Bolt torque fastening method for pneumatic valve diaphragm

Abstract

The bolt torque fastening method for the pneumatic valve diaphragm comprises the steps that the thickness of the diaphragm before fastening is measured, and the original thickness of the diaphragm is obtained; based on the original thickness of the diaphragm, performing fastening and compression ratio calculation on the diaphragm to obtain the compression ratio; when the compression ratio reaches a preset compression threshold value, recording the fastening torque value at the moment and correcting the fastening torque value to obtain a corrected fastening torque; determining a preliminary fastening torque, a final fastening torque, a preliminary re-tightening torque and a final re-tightening torque based on the corrected fastening torque; carrying out primary symmetrical fastening on the bolt of the diaphragm by using the primary fastening torque; the bolt of the diaphragm is symmetrically fastened for the second time through the final fastening torque; performing primary circumferential re-tightening on the bolt of the diaphragm by using the primary re-tightening torque; and after the first-time circumferential re-tightening is completed, waiting for a preset time length, and performing second-time circumferential re-tightening on the bolt of the diaphragm by using the final re-tightening torque. By means of the method, the long-term reliability of diaphragm sealing is improved, and the service life of key components is prolonged.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of nuclear power operation and maintenance, and particularly relates to a bolt torque fastening method for a diaphragm of a pneumatic valve. BACKGROUND

[0002] The pneumatic valve is a widely used flow control device in a nuclear power plant, and plays a key role in flow regulation in various systems of the nuclear power plant. The working principle is that driving force is provided by compressed air to push the rubber diaphragm to overcome the spring force to generate displacement, and then the diaphragm drives the valve stem to move, and finally the precise regulation of the valve opening is realized to meet the flow control requirements of the nuclear power plant under different working conditions.

[0003] The rubber diaphragm is a core and key component of the pneumatic valve, and its sealing performance directly determines whether the pneumatic valve can stably and reliably operate, and has an important influence on maintaining the normal working condition of the related system of the nuclear power plant. However, the rubber material itself has inherent creep and stress relaxation characteristics, which makes the bolt torque gradually decrease after the rubber diaphragm is installed and fastened by bolts and operated for a period of time. In recent years, there have been many feedbacks of individual bolts completely relaxing. It is found through research that during the installation and subsequent operation of the diaphragm, the bolt and the nut do not rotate relative to each other, but the pre-tightening force of the bolt cannot meet the use requirements, showing the characteristics of non-rotating relaxation. The essence of this phenomenon is that the rubber diaphragm will produce irreversible deformation with time under the continuous compression of the diaphragm box. Under the condition that the deformation remains unchanged, the stress inside the rubber will gradually decrease, directly leading to the decrease of the axial pre-tightening force of the diaphragm box bolt, and finally showing the relaxation of the bolt fastening torque. The root cause is closely related to the creep characteristics of the rubber diaphragm.

[0004] At the same time, the bolt fastening torque of the diaphragm is not the larger the better. If the fastening torque is too large, it will cause permanent deformation of the diaphragm, and even cause damage and tearing of the diaphragm, thereby affecting the normal function of the pneumatic valve. In recent years, some cases of diaphragm damage and tearing have shown that there may be a correlation with the excessive fastening torque. At present, the fastening torque of the diaphragm of the pneumatic valve in the industry is mainly dependent on the experience data of the valve manufacturer, and lacks scientific and quantitative basis based on the characteristics of the rubber material and the working state of the diaphragm, resulting in insufficient stability of the fastening effect under different scenes. It is difficult to effectively avoid the loosening and leakage problem caused by insufficient torque, and it is also difficult to reliably prevent the damage risk of the diaphragm caused by excessive torque, which brings hidden dangers to the long-term stable operation of the pneumatic valve of the nuclear power plant. SUMMARY

[0005] The purpose of the present application is to provide a bolt torque fastening method for a diaphragm of a pneumatic valve, which aims to solve the problem of unstable fastening effect in the prior art.

[0006] This invention provides a method for tightening bolts to a pneumatic valve diaphragm using a specific torque, comprising: Measure the thickness of the diaphragm before tightening to obtain the original thickness of the diaphragm; Based on the original thickness of the diaphragm, the diaphragm is tightened and the compression ratio is calculated to obtain the compression ratio; When the compression ratio reaches the predetermined compression threshold, the tightening torque value at this time is recorded and corrected to obtain the corrected tightening torque; Based on the corrected tightening torque, the initial tightening torque, the final tightening torque, the initial re-tightening torque, and the final re-tightening torque are confirmed; The bolts of the diaphragm are tightened symmetrically for the first time using the initial tightening torque; The bolts of the diaphragm are then tightened a second time symmetrically using the final tightening torque. The bolts of the diaphragm are circumferentially tightened for the first time using the initial tightening torque; After the first circumferential tightening is completed, wait for a predetermined time, and then tighten the bolts of the diaphragm a second time with the final tightening torque.

[0007] Furthermore, measuring the thickness of the diaphragm before tightening to obtain the original thickness of the diaphragm includes: Multiple points were selected on the surface of the diaphragm for thickness measurement, resulting in multiple measurement data. The original thickness of the diaphragm is obtained by calculating the average value of multiple measurement data.

[0008] Furthermore, based on the original thickness of the diaphragm, the clamping and compression ratio of the diaphragm are calculated to obtain the compression ratio, which includes: The diaphragm is tightened using a predetermined initial torque; Based on the predetermined initial torque, the diaphragm is tightened by iteratively increasing the torque according to a predetermined first amplitude, and the corresponding compression ratio is measured and calculated after each tightening to obtain the first compression ratio, until the first compression ratio reaches the predetermined initial threshold. When the first compression ratio reaches the predetermined initial threshold, the diaphragm is tightened by iteratively increasing the torque according to the predetermined second amplitude. After each tightening, the corresponding compression ratio is measured and calculated to obtain the second compression ratio. When the second compression ratio reaches the predetermined compression threshold, the tightening torque value at this time is recorded.

[0009] Furthermore, when the compression ratio reaches a predetermined compression threshold, recording the tightening torque value at this time and correcting it to obtain the corrected tightening torque includes: When the compression ratio reaches the predetermined compression threshold, the tightening torque value at this time is recorded to obtain the theoretical tightening torque; The average number of diaphragm movements per minute is obtained, and a frequency correction factor is calculated based on the average number of diaphragm movements per minute. The theoretical fastening torque is corrected based on the frequency correction factor to obtain the corrected fastening torque.

[0010] Furthermore, the frequency correction factor calculated based on the average number of movements per minute of the diaphragm includes: The frequency correction factor is calculated using the following formula: D = 1 + 0.2 × (1+f) Where D represents the frequency correction factor; f represents the average number of diaphragm movements per minute.

[0011] Furthermore, the step of confirming the initial tightening torque, final tightening torque, initial re-tightening torque, and final re-tightening torque based on the modified tightening torque includes: Set the initial tightening torque to 40%-60% of the corrected tightening torque; Set the final tightening torque to 100% of the corrected tightening torque; Set the initial tightening torque to 100% of the corrected tightening torque; Set the final tightening torque to 75%-85% of the corrected tightening torque.

[0012] Furthermore, symmetrical fastening includes: Number all bolts distributed around the circumference in a clockwise or counterclockwise order, and set the same number for two symmetrical bolts; Tighten all bolts sequentially, tightening one bolt at a time and then another bolt with the same number.

[0013] Furthermore, circumferential compressibility includes: Tighten all the bolts distributed around the circumference in a clockwise or counterclockwise order.

[0014] Furthermore, it also includes: Obtain the tensile deformation data of the bolt; Calculate the bolt torque value based on the tensile deformation data; Plot a torque-time relationship curve based on the bolt torque values; The torque variation trend is analyzed based on the torque-time relationship curve, and the torque relaxation rate is calculated based on the torque variation trend. An early warning is triggered when the torque relaxation rate exceeds a preset threshold.

[0015] Furthermore, it also includes: Data on the diaphragm fastening process and multi-dimensional attribute information are collected. The fastening process data includes the final compression rate and the final torque. The multi-dimensional attribute information includes the diaphragm's manufacturer, batch, model, material, service life, disassembly inspection results, and experience feedback information. A diaphragm performance database is established, and the fastening process data is associated with the corresponding multi-dimensional attribute information and stored in the diaphragm performance database; Cluster analysis was performed on historical data of diaphragms of the same model and material in the diaphragm performance database to generate data distribution characteristics; The correlation between different compression rates and the service life of the diaphragm was identified using artificial intelligence analysis tools. Based on the aforementioned correlation, the optimal compression ratio range for diaphragms of the same model and material is calculated.

[0016] This invention discloses a bolt torque tightening method for a pneumatic valve diaphragm, comprising: measuring the thickness of the diaphragm before tightening to obtain the original thickness of the diaphragm; calculating the tightening and compression ratio of the diaphragm based on the original thickness of the diaphragm to obtain the compression ratio; recording the tightening torque value at this time when the compression ratio reaches a predetermined compression threshold and correcting it to obtain a corrected tightening torque; confirming the initial tightening torque, final tightening torque, initial re-tightening torque, and final re-tightening torque based on the corrected tightening torque; performing a first symmetrical tightening of the diaphragm bolts with the initial tightening torque; performing a second symmetrical tightening of the diaphragm bolts with the final tightening torque; performing a first circumferential re-tightening of the diaphragm bolts with the initial re-tightening torque; and after the first circumferential re-tightening is completed, waiting for a predetermined time, and then performing a second circumferential re-tightening of the diaphragm bolts with the final re-tightening torque. This invention introduces a scientific torque determination method based on compression ratio testing, replacing the traditional approach that relies on manufacturer experience. This precisely matches the creep characteristics of rubber materials, solving the problem of insufficient basis for torque setting. A four-step tightening process is employed, using step-by-step loading, symmetrical tightening, and delayed re-tightening to effectively compensate for stress relaxation, ensuring uniform and durable pre-tightening force and avoiding the dual risks of damage from over-tightening and leakage from over-loosening. This significantly improves the long-term reliability of the diaphragm seal, reduces the risk of unplanned reactor shutdowns, and extends the service life of critical components. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1A schematic diagram of the bolt torque tightening method for pneumatic valve diaphragms; Figure 2 This is a schematic diagram showing the installation of the diaphragm inside the diaphragm box; Figure 3 This is a schematic diagram of the diaphragm structure; Figure 4 This is a schematic diagram of the cross-section of the diaphragm. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] It should be understood that, when used in this specification and the appended claims, the terms “comprising” and “including” indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0021] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0022] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0023] Please see Figures 1-4 This embodiment provides a method for tightening bolt torque of a pneumatic valve diaphragm, including: S101: Measure the thickness of the diaphragm before tightening to obtain the original thickness of the diaphragm; In this embodiment, measuring the thickness of the diaphragm before tightening to obtain the original thickness of the diaphragm includes: Multiple points were selected on the surface of the diaphragm to measure its thickness, resulting in multiple measurement data. The original thickness of the diaphragm is obtained by averaging the multiple measurement data.

[0024] Specifically, the new diaphragm to be installed is placed flat on the measuring platform, ensuring it is in a natural, stress-free state. A high-precision digital micrometer or ultrasonic thickness gauge is used as the measuring tool, and calibrated according to metrological standards before use. The selection of measurement points should be representative, typically within the effective pressure area of ​​the diaphragm sealing surface, following the principle of uniform distribution, selecting multiple symmetrical locations. At each selected point, the measuring tool probe is brought into perpendicular contact with the diaphragm surface, and a stable and slight pressure is applied. The displayed thickness value is read and recorded. After completing the measurement at all selected points, the multiple thickness measurements are summed and divided by the total number of measurement points to calculate the arithmetic mean, which is recorded as the original thickness of the diaphragm. This process eliminates errors caused by local thickness inhomogeneities, providing accurate and reliable initial input parameters for subsequent calculations of the compression ratio, thus ensuring that the subsequently determined tightening torque value is based on objective and quantitative data.

[0025] S102: Based on the original thickness of the diaphragm, the diaphragm is tightened and the compression ratio is calculated to obtain the compression ratio; In this embodiment, based on the original thickness of the diaphragm, the diaphragm is tightened and its compression ratio is calculated, and the compression ratio is obtained as follows: The diaphragm is tightened using a predetermined initial torque; Based on the predetermined initial torque, the torque is iteratively increased according to the predetermined first amplitude to tighten the diaphragm, and the corresponding compression ratio is measured and calculated after each tightening to obtain the first compression ratio, until the first compression ratio reaches the predetermined initial threshold. When the first compression ratio reaches the predetermined initial threshold, the torque is increased iteratively according to the predetermined second amplitude to tighten the diaphragm. After each tightening, the corresponding compression ratio is measured and calculated to obtain the second compression ratio. When the second compression ratio reaches the predetermined compression threshold, the tightening torque value at this time is recorded.

[0026] Specifically, after obtaining the initial thickness of the diaphragm, the diaphragm is installed on a valve diaphragm box or a dedicated test bench, and initially tightened using a predetermined initial torque, typically 10 to 20 Nm. Immediately after tightening, the thickness of the diaphragm under pressure is measured, and the initial compression ratio is calculated using the compression ratio calculation formula. Subsequently, based on the predetermined initial torque, the tightening torque is iteratively increased in predetermined first increments of 5-15 Nm each time. After each torque increase, the diaphragm thickness is remeasured and the corresponding compression ratio is calculated, resulting in a series of first compression ratio values. This iterative process continues until the calculated compression ratio reaches a predetermined initial threshold of 15% to 20%. When the compression ratio reaches this predetermined initial threshold, the tightening torque is iteratively increased in predetermined second increments of 1 to 2 Nm each time. Again, the thickness is measured and the compression ratio is calculated after each increase, resulting in a second compression ratio value. This fine-tuning process continues until the compression ratio meets a predetermined compression threshold of 20% to 25%. Once the final compression threshold is reached, the applied tightening torque value is recorded. This torque value is the theoretical tightening torque of the diaphragm determined through the compression ratio test, providing key input parameters for subsequent corrections and actual tightening operations.

[0027] The formula for calculating the compression ratio is as follows: Compression ratio = (original thickness - tightened thickness) / original thickness * 100%.

[0028] S103: When the compression ratio reaches the predetermined compression threshold, record the tightening torque value at this time and make corrections to obtain the corrected tightening torque; In this embodiment, when the compression ratio reaches a predetermined compression threshold, the tightening torque value at this time is recorded and corrected. The corrected tightening torque is then obtained by: When the compression ratio reaches the predetermined compression threshold, record the tightening torque value at this time to obtain the theoretical tightening torque; The average number of diaphragm movements per minute is obtained, and the frequency correction factor is calculated based on the average number of diaphragm movements per minute. The theoretical tightening torque is corrected by adjusting the frequency correction factor to obtain the corrected tightening torque.

[0029] Specifically, when the diaphragm compression rate reaches a predetermined compression threshold of 20% to 25%, the tightening torque value displayed on the test bench or torque wrench at this moment is recorded; this value is defined as the theoretical tightening torque. Subsequently, the average number of actuations per minute of the pneumatic valve diaphragm under actual operating conditions needs to be obtained. This data is typically obtained from the valve's historical operating logs or statistical information from the control system. Next, this average number of actuations per minute is substituted into a preset empirical formula for calculation; the result is the frequency correction factor. Finally, the theoretical tightening torque obtained above is multiplied by the frequency correction factor; the product is the corrected tightening torque after operating condition correction. This correction process quantifies the impact of dynamic load differences caused by varying diaphragm actuation frequency on preload retention, ensuring that the final determined tightening torque value is not only based on static compression characteristics but also closely matches the actual dynamic service conditions of the diaphragm, providing more scientific and personalized target parameters for subsequent installation and tightening.

[0030] In this embodiment, the frequency correction factor calculated based on the average number of times the diaphragm moves per minute includes: The frequency correction factor is calculated using the following formula: D = 1 + 0.2 × (1+f) Where D represents the frequency correction factor; f represents the average number of diaphragm movements per minute.

[0031] S104: Based on the modified tightening torque, confirm the initial tightening torque, final tightening torque, initial re-tightening torque, and final re-tightening torque; In this embodiment, determining the initial tightening torque, final tightening torque, initial re-tightening torque, and final re-tightening torque based on the corrected tightening torque includes: Set the initial tightening torque to 40%-60% of the corrected tightening torque; Set the final tightening torque to 100% of the corrected tightening torque; Set the initial re-tightening torque to 100% of the corrected tightening torque; Set the final tightening torque to 75%-85% of the corrected tightening torque.

[0032] S105: Tighten the bolts of the diaphragm symmetrically for the first time with the initial tightening torque; In this embodiment, symmetrical fastening includes: Number all bolts distributed around the circumference in a clockwise or counterclockwise order, and set the same number for two symmetrical bolts; Tighten all bolts sequentially, tightening one bolt at a time and then another bolt with the same number.

[0033] Specifically, all bolts evenly distributed around the diaphragm box are systematically numbered, proceeding clockwise or counterclockwise sequentially. The key rule is to assign the same number to two bolts that are directly opposite each other (180 degrees apart). For example, bolts at the 12 o'clock and 6 o'clock positions are designated as number A, and bolts at the 3 o'clock and 9 o'clock positions are designated as number B. If there are eight bolts in total, the bolts at the 1 o'clock and 7 o'clock positions may be designated as number C, the bolts at the 4 o'clock and 10 o'clock positions as number D, and so on, until all bolts are numbered. The tightening operation is strictly performed according to this numbering system. The operation sequence is as follows: First, apply torque to one of the bolts numbered A (e.g., the 12 o'clock position) to the target value and tighten it. Immediately afterwards, tighten the other bolt numbered A (i.e., the 6 o'clock position). After both symmetrical bolts corresponding to number A are tightened, move to number B and tighten its two corresponding symmetrical bolts in the same way (e.g., the 3 o'clock and 9 o'clock positions). The subsequent bolts are tightened in the order of C, D, etc. This method, by assigning the same identification to symmetrical bolts and tightening them in pairs, enforces symmetrical and balanced application of force from an operational logic perspective, effectively avoiding uneven stress on the diaphragm caused by improper tightening sequence.

[0034] In some embodiments, all bolts are tightened sequentially in the following manner: tightening one bolt 1, tightening another bolt 2 with the same number, then tightening bolt 3 (bolt 3 is located between bolt 1 and bolt 2, the number of bolts between bolt 3 and bolt 1 is the same as or differs by only one from the number of bolts between bolt 3 and bolt 2), and then tightening bolt 4 with the same number as bolt 3.

[0035] Specifically, all the bolts distributed around the circumference are treated as clockwork, and they are numbered sequentially in the order of "12 o'clock → 6 o'clock → 3 o'clock → 9 o'clock → 1 o'clock → 7 o'clock → 4 o'clock → 10 o'clock → 2 o'clock → 8 o'clock → 5 o'clock → 11 o'clock".

[0036] Then, using a pre-set target torque wrench, first align it with the bolt at the 12 o'clock position and apply force to the specified torque; after completion, immediately move to the bolt at the 6 o'clock position directly opposite and tighten it; then, tighten the bolt at the 3 o'clock position, and then tighten the bolt at the 9 o'clock position directly opposite. After completing this first set of four perfectly symmetrical points, the process does not end, but continues to follow the core principle of tightening one point at a time, that is, tightening its corresponding point, and continues to operate in the numbered sequence: that is, immediately tighten the bolt at the 1 o'clock position, and then immediately tighten the bolt at the 7 o'clock position directly opposite; then tighten the bolt at the 4 o'clock position, and then tighten the bolt at the 10 o'clock position directly opposite; then tighten the bolt at the 2 o'clock position, and then tighten the bolt at the 8 o'clock position directly opposite; finally tighten the bolt at the 5 o'clock position, and complete the tightening of the bolt at the 11 o'clock position directly opposite. This carefully designed sequence ensures that the stress generated by each bolt on the circumference is immediately balanced by the force exerted by the bolts on the opposite diameter when it is loaded. This maintains the diaphragm's central position and parallelism with the flange face to the greatest extent possible during loading, effectively preventing bias pressure, unilateral warping, and the resulting uneven sealing. This lays a crucial foundation for establishing a uniform and stable initial sealing stress field.

[0037] S106: Tighten the bolts of the diaphragm a second time symmetrically with the final tightening torque; S107: Tighten the bolts of the diaphragm for the first time in a circumferential manner with the initial tightening torque; In this embodiment, circumferential tightening includes: Tighten all the bolts distributed around the circumference in a clockwise or counterclockwise order.

[0038] Specifically, during retightening, a single clockwise or counterclockwise direction is chosen as the tightening path. For example, starting with the bolt at the 12 o'clock position, then sequentially tightening the bolts at the 1 o'clock, 2 o'clock, and 3 o'clock positions until all bolts are tightened in a circumferential sequence. During this process, a torque wrench with a preset target torque value is used to individually tighten each bolt to the specified torque, ensuring that each bolt undergoes a complete and independent load. This circumferential retightening method systematically eliminates the drawbacks of mutual constraint between bolts or the potential for human error during symmetrical tightening when there are many bolts. This further balances and stabilizes the clamping force across the entire diaphragm circumference, thereby consolidating the overall contact state of the sealing surface and improving the long-term reliability of the seal.

[0039] S108: After the first circumferential tightening is completed, wait for a predetermined time, and then tighten the bolts of the diaphragm a second time with the final tightening torque.

[0040] Specifically, after the first circumferential tightening based on 100% of the corrected tightening torque value, a predetermined stress relaxation waiting period is required. This predetermined duration is set to 4 to 24 hours based on the creep characteristics of the rubber material. During this period, the internal stress of the initially tightened diaphragm naturally relaxes over time under constant compression deformation, and the axial preload of each bolt will decrease to varying degrees. After the waiting period, the operator uses a calibrated torque wrench to perform a second circumferential tightening. The target torque used for this tightening is the final tightening torque, which is set to 75%-80% of the initial installation torque, i.e., the corrected tightening torque. The tightening operation also follows a circumferential sequence. Starting from any bolt, torque is applied individually to each bolt on the circumference along a pre-selected clockwise or counterclockwise path consistent with the first tightening direction until the wrench reaches the set final tightening torque value and issues a prompt. This step, by applying a torque slightly lower than the initial installation value, aims to gently raise and level the bolt preload, which has generally decreased due to stress relaxation, to a reasonable and uniform level. This avoids potential diaphragm overload damage caused by excessive secondary loading torque and effectively compensates for preload loss caused by material creep. Thus, a stable, reliable, and durable sealing preload is established before the diaphragm is subjected to long-term compression service, completing the final key step in the entire fastening process.

[0041] This embodiment also includes: Obtain the tensile deformation data of the bolt; Calculate the bolt torque value based on the tensile deformation data; Plot the torque-time relationship curve based on the bolt torque value; The torque variation trend is analyzed based on the torque-time relationship curve, and the torque relaxation rate is calculated based on the torque variation trend. An early warning is triggered when the torque relaxation rate exceeds a preset threshold.

[0042] Specifically, the tensile deformation of the bolts is monitored in real time using high-precision strain sensors or ultrasonic testing devices mounted on the bolts. The sensors convert the tensile deformation into optical signals, which are then converted into digital signals by a demodulator and transmitted to the monitoring server, forming a continuous sequence of tensile deformation data. Based on Hooke's Law and the bolt material properties, the tensile deformation data is converted into corresponding bolt torque values ​​using the formula Bolt torque value = Tensile deformation × Elastic modulus × Cross-sectional area × Bolt radius / Pitch. Temperature compensation coefficients are considered to correct for thermal expansion effects during the calculation. The monitoring system automatically associates the calculated torque values ​​with corresponding timestamps, smooths them using a cubic spline interpolation algorithm, and plots the torque-time relationship curve in real time on the monitoring interface. Simultaneously, the original data and processing results are stored in the diaphragm performance database. The system performs a slope analysis on the torque-time curve every 24 hours, using the least squares method to fit the data points of the most recent 72 hours, calculating the torque relaxation rate as (initial torque value - current torque value) / initial torque value / time difference × 100%, in % / 24h. When the calculated torque relaxation rate exceeds the preset 5% / 24h threshold, the system automatically triggers a three-level early warning mechanism: a red warning icon is displayed on the monitoring interface, a warning SMS is sent to the equipment management personnel's mobile phone, and a pending work order is generated in the maintenance management system. The warning information includes the valve location, bolt number, current relaxation rate, and historical trend comparison data, reminding maintenance personnel to promptly perform on-site re-tightening or diaphragm replacement. The system also records the trigger time and processing result of each warning for optimization of the warning threshold and prediction model parameters.

[0043] This embodiment also includes: Data on the diaphragm fastening process and multi-dimensional attribute information are collected. The fastening process data includes the final compression rate and final torque, while the multi-dimensional attribute information includes the diaphragm's manufacturer, batch, model, material, service life, disassembly inspection results, and experience feedback. Establish a diaphragm performance database, associate the fastening process data with the corresponding multi-dimensional attribute information and store them in the diaphragm performance database; Cluster analysis was performed on historical data of diaphragms of the same model and material in the diaphragm performance database to generate data distribution characteristics; Artificial intelligence analysis tools were used to identify the correlation between different compression rates and the service life of the diaphragm based on the data distribution characteristics. The optimal compression ratio range for diaphragms of the same model and material is calculated based on the correlation.

[0044] Specifically, during each diaphragm installation or replacement operation, the system collects detailed tightening process data, including the final compression ratio and final torque determined through testing. It also records multi-dimensional attribute information of the diaphragm, covering manufacturer, product batch, model specifications, material composition, service life, detailed information from previous disassembly inspections, and relevant experience feedback. Subsequently, the system automatically structures and correlates this tightening process data with its corresponding multi-dimensional attribute information, forming a complete and traceable digital record, which is stored long-term in a dedicated diaphragm performance database. Based on this, cluster analysis is performed on all historical records of diaphragms of the same specific model and material accumulated in the diaphragm performance database, extracting data distribution characteristics of key parameters such as compression ratio, torque, and service life from the massive amount of data. Then, artificial intelligence analysis tools are used to deeply mine these data distribution characteristics, identifying the potential correlations and influence patterns between different compression ratio levels and the actual service life of the diaphragm through machine learning algorithms. Ultimately, based on the identified strong correlation, the system calculates and analyzes a compression ratio range that ensures reliable sealing while extending service life. This range is determined as the optimal compression ratio range for this specific model and material diaphragm, providing a direct and scientific basis for torque setting of subsequent diaphragms of the same type.

[0045] In some embodiments, establishing a diaphragm performance database, and associating and storing fastening process data with corresponding multi-dimensional attribute information in the diaphragm performance database includes: Data on the diaphragm fastening process and multi-dimensional attribute information are collected. The fastening process data includes the final compression rate and final torque, while the multi-dimensional attribute information includes the diaphragm manufacturer, batch, model, material, service life, disassembly inspection results, and experience feedback. A unique identifier for the diaphragm is generated based on the diaphragm manufacturer, batch, and model information from multi-dimensional attribute information. Design the table structure of the diaphragm performance database. The table structure includes a main table and a sub-table. The main table stores the diaphragm's unique identification code and basic attribute fields, while the sub-table stores the fastening process data fields. The main table and the sub-table are associated through the diaphragm's unique identification code. The unique identification code of the diaphragm is mapped to the corresponding fastening process data and multi-dimensional attribute information into a structured data record; Structured data records are written into the diaphragm performance database according to the table structure. The consistency of data between the main table and the sub-table is ensured through the database transaction mechanism, forming an association between the diaphragm's unique identification code, the fastening process data, and multi-dimensional attribute information.

[0046] Specifically, comprehensive data and multi-dimensional attribute information of the diaphragm tightening process are collected. The final compression rate is calculated by the difference between the original thickness of the diaphragm and the thickness after the last tightening. The final torque is the determined value after frequency correction and re-tightening verification. Among the multi-dimensional attribute information, the diaphragm manufacturer, batch, model, and material are verified and extracted from the equipment purchase contract and warehousing acceptance ledger. The service life is calculated based on the diaphragm installation and commissioning date and the current operating time. The disassembly inspection results are recorded in detail, including the diaphragm surface integrity and the degree of wear on the sealing surface. Experience feedback information is summarized, including torque relaxation and potential leakage hazards that occurred during operation. Subsequently, a unique identification code is generated based on the collected diaphragm manufacturer, batch, and model. The coding rule adopts the combination of the manufacturer's English abbreviation + 6-digit batch number + model code. For example, if the manufacturer's abbreviation is XM, the batch number is 202405, and the model is GM-10, then the unique identification code is XM202405GM10, ensuring that each identification code corresponds to a unique diaphragm. Next, the table structure of the diaphragm performance database was designed. The main table includes basic attribute fields such as unique identifier, manufacturer, batch, model, material, and service life, with the unique identifier as the primary key. The sub-table includes fastening process data fields such as unique identifier, final compression ratio, and final torque, with the unique identifier used as a foreign key to establish a relationship with the main table. Then, the generated diaphragm unique identifier is matched one-to-one with the corresponding fastening process data and multi-dimensional attribute information, mapping them to structured data records that meet the table structure requirements, ensuring that the field information of each record is complete and accurately corresponds to the unique identifier. Finally, the database transaction mechanism is initiated. First, the unique identifier and basic attribute information from the structured data records are written to the main table. After the main table is successfully written, the unique identifier and corresponding fastening process data are written to the sub-table. If any step fails to write, a transaction rollback is triggered to ensure that the data in the main table and sub-table are synchronized and consistent, ultimately achieving the associated storage of the diaphragm unique identifier, fastening process data, and multi-dimensional attribute information.

[0047] In some embodiments, cluster analysis is performed on historical data of diaphragms of the same model and material in the diaphragm performance database to generate data distribution characteristics including: Historical data of diaphragms of the same model and material were retrieved from the diaphragm performance database. The historical data included the final compression ratio, final torque value, service life and failure mode records. Outlier filtering and missing value imputation are performed on the retrieved historical diaphragm data to generate a standardized historical dataset; The K-means algorithm is applied to cluster the standardized historical dataset according to the feature dimension, generating the initial cluster center coordinates and sample affiliation labels; Calculate the sum of squared Euclidean distances between the initial cluster center coordinates and each sample point to generate a clustering error sum of squares index. Compare the sum of squared clustering errors index with the preset convergence threshold. When the sum of squared clustering errors index is greater than the preset convergence threshold, adjust the clustering number parameter of the K-means algorithm. The K-means clustering algorithm is re-executed using the adjusted cluster size parameter to generate optimized clustering results, which include the coordinates of the cluster centroids and the final label of each sample. The statistical optimization of the clustering results generates data distribution characteristics by analyzing the sample size distribution ratio of each cluster, the mean of the final compression rate, and the standard deviation of the usage years.

[0048] Specifically, the process of performing cluster analysis on historical data in the diaphragm performance database to generate data distribution characteristics is as follows: First, based on the specified diaphragm model and material constraints, all historical data records meeting the conditions are retrieved and extracted from the established diaphragm performance database. These records contain key process and service parameters, such as final compression ratio, final torque value, actual service life, and failure mode information. Next, data cleaning is performed on the extracted raw historical data, including identifying and filtering outliers that significantly deviate from the normal range, and filling in missing values ​​in some fields using interpolation or methods based on the mean of similar data, thereby generating a clean and complete standardized historical dataset. Subsequently, the K-means clustering algorithm is applied to analyze this standardized historical dataset. The algorithm first randomly initializes multiple cluster centers and assigns initial labels to each sample point based on the characteristic distance between each sample point and these centers, while simultaneously calculating the initial coordinates of the cluster centers. Based on this initial partitioning, the system calculates the sum of squared Euclidean distances from all sample points to their respective cluster centers, obtaining the key indicator of the sum of squared clustering errors. The system compares this indicator with a preset convergence threshold. If the indicator value is greater than the threshold, it indicates that the current clustering is not optimal, and the algorithm will automatically adjust the number of clusters. Using the adjusted parameters, the K-means algorithm is re-executed, and after iterative optimization, the final optimized clustering result is generated. This result includes more accurate cluster center coordinates and the final assignment label for each sample point. Finally, based on this optimized clustering result, the system performs statistical calculations on each formed cluster. Specifically, this includes calculating the proportion of samples in each cluster to the total dataset, calculating the arithmetic mean of the final compression rate of samples within the cluster, and calculating the standard deviation of the service life of samples within the cluster. Combining these statistics constitutes a data distribution characteristic that can quantitatively reflect the historical performance of a specific model and material diaphragm, providing structured input for subsequent correlation analysis and process optimization.

[0049] In some embodiments, identifying the correlation between different compression rates and diaphragm service life using artificial intelligence analysis tools includes: Obtain the data distribution characteristics generated by cluster analysis in the diaphragm performance database. The data distribution characteristics include the sample size distribution ratio of each cluster, the mean of the final compression ratio, and the standard deviation of the service life. A multi-dimensional training sample set is constructed based on the data distribution characteristics. The multi-dimensional training sample set contains the mapping relationship between the final compression rate interval label and the corresponding usage years statistical value. Design a long short-term memory neural network structure. The neural network structure includes an input layer, two hidden layers and an output layer. The number of nodes in the input layer matches the dimension of the final compression rate range, and the number of nodes in the output layer matches the dimension of the usage period prediction. A multi-dimensional training sample set is input into a long short-term memory neural network structure for iterative training to generate an initial association prediction model. Calculate the root mean square error of the initial correlation prediction model and generate a model accuracy evaluation index; Compare the model accuracy evaluation index with the preset accuracy threshold. When the model accuracy evaluation index is greater than the preset accuracy threshold, adjust the number of hidden layer nodes and the learning rate parameter of the long short-term memory neural network. The long short-term memory neural network was retrained using the adjusted parameters to generate an optimized association prediction model. By optimizing the correlation prediction model, the influence weights of different final compression ratio ranges on the service life of the diaphragm are analyzed, and the nonlinear correlation function of compression ratio and service life is output.

[0050] Specifically, the structured data distribution characteristics previously generated through cluster analysis are retrieved from the diaphragm performance database. These characteristics are not simply a list of raw measurements, but rather refined statistical summaries. The core content includes the sample size distribution ratio of each cluster formed from historical data for specific diaphragm models and materials, the arithmetic mean of the final compression ratio of all samples within each cluster, and the standard deviation of the corresponding service life. These characteristics collectively characterize the statistical regularity and dispersion of the diaphragm group's lifespan performance at different compression ratio levels, providing high-quality, low-noise input information for subsequent correlation analysis.

[0051] Next, based on the aforementioned data distribution characteristics, the system needs to construct a multi-dimensional training sample set that can be understood and learned by the machine learning model. The construction logic is to treat each cluster as an independent data sample. Specifically, the input feature label for this sample is a predefined compression rate range (e.g., 20%-21%) into which the average final compression rate of the cluster falls. Simultaneously, the cluster's usage age statistics, such as average or median usage age, are used as the target output value for this sample. The system iterates through all clusters, establishing a one-to-one mapping between the compression rate range label of each cluster and the corresponding usage age statistics, thus forming a well-formatted, multi-dimensional training sample set with clearly defined features and targets. This step transforms abstract distribution characteristics into training data with explicit supervisory signals.

[0052] Subsequently, to capture the complex and nonlinear temporal or cumulative effects that may exist between compression rate and service life, this embodiment selects a Long Short-Term Memory (LSTM) neural network as the core architecture of the analysis tool. This network structure specifically includes an input layer, two cascaded hidden layers, and an output layer. The number of nodes in the input layer is strictly matched to the encoding dimension of the compression rate interval labels in the training sample set, ensuring that the input information can be fed into the network without loss. The number of nodes in the output layer is set to 1, directly corresponding to the single continuous value of the diaphragm service life to be predicted. The design of the two hidden layers aims to give the network sufficient nonlinear expressive power to learn potential high-order correlation patterns in the data.

[0053] The constructed multi-dimensional training sample set is then divided into a training set and a validation set according to a certain ratio, and input into the designed long short-term memory neural network structure for iterative training. The training process uses backpropagation, a common time series algorithm, with mean squared error as the loss function. By iteratively optimizing the network weight parameters, an initial correlation prediction model that can predict the corresponding service life based on the input compression rate range is finally generated.

[0054] Next, the initial association prediction model is evaluated using independent test sets or reserved validation set data. The core evaluation metric is the root mean square error (RMSE) between the model's predicted values ​​and the actual statistical values; this value is used as a key evaluation metric to quantify the model's prediction accuracy. The system compares this calculated model accuracy evaluation metric with a preset accuracy threshold set according to engineering application requirements. This threshold represents the upper limit of acceptable prediction error.

[0055] If the evaluation metrics show that the root mean square error of the initial model is greater than the preset accuracy threshold, indicating that the model's prediction accuracy has not met the requirements, the system will proceed to the model parameter tuning and optimization phase. The system will automatically execute hyperparameter tuning strategies. Typical adjustments include increasing or decreasing the number of nodes in the hidden layers of the Long Short-Term Memory (LSTM) neural network to change the model capacity, and may also adjust the learner's learning rate parameter to improve the convergence of the training process. These adjustments aim to enhance the model's learning ability and generalization performance.

[0056] After parameter tuning, the system enters the model retraining phase. Using the adjusted number of hidden layer nodes and learning rate, the system reinitializes and trains a new Long Short-Term Memory (LSTM) neural network on the same training sample set. Because the parameters have been specifically optimized, this training aims to obtain a model with superior performance. After a complete training cycle, an optimized association prediction model is generated.

[0057] Finally, by analyzing the activation patterns of the model's hidden layer when receiving inputs from different compression ratio ranges, or by employing gradient analysis, the system can calculate and quantify the influence weights of different final compression ratio ranges on the predicted diaphragm service life. Based on these weight analyses and the mathematical structure of the model itself, the system can fit or derive an explicit mathematical expression, which is the output nonlinear correlation function between compression ratio and service life. This function, in a clear and analytical form, quantitatively describes how compression ratio, as the independent variable, affects service life, the dependent variable. This completes the entire intelligent identification process from data features to a clear physical correlation, providing a direct scientific basis for subsequent optimal range calculations.

[0058] In some embodiments, calculating the optimal compression ratio range for diaphragms of the same type and material based on correlation includes: Obtain the nonlinear correlation function between compression ratio and service life generated by artificial intelligence analysis tools. The nonlinear correlation function characterizes the influence of different compression ratios on the service life of the diaphragm. An objective function is constructed based on a nonlinear correlation function between compression ratio and service life, with the objective function aiming to maximize the service life of the diaphragm. Set compression ratio boundary constraints, including minimum sealing compression ratio threshold and maximum material deformation compression ratio threshold; The objective function is solved under boundary constraints using a sequential quadratic programming algorithm to generate the theoretically optimal compression ratio point. Calculate the absolute value of the partial derivative of the service life with respect to the compression ratio at the theoretically optimal compression ratio point to generate a compression ratio sensitivity index; Compare the compression ratio sensitivity index with the preset sensitivity threshold. When the compression ratio sensitivity index is less than the preset sensitivity threshold, expand the neighborhood range of the theoretically optimal compression ratio point. Based on the theoretical optimal compression ratio point and its neighborhood range, the lower limit and upper limit of the compression ratio are determined to form the optimal compression ratio range.

[0059] Specifically, the system obtains a trained nonlinear correlation function between compression ratio and service life. This function, modeled using a deep neural network, accurately describes the variation in service life of diaphragms of specific models and materials under different compression ratios. Based on this nonlinear correlation function, a mathematical optimization objective function is constructed. The objective function is defined as a continuously differentiable function of diaphragm service life with respect to compression ratio, and the optimization direction is set to find the compression ratio value that maximizes the service life. The system also sets compression ratio boundary constraints, including a minimum sealing compression ratio threshold determined based on sealing performance requirements and a maximum material deformation compression ratio threshold determined based on the material's elastic limit, forming a closed feasible region. Then, a sequential quadratic programming algorithm is applied to iteratively solve for the extreme points of the objective function under the set boundary constraints. Through gradient descent and Hessian matrix correction, the algorithm converges to the theoretically optimal compression ratio point. Subsequently, the absolute value of the first-order partial derivative of the service life function with respect to compression ratio at this theoretically optimal compression ratio point is calculated to generate a compression ratio sensitivity index characterizing the impact of small changes in compression ratio on service life. The system compares the compressibility sensitivity index with a preset engineering sensitivity threshold. When the sensitivity index is less than the threshold, it indicates the existence of a wide, flat region near the optimal compressibility point. The system automatically expands the neighborhood of the theoretical optimal compressibility point, calculating all compressibility boundaries with a service life not less than 95% of the maximum value. Finally, the system combines the theoretical optimal compressibility point and the expanded neighborhood to determine the lower and upper limits of the compressibility that meet the service life requirements, forming a complete optimal compressibility range. This range is marked and stored in the diaphragm performance database for the corresponding model and material's process parameter table, providing precise parameter guidance for subsequent diaphragm fastening processes.

[0060] In some embodiments, the sequential quadratic programming algorithm is applied to the diaphragm compression ratio optimization problem. The objective function is defined as a nonlinear function f(x) of the service life with respect to the compression ratio, where x represents the compression ratio variable, and the function exhibits a single-peak characteristic curve. The boundary constraints are set as a closed interval [a, b], where a is the minimum sealing compression ratio threshold to ensure the diaphragm sealing performance, and b is the maximum material deformation compression ratio threshold to prevent excessive material deformation failure. During the initialization of the optimization process, the midpoint of the interval is selected as the initial iteration point x0, and the objective function value f(x0), the first derivative g(x0), and the second derivative H(x0) at this point are calculated. In each iteration, the algorithm constructs a local quadratic programming subproblem, with the objective of minimizing the quadratic approximation function q(d) = f(x0). k )+g(x k )d+0.5H(x k d, where the constraint is a≤x k+d≤b, where d is the search direction. The search direction d is obtained by solving this quadratic programming subproblem. k The step size α is determined using Armijo backtracking search (Armijo backtracking search is a classic step size determination method widely used in nonlinear optimization algorithms to ensure that the objective function decreases sufficiently during iteration). k To ensure the objective function decreases sufficiently, update the iteration point x. k+1 =x k +α k d k Simultaneously monitor the change in iteration point ||x k+1 -x k ||and the gradient norm of the objective function||g(x) k+1 Convergence is determined when both the objective function and the constraint are less than the preset convergence tolerance. During the iteration process, if the search direction points outward from the boundary, the algorithm automatically activates the constraint handling mechanism, taking the boundary point as a candidate solution and calculating the degree of satisfaction of the KKT conditions (KKT conditions are an extension of the Lagrange multiplier method, specifically for handling optimization problems that simultaneously contain equality and inequality constraints. When the objective function and constraint conditions satisfy convexity, the KKT conditions are both necessary and sufficient conditions). After multiple iterations, the algorithm converges to x*=22.7%, which satisfies the KKT optimality condition. The first derivative is close to zero, and the second derivative is negative, confirming it as a local maximum point. Simultaneously, it lies within the constraint interval, with no boundary constraints activated. This point is determined as the theoretically optimal compression ratio point and used as the benchmark input for subsequent sensitivity analysis.

[0061] In some embodiments, a previously calculated compression ratio sensitivity index is obtained, which quantifies the sensitivity of the service life at the theoretical optimal compression ratio point to changes in compression ratio. The system simultaneously reads a preset sensitivity threshold parameter, determined based on diaphragm material characteristics and engineering experience, and stored in the process parameter configuration file. A numerical comparison operation is then performed, comparing the compression ratio sensitivity index with the preset sensitivity threshold. When the compression ratio sensitivity index is determined to be less than the preset sensitivity threshold, the system identifies a relatively flat performance plateau region near the theoretical optimal compression ratio point, and activates the neighborhood range expansion algorithm. The expansion algorithm uses the theoretical optimal compression ratio point as the reference center, defines an initial expansion step size, and employs an adaptive stepping strategy to search simultaneously in both directions of increasing and decreasing compression ratio. In each search direction, the predicted service life corresponding to the current compression ratio is calculated and compared with the maximum service life at the theoretical optimal compression ratio point. When the service life decreases to 95% of the maximum value, this position is recorded as a candidate neighborhood boundary point. To ensure the accuracy of the boundary points, the algorithm uses quadratic interpolation to perform a refined search in the critical region, gradually reducing the search step size to an accuracy level of 0.1%. Simultaneously, the system monitors the rate of change of material stress during the expansion process. When an abnormal increase in the rate of change of stress is detected, the boundary position is automatically adjusted to avoid areas of abrupt changes in material properties. The expansion process also considers operating condition fluctuations, introducing a 5% safety margin coefficient based on the determined boundary to moderately expand the neighborhood range. Ultimately, the algorithm generates an expanded neighborhood range centered on the theoretical optimal compression ratio point, with clearly defined upper and lower boundaries. Compression ratio values ​​within this range can ensure that the diaphragm's service life remains at a high level, providing a crucial data foundation for subsequently determining the optimal compression ratio range.

[0062] In some embodiments, it also includes: A mathematical model for material stress relaxation is established, which includes the stress decay time constant and the material creep property coefficient. Once the diaphragm is tightened to the target compression ratio, the real-time bolt torque value sequence and the real-time diaphragm compression ratio value sequence within the preset time window are collected to generate a stress relaxation monitoring dataset. The stress relaxation monitoring dataset is compared and analyzed with the material stress relaxation mathematical model. The difference between the measured stress attenuation rate and the model predicted stress attenuation rate is calculated, and stress relaxation error parameters are generated. Based on stress relaxation error parameters and material stress relaxation mathematical models, the cumulative stress loss within a preset operating time is predicted, and the compensation torque value is calculated. The compensation torque value is directly proportional to the cumulative stress loss. A secondary tightening command is generated based on the compensation torque value, driving the tightening device to perform a secondary tightening operation on the bolts. The torque parameter of the secondary tightening operation is equal to the sum of the initial tightening torque value and the compensation torque value.

[0063] Specifically, a mathematical model for material stress relaxation is constructed. This model is based on the Boltzmann superposition principle (a fundamental theory in viscoelasticity used to describe the response characteristics of materials under complex load histories). It introduces Maxwell elements (a fundamental model in viscoelasticity used to describe the stress relaxation behavior of materials) and Kelvin-Voigt elements (a fundamental model in viscoelasticity used to describe the viscoelastic behavior of materials) to describe the viscoelastic behavior of the diaphragm material. The stress decay time constant is determined through tensile testing, reflecting the rate of stress decay over time. The creep coefficient is obtained through stepped loading tests, characterizing the cumulative deformation effect of the material under sustained load. The model parameters are optimized by fitting experimental data using the least squares method to ensure prediction accuracy. Once the diaphragm tightening process reaches the preset target compression rate of 23.5%, the system automatically triggers the monitoring mode. Using high-precision torque and displacement sensors, it continuously collects real-time bolt torque and diaphragm compression rate values ​​within a 10-minute time window at a sampling frequency of 1Hz. Simultaneously, ambient temperature and medium pressure are recorded as auxiliary parameters. Wavelet denoising is applied to the raw signals to eliminate high-frequency interference, generating a stress relaxation monitoring dataset containing 320 valid data points. This dataset is then input into the material stress relaxation mathematical model. The relative deviation between the measured stress decay rate and the model-predicted stress decay rate at each time point is calculated. A weighted average method is used to synthesize the deviation values ​​at each time point, generating a stress relaxation error parameter characterizing the model's prediction accuracy. When the error parameter exceeds a 5% threshold, the system automatically corrects the material creep coefficient to improve model adaptability. Then, combining the corrected material stress relaxation mathematical model and stress relaxation error parameters, the cumulative stress loss of the equipment during the preset 720-hour operating time is extrapolated. Based on the material's elastic modulus and contact area parameters, the stress loss is converted into a torque loss value, which is then multiplied by a safety factor of 1.15 to obtain the compensation torque value, ensuring that the compensation amount is slightly greater than the predicted loss to offset uncertainties. Next, a secondary tightening command is automatically generated based on the compensation torque value, and the secondary tightening operation is finally completed to ensure that the diaphragm maintains stable sealing performance throughout the entire operating cycle.

[0064] In some embodiments, a material stress relaxation mathematical model is established, which includes a stress decay time constant and material creep characteristic coefficients, including: An integral equation framework based on the Boltzmann superposition principle is established, which characterizes the influence of historical strain on the current stress state. A series combination structure of Maxwell elements and Kelvin-Voigt elements is configured in the integral equation framework to form a viscoelastic basic model, which includes the parameter set of spring elements and sticky pot elements. Perform tensile tests on the material, measure the stress decay curve of the diaphragm material over time under constant strain conditions, match the decay curve with the Maxwell element response in the viscoelastic basic model, and calculate the stress decay time constant. A stepped loading test was conducted, and the cumulative deformation response of the diaphragm material under multi-level incremental loads was recorded. The cumulative deformation response data was then input into the Kelvin-Voigt element equation in the viscoelastic basic model to solve for the material creep coefficient. Substitute the stress decay time constant and the material creep coefficient into the viscoelastic basic model to generate the initial stress relaxation mathematical model. The initial stress relaxation mathematical model outputs the predicted stress decay curve. Multiple sets of actual stress relaxation experimental data under different temperatures and strain levels were collected. An error function was constructed to evaluate the difference between the predicted stress decay curve and the actual stress relaxation experimental data. The least squares method was applied to iteratively adjust the component parameters in the viscoelastic basic model to generate an optimized material stress relaxation mathematical model.

[0065] Specifically, an integral equation framework based on the Boltzmann superposition principle is constructed. This framework uses convolutional integrals to express the dependence between the current stress state and the historical strain history. The integral kernel function represents the relaxation modulus of the material, and the time variable extends from the initial loading moment to the current moment, forming a complete stress-strain history mapping mechanism. Next, a three-element model structure is configured within the integral equation framework, consisting of a cascaded synthesis of a Maxwell element and a Kelvin-Voigt element, forming a viscoelastic basic model. The parameter set of this model includes the elastic modulus E1 and viscosity coefficient η1 of the Maxwell element, and the elastic modulus E2 and viscosity coefficient η2 of the Kelvin-Voigt element. Then, during the material tensile test, the stress decay process over time is recorded using a high-precision load sensor at a sampling frequency of 10 times per second, obtaining a complete stress-time decay curve. This experimental curve is then nonlinearly fitted to the theoretical response function σ(t) = σ0exp(-t / τ) of the Maxwell element, where τ is the stress decay time constant. The optimal fitting parameters are iteratively calculated using the Levenberg-Marquardt algorithm to determine the value of τ = η1 / E1. During the stepped loading test, new specimens were subjected to four strain levels of 5%, 10%, 15%, and 20% sequentially under the same temperature control environment, with each level held for 30 minutes and stress changes recorded. Particular attention was paid to the residual deformation during the unloading phase. The cumulative deformation response data was input into the differential equation solution process of the Kelvin-Voigt element. The Runge-Kutta numerical method was used to calculate the cumulative deformation rate of the material under sustained load, and the characteristic coefficients characterizing the long-term creep behavior of the material were obtained through inversion. The determined stress decay time constant and material creep characteristic coefficients were substituted into the parameter set of the viscoelastic fundamental model to construct an initial stress relaxation mathematical model. This model can generate corresponding stress response predictions for any strain history input and output theoretical stress decay curves. To optimize model accuracy, nine sets of independent stress relaxation experimental data of the diaphragm material were collected at three different predetermined temperatures (e.g., 15°C, 25°C, and 40°C) and different strain levels (e.g., 10%, 15%, and 20%). A weighted error function was constructed to quantify the sum of squared Euclidean distances between the predicted curve and the experimental data points. The four core parameters E1, η1, E2, and η2 were iteratively adjusted using the least squares method combined with a gradient descent strategy. The parameter update direction and step size were calculated in each iteration. Optimization was terminated when the error function value decreased by less than 0.1% for five consecutive iterations, generating the final mathematical model of material stress relaxation. This model exhibited an average relative error of less than 3% on the validation dataset and can accurately predict the long-term stress relaxation behavior of the diaphragm material under actual working conditions.

[0066] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section. It should be noted that those skilled in the art can make various improvements and modifications to this invention without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this invention.

[0067] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusivity.

[0068] The term "comprises" implies that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprises a..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

Claims

1. A method for tightening bolts to a pneumatic valve diaphragm, characterized in that, include: Measure the thickness of the diaphragm before tightening to obtain the original thickness of the diaphragm; Based on the original thickness of the diaphragm, the diaphragm is tightened and the compression ratio is calculated to obtain the compression ratio; When the compression ratio reaches the predetermined compression threshold, the tightening torque value at this time is recorded and corrected to obtain the corrected tightening torque; Based on the corrected tightening torque, the initial tightening torque, the final tightening torque, the initial re-tightening torque, and the final re-tightening torque are confirmed; The bolts of the diaphragm are tightened symmetrically for the first time using the initial tightening torque; The bolts of the diaphragm are then tightened a second time symmetrically using the final tightening torque. The bolts of the diaphragm are circumferentially tightened for the first time using the initial tightening torque; After the first circumferential tightening is completed, wait for a predetermined time, and then tighten the bolts of the diaphragm a second time with the final tightening torque.

2. The bolt torque tightening method for the pneumatic valve diaphragm according to claim 1, characterized in that, The measurement of the diaphragm thickness before tightening to obtain the original diaphragm thickness includes: Multiple points were selected on the surface of the diaphragm for thickness measurement, resulting in multiple measurement data. The original thickness of the diaphragm is obtained by calculating the average value of multiple measurement data.

3. The bolt torque tightening method for the pneumatic valve diaphragm according to claim 1, characterized in that, Based on the original thickness of the diaphragm, the clamping and compression ratio of the diaphragm are calculated, and the compression ratio is obtained including: The diaphragm is tightened using a predetermined initial torque; Based on the predetermined initial torque, the diaphragm is tightened by iteratively increasing the torque according to a predetermined first amplitude, and the corresponding compression ratio is measured and calculated after each tightening to obtain the first compression ratio, until the first compression ratio reaches the predetermined initial threshold. When the first compression ratio reaches the predetermined initial threshold, the diaphragm is tightened by iteratively increasing the torque according to the predetermined second amplitude. After each tightening, the corresponding compression ratio is measured and calculated to obtain the second compression ratio. When the second compression ratio reaches the predetermined compression threshold, the tightening torque value at this time is recorded.

4. The bolt torque tightening method for the pneumatic valve diaphragm according to claim 1, characterized in that, When the compression ratio reaches a predetermined compression threshold, the tightening torque value at this time is recorded and corrected. The corrected tightening torque then includes: When the compression ratio reaches the predetermined compression threshold, the tightening torque value at this time is recorded to obtain the theoretical tightening torque; The average number of diaphragm movements per minute is obtained, and a frequency correction factor is calculated based on the average number of diaphragm movements per minute. The theoretical fastening torque is corrected based on the frequency correction factor to obtain the corrected fastening torque.

5. The bolt torque tightening method for the pneumatic valve diaphragm according to claim 4, characterized in that, The frequency correction factor, calculated based on the average number of movements of the diaphragm per minute, includes: The frequency correction factor is calculated using the following formula: D=1+0.2× (1+f) Where D represents the frequency correction factor; f represents the average number of diaphragm movements per minute.

6. The bolt torque tightening method for the pneumatic valve diaphragm according to claim 1, characterized in that, The process of confirming the initial tightening torque, final tightening torque, initial re-tightening torque, and final re-tightening torque based on the modified tightening torque includes: Set the initial tightening torque to 40%-60% of the corrected tightening torque; Set the final tightening torque to 100% of the corrected tightening torque; Set the initial tightening torque to 100% of the corrected tightening torque; Set the final tightening torque to 75%-85% of the corrected tightening torque.

7. The bolt torque tightening method for the pneumatic valve diaphragm according to claim 1, characterized in that, Symmetrical fastening includes: Number all bolts distributed around the circumference in a clockwise or counterclockwise order, and set the same number for two symmetrical bolts; Tighten all bolts sequentially, tightening one bolt at a time and then another bolt with the same number.

8. The bolt torque tightening method for the pneumatic valve diaphragm according to claim 1, characterized in that, The circumferential compressive includes: Tighten all the bolts distributed around the circumference in a clockwise or counterclockwise order.

9. The bolt torque tightening method for the diaphragm of a pneumatic valve according to claim 1, characterized in that, Also includes: Obtain the tensile deformation data of the bolt; Calculate the bolt torque value based on the tensile deformation data; Plot a torque-time relationship curve based on the bolt torque values; The torque variation trend is analyzed based on the torque-time relationship curve, and the torque relaxation rate is calculated based on the torque variation trend. An early warning is triggered when the torque relaxation rate exceeds a preset threshold.

10. The bolt torque tightening method for the pneumatic valve diaphragm according to claim 1, characterized in that, Also includes: Data on the diaphragm fastening process and multi-dimensional attribute information are collected. The fastening process data includes the final compression rate and the final torque. The multi-dimensional attribute information includes the diaphragm's manufacturer, batch, model, material, service life, disassembly inspection results, and experience feedback information. A diaphragm performance database is established, and the fastening process data is associated with the corresponding multi-dimensional attribute information and stored in the diaphragm performance database; Cluster analysis was performed on historical data of diaphragms of the same model and material in the diaphragm performance database to generate data distribution characteristics; The correlation between different compression rates and the service life of the diaphragm was identified using artificial intelligence analysis tools. Based on the aforementioned correlation, the optimal compression ratio range for diaphragms of the same model and material is calculated.

Images