A method for predicting the lifespan of a titanium alloy intermediate casing
By establishing a model of the microstructure and performance degradation of titanium alloy intermediate casings, and combining grey relational analysis and non-destructive testing, the problems of non-destructive assessment and multi-level life prediction of titanium alloy intermediate casings were solved, achieving high-precision life prediction and maintenance decision support.
Patent Information
- Application Number
- CN202611589799.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-14
- Publication Date
- 2026-11-10
AI Technical Summary
Existing technologies lack a dedicated life database for titanium alloy intermediate casings, making it difficult to perform non-destructive testing and predict effective remaining life. Furthermore, they fail to consider the differences in stress regions, resulting in significant deviations between life prediction results and actual conditions.
By statistically analyzing the grain size and tensile strength of each region of the component, a quantitative correlation model between microstructure characteristics and performance degradation is established. Grey relational analysis is used to screen health indicators, and combined with non-destructive testing and multi-level life index setting, non-destructive assessment and multi-level life prediction are achieved.
It improves the scientific rigor and accuracy of life prediction, enables non-destructive remaining life prediction of in-service components, provides multi-level life indicators to support maintenance decisions, and enhances the flexibility and precision of prediction.
Smart Images

Figure CN123069595A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of metal parts health management and life assessment technology, specifically relating to a method for predicting the life of titanium alloy intermediate casings based on microstructure characteristics and performance degradation. Background Technology
[0002] The service life of castings determines the service life of the entire aero-engine system. As a key load-bearing component of the engine, the service life assessment of the titanium alloy intermediate casing under high temperature and complex stress conditions is a core technical issue for ensuring the safe and reliable operation of the engine.
[0003] In existing technologies, the life prediction of key components mainly faces the following problems: (1) There is a lack of a dedicated life database for titanium alloy intermediate casings, making it difficult to directly assess their service life; (2) Traditional life prediction methods mainly rely on component dissection analysis, but components after installation and commissioning are often unable to be destructively tested, resulting in a lack of effective means to predict the remaining life of in-service components. (3) Existing methods have failed to establish a quantitative correlation model between tissue characteristics and performance degradation, making it difficult to achieve lifetime prediction based on non-destructive testing; (4) The difference in the contribution of different stress regions to life was not fully considered, resulting in a large deviation between the life prediction results and the actual situation.
[0004] Therefore, there is an urgent need for a method to predict the life of titanium alloy intermediate casings that can combine anatomical analysis and non-destructive assessment and take into account the differences in stress regions. Summary of the Invention
[0005] The technical problem to be solved by this invention: (1) How to establish a quantitative relationship between the microstructure characteristics of titanium alloy intermediate casing and the degradation of tensile strength; (2) How to predict the remaining life of in-service components based on non-destructive testing methods (such as surface pickling and grain size statistics); (3) How to reasonably set multi-level life indicators such as stress-sensitive life, overhaul life, failure life and ultimate life based on the differences in different stress areas.
[0006] The technical solution of this invention: A method for predicting the lifespan of a titanium alloy intermediate casing, characterized by the following steps: A. Statistical analysis of grain size in each region of the component to construct a sequence of microstructure characteristics; B. Measure the tensile strength of each region of the component and construct a reference series; C. Calculate the correlation coefficient between the tissue characteristic series and the reference series at each point, and take the tissue characteristic corresponding to the maximum value of the correlation coefficient as the tissue health index H of the component; D. The tensile strength and the tissue health index H were linearly fitted to obtain the fitting relationship; E. Statistically analyze the service life of components with different service cycles, analyze the tensile strength of each region of each component, and mathematically fit the tensile strength of each region with the service life to obtain the performance degradation fitting curve. The intersection of the performance degradation fitting curve and the service stress level line is the fixed life of that region. F. Identify the surface microstructure features of the component to be predicted to obtain the grain size data of the dangerous area of the component, extract the microstructure health index H and substitute it into the fitting relationship obtained in step D to calculate the predicted tensile strength; substitute the predicted tensile strength into the performance degradation fitting curve in step E to calculate the remaining life.
[0007] In step A, the grain size of each region of the component is statistically analyzed, and the resulting microstructure features include: mean grain size (Mean), expected grain size distribution (Xc), full width at half maximum (FWHM) of grain size distribution, and the grain size corresponding to the cumulative probability levels P50, P60, P70, P80, P90, P95, and P99.
[0008] In step C, the organizational feature series and the reference series are first dimensionless. Then, grey relational analysis is used to calculate the degree of correlation between the two series at each point. The average correlation coefficient of each organizational feature is calculated, and the organizational feature with the largest average correlation coefficient is taken as the component organizational health index H.
[0009] In step E, a performance degradation fitting curve is obtained by mathematical fitting using a single exponential decay function model, a quadratic polynomial model, or a linear model.
[0010] In step E, the high stress zone, second-highest stress zone, second-lowest stress zone and low stress zone of the component are determined according to the set of service stress levels of the component, and the stress-sensitive life, overhaul life, failure life and ultimate life of the component are set accordingly.
[0011] For the high stress level region of a component, the x-coordinate of the intersection point of the degradation curve and the corresponding threshold level line is taken as the stress-sensitive life of the component; for the second highest stress level region of a component, there are multiple x-coordinates of the intersection point of the degradation curve and the threshold level line, and the service life with the smallest x-coordinate is taken as the overhaul life; for the second lowest stress level region of a component, there are multiple x-coordinates of the intersection point of the degradation curve and the threshold level line, and the service life with the smallest x-coordinate is taken as the failure life; for the lowest stress level region of a component, there are multiple x-coordinates of the intersection point of the degradation curve and the threshold level line, and the service life with the largest x-coordinate is taken as the ultimate life.
[0012] The beneficial effects of this invention are: (1) By using grey relational analysis, the tissue health indicators most closely related to tensile strength are automatically selected from multiple tissue characteristic indicators, avoiding the bias caused by subjective selection and improving the scientificity and accuracy of life prediction. (2) A two-stage life prediction framework of “anatomical modeling + non-destructive evaluation application” was established, which enables the remaining life prediction of in-service components without destructive testing, and solves the engineering problem that the installed components cannot be dissected. (3) Based on the differences in different stress areas, multi-level life indicators such as stress-sensitive life, overhaul life, failure life and ultimate life were set, providing differentiated basis for different maintenance decision scenarios (such as condition-based maintenance, timed overhaul and ultimate scrapping). (4) It provides a variety of performance degradation fitting models (single exponential decay, quadratic polynomial, linear), and users can select the optimal model according to the actual data characteristics, which improves the flexibility and accuracy of prediction. Attached Figure Description
[0013] Figure 1 This is a flowchart of the present invention.
[0014] Figure 2 This is a schematic diagram of the performance degradation curve. Detailed Implementation
[0015] Example 1: This invention establishes a lifetime prediction model based on the following physical assumptions: (1) The grain size distribution in different regions of components with different service durations conforms to a skewed distribution; (2) During service, the residual stress generated by the component is small, and the residual stress and its redistribution have almost no impact on the service life of the component; (3) During the service process, the stress distribution in each region of the component is consistent with the numerical simulation analysis results and does not change with the evolution of service environment, material properties and other conditions; (4) Using tensile strength as the performance judgment index for component life prediction, when the tensile strength of a certain area degrades to the service stress level of that area during service, that area is judged to be in failure.
[0016] Step A: Statistically analyze the grain size of each region of the component; the arithmetic mean of these values is the average grain size. The calculation method is as follows:
[0017] In the formula, Represents the equivalent diameter of a grain, in μm; The number of grains counted in this region.
[0018] The grain size distribution in each region is fitted using a Gaussian function, and the grain size distribution can be represented by the following function:
[0019] In this function, the horizontal axis x represents the grain size, and the vertical axis y represents the probability density or frequency of that grain size appearing in the entire statistical sample. Here, y0, A, x0, and w are parameters obtained from Gaussian fitting. Based on the characteristics of Gaussian distribution and the physical meaning of its function, y0 represents the baseline of the distribution curve, A represents the amplitude of the peak value, x0 represents the expected value of the statistical sample, and w represents the standard deviation of the statistical sample. Therefore, the expected value of the grain size distribution... The calculation method is as follows:
[0020] The method for calculating the full width at half maximum (FWHM) of grain size distribution is as follows:
[0021] The grain size corresponding to different levels of the cumulative probability of grain size distribution in each region. To obtain this, the grain size first needs to be determined. The sequence is sorted from smallest to largest to determine the position corresponding to the cumulative probability level of a certain grain size distribution. If the position is a decimal, it is rounded to the nearest integer to extract the grain size corresponding to that position. The calculation method is as follows:
[0022] In the formula, represents a mathematical sequence of grain sizes arranged in ascending order, in μm; xx represents the cumulative probability level of the grain size distribution. This represents the number of grains counted in this region. Based on this, indices such as P50, P60, P70, P80, P90, P95, and P99 can be calculated.
[0023] Step B selects tensile strength as the reference series Y, measures the tensile strength of each region of the component, and constructs the reference series.
[0024] Step C first involves dimensionless transformation of the organizational feature series and the reference series, then using grey relational analysis to calculate the degree of correlation between the two series at each point, calculating the average correlation coefficient of each organizational feature, and taking the organizational feature with the largest average correlation coefficient as the component's organizational health index H.
[0025] If we take the tissue feature sequence as the comparison sequence X, then the reference sequence is: The sequences to be compared are:
[0026] In the formula x 11 ~x nm To compare the elements in sequence X, y1~y n These are elements in the reference sequence Y, where the subscripts n and m represent the row and column numbers, respectively.
[0027] Perform dimensionless transformation of variables and organize the characteristic sequence. With tensile strength series The data in this paper may be difficult to directly compare and analyze between indicators due to differences in units of measurement. This paper uses an interval generation method to make the variables dimensionless. The interval generation formula is shown below: It is the dimensionless y1~y n value, x is the dimensionless version 11 ~x nm The values are represented by the subscripts i and j, which refer to the row and column, respectively, and the subscripts max and min, which represent the maximum and minimum values, respectively.
[0028] In the formula, This represents the minimum difference between two levels, that is, the minimum value among all absolute differences. This represents the maximum difference between two levels, that is, the maximum value among all absolute differences. The resolution coefficient, A smaller value corresponds to a greater resolution, generally speaking. The value is 0.5.
[0029] Calculate the correlation coefficient. It is an organizational characteristic sequence With tensile strength sequence When comparing the correlation values at various points, it is necessary to consider the sequence of organizational characteristics. correlation coefficient By integrating these values, i.e., taking their average, a certain organizational characteristic is obtained. correlation The calculation formula is: The set of correlation degrees corresponding to different organizational feature sequences for: In the formula,j =1,2,3……10, corresponding to tissue health indicators in turn. Mean , Xc , FWHM , P 50. P60、P70、P80、 P90、P95、P99 Rank the relevance. The magnitude of the values reflects the comparison of sequences. With reference sequence The degree of closeness of the connection, The larger the value, the more likely it is to represent the j-th feature in the organizational feature sequence. The closer the correlation with tensile strength, the higher the correlation. The organizational characteristics corresponding to the maximum value are used as an indicator of the health of the component tissue. H The mathematical expression is: Step D involves linearly fitting the tensile strength to H to obtain the fitting relationship. , H n This is a tissue health indicator for the nth component, expressed in μm. k n 、q n The parameters are used for the linear fitting of the corresponding tissue-performance relationship.
[0030] Step E: Let the set of service durations for components with different service cycles be . t Then we have: In the formula, This represents the service duration of the nth component, measured in hours (h).
[0031] Let the set of tensile strengths of all anatomical regions of a component at different service cycles be . s Then we have: In the formula, This represents the tensile strength of the m-th region of the n-th component obtained through anatomical analysis, expressed in MPa.
[0032] A single exponential decay function is used to fit the degradation characteristics of the tensile strength of each region of the component during service. The expression of the single exponential decay model is as follows: In the formula, x The service life of the component is substituted into the set. t The unit is h; y For the tensile strength of a certain region of the component, that is, substitute into the set s A certain column The unit is MPa; f , g , h These are the fitting parameters for a single exponential decay function.
[0033] A quadratic polynomial is used to fit the degradation characteristics of the tensile strength of each region of the component during service. The expression of the quadratic polynomial model is as follows: In the formula, x The service life of the component is substituted into the set. t The unit is h; y For the tensile strength of a certain region of the component, that is, substitute into the set s A certain column The unit is MPa; c , d , e These are the fitting parameters for a quadratic polynomial. For example... Figure 2 This demonstrates the performance degradation curves obtained by fitting quadratic functions to different regions.
[0034] A linear model is used to fit the degradation characteristics of the tensile strength of each region of the component during service. The linear model expression is as follows: In the formula, x The service life of the component is substituted into the set. t The unit is h; y For the tensile strength of a certain region of the component, that is, substitute into the set s A certain column The unit is MPa; a , b These are the parameters for linear fitting.
[0035] Let the set of service stress levels for all anatomical regions of the component be . s d Then we have: In the formula, This represents the service stress level of the m-th region, in MPa.
[0036] The x-coordinate of the intersection point of the performance degradation curve of a certain region of a component and its corresponding service stress level line represents the service life duration of that region. When using a single exponential decay function model, the set of service life durations for each region of the component is... have: In the formula, , , The parameters are obtained by fitting the single exponential decay function to the performance degradation curve of the m-th region. The duration is given in hours (h) for determining the lifetime of the m-th region using a single exponential decay function model.
[0037] The x-coordinate of the intersection point of the performance degradation curve of a certain region of a component and its corresponding service stress level line represents the service life duration of that region. When a quadratic function model is used, the set of service life durations for each region of the component is... have: In the formula, , , The parameters are obtained by fitting a quadratic function to the performance degradation curve of the m-th region. The duration is the time obtained by using a quadratic function model to determine the lifetime of the m-th region, in hours.
[0038] The x-coordinate of the intersection point of the performance degradation curve of a certain region of a component and its corresponding service stress level line represents the service life duration of that region. When a linear function model is used, the set of service life durations for each region of the component is... have: In the formula, , The parameters are obtained by fitting a linear function to the performance degradation curve of the m-th region. The duration is the lifetime obtained by using a linear function model to determine the lifetime of the m-th region, in hours.
[0039] Based on the set of component service stress levels s d The expressions for the service stress levels of the high-stress zone, the second-highest stress zone, the second-lowest stress zone, and the low-stress zone of a component are as follows: , , , Based on this, the stress-sensitive life, overhaul life, failure life, and ultimate life of the components are set accordingly, and the characteristic life expressions are as follows: , 、 , .
[0040] For the high-stress region (Region I) of the component, the x-coordinate of the intersection point of the degradation curve and the corresponding threshold level line is taken as the stress-sensitive lifetime of the component. Taking the single exponential decay function model for lifetime determination as an example, we have: For the second-highest stress level region (region F, region L) of a component, there are multiple x-coordinates at the intersection of its degradation curve and the threshold level line. To ensure the safe operation of the component, the service life with the smallest x-coordinate is taken as the overhaul life. Taking the single exponential decay function model for life determination as an example, we have: For the second lowest stress level region of the component (regions E and J), the method for determining the overhaul life is the same as that for the overhaul life. The service life with the smallest abscissa is taken as the failure life. Taking the single exponential decay function model for life determination as an example, we have: For the low-stress level region (other regions) of a component, there are multiple x-coordinates at the intersection of its degradation curve and the threshold level line. To determine the service life limit of the component, the service life with the largest x-coordinate is taken as the limit life. Taking the single exponential decay function model for life determination as an example, we have: When using quadratic function and sequential function models for lifetime determination, the calculation method is consistent with that of the single exponential decay function model, and will not be repeated here. Summarizing the characteristic lifetimes obtained from the three function models yields the component characteristic lifetime set based on anatomical analysis. for: Step F: After large and complex titanium alloy casing components are actually installed and put into service, they often cannot be dissected and analyzed. However, the microstructure of the components can be non-destructively monitored through surface pickling. By combining the microstructure health indicators under different service durations and the corresponding microstructure-property relationships, the tensile strength of the corresponding area can be predicted. Therefore, the predicted tensile strength of a certain area of a component with a certain service duration under non-destructive monitoring conditions is... The calculation method is as follows: In the formula, This is a tissue health indicator for the nth component, expressed in μm. , The corresponding tissue-property relationship linear fitting parameters are used. The tensile strength prediction value of the component obtained from dissection is replaced by the tensile strength test value obtained from non-destructive evaluation, and the result is substituted into step E to obtain the performance degradation curve, thus realizing the prediction of the service life of the component based on non-destructive evaluation.
[0041] Example 2: This example takes the ZTC4 cast titanium alloy intermediate casing of a certain type of aero-engine as an example to describe in detail the specific implementation process of the titanium alloy intermediate casing life prediction method of the present invention.
[0042] Step A: Organizational Feature Data Collection and Organizational Feature Series Construction Five ZTC4 titanium alloy intermediate casing castings from the same batch were selected as samples. Each intermediate casing was divided into multiple regions according to stress distribution characteristics, including the outer ring thick-walled region, the support plate rounded corner region, the inner ring thin-walled region, the gating region, the riser region, and the normal regions of the front and rear end faces. Among them, the support plate rounded corners are the main areas of stress concentration. Metallographic specimens were cut from each region, and after mounting, grinding, polishing, and etching, the microstructure was observed using an optical microscope and a scanning electron microscope (SEM).
[0043] Image analysis software was used to quantitatively analyze the grain size in each region. For each region, 10 fields of view were randomly selected, and the size data of no less than 200 grains were collected for each field of view. Based on the statistical results, the following microstructural characteristics were constructed: mean grain size (Mean), expected grain size distribution (Xc), full width at half maximum (FWHM) of grain size distribution, and the grain sizes corresponding to cumulative probability levels P50, P60, P70, P80, P90, P95, and P99.
[0044] Taking a certain region as an example, the organizational characteristic series obtained by statistics is as follows: X=(Mean,Xc,FWHM,P50,P60,P70,P80,P90,P95,P99).
[0045] Step B: Tensile strength measurement and reference series construction In adjacent areas at the same slice location, the tensile strength of each region is measured. Three parallel specimens are tested for each condition, and the average value is taken as the tensile strength value of that region, thus constructing a tensile strength sequence. The tensile strength of each region of all samples is measured to construct a complete reference sequence set Y.
[0046] Step C: Grey Relational Analysis and Determination of Organizational Health Indicator H The grey relational analysis method was used to calculate the correlation between various tissue characteristic indicators and tensile strength. The specific calculation process is as follows: First, the number series are dimensionless using formulas (8) and (9) in Example 1. The mean value is used to calculate the absolute difference between each comparison series and the reference series at each point. The correlation coefficient is calculated using formula (10). The gray correlation degree ranking of each organizational characteristic index and tensile strength is shown in Table 1.
[0047] Table 1. Grey Relationship between Various Tissue Characteristic Indicators and Tensile Strength
[0048] As shown in Table 1, the correlation between grain size and tensile strength is the highest for P99 (0.892). Therefore, the grain size corresponding to P99, which corresponds to the maximum correlation coefficient, is taken as the component's structural health indicator H.
[0049] Step D: Linear fitting of tensile strength and tissue health index H Using the tissue health index H determined in step C as the independent variable x, and the tensile strength σ as the variable... b Let y be the dependent variable. Perform linear regression fitting on all sample data (Equation 18) to obtain the fitted relationship σ. b =-3.86×H+1245.6.
[0050] Step E: Determination of performance degradation fitting curve and service life duration Twelve ZTC4 titanium alloy intermediate casings with different service periods were selected, namely 0h, 500h, 1000h, 1500h, 2000h, 2500h, 3000h, 3500h, 4000h, 4500h, 5000h, and 5500h. Each intermediate casing was fully dissected, and the tensile strength of each region was measured according to step B. A single exponential decay function model (Equation 16) was used to mathematically fit the tensile strength of each region to the service period, yielding the following fitted relationship: .
[0051] Regional Classification: The service stresses of each region are ranked and classified into high-stress, second-highest-stress, second-lowest-stress, and low-stress regions. The specific method for regional classification is as follows: The maximum service stress σ in each region is taken. max The service stress in the high-stress zone is ≥0.8σ. max , 0.8σ max >The service stress in the second highest stress zone is ≥0.6σ max , 0.6σ max >The service stress in the second lowest stress zone is ≥0.4σ max , 0.4σ max >Service stress in low-stress areas.
[0052] Step F: For the ZTC4 titanium alloy intermediate casing to be predicted in service, use a portable metallographic microscope or replica film technology to identify the microstructure characteristics of the component surface, focusing on obtaining grain size data of the critical areas of the component (i.e., high stress areas, at the rounded corners of the support plate).
[0053] The grain size corresponding to P99 in the dangerous area of a certain component to be predicted is measured to be H=142.5μm.
[0054] (1) Extract the tissue health index H and substitute it into the fitting relationship in step D to calculate the predicted tensile strength σ.pred : σ pred =-3.86×142.5+1245.6=695.6MPa (2) The predicted tensile strength σ pred Substitute the performance degradation fitting curve of this region (high stress zone) from step E: σ(t) = 520 + 372·exp(-t / 2820) Let σ(t) = σ pred =695.6MPa, solve for t: 695.6 = 520 + 372·exp(-t / 2820) t=2118 The result indicates that the current effective service life of the component to be predicted (adjusted for performance degradation) is approximately 2118 hours.
[0055] (3) Calculate the remaining lifespan: The remaining life is calculated based on the stress-sensitive lifetime (2650h) of this region: Remaining service life = Stress-sensitive service life - Current equivalent service life = 2650 - 2118 = 532h This means that the component to be predicted can continue to serve safely for about 532 hours in the high-stress area of the support plate's rounded corners, and should be inspected when it reaches 2650 hours.
[0056] If the failure life (5200h) is used as the benchmark, then the remaining life = 5200 - 2118 = 3082h. In practical engineering applications, the appropriate life benchmark can be selected for remaining life assessment based on different maintenance strategies.
Claims
1. A method for predicting the lifespan of a titanium alloy intermediate casing, characterized in that... Includes the following steps: A. Statistical analysis of grain size in each region of the component to construct a sequence of microstructure characteristics; B. Measure the tensile strength of each region of the component and construct a reference series; C. Calculate the correlation coefficient between the tissue characteristic series and the reference series at each point, and take the tissue characteristic corresponding to the maximum value of the correlation coefficient as the tissue health index H of the component; D. The tensile strength and the tissue health index H were linearly fitted to obtain the fitting relationship; E. Statistically analyze the service life of components with different service cycles, analyze the tensile strength of each region of each component, and mathematically fit the tensile strength of each region with the service life to obtain the performance degradation fitting curve. The intersection of the performance degradation fitting curve and the service stress level line is the fixed life of that region. F. Identify the surface microstructure features of the component to be predicted to obtain the grain size data of the dangerous area of the component, extract the microstructure health index H and substitute it into the fitting relationship obtained in step D to calculate the predicted tensile strength; substitute the predicted tensile strength into the performance degradation fitting curve in step E to calculate the remaining life.
2. The method for predicting the lifespan of a titanium alloy intermediate casing according to claim 1, characterized in that: In step A, the grain size of each region of the component is statistically analyzed, and the resulting microstructure features include: mean grain size (Mean), expected grain size distribution (Xc), full width at half maximum (FWHM) of grain size distribution, and the grain size corresponding to the cumulative probability levels P50, P60, P70, P80, P90, P95, and P99.
3. The method for predicting the lifespan of a titanium alloy intermediate casing according to claim 1, characterized in that: In step C, the organizational feature series and the reference series are first dimensionless. Then, grey relational analysis is used to calculate the degree of correlation between the two series at each point. The average correlation coefficient of each organizational feature is calculated, and the organizational feature with the largest average correlation coefficient is taken as the component organizational health index H.
4. The method for predicting the lifespan of a titanium alloy intermediate casing according to claim 1, characterized in that: In step E, a performance degradation fitting curve is obtained by mathematical fitting using a single exponential decay function model, a quadratic polynomial model, or a linear model.
5. The method for predicting the lifespan of a titanium alloy intermediate casing according to claim 1, characterized in that: In step E, the high stress zone, second-highest stress zone, second-lowest stress zone and low stress zone of the component are determined according to the set of service stress levels of the component, and the stress-sensitive life, overhaul life, failure life and ultimate life of the component are set accordingly.
6. The method for predicting the lifespan of a titanium alloy intermediate casing according to claim 5, characterized in that: For the high stress level region of a component, the x-coordinate of the intersection point of the degradation curve and the corresponding threshold level line is taken as the stress-sensitive life of the component; for the second highest stress level region of a component, there are multiple x-coordinates of the intersection point of the degradation curve and the threshold level line, and the service life with the smallest x-coordinate is taken as the overhaul life; for the second lowest stress level region of a component, there are multiple x-coordinates of the intersection point of the degradation curve and the threshold level line, and the service life with the smallest x-coordinate is taken as the failure life; for the lowest stress level region of a component, there are multiple x-coordinates of the intersection point of the degradation curve and the threshold level line, and the service life with the largest x-coordinate is taken as the ultimate life.
