ultra-high-speed reversing valve, after receiving the control command, the high-pressure medium is instantaneously injected into the valve body outlet, forming a rapidly rising high pressure difference in the valve cavity, so that the valve disc completes dynamic impact closure.
[0007] Preferably, the leakage rate value measured in step S2 according to the standard pressure holding test procedure is obtained by means of: After the dynamic impact shutdown process ends and the pressure stabilizes, apply the specified pressure to the valve body inlet or outlet and maintain the pressure, then measure the leakage rate.
[0008] Preferably, the correlation analysis in step S3 specifically includes: From the valve disc motion state parameters and the valve disc contact state parameters, calculate the normal impact velocity and impact kinetic energy of the valve disc hitting the valve seat at the moment of valve closing, and use them as the first dynamic characteristic parameter. Time-frequency domain analysis is performed on the transient acoustic signal to extract the main frequency energy of the acoustic emission signal, which characterizes the collision energy and stability of the sealing surface, as the second dynamic characteristic parameter; The extracted first and second dynamic feature parameters are fitted with the leakage rate using machine learning to establish a check valve sealing correlation model that characterizes the quantitative relationship between the dynamic process and the static result.
[0009] Preferably, the check valve sealing correlation model established in step S3 is used for: To determine whether the dynamic sealing performance of the check valve is qualified under a single impact, the judgment method is: whether the characteristic parameters are within the preset empirical safety threshold range, and whether the static sealing result meets the standard requirements; Predicting the sealing life of a check valve: Perform multiple cyclic tests of steps S1-S3 on the same valve body, track the evolution trend of characteristic parameters and static sealing results with the number of impacts, and predict the number of failure cycles based on the performance degradation law.
[0010] Preferably, the calculation method for the check valve sealing correlation model is as follows:
[0011] in, This is the predicted leakage rate. Normal impact velocity, For impact kinetic energy, For the peak impact force, The characteristic value of acoustic emission signal energy. For error terms; function Obtained by fitting the training dataset.
[0012] Preferably, the method for calling the check valve sealing performance evaluation model is as follows: The feature parameters extracted in this test include the normal impact velocity. Impact kinetic energy Peak impact force Harmony emission signal energy eigenvalues The input is fed into a pre-trained check valve sealing performance evaluation model for calculation, and the dynamic sealing quality index is output: Through formula Calculate the dynamic sealing quality index ; in, Tangential velocity, referring to the velocity perpendicular to the vector. The impact velocity component on the plane, This is a weighting coefficient for the influence of tangential velocity relative to normal impact velocity; The duration of the impact force; For the valve disc equivalent mass Eigenvalues of acoustic emission signal energy The normalization function will convert the acoustic emission signal energy eigenvalues Mapped to Within the interval, , , These are the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient, respectively.
[0013] Preferably, the methods for constructing a check valve sealing performance evaluation model include: Collect the characteristic parameters and corresponding leakage values of check valves of different specifications and operating conditions obtained from historical tests to form a sample dataset; Using the sample dataset, a neural network is trained to obtain the mapping relationship from feature parameters to predicted leakage rate and performance score, thereby generating a check valve sealing performance evaluation model.
[0014] Preferably, step S4 includes: Multiple consecutive dynamic sealing quality indices are preset. The threshold range, each threshold range corresponds to a performance level; The calculated The value is compared with the threshold range to determine the performance level of the check valve; the performance level includes at least: excellent, good, qualified, and unqualified. Among them, the performance level corresponding to "excellent" is The highest value indicates a smooth dynamic impact process, good alignment, moderate energy, and optimal predicted sealing performance; the performance level corresponds to the unqualified level. The lowest value indicates a serious problem in the dynamic impact process, such as excessive impact energy, poor centering, or abnormal collision. Output quantized performance score , ,in For the first One characteristic parameter, For normalization function, For the corresponding weights.
[0015] The beneficial effects of this invention are: (1) The test method of the present invention breaks through the limitations of traditional static test. By simulating the water hammer impact phenomenon caused by the high-speed backflow of the medium in actual operation, the early failure of the valve sealing surface is caused by the test pipeline. By precisely controlling the accumulator pressure, the opening time of the reversing valve, and the initial opening degree of the valve disc, highly consistent and repeatable impact conditions can be generated. By issuing instructions through the control system, the simulated dynamic impact closing process combining dynamic test and static analysis is completed.
[0016] (2) By performing correlation analysis on the static results and dynamic process parameters in the test process, the complex test parameters such as normal impact velocity, impact kinetic energy and acoustic emission main frequency energy extracted from multiple tests are integrated. By constructing a correlation model, the correlation calculation of dynamic characteristic parameters and static sealing results is completed. In a single test, the first dynamic characteristic parameter and the second dynamic characteristic parameter are input into the evaluation model pre-trained with historical data to output the dynamic sealing quality index, thus completing the evaluation and quantification of the dynamic sealing performance results. Based on the correlation analysis results, the check valve sealing performance evaluation model is called to evaluate and quantify the dynamic sealing performance of the check valve under high pressure differential impact. The measurement and calculation process of the new valve closing process is realized. Only the dynamic impact process parameters need to be measured to realize the model to accurately predict the potential leakage rate level and complete the optimization of the new valve impact process prediction.
[0017] Of course, any product implementing this invention does not necessarily need to achieve all the advantages described above at the same time. Attached Figure Description
[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 This is a step diagram of a method for testing the sealing performance of a check valve under high pressure differential conditions according to the present invention. Figure 2 This is a step diagram of the method for simulating the dynamic impact shutdown process in step S1 of the present invention; Figure 3 This is a schematic diagram of a high-pressure differential check valve according to the present invention.
[0020] Reference numerals in the attached drawings: 1. Valve body; 2. Valve disc; 3. Valve seat; 4. Valve cavity; 5. Guide mechanism; 6. Operating mechanism; 7. Rocker arm. Detailed Implementation
[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0022] Please see Figure 1 , 3As shown, this invention provides a method for testing the sealing performance of a check valve under high pressure differential conditions. The check valve includes a valve body 1, a valve disc 2, a valve seat 3, a valve cavity 4, a guide mechanism 5, an operating mechanism 6, and a rocker arm 7. The testing method includes: S1. Construct a high pressure differential impact flow field in the valve cavity 4 between the inlet and outlet of the check valve body 1, and simulate the dynamic impact closing process; S2. Synchronously collect process parameters within the target time period before and after the dynamic impact closing process, and collect the leakage rate of the check valve body 1 in the closed state; the process parameters include the movement state parameters of the valve disc 2, the contact state parameters between the valve disc and the valve seat, and the transient acoustic signal. S3. Construct a check valve sealing correlation model by performing correlation analysis on process parameters and leakage rate, and determine whether the correlation results are qualified based on the check valve sealing correlation model; S4. Based on the correlation analysis results, call the check valve sealing performance evaluation model to evaluate and quantify the dynamic sealing performance of the check valve under high pressure differential impact.
[0023] In the above technical solution, the main structure of the check valve includes a valve body 1, a valve disc 2, a valve seat 3, a valve cavity 4, a guide mechanism 5, an operating mechanism 6, and a rocker arm 7. The guide mechanism 5 ensures that the valve disc 2 is accurately aligned when closed, and the seal formed by the valve disc 2 and the valve seat 3 is the key part to prevent backflow and leakage of the medium. The operating mechanism 6 (such as a counterweight) is used to balance or assist the closing of the valve disc 2, so that it can close quickly and forcefully under extremely high system pressure differentials to prevent backflow of the medium. However, this also makes the valve disc 2 have huge kinetic energy at the moment of closing, and its impact process with the valve seat 3 becomes the most critical link affecting the life and reliability of the sealing pair. Therefore, in the high pressure differential working condition simulation test, by instantaneously injecting high pressure medium at the outlet end and forcing valve disc 2 to close rapidly under huge pressure difference, this dynamic impact process will generate strong impact kinetic energy and transient acoustic signals. The system analyzes the dynamic characteristics such as impact velocity, kinetic energy, peak impact force and acoustic emission energy when valve disc 2 closes by collecting these process parameters and subsequent static leakage rate, and then establishes a correlation model to quantitatively evaluate and classify the dynamic sealing performance of the valve. The specific testing method is as follows: First, in step S1, a high-pressure impact flow field is constructed within the valve cavity 4 formed between the inlet and outlet of the check valve body 1. The dynamic impact closing process is simulated based on the high-pressure impact flow field. Then, in step S2, during the entire brief process of impact closing via valve disc 2, i.e., the target time period before, during, and after the impact, high-frequency sensors deployed at key parts of the valve work synchronously to collect the leakage rate of the check valve body 1 in the valve closed state. The specific leakage rate includes valve disc motion state parameters, valve disc and valve seat contact state parameters, and transient acoustic signals. The valve disc motion state parameters are mainly obtained by installing a miniature inertial measurement unit on the valve disc to collect angular velocity and acceleration, which can be used to calculate impact velocity, kinetic energy, and trajectory. The valve disc contact state parameters are obtained by installing a dynamic force sensor on the support structure of the valve seat 3 to directly measure the peak force at the moment of impact. The transient acoustic signal is obtained by installing an ultrasonic sensor on the outer wall of the valve near the valve seat 3 to collect the acoustic emission signal generated when the valve disc 2 impacts the valve seat 3. Its spectrum and energy reflect the microscopic characteristics of the collision. Next, in step S3, a correlation analysis is performed on the process parameters including valve disc motion state parameters, valve disc-seat contact state parameters, transient acoustic signal parameters, and leakage rate. This correlation analysis is performed by extracting specific characteristic parameters that can represent the valve disc state from the valve disc-seat contact state parameters obtained from the sensor. The normal impact velocity and impact kinetic energy are extracted as the first dynamic characteristic parameters. The acoustic domain signal is extracted from the transient acoustic signal, and the acoustic emission signal main frequency energy that characterizes the collision energy and stability of the sealing surface is extracted as the second dynamic characteristic parameter. The first dynamic characteristic parameter, the second dynamic characteristic parameter, and the leakage rate are correlated and analyzed. A mathematical model is constructed using a machine learning algorithm. This model can describe the quantitative relationship between the severity of the impact process and the leakage amount. This model is defined as the check valve sealing correlation model. The dynamic sealing performance and static sealing results are judged to be qualified by the check valve sealing correlation model. If they are qualified, steps S1-S3 are continued; otherwise, step S4 is performed for further analysis. Finally, in step S4, a sealing performance evaluation model is constructed by training based on a large amount of historical data. In a single test, the first dynamic feature parameter and the second dynamic feature parameter are input into the evaluation model pre-trained with historical data to output the dynamic sealing quality index, thereby realizing the construction of the check valve sealing performance evaluation model and completing the process of evaluating and quantifying the dynamic sealing performance results.
[0024] Please see Figure 2 , 3 As shown, in one embodiment of the present invention, step S1, simulating the dynamic impact shutdown process, includes: S11. Install the check valve into the test pipeline and position its valve disc 2 at the specified opening. S12. A stable flow field is provided to the inlet of valve body 1 through a positive impact flow path to simulate the working state of the valve. S13. After receiving the control command, the high-pressure medium is instantaneously injected into the outlet of valve body 1 through the reverse impact flow path containing the high-pressure accumulator and the ultra-high-speed reversing valve, forming a rapidly rising high pressure difference in the valve cavity 4, so that the valve disc 2 completes dynamic impact closure.
[0025] In the above technical solution, a high-pressure impact flow field is constructed within the valve cavity 4 formed between the inlet and outlet of the check valve body 1. The dynamic impact closing process is simulated based on this high-pressure impact flow field: the check valve to be tested is installed on a dedicated test bench; through external drive, its valve disc 2 is stabilized at a preset opening degree by the action of the rocker arm 7 and the operating mechanism 6, simulating the working state of the valve in the pipeline; and at the valve inlet end ( Figure 3 A stable low-pressure and medium-pressure flow field is applied to the right side, and this pressure can maintain the open state of valve disc 2, simulating the upstream pressure during normal valve operation, and at the valve outlet end ( Figure 3 (Left side) An accumulator pre-connected with ultra-high pressure medium is released instantaneously through an ultra-high speed reversing valve. The high pressure medium is injected into valve chamber 4 from the valve outlet in milliseconds. In extreme events, valve chamber 4 forms a huge pressure difference where the outlet pressure is much higher than the inlet pressure. The force generated by this pressure difference overcomes the holding force of operating mechanism 6, the positive flow force of the medium, and friction, and drives valve disc 2 to rotate rapidly around the pointer of the remote rod to close. Under the constraint of guide mechanism 5, valve disc 2 impacts the sealing surface of valve seat 3 at a very high speed and energy, completing a simulated real accident shutdown, causing operating mechanism 6 (commonly equipped with water hammer) to close by impact.
[0026] In one embodiment of the present invention, the leakage rate value measured according to the standard pressure holding test procedure in step S2 is obtained in the following manner: After the dynamic impact shutdown process ends and the pressure stabilizes, apply the specified pressure to the inlet or outlet of valve body 1 and maintain the pressure, then measure the leakage rate.
[0027] In the above technical solution, after the dynamic impact process ends and the valve is fully closed and the internal pressure of valve cavity 4 stabilizes, a false test pressure is applied to the valve inlet or outlet, and this pressure is maintained for a period of time. The leakage rate during this period is measured using a high-precision instrument. Various methods can be used to calculate the leakage rate, such as directly measuring the volume or mass flow rate of the leaking medium using a flow meter; indirectly calculating the leakage rate by monitoring the rate of pressure decrease over time in a closed system using the pressure drop method; or evaluating the leakage by applying soap solution or immersing the valve in water at observable leak points using the bubble detection method, and observing and counting the generated bubbles. In this embodiment, the pressure drop method is used to measure the leakage rate. The leakage rate is used as the Jintai result and is correlated with subsequent dynamic process parameters to verify and calibrate the effectiveness of the dynamic process.
[0028] As one embodiment of the present invention, please refer to Figure 3 As shown, the correlation analysis in step S3 specifically includes: The normal impact velocity and impact kinetic energy of valve disc 2 impacting valve seat 3 at the moment of closing are calculated from the valve disc motion state parameters and valve disc contact state parameters, and used as the first dynamic characteristic parameter. Time-frequency domain analysis of transient acoustic signals is performed to extract the main frequency energy of acoustic emission signals, which characterizes the collision energy and stability of the sealing surface, as the second dynamic characteristic parameter; The extracted first and second dynamic feature parameters are fitted with the leakage rate using machine learning to establish a check valve sealing correlation model that characterizes the quantitative relationship between the dynamic process and the static result.
[0029] In the above technical solution, physical quantities representing the severity of the impact are calculated from the valve disc motion state parameters, which are usually extracted from IMU data and high-speed image data: normal impact velocity and impact kinetic energy, and used as the first dynamic characteristic parameters; then, the signal energy at the main frequency is extracted from the transient acoustic signal through time domain analysis. The frequency of the extracted main frequency energy is mainly related to the collision and friction of the sealing surface, and is used as the second characteristic parameter reflecting the contact state of the sealing surface.
[0030] Next, the normal impact velocity, impact kinetic energy, and acoustic emission frequency energy extracted from multiple tests are used as input features, and the leakage rate obtained from the corresponding tests is used as the target value. A machine learning algorithm is used, which employs multivariate regression calculation to train the data and fit a function. This fitted function is then used as the check valve sealing correlation model.
[0031] As one embodiment of the present invention, the check valve sealing correlation model established in step S3 is used for: To determine whether the dynamic sealing performance of the check valve is qualified under a single impact, the judgment method is: whether the characteristic parameters are within the preset empirical safety threshold range, and whether the static sealing result meets the standard requirements; Predicting the sealing life of a check valve: Perform multiple cyclic tests of steps S1-S3 on the same valve body 1, track the evolution trend of characteristic parameters and static sealing results with the number of impacts, and predict the number of failure cycles based on the performance degradation law.
[0032] In the above technical solution, step S3 uses a check valve sealing correlation model to judge the dynamic sealing performance under a single impact. The judgment method involves checking whether the characteristic parameters—the first and second dynamic characteristic parameters and the static leakage rate—meet their respective preset safety threshold ranges. Then, it simultaneously checks whether the dynamic characteristic parameters exceed the limit or whether the static results are qualified. Only when both are met is it considered qualified. Furthermore, steps S1-S3 are repeated until the cyclic impact test is completed. The dynamic characteristic parameters and static results are recorded for each test, and corresponding change curves are plotted. Through curve fitting, the number of impacts the valve can withstand when the leakage rate exceeds a reasonable threshold is predicted, thus predicting its impact cycle life. This life prediction function provides data for determining the preventative replacement cycle of valves in critical pipelines, provides an accelerated testing method for evaluating the long-term durability of valves made of different materials and with different structures, significantly shortens the R&D cycle, and reduces testing costs.
[0033] As one embodiment of the present invention, the calculation method of the check valve sealing correlation model is as follows:
[0034] in, This is the predicted leakage rate. Normal impact velocity, For impact kinetic energy, For the peak impact force, The characteristic value of acoustic emission signal energy. For error terms; function Obtained by fitting the training dataset.
[0035] In the above technical solution, for each sample valve, a high-pressure differential dynamic impact test is strictly performed according to the aforementioned steps S1-S2; the complete process parameters of each test are recorded synchronously and with high precision. (Original and derived data) and the final measured static leakage rate. By extracting the normal impact velocity from multiple tests Impact kinetic energy Peak impact force and acoustic emission main frequency energy These input features form the training dataset, which consists of the input features for each test. and the corresponding output labels Pair them up to form a data sample. The set of all samples constitutes the original training dataset. The preprocessed dataset The dataset is randomly divided into training and test sets in a ratio (e.g., 8:2); the selected algorithm is trained using the training set data; and the model's internal parameters are adjusted by optimizing the algorithm (e.g., gradient descent) to improve the model's predictive output. Compared with the actual leakage rate The system minimizes the error (commonly mean square error, MSE) between different values; and during online prediction, for a new valve under test, the system automatically extracts the feature values from the current test after performing the dynamic impact test and inputs them into the deployed model. The leakage rate prediction value is calculated; a quantitative mathematical relationship is established between quantities related to dynamic impact behavior and sealing performance, and the leakage rate prediction value is also used. This model enables accurate prediction of potential leakage rates; it should be noted that... The error term is set based on historical experience; the function The data was obtained by fitting the training dataset, and the fitting function was a multiple linear regression function.
[0036] As one embodiment of the present invention, the method of calling the check valve sealing performance evaluation model is as follows: The feature parameters extracted in this test include the normal impact velocity. Impact kinetic energy Peak impact force Harmony emission signal energy eigenvalues The input is fed into a pre-trained evaluation model for calculation, which outputs the dynamic sealing quality index. Through formula Calculate the dynamic sealing quality index ; in, Tangential velocity, referring to the velocity perpendicular to the vector. The impact velocity component on the plane, This is a weighting coefficient for the influence of tangential velocity relative to normal impact velocity; The duration of the impact force; For the valve disc equivalent mass Eigenvalues of acoustic emission signal energy The normalization function will convert the acoustic emission signal energy eigenvalues Mapped to Within the interval, , , These are the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient, respectively.
[0037] In the above technical solution, a comprehensive score is generated by weighting and combining the feature parameters extracted from the test, and this comprehensive score is used as the dynamic sealing quality value score. The formula for the calculation step is as follows: Calculate the dynamic sealing quality index ,in This item represents a scoring item indicating the degree to which the impact velocity direction deviates from the ideal normal during the impact test, measured by tangential velocity. The percentage is determined by calculation, among which The first weighting coefficient is a weighting factor for the influence of tangential velocity relative to normal impact velocity. This process is pre-set based on the relative importance of tangential velocity to impact velocity. This indicates the importance of the weighting factor, which is usually preset based on historical data; The item is a scoring item for the rationality of impact force and energy conversion. The total duration of the impact force. For the peak impact force, when Approximately impulse For the valve disc equivalent mass The term is related to the change in kinetic energy; the ratio of the two represents the efficiency of impact force and energy transfer, and ideally, it should be within a reasonable range. The second weighting coefficient... This indicates the importance of the weighting factor, which is usually preset based on historical data; The contact state score is represented by the characteristic value of the acoustic emission signal energy, while the normalized acoustic emission energy... The function reflects the severity of the collision and potential microscopic damage; the sum of these three values integrates the numerical representations of the angle deviation factor, impact force eccentricity factor, and acoustic emission signal uniformity factor, which are significantly related to changes in sealing performance in the test characteristics. This allows complex multidimensional physical parameters to be aggregated into a single, comparable value, namely the dynamic sealing quality index. This allows the dynamic sealing quality index to be used during the testing process. The size determines the sealing performance during actual testing, and has strong diagnostic value for valve testing.
[0038] As one embodiment of the present invention, the method for constructing a check valve sealing performance evaluation model includes: Collect the characteristic parameters and corresponding leakage values of check valves of different specifications and operating conditions obtained from historical tests to form a sample dataset; Using a sample dataset, a neural network is trained to obtain the mapping relationship from feature parameters to predicted leakage rates and performance scores, thereby generating a check valve sealing performance evaluation model.
[0039] In the above technical solution, multi-dimensional feature parameters obtained from the data of each test sample are used to form a feature vector X. After this test, the actual leakage value is accurately measured according to standard methods, and a historical performance score is determined based on the comprehensive performance. Next, a neural network model is trained and configured, using a multi-layer feedforward neural network. The number of input layer nodes equals the number of feature parameters, and the output layer is set to a dual-head output. One head is used for regression prediction of the leakage rate, and the other head is used for regression calculation of the predicted performance score. Several hidden layers are included in the middle to automatically learn the higher-order nonlinear relationship between the data and the target. The sample dataset is then divided according to a specific ratio. The model is divided into training, validation, and test sets. The training set minimizes the combined loss between the model's bi-headed prediction output and the true label. The validation set prevents overfitting during training and is used to select the optimal number of training iterations. Finally, the trained model is evaluated using a test set that was not used in the training process to assess its accuracy in predicting leakage rates and performance scores. This ensures the model has strong generalization ability and can reliably evaluate new valves. The constructed check valve sealing performance evaluation model can provide direct data input for subsequent quantitative grading, realizing an automated and intelligent evaluation process.
[0040] As one embodiment of the present invention, step S4 includes: Multiple consecutive dynamic sealing quality indices are preset. The threshold range, each threshold range corresponds to a performance level; The calculated The value is compared with the threshold range to determine the performance level of the check valve; the performance level includes at least: excellent, good, qualified, and unqualified. Among them, the performance level corresponding to "excellent" is The highest value indicates a smooth dynamic impact process, good alignment, moderate energy, and optimal predicted sealing performance; the performance level corresponds to the unqualified level. The lowest value indicates a serious problem in the dynamic impact process, such as excessive impact energy, poor centering, or abnormal collision. Output quantized performance score , ,in For the first One characteristic parameter, For normalization function, For the corresponding weights.
[0041] In the above technical solution, step S4 completes the quantitative output and intelligent decision-making process after dynamic performance evaluation calculation. Specifically, it establishes a performance level scale to classify different levels according to preset thresholds based on grading standards. This process is determined based on a large amount of historical test data and relevant engineering experience, analyzing the distribution of dynamic sealing quality indices calculated from a large number of valve samples. The threshold is set within the range of [insert range here]. Within the value range, a series of key threshold points are set, and these key threshold points are divided with interval breakpoints to divide continuous threshold intervals. For example, [0.85, 1.0] corresponds to "Excellent", [0.70, 0.85) corresponds to "Good", [0.60, 0.70) corresponds to "Pass", and [0, 0.60) corresponds to "Fail". After completing steps S1-S3, the shift calculation is performed to determine the value of this test. The value, and the system will The value is automatically compared within a preset threshold range, and based on... If the value falls within a certain range, the system automatically determines and outputs the final performance level of the valve, and simultaneously calculates the performance score based on the output level using the formula. Normalization function Mapped to a uniform dimension (e.g., 0-1 points, where a higher score indicates better performance). This is the weight of the parameter, reflecting its importance in the entire evaluation system.
[0042] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, and therefore described more simply; relevant parts can be referred to the descriptions of the method embodiments.
[0043] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this application, they should all fall within the protection scope of the present invention.
Claims
1. A method for testing the sealing performance of a check valve under high pressure differential conditions, wherein the check valve comprises a valve body, a valve disc, a valve seat, a valve cavity, a guide mechanism, an operating mechanism, and a rocker arm, characterized in that, The testing method includes: S1. Construct a high pressure differential impact flow field in the valve cavity between the inlet and outlet of the check valve body, and simulate the dynamic impact closing process; S2. Synchronously collect process parameters within the target time period before and after the dynamic impact closing process, and collect the leakage rate of the check valve body in the closed state; the process parameters include valve disc motion state parameters, valve disc and valve seat contact state parameters, and transient acoustic signals; S3. Construct a check valve sealing correlation model by performing correlation analysis on process parameters and leakage rate, and determine whether the correlation results are qualified based on the check valve sealing correlation model; S4. Based on the correlation analysis results, call the check valve sealing performance evaluation model to evaluate and quantify the dynamic sealing performance of the check valve under high pressure differential impact.
2. The method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 1, characterized in that, The step S1, simulating the dynamic impact shutdown process, includes: S11. Install the check valve into the test pipeline and position its valve disc at the specified opening degree. S12. A stable flow field is provided to the valve body inlet through a positive impact flow path to simulate the valve's working state; S13. Through the reverse impact flow path containing a high-pressure accumulator and an ultra-high-speed reversing valve, after receiving the control command, the high-pressure medium is instantaneously injected into the valve body outlet, forming a rapidly rising high pressure difference in the valve cavity, so that the valve disc completes dynamic impact closure.
3. The method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 1, characterized in that, The leakage rate value measured according to the standard pressure holding test procedure in step S2 is obtained by the following method: After the dynamic impact shutdown process ends and the pressure stabilizes, apply the specified pressure to the valve body inlet or outlet and maintain the pressure, then measure the leakage rate.
4. The method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 1, characterized in that, The correlation analysis in step S3 specifically includes: From the valve disc motion state parameters and the valve disc contact state parameters, calculate the normal impact velocity and impact kinetic energy of the valve disc hitting the valve seat at the moment of valve closing, and use them as the first dynamic characteristic parameter. Time-frequency domain analysis is performed on the transient acoustic signal to extract the main frequency energy of the acoustic emission signal, which characterizes the collision energy and stability of the sealing surface, as the second dynamic characteristic parameter; The extracted first and second dynamic feature parameters are fitted with the leakage rate using machine learning to establish a check valve sealing correlation model that characterizes the quantitative relationship between the dynamic process and the static result.
5. The method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 1, characterized in that, The check valve sealing correlation model established in step S3 is used for: To determine whether the dynamic sealing performance of the check valve is qualified under a single impact, the judgment method is as follows: whether the characteristic parameter is within the preset empirical safety threshold range, and whether the static sealing result meets the standard requirements; Predicting the sealing life of the check valve: Perform multiple cyclic tests of steps S1-S3 on the same valve body, track the evolution trend of the characteristic parameters and the static sealing results with the number of impacts, and predict the number of failure cycles based on the performance degradation law.
6. The method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 1, characterized in that, The calculation method for the check valve sealing correlation model is as follows: in, This is the predicted leakage rate. Normal impact velocity, For impact kinetic energy, For the peak impact force, The characteristic value of acoustic emission signal energy. For error terms; function It is obtained by fitting the training dataset.
7. The method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 6, characterized in that, The method for invoking the check valve sealing performance evaluation model is as follows: The feature parameters extracted in this test include the normal impact velocity. Impact kinetic energy Peak impact force Harmony emission signal energy eigenvalues The input is fed into a pre-trained check valve sealing performance evaluation model for calculation, and the dynamic sealing quality index is output: Through formula Calculate the dynamic sealing quality index ; in, Tangential velocity, referring to the velocity perpendicular to the vector. The impact velocity component on the plane, This is a weighting coefficient for the influence of tangential velocity relative to normal impact velocity; The duration of the impact force; For the valve disc equivalent mass Eigenvalues of acoustic emission signal energy The normalization function will convert the acoustic emission signal energy eigenvalues Mapped to Within the interval, , , These are the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient, respectively.
8. The method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 7, characterized in that, The methods for constructing the check valve sealing performance evaluation model include: Collect the characteristic parameters and corresponding leakage values of check valves of different specifications and operating conditions obtained from historical tests to form a sample dataset; Using the sample dataset, a neural network is trained to obtain the mapping relationship from feature parameters to predicted leakage rate and performance score, thereby generating a check valve sealing performance evaluation model.
9. A method for testing the sealing performance of a check valve under high pressure differential conditions according to claim 8, characterized in that, Step S4 includes: Multiple consecutive dynamic sealing quality indices are preset. The threshold range, each threshold range corresponds to a performance level; The calculated The value is compared with the threshold range to determine the performance level of the check valve; the performance level includes at least: excellent, good, qualified, and unqualified. Among them, the performance level corresponding to "excellent" is The highest value indicates a smooth dynamic impact process, good alignment, moderate energy, and optimal predicted sealing performance; the performance level corresponds to the unqualified level. The lowest value indicates a serious problem in the dynamic impact process, such as excessive impact energy, poor centering, or abnormal collision. Output quantized performance score , ,in For the first One characteristic parameter, For normalization function, For the corresponding weights.