CN-121999940-A - Mechanical property test life prediction method and system

CN121999940ACN 121999940 ACN121999940 ACN 121999940ACN-121999940-A

Abstract

The disclosure relates to the technical field of mechanical property testing of materials, in particular to a mechanical property testing life prediction method and a mechanical property testing life prediction system. The method and the device realize full utilization of the overflow point data through the synergistic effect of hierarchical fitting, normal distribution correction and weighting treatment, and solve the problems of overflow point data waste and low treatment precision in the prior art. The method and the device have the advantages that the method and the device do not need to rely on experience parameters of specific materials, are wide in application range, and can be suitable for mechanical property test service life data processing scenes of different types of materials and different stress levels. The processing steps disclosed by the invention are clear in logic and strong in operability, are convenient for engineering technicians to apply practically, can effectively reduce the test cost, improve the reliability of life prediction, and have obvious engineering application value and economic value.

Inventors

LI YING
WANG JIAYOU
JIAO ZEHUI
XU RUIDA
HE YUHUAI
Nie Feiqing

Assignees

中国航发北京航空材料研究院

Dates

Publication Date: 20260508
Application Date: 20260130

Claims (10)

1. A mechanical property test life prediction method, the method comprising: Testing a plurality of samples under different stress levels, and taking logarithms of fracture life and overflow point data obtained by the test to obtain a fracture life logarithmic value and an overflow point data logarithmic value; Performing linear fitting on the numerical values of the stress fracture life of each stage with complete fracture to obtain the numerical value of the initial expected fracture life of the stress stage with overflow point data; Constructing a normal distribution function by taking the initial expected fracture life logarithmic value as a mean value and the logarithmic life standard deviation as a standard deviation; Calculating the total probability of a low region lower than the overflow data logarithmic value in the normal distribution function based on the position of the overflow data logarithmic value in the normal distribution function, and correcting the normal distribution function based on the total probability of the low region to obtain a corrected normal distribution function; calculating the total probability of a high area higher than the overflow point data logarithmic value in the corrected normal distribution function based on the position of the overflow point data logarithmic value in the corrected normal distribution function; Calculating an interval upper limit by taking the overflow point data logarithmic value as an interval lower limit, taking half of the total probability of a high area as an interval probability, taking the interval upper limit as the residual life logarithmic value of the highest probability of the overflow point data, And adding the overflow point data logarithmic value and the overflow point data highest probability residual life logarithmic value to obtain the overflow point data expected fracture life logarithmic value and obtain the predicted life.
2. The method for predicting mechanical property test life according to claim 1, wherein, Several samples were tested at different stress levels, including: Dividing a plurality of samples into a plurality of groups, respectively carrying out mechanical property test under different stress levels of multiple stages, and keeping the same stress level for each group of samples to obtain fracture life or overflow point data, wherein the number of samples under the stress level of at least two stages in the multiple stress levels is larger than or equal to the threshold number.
3. The method for predicting mechanical property test life according to claim 1, wherein, Linear fitting of the values of stress fracture life at each stage of the complete fracture to obtain initial expected fracture life log values for stress levels at which spill point data occurs, comprising: Based on fracture life data under different stresses, which are all fracture life logarithmic values, performing linear fitting by taking a stress level as an abscissa and fracture life as an ordinate to obtain a fitting equation; substituting the stress level of the data of the overflow points into a fitting equation to obtain the corresponding initial expected fracture life logarithmic value.
4. The method for predicting mechanical property test life according to claim 1, wherein, Performing linear fitting on the logarithmic standard deviation of the stress fracture life of each stage of complete fracture to obtain the logarithmic standard deviation of the stress level of the overflow point, wherein the method comprises the following steps: Based on the logarithmic life standard deviation under different stresses, which are all fracture life logarithmic values, performing linear fitting by taking the stress level as an abscissa and the logarithmic life standard deviation as an ordinate to obtain a fitting equation; Substituting the stress level of the data with the overflow points into a fitting equation to obtain the corresponding logarithmic life standard deviation.
5. The method for predicting mechanical property test life according to claim 1, wherein, Based on the position of the overflow data logarithmic value in the normal distribution function, calculating the total probability of the low region lower than the overflow data logarithmic value in the normal distribution function, comprising: Substituting the overflow data logarithmic value into a normal distribution function, and calculating an integral value lower than the overflow data logarithmic value in the normal distribution function to obtain the total probability of the low region.
6. The method for predicting mechanical property test life according to claim 1, wherein, Correcting the normal distribution function based on the total probability of the low region to obtain a corrected normal distribution function, comprising: and multiplying the normal distribution function by a factor to obtain a corrected normal distribution function, wherein the factor is equal to the total probability of the 1+ low region.
7. The method for predicting mechanical property test life according to claim 1, wherein, Calculating a total probability of a high region above the overflow point data logarithmic value in the modified normal distribution function, comprising: substituting the overflow data logarithmic value into a corrected normal distribution function, and calculating an integral value higher than the overflow data logarithmic value in the corrected normal distribution function to obtain the total probability of the high region.
8. The method for predicting mechanical property test life according to claim 1, wherein, Calculating an interval upper limit by taking the overflow point data logarithmic value as an interval lower limit and taking half of the total probability of a high area as interval probability, and taking the interval upper limit as the residual life logarithmic value of the highest probability of the overflow point data, wherein the method comprises the following steps: And taking the overflow point data logarithmic value as a lower interval limit, taking half of the total probability of the high area as the interval probability, and inquiring a normal distribution table or obtaining an upper interval limit value through integral calculation, namely the residual life logarithmic value of the highest probability of the overflow point data.
9. The method for predicting mechanical property test life according to claim 1, wherein, Further comprises: and (3) taking the stress level as an abscissa and the fracture life logarithmic value as an ordinate, and performing linear regression fitting to obtain a life prediction equation for predicting the life of the material under different stress levels.
10. A mechanical property test life prediction system, comprising: the device comprises a test unit, a normal distribution function construction unit, a normal distribution function correction unit, a highest probability residual life logarithmic value calculation unit and a predicted life calculation unit; The test unit is used for testing a plurality of samples under different stress levels, taking the logarithm of fracture life and overflow point data obtained by the test, and obtaining a fracture life logarithmic value and an overflow point data logarithmic value; The normal distribution function construction unit is used for carrying out linear fitting on the numerical values of the stress fracture life of each stage with complete fracture to obtain the numerical value of the initial expected fracture life of the stress stage with overflow point data; The normal distribution function construction unit is also used for constructing a normal distribution function by taking the initial expected fracture life logarithmic value as a mean value and the logarithmic life standard deviation as a standard deviation; The normal distribution function correction unit is used for calculating the total probability of a low region lower than the overflow point data logarithmic value in the normal distribution function based on the position of the overflow point data logarithmic value in the normal distribution function, and correcting the normal distribution function based on the total probability of the low region to obtain a corrected normal distribution function; the maximum probability remaining life logarithmic value calculating unit calculates the total probability of a high area higher than the overflow point logarithmic value in the corrected normal distribution function based on the position of the overflow point data logarithmic value in the corrected normal distribution function; the maximum probability remaining life logarithmic calculation unit is also used for calculating an interval upper limit by taking the overflow point data logarithmic value as an interval lower limit and taking half of the total probability of the high area as interval probability, taking the interval upper limit as the overflow point data maximum probability remaining life logarithmic value, And the predicted life calculation unit is used for adding the overflow point data logarithmic value and the overflow point data highest probability residual life logarithmic value to obtain the overflow point data expected fracture life logarithmic value and predicting the life based on the overflow point data expected fracture life logarithmic value.

Description

Mechanical property test life prediction method and system Technical Field The disclosure relates to the technical field of mechanical property testing of materials, in particular to a mechanical property testing life prediction method and a mechanical property testing life prediction system. Background In the field of mechanical property testing of materials, the fracture life of the materials is a key index for evaluating the reliability and the service life of the materials, and the materials are widely applied to the scenes of severe requirements on the performance of the materials such as aerospace, automobile manufacturing, engineering machinery and the like, and the accuracy of test data directly influences the rationality of product design, safety evaluation and service life prediction. In the prior art, overflow points with the service life not broken and interrupted in advance in the mechanical property test are commonly found in high cycle fatigue and endurance tests of materials, and refer to test data points corresponding to the fact that a sample is not broken and stopped in advance after reaching a preset cycle (such as 10 7 times). In such tests, if the materials are tested according to the actual breaking cycle times, part of the high-toughness materials may need to undergo more than 10 7 cycles, the test period may be prolonged to several months, and the test cost and the time cost are greatly increased. The industry therefore generally presets a cycle threshold value, which is reached by the test specimen, which interrupts the test in advance without breaking, thus creating an overflow point. Meanwhile, temperature and humidity fluctuation of a test environment, sudden faults of a loading device and the like can force the test to be interrupted in advance, unexpected overflow points are formed, and the overflow points also need to be included in the data processing category. In the conventional technology, aiming at the processing mode of overflow points, which is simple and rough, obvious defects exist, and the requirement of engineering on data accuracy is difficult to meet, so that the method becomes a core driving force for development of the related background technology, namely, the problem of resource waste of a direct discarding method is solved, and the method considers that the overflow points are not actually broken for service life and are directly removed. However, the overflow point usually comes from a low-stress test working condition with high cost and long period, and contains anti-fatigue information of the material under the high cycle, and the discarding of the material can cause the waste of data resources, and can further enable the life evaluation of the corresponding stress level to be far lower than an actual value, so that the subsequent performance analysis is influenced. The problem of estimation deviation of the equivalent breaking point method is that the method treats the overflow point as a broken data point at a preset threshold value together with the breaking point. However, the actual break life corresponding to the overflow point may be between a preset threshold and infinity, and such equivalent treatment may significantly underestimate the material fatigue limit. For example, the S-N curve fitted according to the method has a fatigue life corresponding to the lowest stress level often lower than the preset 10 7 weeks, and if the aeroengine component is designed according to the data, the structural design is too conservative, and the running efficiency of the equipment is reduced. The severe requirements of engineering application on data precision are extremely high, and the reliability requirements of various key fields on material mechanical property data are extremely high, so that the overflow point data processing technology is upgraded. In the aerospace field, the fatigue life of parts such as an aircraft engine blade, a fuselage frame and the like are directly related to flight safety, and in the building fields such as bridges, wind power and the like, the fatigue performance of materials determines the service cycle and maintenance cost of a structure. If the fatigue life assessment is distorted due to improper treatment of the overflow points, safety accidents are caused or economic loss of over design is caused. For example, if the fatigue limit of the material is underestimated, thicker materials may be selected for design, the dead weight of the aircraft or the construction cost of the bridge are increased, and if the fatigue limit of the material is overestimated, the structure may fail in advance. In summary, the above-mentioned prior art solutions cannot meet the requirements of high-efficiency utilization, high correction accuracy and wide application range of the overflow point data at the same time, so a life overflow point data processing method capable of fully mining the overflow point data value, improving the data processing